Derek Kennet, J. Varaprasada Rao, and M.

Kasturi Bai
Excavations at Paithan, Maharashtra
Beyond Boundaries

Religion, Region, Language and the State

Edited by
Michael Willis, Sam van Schaik
and Lewis Doney

Volume 5
Derek Kennet, J. Varaprasada Rao and M. Kasturi Bai

Excavations at Paithan,
Maharashtra

Transformations in Early Historic and Early Medieval India

With contributions by
Shailendra Bhandare, Dorian Fuller, Jason Hawkes,
Peter Rowley-Conwy, P. K. Thomas and Michael Willis
Published with support of the European Research Council
Beyond Boundaries: Religion, Region, Language and the State
(Project No. 609823)

ISBN 978-3-11-065353-3
e-ISBN (PDF) 978-3-11-065354-0
ISSN 2510-4446

Library of Congress Control Number: 2019952751

Bibliographic information published by the Deutsche Nationalbibliothek

The Deutsche Nationalbibliothek lists this publication in the Deutsche
Nationalbibliografie; detailed bibliographic data are available in the Internet
at http://dnb.dnb.de.

© 2020 Walter de Gruyter GmbH, Berlin/Boston

Cover: “Lintel,” circa 475 CE, in Sārnāth, Uttar Pradesh, India. Photograph by Michael Willis
Typesetting: Michael Peschke, Berlin
Printing and binding: CPI books GmbH, Leck

www.degruyter.com
This volume is dedicated to the fond memory of Shri Ajai Shankar (1943–2000), Director General of the Archaeological
Survey of India from September 1996, who encouraged and supported this project. His dear friendship, open mind, and
deep love of Indian culture are remembered and missed by those who knew him.
Preface
This volume reports on the excavations carried out at seasons of yearly excavation took place from 1996 until
the Early Historic and Medieval site of Paithan in Maha- 1999.
rashtra between 1996 and 1999 as part of a collabora-
tive project between the Archaeological Survey of India
(ASI) and the Society for South Asian Studies (SSAS) of Related Publications
the British Academy (now the British Association for
South Asian Studies or BASAS). In the course of the Paithan project, two lengthy unpub-
The town of Paithan is known to have been an impor- lished interim reports were produced and circulated to
tant Satavahana centre and must also have had something relevant parties (Rao et al. 1998; Rao and Kennet 1999).
of an international reputation in the Early Historic period These incorporate the field reports written by the trench
as it is one of the few inland sites in India mentioned in supervisors at the end of each field season. Copies
the mid 1st-century AD Periplus of the Erythraean Sea. It is of these reports are lodged with the ASI in Delhi and
also known to have been an important centre in the Medi- also at the Goa Circle of the ASI, and are also lodged in
eval period. This is reflected in the fact that it remains a Durham University library in the UK. They provide the
centre of religious pilgrimage to the present day, attract- basis of much of Part I of the present monograph which
ing pilgrims from across Maharashtra and beyond to its describes the site and the excavations. After the end of
annual festival. the project a short paper was published by Kennet and
Some limited archaeological excavation had been Rao (2003) in South Asian Studies summarizing the
conducted at the site in 1937 and later in 1965, resulting development of the two Early Hindu brick temples at the
in three very brief published reports, but there has been site. In addition, a number of spin-off publications has
no further work published and no systematic attempts been produced related to the problems of investigating
have been made at investigating the site’s archaeology. the nature of Medieval urbanism in India, for example
The importance of the site and the opportunity it pro- Kennet 2004b and Kennet 2013.
vides to investigate a number of key questions relating
to the archaeology of the Early Historic and Early Medi-
eval periods in India led, in 1996, to the conception of a Report Structure and Format
joint Indian-British excavation project. At the broadest
level, the project was aimed at providing more informa- This volume is made up of fifteen chapters divided into
tion on the nature, chronology and development of an five parts. Part I includes an introduction to the project’s
Early Historic and Early Medieval town in central India, research aims and their background along with a descrip-
an issue that is key to our understanding of these periods tion of the site’s geographical location and context
but about which, up to now, there is a real paucity of (Chapter 1). This is followed by a discussion of some
concrete archaeological evidence. key historical texts related to Paithan (Chapter 2) and
The project for the excavation of Paithan was there- a summary of previous archaeological work at the site
fore initiated by a formal agreement between the Archae- (Chapter 3). Part II covers the 1996–1998 excavations,
ological Survey of India and the Society for South Asian beginning with Trench A, the largest and most signifi-
Studies of the British Academy, which was implemented cant trench (Chapter 4), followed by the smaller trenches
on 15 February 1996. The project is one of a series of col- that were excavated at various locations across the site
laborations between these two organizations, others (Chapter 5). Finally, Chapter 6 summarizes the site’s
include the excavations of the Buddhist stupa at San- phasing and chronology. Part III covers the finds from the
nathi in Gulbarga District, Karnataka, from 1986 to 1989 excavations, including pottery (Chapter 7), small finds
(Howell 1995) and the excavations at Sopara in Thane and glass (Chapter 8), coins (Chapter 9) and a summary
District, Maharashtra, carried out in 1992. The Paithan of the key finds from the first two seasons of excavations
agreement outlined a five-year excavation project, to (Chapter 10). In addition Chapter 11 describes significant
be carried out under the co-directorship of the Super- finds from the Balasaheb Patil Government Museum
intending Archaeologist of the Aurangabad Circle, the at Paithan, many of which come from the site and the
Archaeological Survey of India, and the Research Fellow immediately surrounding region. Part IV presents the
of the Society for South Asian Studies. In the event, four environmental evidence, including a report on the large

https://doi.org/10.1515/9783110653540-201
viii   Preface

archaeobotanical assemblage (Chapter 12), and a report Conventions

on the faunal remains (Chapter 13). Part V contains the
conclusions which bring together and contextualise the Transliteration: aside from Chapter 2, diacritical marks
key themes and results of the project. A single appendix have not been used in the transliteration of words from
lists the excavated contexts described by the project. Sanskrit or other languages. This includes architectural
Further stratigraphical detail whose incorporation into terms as well as other terms and the names of people,
a printed volume such as this would be impractical, for places, dynasties, and texts. Although some readers may
example data tables and Harris matrices, are deposited find this unscholarly, it was decided that the inclusion of
with the Archaeological Data Service and are available diacriticals, whilst presenting certain technical difficul-
online (http://dx.doi.org/10.5284/1017460). ties, would have added nothing to the clarity or content
of the text and that their absence did not risk creating any
confusion or obfuscation of meaning at any point.
Authorship Terminology: for the purposes of the present volume,
the terms ‘Early Historic’ and ‘Early Medieval’ will be used
A number of specialists and collaborators has been to refer to the following periods:
involved in the present project in various ways. Author- –– Early Historic: 5th century BC to 5th century AD,
ship is indicated on each of the chapters throughout the –– Early Medieval: 6th century AD to 9th century AD.
volume. The role of J. Varaprasda Rao in the project and
its final publication needs to be explained. He was the These dates are not precise, however, and acknowledge a
joint director of the project, along with D. Kennet, during degree of overlap between the two periods. The authors
the 1998 and 1999 seasons and took an equal part in key see no value in entering into a detailed discussion of the
decisions about strategy, priorities, and research ques- definition of these historical periods at this point. On the
tions related to the excavations. He also played a major use and meaning of these terms in Indian history, see
role in the study season of February 2000 and traveled Chattopadhyaya 1994.
to Durham University, together with M. Kasturi Bai, for
three weeks during September 2002 in order to advance
work on the final publication. It was during this time that Acknowledgements
the final format of the publication and many of the key
themes were agreed. He was also closely involved in the The authors would like to express their thanks to all of the
study of the small finds, coins, and the Balasaheb Patil Directors General of the ASI that have held post since the
Collection. Due to his tragic death in a car accident on project was begun in 1996. Most important amongst these
12 June 2006 the final writing up of the project – much is Shri Ajai Shankar, who was Director General during the
of which took place in Dharwad, Karnataka, between key 1998 and 1999 seasons and who provided enthusias-
October and December 2006 – very sadly took place tic support and assistance. It is to his memory that the
without his wisdom, knowledge, and huge experience. It authors have dedicated this volume.
is for this reason that none of the final chapters are cred- The following senior officers of the ASI were of par-
ited to his authorship, although he made a major intel- ticular help to the project: Smt Achala Moulik, Dr R. S.
lectual and practical contribution to every one of them as Bisht, Dr Amarendra Nath, Dr S. V. Venkateshaiah and
well as to the fieldwork and research that underlies them. Mr P. N. Kamble. Thanks should also be expressed to Mr
His contribution to the final production of this volume Colin Perchard, former Minister (Cultural Affairs) of the
was much missed. British Council, Delhi, and Mrs. Sushma Bahl and Mrs.
Parts of Chapters 4 and 5 of this report have been com- Chandrika Grover both of the British Council, Delhi, for
piled from edited versions of original field reports written constant help and advice during the time of the project.
by the excavators: E. Eastaugh, G. Halder, M. Kasturi Bai, In the UK the Society for South Asian Studies, now
D. Kennet, L. Lloyd-Smith, J. Martin, J. Varaprasada Rao, the British Association for South Asian Studies, provided
M. Saxena and N. Wells. generous funding, without which the project would not
have gone ahead. The following officers of the then SSAS
were particularly involved with the work: Dr G. Tillotson,
Sir Oliver Forster, and Dr D. MacDowall. We would also
like to thank Mr James Howell for his support and advice
Preface   ix

and Mr R. Knox, former Keeper of the Dept. of Oriental During the October to December 2006 writing up
Antiquities of the British Museum. season Basavaraj Badiger, Basavaraj R. Mayachari, Anita
Throughout the project Dr Michael Willis was a S. Hegde and Prashant L. Gadre made an important con-
guiding light in all aspects of work from the organiza- tribution to the work; the kind assistance of Dr S. V. Ven-
tional and administrative to the academic. A great debt of kateshaiah was once again very much appreciated.
gratitude is owed him. We would also like to thank Professor Hans Bakker
We are grateful to all the members of the excavation for his advice on certain aspects of Medieval texts, Dr Paul
teams, especially Mr Manoj Saxena, Mr Goutam Haldar, Dundas for his advice on Jain texts and Ran Zhang for his
Shri M. Mahadeviah (Dy. S. A.), Shri M. R. Dhekane and comments on the Chinese pottery from the site. Thanks
Dr D. N. Sinha (assistant archaeologists), Shri D. T. Asar are due to Dr Michael Willis, Prof. Chris Gerrard and Prof.
and Shri S. B. Rindhe (surveyors), Shri D. L. Sirdeshpande Graham Philip for reading and commenting on drafts of
and Shri R. D. Ingle (draughtsmen), Shri G. L. Gaikwad some or all of the chapters.
and Shri V. R. Satbhai (photographers), Shri R. L. Shaik Many local people in Paithan and Aurangabad helped
(L.D.C.), Mr Nazimuddin and Mr Samad, Shri B. S. Tadvi us with this work and we would like to thank all of them
and Shri S. L. Magre (attendants) and Shri Sheik Chand for their friendly welcome and kind assistance. We hope
(monument attendant) from Archaeological Survey of that they will feel that this study does justice to the history
India, and to Nicholas Perchard, Lindsay Lloyd-Smith, Ed of their town.
Eastaugh, John Martin and Nick Wells from the Society for
South Asian Studies. All team members worked with ded- Derek Kennet
ication, commitment and high levels of professional skill. Durham University
Table of Contents
Preface   vii

List of Tables   xiii

List of Figures   xv

Part I Background to the project

Derek Kennet
Chapter 1 The Scope of the Project and Its Background   3

Shailendra Bhandare
Chapter 2 Historical Context   11

Derek Kennet
Chapter 3 Previous Archaeological Work at Paithan   18

Part II The Excavations

Derek Kennet
Introductory Remarks on the Excavations   22

Derek Kennet
Chapter 4 The Brick Temples in Trench A   23

J. Varaprasada Rao and Derek Kennet

Chapter 5 The Excavation of Trenches B, C, D, E and F   81

Derek Kennet
Chapter 6 Periodisation and Dating of the Sequence   90

Derek Kennet
Summary Remarks on the Excavations   98

Part III The Finds

Derek Kennet
Brief Remarks on the Finds   100

Derek Kennet
Chapter 7 The Pottery   101

M. Kasturi Bai
Chapter 8 The Small Finds and Glass   158
xii   Table of Contents

M. Kasturi Bai and Shailendra Bhandare (with an appendix by D. Kennet)

Chapter 9 The Coins   215

Derek Kennet
Appendix 9.1 Coin Deposition in the Temple Foundation Deposits and Coin Residuality   229

M. Kasturi Bai
Chapter 10 Summary of the Finds from the 1996 and 1997 Excavations   232

M. Kasturi Bai
Chapter 11 Objects Displayed in the Balasaheb Patil Government Museum   234

Part IV Environmental Evidence

Derek Kennet
Introductory Remarks on the Environmental Evidence   244

Dorian Q. Fuller
Chapter 12 Archaeobotany   245

Peter Rowley-Conwy, with a contribution by P. K. Thomas

Chapter 13 The Faunal Remains   281
Appendix 13.1 Minimum Animal Units (MAU) Tables   290
Appendix 13.2 Ageable jaw fragments of the three main species at Paithan   292
Appendix 13.3 Measurements of animal bones from Paithan   293

Part V Conclusions

Derek Kennet
Chapter 14 Paithan in Its Regional, Historical and Cultural Context   299

Derek Kennet, Jason Hawkes and Michael Willis

Chapter 15 Paithan Excavation: Historical, Archaeological, Geographical and Epigraphical Contexts   324

Reference list   336

Appendix I List of recorded contexts   348

General Author Index   381

General Subject Index   382

List of Tables
Table 4.1 List of test pits and areas excavated in Trench A   27
Table 4.2 The proposed relative and absolute chronology of the two temples   59
Table 4.3 Stages in the development and use of the shrine within the mandapa of the North Temple   68
Table 4.4 Detailed description of the principal walls in Trench A   75
Table 6.1 Occurrence of glazed wares in the Nevasa sequence   90
Table 6.2 Percentage occurrence of Type 1 in the lower parts of the sequences from Trenches B, D1 and D2   94
Table 6.3 C14 dates from the temple foundations   96
Table 7.1 Sherd numbers for the main sequences and secondary assemblages included in the final pottery study   102
Table 7.2 The occurrence of the main fabrics by period, by rim-sherd count above, and by rim-sherd count as a percentage of the total
period rim-sherd count below   103
Table 7.3 Summary of selected ceramic classes from Paithan by period (rim-sherd counts above; percentages of total period
rim-sherd assemblages below)   109
Table 7.4 Summary of body sherd counts of by period for those wares for which body sherds were counted separately   109
Table 7.5 Types by period based on rim-sherd count (left) and percentages of total rim-sherd count (right)   114
Table 7.6 Types from Trench A: Test Pit 1 by rim-sherd count and layer in stratigraphic sequence   140
Table 7.7 Types from Trench B by rim-sherd count and layer in stratigraphic sequence   142
Table 7.8 Types from Trench C by rim-sherd count and layer in stratigraphic sequence   146
Table 7.9 Types from Trench D1 by rim-sherd count and layer in stratigraphic sequence   146
Table 7.10 Types from Trench D2 by rim-sherd count and layer in stratigraphic sequence   147
Table 7.11 Types from Trench F by rim-sherd count and layer in stratigraphic sequence   148
Table 7.12 The number of sherds and the amount of earth excavated from a sample of period-assemblages from across the site (Trench
A Test Pit 1 (TP1) Periods 1–4; Trench B (TrB) Periods 1, 2 and 4; Trench D1 (TrD1) Period 1; Trench D2 (TrD2) Periods 1 and 2;
Trench F (TrF) Periods 1, 2 and 3; Trench A Garbha Griha of the North Temple (GGN), Period 3-Temple and Trench A Test Pit 5
(TP5) Period 3-Temple)   150
Table 7.13 The density of sherds per cubic metre from 15 period-assemblages from across the site   150
Table 7.14 Average density of sherds per meter cubed based on the data from Table 7.13 (not including the anomalously high figure for
TP1–1 of 10,938 per m³)   151
Table 7.15 Catalogue of unique sherds   152
Table 8.1 The main categories of small finds   158
Table 8.2 Jewellery objects by period   160
Table 8.3 Summary of beads by material and period   160
Table 8.4 Summary of non-terracotta beads by shape and period   163
Table 8.5 Summary of terracotta beads by shape and period   164
Table 8.6 Summary of bangles by material and period   171
Table 8.7 Summary of metal objects by type of metal and period   184
Table 8.8 Summary of copper objects by type of object and period   192
Table 8.9 Summary of iron objects by type of object and period   196
Table 8.10 Summary of play objects by type of object and period   202
Table 8.11 Glass fragments   214
Table 9.1 The occurrence of coins by period   215
Table 9.2 Occurrence of coins by period as an approximate percentage of the total number of coins found in
each period   226
Table 9.3 The coins from Trench A summarized by period   229
Table 9.4 The number of coins, sherds and the amount of earth excavated from a sample of period-sequences from across
the site   229
Table 9.5 Coins per cubic metre, sherds per cubic metre and the ratio of coins to sherds from a sample of period-sequences from
across the site (based on the data in Table 9.4)   230
Table 12.1a Archaeobotanical data from Paithan. P: present; T: total   248
Table 12.1b Archaeobotanical data from Paithan. P: present; T: total   250
Table 12.2 The presence/absence of cereal crops on peninsular Indian sites of Iron Age or Early Historic date. Sources: Cooke et al.
(2005), Kajale (1974, 1977b, 1982, 1984, 1989, 1994), Fuller et al. (2004), Fuller (unpublished data) (Piklihal), Vishnu-Mittre
(1968), Vishnu-Mittre and Gupta (1968a,b) and Vishnu-Mittre et al. (1971, 1984)   254
Table 12.3 The occurrence of kodo millet in South Asian archaeology   256
Table 12.4 The presence/absence of pulse crops on peninsular Indian sites of Iron Age or Early Historic date. Sources as for
Table 12.2   264
Table 12.5 Archaeobotanical and textile remains of cotton from South Asia   271

https://doi.org/10.1515/9783110653540-203
xiv   List of Tables

Table 13.1 Number of Identified Specimens (NISP) of the main domestic taxa at Paithan, divided by layer and expressed as
percentages of the layer total   282
Table 13.2 Number of sheep/goat specimens definitely identified as either sheep or goat   283
Table 13.3 Other specimens identified by the author, not included in Table 13.1   283
Table 13.4 Dr P. K. Thomas’ identifications of bones and scute fragments from Paithan   284
Table 13.5 The unidentified bone fragments from Paithan, divided into size class and anatomical region where
possible   285
Table 13.6 Minimum Animal Units (MAU) of the main domestic taxa at Paithan, divided by layer and expressed as percentages of
the layer total (see text for the method of deriving the MAU)   287
Table 14.1 Tentative correlation of the excavated sequences from Yusuf (1938; 1939), Dikshit (1973) and the present
excavations   303
Table 14.2 Coins and cowrie shells from the Paithan sequence (summarized from Table 9.1)   319
Table 14.3 A summary of the development of key aspects of the Paithan sequence   322
List of Figures
Fig. 1.1 Paithan and other Early Historic sites mentioned in the text   7
Fig. 1.2 Paithan and surrounding area   7
Fig. 1.3 Topographic map of Paithan with contours at 1-m intervals. The location of the section shown in Fig. 1.7 is also
shown   8
Fig. 1.4 Map of Paithan showing key landmarks, the ASI protected area and the location of the old town   8
Fig. 1.5 Satellite image of Paithan (Google Earth)   9
Fig. 1.6 A view of the ASI protected area at Paithan, looking north from the south bank of the Godavari   9
Fig. 1.7 Section through the ASI protected area and the main part of the old town   10
Fig. 4.1 Plan showing the central mound at Paithan and the location of the main trenches. Note that the trenches are not
shown to scale in this plan   24
Fig. 4.2 Satellite image of the ASI protected area showing the location of the trenches   24
Fig. 4.3 View of Trench A from the south-west at the end of the 1998 season   25
Fig. 4.4 Yusuf’s photograph of the 1937 excavations showing the two brick temples from the south (from Yusuf 1938: pl. v; 1939: pl.
XVIIb)   25
Fig. 4.5 Plan of Trench A showing the location of test pits and excavation areas   26
Fig. 4.6 General view of the temples from the north at the end of the 1998 season   28
Fig. 4.7 Plan of the North Temple showing the locations of the main walls   29
Fig. 4.8 The North Temple from the north-east   29
Fig. 4.9 Dimensions of the North Temple   30
Fig. 4.10 Phases of the North Temple   30
Fig. 4.11 The west end of the North Temple from the west   31
Fig. 4.12 Schematic section through the garbhagriha of the North Temple from the north   32
Fig. 4.13 The Phase I structure of the North Temple showing wall 554 and the simple ‘Ramtek-like’ moulding exposed in a
post-Medieval pit cut   33
Fig. 4.14 The garbhagriha of the North Temple after excavation to the lowest foundation fill of boulders. Note the vertical incised
lines in the middle of each wall beneath the level of the blackboard   33
Fig. 4.15 Isometric reconstruction of Phases I, II and III of the North Temple from the south-west   34
Fig. 4.16 Elevation of walls 555 and 556   34
Fig. 4.17 The robbing on the north side of the North Temple from the east   34
Fig. 4.18 Details of plinth mouldings from various walls from both the North and South Temples   35
Fig. 4.19 East-west section through wall 514 showing extra face 559/1213   36
Fig. 4.20 Sketch of the joint between walls 556 and 557 at the southwest corner of the North Temple from the south-west. Arrows
indicate where a brick abuts another   37
Fig. 4.21 The joint between the Phase II and Phase III brickwork of the North Temple, looking west in the southern half of the
mandapa, with wall 558 to the immediate right   38
Fig. 4.22 Test Pit 9 sections. (A) Composite section across the whole test pit and the mandapa of the North Temple looking west. (B)
East-west section against wall 556 below the southern part of wall 558 where it crosses the test pit. (C) The western section
of the test pit to the north of the Phase I structure showing the edge of foundation cut 1171   38
Fig. 4.23 Feature 1137, the earliest evidence for a possible shrine in the west end of the mandapa of the North Temple   39
Fig. 4.24 North Temple, showing the remains of a possible simple shrine 983 against the western wall of the Phase III
mandapa   39
Fig. 4.25 Photo of the rectangular depression in layer 551 that might indicate the location of a shrine   40
Fig. 4.26 Wall 383/544 of Phase IV in the North Temple before the excavation of Test Pit 9   40
Fig. 4.27 Schematic section/view of the garbhagriha wall of the North Temple looking north showing the extent of later
rebuild 971   41
Fig. 4.28 The garbhagriha of the North Temple from the west showing possible evidence of late use   42
Fig. 4.29 General view of the South Temple from the southeast during excavation   42
Fig. 4.30 Plan of the South Temple showing the location of the main walls   45
Fig. 4.31 Dimension plan of the South Temple   45
Fig. 4.32 Phases of the South Temple   46
Fig. 4.33 West-facing composite section across the South Temple showing foundations   47
Fig. 4.34 View of the eastern section of Test Pit 5 across the interior of the South Temple looking south-east showing the successive
foundation fills   47
Fig. 4.35 Interior of the South Temple during excavation of Test of the South Temple looking south-east showing the successive Pit 5
showing a level of foundation fill consisting of small boulders   47
Fig. 4.36 Composite east-west section through the west end of the South Temple looking north   48

https://doi.org/10.1515/9783110653540-204
xvi   List of Figures

Fig. 4.37 The wall between the mandapa and the garbhagriha in the South Temple looking west during the excavation of Test Pit 5.
The Phase I gap in this wall can be clearly seen below the Phase II wall   48
Fig. 4.38 Western section of Test Pit 5 across the South Temple in front of the wall between the garbhagriha and the
mandapa   49
Fig. 4.39 Cross-section of walls 1247 and 1224 South Temple   49
Fig. 4.40 View of the corner formed by walls 1240/1251 and 1224 in the southwest corner of Test Pit 5 looking southwest. The arrows
show the boundaries between the upper and lower walls. The inter-bonding between walls 1240 and 1251 is clearly visible
just to the right of the corner and the difference in the quality of the brickwork of the two walls is also clear   49
Fig. 4.41 Remains of mouldings in wall 1240 at the eastern end of the South Temple looking west   50
Fig. 4.42 View showing the eastern end of the south wall of the South Temple looking north (see Fig. 4.46 for a key to
wall numbers)   50
Fig. 4.43 South-facing section of Test Pit 8 showing the sequence of floors that abut the eastern face of the Phase I – III walls of the
South Temple   51
Fig. 4.44 Elevation of the western wall of the South Temple above the eastern section of Test Pit 1 showing the fills of the
temple foundations   51
Fig. 4.45 The robbed-out wall 352 at the east end of the South Temple looking south   52
Fig. 4.46 Elevation of the southern face of the southern wall of the South Temple (wall 560)   52
Fig. 4.47 Foundations of the Early Medieval perimeter wall 699 in the South Area of Trench A looking north. The stones of wall 691 are
visible immediately to the west. At the top of the photograph, the foundations of 699 turn to butt the south-west corner of
the South Temple   54
Fig. 4.48 Trench A showing the remains of the possible late enclosure walls 578/699 and 691 and associated features to the
south of the South Temple and feature 617 to the west of the North Temple   55
Fig. 4.49 A composite section through the two temples and their foundations   56
Fig. 4.50 West-facing section of Test Pit 3 showing the foundation cuts for both buildings   59
Fig. 4.51 South-facing section of Test Pit 1, Trench A, showing the foundation cut for the South Temple   60
Fig. 4.52 Plinth mouldings from Ramtek: (a) Rudra-Narasimha temple; (b) Kevla-Narasimha temple; (c) Bhogarama temple
(after Meister et al. 1988: fig. 40)   61
Fig. 4.53 Plinth mouldings on the main temple at Chejerla (from Sarma 1982: fig. 18)   61
Fig. 4.54 Eastern section of Test Pit 1 showing the foundation fills under the west end of the South Temple   63
Fig. 4.55 South-facing section, Test Pit 6   66
Fig. 4.56 View of wall 360, built from re-used stone architectural fragments above wall 352 at the east end of the South Temple from
the north-west   74
Fig. 4.57 Drawing of the north face of the north wall of the South Temple with an interpretive overlay   74
Fig. 5.1 Trench B, view of the south and east sections of the trench at the end of excavation   82
Fig. 5.2 Trench B showing wall 401 and related pillar bases to the west   82
Fig. 5.3 Trench B, north-facing section   83
Fig. 5.4 Trench B, west-facing section   83
Fig. 5.5 Trench C section (155/450, south-facing)   85
Fig. 5.6 Trench D1, east-facing section   86
Fig. 5.7 Trench D2, south-facing section, as dug   87
Fig. 5.8 Trench F, east-facing section   88
Fig. 5.9 Trench F, view of the west-facing section   88
Fig. 6.1 Graph showing the percentage of Type 1 through the lower parts of the sequences from Trenches B, D1 and D2 (sherd count
as % of total sherd assemblage)   94
Fig. 7.1 Northern Black Polished ware sherds (NBP) (scale in cm)   106
Fig. 7.2 Classes NBP, RPW, SLIP and CBW   107
Fig. 7.3 Red Polished ware sherds (RPW)   110
Fig. 7.4 Chinese Blue and White porcelain sherds (CBW)   110
Fig. 7.5 Blue and White Frit sherds (PERS)   111
Fig. 7.6 Slip-painted ware sherds (SLIP)   111
Fig. 7.7 White and Red Painted ware sherds (WARP)   111
Fig. 7.8 Types 1 to 7   117
Fig. 7.9 Types 8 to 11   120
Fig. 7.10 Types 12 to 15   121
Fig. 7.11 Types 16 to 22   123
Fig. 7.12 Types 23 to 27   126
Fig. 7.13 Types 29 to 33   128
Fig. 7.14 Types 34 to 39   130
Fig. 7.15 Types 40 to 44   132
 List of Figures   xvii

Fig. 7.16 Types 45 to 48   134

Fig. 7.17 Types 49 to 51   135
Fig. 7.18 Types 52 to 58   138
Fig. 7.19 Ranked analysis of the density of sherds per metre cubed from 15 period-assemblages from across the site   150
Fig. 7.20 Sherds with incised marks: P863, P1766, P2552 and sherd P2553, Roman amphora   151
Fig. 7.21 Unique sherds, Periods 1 to 3   153
Fig. 7.22 Unique sherds, Periods 3 and 4   154
Fig. 7.23 Occurrence profiles of Types 1, 2, 3, 4, 7 and 8. The histograms show occurrence as a percentage of the total period
rim-sherd assemblages. Note that vertical scales differ on the histograms   155
Fig. 7.24 Occurrence profiles of Types 9, 10, 12, 14, 16, 20, 24 and 25. The histograms show occurrence as a percentage of the total
period rim-sherd assemblages. Note that vertical scales differ on the histograms   156
Fig. 7.25 Occurrence profiles of Types 26, 27, 30, 32, 33, 38, 41 and 51. The histograms show occurrence as a percentage of the total
period rim-sherd assemblages. Note that vertical scales differ on the histograms   157
Fig. 8.1 Microliths   159
Fig. 8.2 Beads of different materials   161
Fig. 8.3 Terracotta beads   162
Fig. 8.4 Beads of different materials   165
Fig. 8.5 Bangles of different materials   172
Fig. 8.6 Copper objects   173
Fig. 8.7 Ear-studs, pendants and other terracotta objects   181
Fig. 8.8 Bone and ivory objects   183
Fig. 8.9 Iron objects   186
Fig. 8.10 Bone, kaolin and porcelain objects   195
Fig. 8.11 Stone objects   197
Fig. 8.12 Small jars, crucibles and lamps   200
Fig. 8.13 Animal terracotta figurines   207
Fig. 8.14 Kaolin figurines   208
Fig. 8.15 Terracotta figurines   209
Fig. 8.16 Human and animal terracotta figurines   210
Fig. 8.17 Stone sculpture 234   211
Fig. 8.18 Stone sculpture 234   212
Fig. 8.19 Glass fragments   214
Fig. 9.1 Coins 1 (574), 3 (505), 5 (537), 2 (504), 4 (506) and 6 (298)   216
Fig. 9.2 Coins 9 (337), 12 (535), 14 (538), 52 (493), 13 (648) and 25 (379)   218
Fig. 9.3 Coins 30 (740), 18 (02), 15 (724), 33 (820), 17 (745) and 19 (188)   220
Fig. 9.4 Coins 22 (46), 21 (55), 29 (492), 20 (57), 24 (390) and 28 (829)   222
Fig. 9.5 Coins 31 (462), 34 (496), 55 (187), 32 (120), 35 (29) and 56 (56)   224
Fig. 9.6 Rank analysis of the ratio of coins to sherds from a sample of period-sequences from across the site. The two Period
3-temple period-sequences (TP5-3T and GGN-3T) are at the lower end of the range   230
Fig. 11.1 Coins from the Balasaheb Patil Collection (1–7)   236
Fig. 11.2 Coins from the Balasaheb Patil Collection (8–11)   237
Fig. 11.3 Coins from the Balasaheb Patil Collection (12–13)   237
Fig. 11.4 Coins from the Balasaheb Patil Collection (14–18)   238
Fig. 11.5 Coin moulds from the Balasaheb Patil Collection (1–3)   239
Fig. 11.6 Pottery and terracotta objects from the Balasaheb Patil Collection (1–9)   240
Fig. 11.7 Figurines from Paithan (from Morwanchikar 1985: pls X–XIII)   241
Fig. 12.1 Relative frequency of major crop groups   246
Fig. 12.2 Ubiquity of major crop groups and selected subdivisions of millets   247
Fig. 12.3 Relative frequency of major crop groups   252
Fig. 12.4 Comparative diversity indices   252
Fig. 12.5 Scanning electron micrograph of charred Paspalum scrobiculatum (kodo millet) from Paithan; dorsal view (left) and
ventral view (right)   254
Fig. 12.6 Rice spikelet base from context 721; lateral view (left) and ventral view (right). Notice the indented scar indicative of
domesticated/non-shattering morphotype   255
Fig. 12.7 Comparison of the frequency of wheat and barley with other taxa   258
Fig. 12.8 Grains of barley (Hordeum vulgare), hulled variety, from context 721; ventral view (above) and dorsal view (below)   259
Fig. 12.9 Grains of free-threshing wheat (Triticum aestivun/durum) from context 1036; dorsal view   259
Fig. 12.10 Grain of emmer wheat type (Triticum diococcum) from context 919; dorsal view (left) and lateral view (right)   259
xviii   List of Figures

Fig. 12.11 Pearl millet grain (Pennisetum glaucum), narrow grain type, from context 930; dorsal view (left) and dorsal view
(right)   260
Fig. 12.12 Sorghum (Sorghum bicolor) of race bicolor type from context 930; dorsal view (right). Sorghum of race durra type
from context 1092; dorsal view (right)   260
Fig. 12.13 Finger millet (Eleusine coracana), fragmented, from context 721; basal view of embryo. Approximately half of the grain is
preserved   261
Fig. 12.14 Grain of common broomcorn millet (Panicum miliaceum) from context 1128; dorsal view   261
Fig. 12.15 Grain of common foxtail millet (Setaria italica) from context 1135; dorsal view (left) and ventral view (right), both with
adhering lemma. Notice on ventral view the indentation of charred away hilum which is relatively short compared to
overall grain length (in contrast to Brachiaria ramosa)   262
Fig. 12.16 Grain of browntop millet (Brachiaria ramosa) from context 1127; ventral view (left) showing adhering palea (centre) and
dorsal view (right). Notice (left) faint traces of hilum, which is relatively long compared to overall grain length
(1/4–1/3 length)   263
Fig. 12.17 Relative frequency of pulses   263
Fig. 12.18 Interior view of the cotyledon of mung bean (Vigna radiata) from context 1150. Notice the relatively long plumule   265
Fig. 12.19 Interior view of the cotyledon of urd bean (Vigna mungo) from context 1150. Notice the shorter, half-seed-length
plumule   265
Fig. 12.20 Comparison of measurements of Paithan Vigna with samples from Neolithic contexts in southern India   266
Fig. 12.21 Moth beans (Vigna aconitifolia) from context 1136   266
Fig. 12.22 Horsegram (Macrotyloma uniflorum) from context 467; complete seed (left) and interior of split cotyledon (right)   266
Fig. 12.23 Pigeonpea (Cajanus cajan) from context 1036; lateral view (left) and interior view of split cotyledon (right)   267
Fig. 12.24 An example of chickpea (Cicer arietinum); ventral view   268
Fig. 12.25 An example of lentil (Lens culinaris); lateral view   268
Fig. 12.26 Seed of mustard (Brassica cf. juncea) from context 985, with part of seed coat cell pattern visible   268
Fig. 12.27 Seeds of fig type (Ficus sp.) from context 619   269
Fig. 12.28 Seed of mouse cucumber (Melothria cf. heterophylla, family Cucurbitaceae) from context 1116   269
Fig. 12.29 Sesame seed (Sesamum indicum) from context 930, whole seed (left) and detail of seed coat (right)   269
Fig. 12.30 Ubiquity of selected crop types and weeds   270
Fig. 12.31 Cotton (Glossypium sp.) remains from context 930, including fragmented seed (left), cross-section of seed coat
(centre) and funicular cap from inferior seed (right)   270
Fig. 12.32 The distribution of archaeological finds of cotton and flax seed in South Asia, indicated by broad time
horizons   272
Fig. 12.33 The spread of cotton beyond South Asia: finds of seeds and textiles   273
Fig. 12.34 Safflower (Carthamus tinctorius) from context 841   274
Fig. 12.35 Lone specimen, poorly preserved, of possible hemp (cf. Cannabis sativa) from context 1128   275
Fig. 12.36 Weedy grass, wrinklegrass (Ischaemum rugosum) from context 1116, dorsal view   276
Fig. 12.37 Weedy grass, Indian goosegrass (Eleusine indica) from context 721, dorsal view   276
Fig. 12.38 Weedy grass, fingergrass (Digitaria sp.) from context 1116, dorsal view   276
Fig. 12.39 Two species of carpetweeds (Aizooaceae) from context 933: Trianthema portulacastrum (above) and Trianthema triquetra
(below), lateral views   278
Fig. 12.40 Buttonweed mericarp (Borreira/Spermacoce) from context 1116: dorsal (left) and ventral view (right)   279
Fig. 12.41 Indeterminate weed type of Caryophyllaceae/Molluginaceae type from context 721, lateral view   279
Fig. 12.42 Weedy mallow type seed (Malvaceae, cf. malva), lateral view   279
Fig. 12.43 Weedy dayflower seed (Commelina cf. benghalensis) from context 486, dorsal view   279
Fig. 13.1 Phalanges of blackbuck (Antilope cervicapra) showing the characteristic etching caused by the digestive acids of dogs
(top) and holes caused by gnawing (both specimens). The specimens were identified to species by Dr P. K. Thomas   282
Fig. 13.2 Fragment of mandible of Crocodilus palustris, identified by Dr P. K. Thomas   284
Fig. 13.3 Fragments of scute of turtle (Trionyx gangeticus) cut with a knife   285
Fig. 13.4 Minimum Animal Units (MAU) of the main domestic taxa at Paithan, divided by layer and expressed as percentages of the
layer total (see text for the method of deriving the MAU)   286
Fig. 13.5 Diagram of the skeleton of an artiodactyl, showing the various bones   287
Fig. 13.6 Metacarpal shaft fragment of sheep or goat, showing two heavy chop marks   288
Fig. 14.1 A schematic outline of the building materials used in the Paithan sequence and at eight other Early Historic sites in the
region   301
Fig. 14.2 Drawing of a shrine in a relief from Amaravati (from Coomaraswamy 1930: fig. 33)   305
Fig. 14.3 Numbers of Indian livestock (1,000s) from the 17th Livestock Census of 2003 (source: FAO 2001–2003)   311
Fig. 14.4 A histogram of coins from Bhokardan, Maheshwar and Nevasa redated by S. Bandhare showing the numbers of coins
deposited per year (from Bhandare and Kennet forthcoming)   318
Fig. 15.1 Eran (Dist. Vidisha, Madhya Pradesh). Inscribed memorial stone of Goparāja, dated AD 510–11   326
 List of Figures   xix

Fig. 15.2 Salkanpur (Dist. Sehore, Madhya Pradesh). Hero-stone memorial with a pinnacle in the shape of a temple spire, with a relief
sculpture of warriors in battle below, circa 12th century   327
Fig. 15.3 Maharashtra, region of ancient Vidarbha, showing the distribution of key copper-plate charters of the
Vakataka period   330
Fig. 15.4 Maharashtra, region of ancient Vidarbha, showing the distribution of known archaeological sites and coin finds   332
Fig. 15.5 Mahurjhari (Dist. Nagpur, Maharashtra). Raw material for carnelian from ancient quarry   333
Fig. 15.6 Mahurjhari (Dist. Nagpur, Maharashtra). Carnelian bead fragments in fields at village   333
Fig. 15.7 Mahurjhari (Dist. Nagpur, Maharashtra) (?). Engraved carnelian with a Buddhist inscription in Vakataka-style script reading
apramāda with corresponding impression. Probably made at Mahurjhari. British Museum, 1892, 1103. 126, see SIDDHAM
IN01102   333
Fig. 15.8 Chammak (Maharashtra). Copper-plate charter of Pravarasena II with seal of the king and a paper note by J. F. Fleet. British
Library Ind Ch 16, see SIDDHAM OB00149   334
Part I
Background to the project
Derek Kennet

Chapter 1 
The Scope of the Project and Its Background

The study of Early Historic and decline of philosophical traditions and the degenera-

Early Medieval India tion of classical, naturalistic art styles into more styl-
ised, ornamental forms alongside the emergence of
more populist forms of religious worship.
The Early Historic to Early Medieval transition is a key
After independence and into the 1960s, a more
formative period in India’s history. This is particularly
analytical and critical approach to Indian history
true in terms of the emerging economic and politi-
emerged, with a closer focus on economic and social
cal role of temples, the nature of urbanism, the role
issues, particularly amongst Marxist historians. This
of Brahmins and the emergence of Hindu kingship,
approach was based on the view that the ancient
but also in relation to broader debates about transfor-
empires were highly centralized, that they incorpo-
mations in economy and society. In recent years, the
rated numerous, large urban centres and were based
study of this transitional period has itself undergone
around a highly monetized economy in which manu-
significant transformation amongst historians (e.g.
facture and trade were very significant. It was argued
Chattopadhyaya 1994; Kulke 1995b; Bakker 1997; Willis
that these structures came to an end during or after
2009; Bronkhorst 2011). At the same time it has been
the Gupta period, at which time manufacturing and the
notable that – aside from the debate about late Early
minting and circulation of coins dwindled whilst urban
Historic urban decline (Sharma 1987) – the archaeologi-
centres declined in size and wealth. These changes
cal contribution to our understanding of the transi-
were accompanied by a proliferation of copper-plate
tion has been relatively thin. It was a keen awareness
inscriptions recording royal grants of land to temples
of this point, as well as a clear vision of the potential
and Brahmins, reflecting – it is argued –the decreasing
that archaeological evidence has to address many of
significance of coinage. A number of scholars argued
the key issues that surround the transition, that drove
for the development of feudal economic and political
the two directors of the present project (J. V. P. Rao and
structures in India at this time, partly on the basis that
D. Kennet) to steer the Paithan excavation project in
land grants were indicative of the decline of central-
this direction when they jointly took it over in 1998.
ized political power that was also reflected in the emer-
However, before going into the research aims and
gence of smaller, localised Medieval kingdoms.
objectives that formed the underlying rationale of the
There is no doubt that the emphasis on the study
present excavations, it may be useful to provide a short
of economic and social history was one of the most
background to the study of the Early Historic/Early
important innovations of this phase in Indian histori-
Medieval transition in order to set out the research
ography. At the same time, the notion of a period of
context.
feudalism in Indian history continued to be widely
The traditional scholarly view of Indian history that
debated through the 1960s and 1970s. By the 1980s,
had emerged by the early 20th century divided India’s
however, much of the evidence on which this argument
past into ‘ancient’, ‘medieval’ and ‘modern’ periods,
was based had been challenged by scholars such as D.
effectively replicating the older scheme of ‘Hindu’,
C. Sircar (1974), B. D. Chattopadhyaya (1986) and J. S.
‘Islamic’ and ‘British’ that had been developed by 19th
Deyell (1990), and it is fair to say that the idea has
century Orientalists. Within this ‘ancient-medieval-
now largely lost currency amongst scholars, at least
modern’ conceptualisation of history, the transition
as a formal concept. At the same time, other scholars
from ancient to Medieval was seen as marking the end
began to question the level of centralization and cohe-
of a glorious age of ancient empires and the begin-
sion of the Early Historic states such as the Mauryan
ning of a degenerative Medieval period during which
empire, bringing into doubt the very starting point
the great Mauryan and Gupta empires fragmented into
from which Early Medieval formations are supposed
numerous regional kingdoms. These epochal changes
to have developed.
were generally perceived as being accompanied by the
https://doi.org/10.1515/9783110653540-001
4   Derek Kennet

Emerging from this debate, the work of B. D. Chat- According to the processural model, it was core
topadhyaya was key to the formulation of a new way agrarian areas that formed the nuclear areas of future
of seeing Early Medieval Indian society in the 1980s, Medieval kingdoms. In such areas, the upper Goda-
challenging many of the views held by earlier scholars. vari basin around Paithan being a good example, it
Using epigraphic evidence, he drew attention to the is acknowledged that processes are likely to have been
development of new Medieval urban centres and to the quite different to marginal, forested areas, but little
level of social complexity in and around them (Chat- is known of actual developments and practically no
topadhyaya 1986). Nonetheless, the Early Medieval systematic investigation has been undertaken (Bakker
towns he identified are seen as having been fundamen- 1992: 88–90; Kulke 1995b: 235). It is generally assumed,
tally different to their Early Historic predecessors. As on the basis of general trends, that towns in these areas
Daud Ali has put it, ‘whereas the great cities of ancient experienced something of a contraction in the Early
India were linked “horizontally” in a “thin”, but geo- Medieval period. But apart from occasional epigraphic
graphically dispersed network of regular exchange, evidence of land grants consisting of a field, a well or a
urban centres of post-Gupta India seem to have been disused orchard, there is almost no evidence available
more rooted in regional context and local exchange with which to assess the nature or depth of the trans-
networks’ (Ali 2012, 9). formations that took place at this time. This is clearly
By the late 1980s, a new model of Early Medieval problematic. The danger is that the research agenda
state formation, known as the ‘processural’ model, had is driven almost entirely by the peripheral zones and
become the dominant approach. It developed partly attention is drawn away from what Bakker calls the
from the work of Chattopadhyaya and partly from ‘political, religious and economic fundament’, where
multi-disciplinary research of historians and anthro- many of the key social and political structures were
pologists such as A. Eschmann, H. Kulke and others being negotiated (Bakker 1992: 88–90).
(Eschmann et al. 1978; Kulke 1995b). This model, which From an archaeological perspective, there are
focuses particularly on the spread of Medieval king- numerous ways in which it might be possible to use
doms into south and east India, poses a three-stage excavation or survey evidence to redress this imbal-
development of Early Medieval states, from isolated ance and investigate some of the developments that
nuclear areas to the emergence of imperial kingdoms took place in the core agrarian areas. For example,
through a constant process of agrarian expansion and the quantified results of systematic field surveys in
integration. It involves the deliberate horizontal spread the territories around Early Historic cities might be
of agrarian settlements into peripheral, forested areas expected to record changes in the organization and
incorporating hunter-gatherer groups and their cults intensity of rural settlement that reflect population
into pan-Indian structures involving the settlement of levels and the scale of the agricultural economy. Alter-
Brahmins and the foundation of temples. Because this natively, it might be possible to look at the plans of
model does not assume the existence of highly central- excavated Early Historic and Early Medieval towns to
ized Early Historic imperial structures, Early Medieval consider how their layouts reflect changing standards
society can be seen not as a fragmentation of earlier of municipal authority or levels of urban population
structures but rather the continued development of the and wealth. Published archaeobotanical or palaeo-
same processes which had given rise to Early Historic faunal sequences from excavated Early Historic/Early
states. That is to say that agrarian expansion, urban Medieval sites might be examined for quantitative
change, localisation and regional state formation can changes that reflect changing agricultural strategies.
be seen as productive rather than regressive or frag- In addition, studies of pottery sequences might be
menting developments (Ali 2012: 10). interrogated for quantified information on the chang-
However, whilst Chattopadhyaya and Kulke’s pro- ing nature of production and distribution systems.
cessural model works well to explain the integration of Unfortunately, at present, there is very little evidence
newly cleared forested regions into Early Medieval states, of this nature available from Early Historic/Early Medi-
it works much less well when we turn to regions which eval sites in India, despite the large number of exca-
had been cultivated and urbanized since Early Historic vations that have taken place. There are a number of
times – the core agrarian areas – where the urbanized reasons for this.
and sedentary economic and social structures were Firstly, whilst the archaeological excavation of
deeply embedded in traditions going back 800 years or Early Historic sites in India has been going on for over
more. 100 years and has provided information on hundreds
 Chapter 1  The Scope of the Project and Its Background   5

of such sites, the archaeological investigation of Medieval –– When were the temples built and how did they
sites is still very much in its infancy. Secondly, much of develop architecturally?
the considerable energy that has been devoted to exca- –– Does a quantified analysis of the archaeological
vating Early Historic sites has been concentrated on deep sequence associated with the temples reveal infor-
soundings of limited extent that are designed to eluci- mation about the nature of the economy through
date the stratigraphic sequence of a small part of the site the Early Historic and Early Medieval periods in
rather than investigating its spatial layout and develop- the period before the temples’ foundation, during
ment, and the sequences thus retrieved tend to have their use and after their abandonment?
been pushed ‘cookie-cutter’ style into established –– What was the nature of the agricultural economy?
chronological and typological frameworks (Chakra- Can the archaeological record tell us anything
barti 1997: 267–270; Neuss 2012). Thirdly, systematic about crop cultivation and the nature of landhold-
field survey designed to elucidate rural settlement pat- ing? Can it tell us anything about the relationship
terns is not widely enough practiced in India. Fourthly, of the town to the surrounding agrarian hinterland
quantified studies of archaeobotanical, palaeofaunal and about how this relationship may have changed
or ceramic sequences related to the Early Historic and over time?
Early Medieval periods are almost non-existent. Add –– Is it possible to gain any insights into the pos-
to this the enduring problems of the archaeological sible link between the temples and changes that
chronology of the Early Historic and Early Medieval occurred at the site at the time they were con-
periods, and it starts to become clear why archaeology structed and used?
has had so little to contribute to the issues surround- –– Is it possible to say anything about the nature and
ing the Early Historic/Early Medieval transition. Until extent of occupation at the site during the Early
these problems are resolved there are many questions Historic and Early Medieval periods?
in relation to the transition that will remain obscure.
As B. D. Chattopadhyaya himself has acknowledged, At a more specific level, these questions led the project
however, archaeology is the only discipline that can towards a detailed re-excavation and re-examination
unravel the real story of developments during this time of the two temples, including their structure and what
(Chattopadhyaya 1986: 22). Without it, the evidence for remained of the surrounding archaeological deposits.
Early Medieval towns and their economy is overwhelm- They also led to the establishment of a systematic pro-
ingly based on epigraphic evidence. gramme of environmental sampling and a large-scale
When the two directors of the present project programme of flotation of archaeobotanical samples
took over the Paithan excavations in 1998, they were that was carried out alongside the excavation of strati-
presented with two remarkably well-preserved Early graphic soundings around the temples from which the
Hindu brick temples (partially excavated in 1937) sequence of samples was obtained. Given the relatively
which were located in the midst of an extensive Early small scale of the project, the general strategy was to
Historic and Early Medieval town of regional signifi- make full use of small quantities of good-quality infor-
cance that was located in the middle of the Godavari mation. Careful attention was therefore paid to sys-
valley – without doubt a core agrarian heartland. Early tematic sampling strategies and quantified analysis
Hindu temples being, as has been mentioned above, of all aspects of the post-excavation study of pottery,
a key component of the formation of Early Medieval finds and environmental evidence. In many of these
kingdoms according to the processural model, it was methodologies, the project cuts new ground in Indian
immediately clear that this was a unique and important archaeology using procedures and techniques that
opportunity to apply a wide range of archaeological have never or only rarely been applied to the archae-
techniques to elucidating the changing nature of Early ology of Early Historic and Early Medieval sites in the
Historic and Early Medieval occupation at Paithan in region.
relation to the construction and use of the temples. The research questions set out above were broken
On this basis, the following research questions were down into the following three objectives:
defined:1 –– The excavation, recording and analysis of two
early Hindu brick temples including a detailed
1  Many of the ideas in this section derive from discussions and
circulation of draft papers (particularly Ali 2002) in 2002 amongst sisting, in addition, of Daud Ali, Hans Bakker, Dorian Fuller, Isa-
the Vidisha Research Group organized by Michael Willis and con- belle Onians, Julia Shaw and the present author.
6   Derek Kennet

analysis of the construction, phasing and layout Paithan’s location, geography,

of the temples through their period of use. Very
few temples of this key formative period of Indian
climate and form
temple architecture have ever been excavated,
The town of Paithan is known from historical sources
recorded and analysed to the level of detail that was
to have an ancient heritage dating back possibly as
achieved at Paithan and it was therefore expected
early as the Mauryan period and is believed to have
that an investigation of these structures was likely
been a locality of some importance from at least the
to provide important insights into the early devel-
Satavahana period. The town has been associated with
opment of temple architecture, our understanding
several well-known religious personalities in its long
of which has, to date, been based almost entirely
past, and the Ek Nath mandir in the town is still today a
on the evidence from standing monuments.
regional pilgrimage destination of some significance.
–– The excavation and analysis of a deep, stratified
The town lies on the left bank of the River Goda-
sequence close to the Hindu temples that would
vari about 50 km south of Aurangabad (19°27’45”N
allow the retrieval of a complete and systematic
75°22’55”E) (Figs 1.1 and 1.2). It is a taluka headquarters
environmental and artefact sequence through the
in the Aurangabad District of Maharashtra and is a
Early Historic and Early Medieval periods which
bustling modern town. It is situated at a crossing of the
could be linked to the early development of the site
river in the upper Godavari basin, in one of a number of
and the later construction and use of the temples.
eastward-running valleys that cut the great lava pla-
–– The excavation of a number of smaller trenches at
teaus of the western Deccan into sections. These leave
different locations across the site with the inten-
only the dramatic, flat-topped tablelands of lava that
tion of clarifying the nature and extent of occupa-
mark the watersheds between valleys as the highest
tion.
points in an otherwise homogeneous, wide and open
landscape.
At this point, the flat central part of the Godavari
Background of the project basin is almost 20 km wide. In the valley basin around
Paithan the soils are deep, mature black cotton soils
The first stage of the present project ran for two or regurs, which traditionally supported large areas of
seasons (1996 and 1997) as a small-scale excavation stable cultivation although they are difficult to irrigate
and survey directed by P. N. Kamble, then Superin- as they become very heavy and sticky when wet.
tending Archaeologist of Aurangabad Circle of the The area has unreliable rainfall of around 600 mm
Archaeological Survey of India (ASI), and J. Howell, annually. It lies in the rain shadow of the western Ghats
then Research Fellow of the Society for South Asian outside the eastern limit of the zone where some rain-
Studies. In 1998 the project was reconfigured and the fall regularly spills over from the Arabian Sea, but still
directorship passed to J. V. P. Rao and D. Kennet. A some distance west of the Nagpur area where rainfall
further two seasons of excavation took place in 1998 from the Bay of Bengal monsoons is still a regular
and 1999 under their direction. These were funded by factor.
a major research grant from the British Association for Traditionally the area cultivated jowar, wheat,
South Asian Studies (then the Society for South Asian cotton, linseed and pulses and there was relatively
Studies) and an excavation budget from the Archaeo- little double-cropping. Modern irrigation infrastruc-
logical Survey of India. ture was put in place in the 1960s at the time of the
The aims of the first-stage project (1996–97) were construction of the dam and the creation of the Nath
a preliminary investigation of the site and the excava- Sagar power and irrigation scheme immediately above
tion of a stratigraphic sequence intended to clarify the modern town, and this has profoundly changed the
the chronology of the Early Historic to Early Medieval nature of local agriculture (Spate and Learmonth 1967:
period. By the end of the 1997 season, the initial aims 103–105, 690–700).
of the first stage had been achieved. In 1998 a second-
stage research programme was developed and imple-
mented on the basis that has been set out above.
 Chapter 1  The Scope of the Project and Its Background   7

Fig. 1.1: Paithan and other Early Historic sites mentioned in the text.

Fig. 1.2: Paithan and surrounding area.

8   Derek Kennet

Fig. 1.3: Topographic map of Paithan with contours at 1-m inter- Fig. 1.4: Map of Paithan showing key landmarks, the ASI protected
vals. The location of the section shown in Fig. 1.7 is also shown. area and the location of the old town.

The core of the ancient site is a large mound lying Most of the upper part of the mound is occupied by
immediately on the left bank of the Godavari. In this buildings – mostly housing – belonging to the older
place the river makes a sharp turn towards the north- part of modern Paithan. The only exception is an
east after having run southwards for 2 km or so and the archaeological reserve of about 9.3 hectares located at
archaeological mound nestles tightly into the bend the extreme southern end of the mound which is super-
(Figs 1.2–1.6). The mound itself measures between vised by the Archaeological Survey of India and which
1.1 and 1.3 km from north to south and is about 390 is where most of the present excavation trenches were
m wide from east to west at its widest extent. At its located.
highest point it stands about 15 m above the sur-
rounding plain, although not all of this is formed of
archaeological deposit (Fig. 1.7). On the west side, the
river is eroding into the mound, leaving steep cliffs
Technical matters
directly above the river banks. To the north, a very
steep slope drops down from the top of the mound to Excavation season dates
a narrow ‘V’-shaped gully between the main mound
and a lower, smaller mounded area to the north. To The first season of excavation at Paithan was inaugu-
the east the mound slopes steeply down to the level of rated by Shri Balasahib Patil on 10 March 1996 and
the surrounding plain, whilst to the south, the mound closed on 29 March 1996. The second season was carried
gradually narrows and peters out about 250 m from out in February and March 1997. The third season began
the bend in the river, leaving a lower area of small on 21 February 1998 and ended on 26 March 1998. The
mounds between itself and the river bank. fourth season of excavations at Paithan took place
between 15 January and 15 March 1999.
 Chapter 1  The Scope of the Project and Its Background   9

Fig. 1.5: Satellite image of Paithan (Google Earth).

Fig. 1.6: A view of the ASI protected area at Paithan, looking north from the south bank of the Godavari.
10   Derek Kennet

Fig. 1.7: Section through the ASI protected area and the main part of the old town.

Excavation strategy and recording system Site datum

Where appropriate, the excavation was conducted as an An approximation of mean sea level (MSL) was taken
‘area-excavation’ rather than as a grid of small square from the one inch sheet 47 M/7 (first edition, reprinted
‘box’ trenches separated by baulks. This was particu- 1948) published by the Surveyor General of India in
larly useful in Trench A around the temples, where box- 1917. From this, a local datum was established at the
trenching would have obscured the architectural layout site. It was not possible to check the accuracy of the
of the structures (see the discussion in Barker 1982: measurement in relation to MSL, but all site measure-
44–52). ments were taken from the same datum and their rela-
The excavations were recorded using single-con- tive accuracy is expected to be within 1 cm.
text planning according to the method set out by the
Museum of London Archaeology Service (1994). Such
practice is now standard in most countries and pro- Finds storage, interim reports, data storage
vides the only acceptable record of an archaeological
excavation. Each context was planned on a separate The bulk of the finds from the excavations are stored
sheet of tracing paper, photographed, described, lev- with the Goa Mini Circle of the Archaeological Survey of
elled and given an individual context number against India, although those from the 1996 and 1997 seasons
which all finds and other evidence were recorded. are stored with the Aurangabad Circle. Two unpub-
Phasing of the contexts took place only at the post- lished interim reports were produced (Rao et al. 1998;
excavation stage based on an evaluation of the strati- Rao and Kennet 1999). Copies of these were lodged
graphic and pottery evidence (Chapter 6). with the Archaeological Survey of India, the Society
During excavation, all earth was sieved through a for South Asian Studies and the University of Durham
5-mm mesh for artefact and bone collection. A flotation main library.
sample was taken from all excavated layers in selected Some aspects of the archaeological documentation
trenches in order to make certain that a complete strati- that are not possible to publish here, for example data
fied archaeobotanical sequence was retrieved. tables, Harris matrices, interim reports and images,
A large palaeobotanical sampling programme was will be lodged with the Archaeology Data Service at the
organized during the excavations, and a ‘Siraf ’-type University of York with the following reference: http://
flotation machine was run constantly through the 1999 dx.doi. org/10.5284/1017460 (Kennet 2012).
season, producing what is now the largest palaeobot-
anical assemblage from any Early Historic site in India
(Chapter 12). Together with the detailed faunal report
(Chapter 13), this provides a unique insight into the
development of the agricultural economy during the
Early Historic and Early Medieval periods.
Shailendra Bhandare

Chapter 2 
Historical Context

The name Paithan derives from the town’s Sanskrit along the banks of River Godavari and the only jana-
name ‘Pratishṭhāna’, literally translated as ‘respectful pada to be situated to the south of the Vindhya Moun-
abode’. The name also has Prakrit variants such as tains. Its ‘capital’ was a city named Potali, Potana or
‘Patiṭhāna’, ‘Patiṭṭhāṇa’, ‘Paëṭhāṇa’, etc. Immediately Podana. The same is referred to as ‘Paudanya’ in the
prior to Indian independence in 1947, Paithan was part Mahabharata (1.122; 1.197). Some early scholars like
of the state of the Nizam of Hyderabad. It became part Bhandarkar identified this name with Paithan (see
of the Indian Union in 1948, when the Dominion of below), and this seems to be the reason Paithan is often
the Nizam was attached to the union following a police associated with the Assaka janapada. But judging by
action. the fact that Paithan was known by its Prakrit name
The study of the history of Paithan can be ‘Patiṭṭhāṇa’ and its variations, it is unlikely that Potali
approached mainly from literary and material sources or Potana could have been its historical name.
such as inscriptions. These sources help trace it back to As for the Baveru Jataka mentioned by Morwanchi-
the Early Historic period (c. 350 bc). kar, it is evident that he has confounded this reference
(Morwanchikar 1985: 10, note 24). It is not the Baveru
Jataka but the story of Bāvari, the orthodox philosopher,

Literary sources featuring in the Sutta Nipāta, that mentions Paithan

(Sutta Nipāta 5th chapter, Pārāyaa Vagga; Sircar 1960:
189). According to this story, Bāvari, having heard of
Paithan finds extensive mentions in indigenous genre
the Buddha and his preaching of a new philosophical
of ancient literature such as Buddhist, Jaina and Brah-
doctrine, sent 16 of his chosen disciples to debate it
manic texts. Amongst non-indigenous sources, its
with the Buddha. Bāvari resided in the country ruled
name also features in certain Western texts such as
by the King of Assaka and his students travelled to the
the Periplus of the Erythraen Sea and the Geography of
north, to the city of Sāvatthī, to meet the Buddha. The
Ptolemy. Myths and folklore have also grown up around
intellectual encounter of Bāvari’s students with the
the town, and some of these, that incorporate histori-
Buddha ended in the latter’s victory and he managed
cal strands, are worth a mention.
to convert the students to the very doctrine they had
come to debate. The story outlines the route they took
to travel northwards by mentioning towns that lay en
Buddhist sources
route, and Patiṭṭhāṇa features as the first town in the
list. If Bodhan is regarded as the capital city of Assaka,
It is generally believed that Buddhist texts like the
namely Podana or Potana, this description fits well
jatakas make references to Paithan. Morwanchikar
as Bodhan lies southeast of Paithan and one would
lists the Chulla Kalinga, Baveru and Assaka Jatakas as
indeed travel north to travel from Bodhan to Paithan,
amongst those which contain such references (Mor-
thereby also adding further strength to the argument
wanchikar 1985: 10). A reexamination of these sources
that these are different places.
indicates, however, that the Chulla Kalinga and Assaka
Paithan is often associated with another janapada
Jataka do not mention Paithan. Instead, they mention
named ‘Mūlaka’ (Morwanchikar 1985: 8; Sircar 1960:
the country (janapada) of Assaka and the King of
193, 273, note 1). The janapada-names ‘Aśmaka’ and
Assaka. Morwanchikar has obviously inferred, albeit
‘Mūlaka’ are often referred to conjointly (possibly due to
indirectly, this mention to stand for Paithan.
their occurrence as such in a Satavahana inscription,
Assaka (Skt. Aśmaka) is one of the 16 Mahājanapadas
of a date much later than the Buddha) and it is com-
mentioned in another Buddhist text, the Anguttara Nikāya
monly understood that the mention of Mūlaka, like
(I: 213; IV: 252, 256, 261). It is described as being located
Aśmaka, is also of a Buddhist origin. It has to be empha-

https://doi.org/10.1515/9783110653540-002
12   Shailendra Bhandare

sized that Mūlaka does not feature amongst the 16 Jaina texts are replete with references to the Satava-
Mahājanapadas listed in the Anguttara Nikāya (I: 213; hana-Kshaharata conflict. According to the Brhatkalpa-
IV: 252, 256, 261). Its mention as a janapada is defini- Bhāśsya Pīt.ikā, king Sālivāhana of Paithan lay siege to the
tively from Brahmanic sources, such as Puranas, which Kshaharata capital Bharukachha (Skt Bhṛgukachchha,
date considerably later than the Buddhist canon. Much or modern Bharuch) every monsoon (Deo 1975: 3). A
like Aśmaka, the country of Mūlaka was also appar- text named Tiloya-Pan.n. ati states that a king named
ently located along the banks of the Godavari and ‘Naravahana’ (Nahapana) ruled at Bharukachha and
there is nothing to suggest that Paithan was a part of provides a chronology up to his rule, making a mention
the janapada apart from the fact that it, too, is situ- about his war with the ‘Sālivāhana’ of Paithan (Rajgor
ated on the Godavari. We find, however, that some men- 1991: 34). In the Āvaśyakaniryukti (v. 1304), the Jaina
tions of Paithan are reported entirely out of an implied scholar Bhadrabahu narrates an interesting story
and indirect association with Mūlaka. Yazdani not only involving ‘Nahavana’, the king of Bharukachha,
suggests that Paithan was located in Mūlaka but he whose enemy was ‘Satavahana’, the king of Patiṭṭhāṇa.
also asserts that it was the ‘capital’ of this janapada Satavahana besieged Bharuch for 2 years but could
(Yazdani 1960: 25). D. C. Sircar follows Yazdani in reit- not take it, because Navahana could endure the siege
erating the same contention (Sircar 1960: 189). owing to the riches he had accumulated through trade
It is thus evident that many instances where passing through the port of Bharuch. Satavahana
Paithan is reported as being mentioned in Buddhist then retreated and employed stealth tactics: he sent
literature are inferential and not substantive, the only one of his ministers, named Kharaga, to Nahavana
exception being the story of Bāvari and his students pretending he was defecting. Kharaga sought asylum
from the Sutta Nipāta. Amongst later Buddhist texts, at Bharuch and won over Nahavana’s confidence. He
the Mahāmayurī, a lexicon of Yaksha cults and their influenced Nahavana’s religious beliefs and encour-
locations, mentions Paithan as the abode of a Yaksha aged him to give his wealth away in piety. Nahavana
named Khandaka (Agrawal 1942: 24–25). followed his advice and gave large donations to
various sects. When he was thus impoverished, Sata-
vahana attacked and defeated him.
Jaina sources There are a further two stories involving ‘the Sata-
vahana king’ and Paithan. The first is about the king
The major source of information on Paithan from Jaina and his associate Śūdraka, while the second narrates
sources is J. C. Jain’s Life in Ancient India as Described in origins of the cult of a local goddess named ‘Pīṭhajādevī’,
the Jain Canon and Commentaries (Jain 1984: 400–401). practiced by married women (Jinaprabhasūri 61–64).
The Jaina texts B.rhatkalpabhās.ya (vv. 6244–6249) and Both stories mention many topographical features of
Vasudeva Hindi (354, 8) describe Paithan as the chief the town and incorporate certain elements of the ‘Sat-
city of Maharashtra. The Kalpasūtra mentions Paithan as avahana myth’ (see below).
a seat of learning (4, 90a) and the GacchācārV . rtti (208)
states that great scholars like Bhadrabahu and Vara-
hamihira were residents of Paithan. Many Jaina texts Brahmanic sources
mention Paithan explicitly in connection with the Sata-
vahanas. For example, according to the Kālakāchārya In the Early Medieval period, Paithan emerged as a
Kathānaka, a Satavahana king (Sālivāhana) was ruling pre-eminent centre of Brahmanic religion and thus
at Patit.hān.a when the Jaina monk Kālakāchārya visited finds a mention in several Brahmanic texts. These
him (Morwanchikar 1985: 19). According to another include the early Purāṇas, such as the Matsya and
story that features in a text named Niśīthacūrṇi (v. 3153), BrahmaPurāṇa, and later Purān.as, such as the Padma
king Sālivāhana pleaded with Kālakāchārya to move and LingaPurān.a (Morwanchikar 1985: 1, 11). The epic
the dates of Pajjusan.a, a Jaina festival. A text named Mahabharata mentions a Pratishṭhāna-pura but it has
Pin.d.adaniryukti (v. 494, 498) states that a Jaina monk been identified with Jhusi near Allahabad, known in
named Pādalitta Sūri cured the headache of a ‘Murunda’ Jaina sources as Poyanapura (Dey 1899: 71; Deo 1975:
king of Paithan (Bollée 1994: 29). The Murundas were 1). In the BrahmaPurān.a, the story is told of the son of
an ancient tribe of North India, and their association king Surasena of Pratisthana, who is a Naga or snake
with Paithan is otherwise unknown. (BrahmaPurān.a 111). This son marries the daughter of
king Vijaya, named Bhogavati, who finally discov-
 Chapter 2  Historical Context   13

ers the divine, snake nature of her husband. They both that subsequently became associated with Paithan.
go to the Gautami River to bathe. Eventually, the Naga Yusuf ’s statement that Pliny mentions Paithan as the
rules the country together with his wife Bhogavati. At ‘glorious capital of the Andhras’ may serve as a case
the death of his father (Surasena), he and Bhogavati go to illustrate this (Yusuf 1939: 39). The fact remains
back to Sivapura (heaven). that Pliny mentions only the ‘Andarae’ but not Paithan
It is considered important to perform Brahmanic while describing the peoples of India, following
rites such as Śrāddha for the deceased at Tīrthas (lit- Megasthenes’ description of Indian tribes (Pliny VI,
erally ‘fords’ for the soul to cross over to Moksha, its 22). The ‘Andarae’ are identified with the ‘Andhras’,
escape from the cycle of rebirths) where such rites which in turn is the term that sources like the Purāṇas
could be held under Brahmanic supervision. A number employ to describe the Satavahanas. A general belief
of these evolved along the riverfront at Paithan so that prevails that the Satavahana ‘capital’ was located at
it became a centre of pilgrimage and a ‘SthalaPurān.a’ or Paithan (discussed below). It is evident that Yusuf
religious site-guide named Pratishṭhāna-Mahātmya has conflated all these disparate aspects to attribute a
for Paithan is also known and is included within the mention of Paithan to Pliny.
BrahmaPurāṇa text (Morwanchikar 1985: 11, 58). Specific references to Paithan are found in two
Greek texts: first the Geographike Uphegesis (or General
Geography) of Ptolemy and, second, the Periplus Maris
Secular Prakrit literature Erythraei, a seafaring manual attributed to an unnamed
Hellenic Egyptian who seems to have regularly plied
Paithan is mentioned in the Bṛhatkathā, a prominent the maritime trade route between the Red Sea and the
Prakrit literary work attributed to a poet named Gunāđhya Indian coastline. Both sources are roughly contempo-
and dated to the early Satavahana period, c. 100 BC–AD rary: the Periplus is dated to around ad 60–70, while
100. Gun. āđhya was not only a resident of Paithan but Ptolemy’s geography was written only a few decades
also had a Naga-Brahmana ancestry, similar to what later, in the late 1st to early 2nd centuries ad (Schoff
was mythically claimed for the Satavahanas. Indeed, 1974: 7–15).
the Bṛhatkathā makes a mention of the Satavahana Chapter 51 of the Periplus describes Paithan as
king, resident at Paithan, and contains a famous story follows:
about his ignorance of the Sanskrit language. It was
also the first to incorporate a myth about the origins of Of the trading centres in the region of Dachinabades, two are
the most outstanding: Paithana, twenty days’ travels to the
the Satavahanas into a literary form. From Gun.āđhya’s
south from Barygaza; and from Paithana, about ten days to
Br.hatkathā, the literary lore surrounding Paithan was the east, another very large city, Tagara. From these there
transmitted to Kathā-Saritsāgara of Somadeva and is brought to Barygaza, by conveyance in wagons over very
Bṛhatkathā-Manjirī of Kshemendra, which are dated great road-less stretches, from Paithana large quantities of
roughly to the mid-first millennium ad, and treat the onyx, and from Tagara large quantities of cloth of ordinary
quality, all kinds of cotton garments, garments of Molochinon
Bṛhatkathā as their antecedent (Penzer 1968: 60ff).
and certain other merchandise from the coastal parts that
The Gāthā Saptaśati, or Sattasaë as it is called in finds a market locally there.
Prakrit, is a prominent Prakrit anthology said to have
(Casson 1989: 82–83).
been patronized by the Satavahana ruler Hāla (around
the early 1st century ad). It has references to the River
Here, ‘Dachinabades’ is evidently Dakshinapatha, the
Godavari and urban life that flourished on its banks
traditional Sanskrit term for the ‘Southern Country’.
(Ketkar 1963: 397, 422).
‘Barygaza’ is the Greek version of ‘Bharukachha’, the
Prakrit name of modern Bharuch, and ‘Tagara’ is iden-
tified with Ter, an important emporium located in the
Western texts
Deccan hinterland.
Ptolemy’s Geography (McCrindle 1885: 152) describes
There is a general belief that many classical schol-
‘Baethana’ as the ‘royal seat’ of a ruler named ‘Siro Pole-
ars like Arrian, Strabo and Pliny mention Paithan in
maios’ – conceivably, the Satavahana ruler Vasithiputa
various capacities. A fresh look at these sources con-
Siri Pulumavi. This reference has been widely inter-
firms, however, that no direct reference to Paithan fea-
preted to identify Paithan as a ‘capital’ of the Sata-
tures in them. Many of the so-called references are
vahanas. In the same section, Ptolemy also mentions
mere implications based on mentions of other facts
14   Shailendra Bhandare

two other cities as ‘royal seats’: ‘Ozene’ as the seat Another religious sect that contributed widely to
of ‘Tiastenes’ and ‘Hippokoura’ as that of ‘Baleok- Marathi literature of the Yadava and post-Yadava or
ouros’. These rulers are respectively identified as the Sultanate periods was the egalitarian Varkari sect,
western Kshatrapa ruler Chashtana and a local ruler which surrounded the cult of Vitthala or Vithoba,
named Vilivayakura. Ptolemy thus affords a valuable regarded as a form of Vishnu. A progenitor of the sect
‘isochronism’ linking these three rulers and indicat- and pioneer of Marathi literature, Dnyaneshwara, was
ing their contemporaneity. While ‘Ozene’ is certainly a resident of Paithan in his early life, where his father
Ujjayini or modern Ujjain, the identity of ‘Hippokoura’ held a tenurial office of the ‘Kulkarni’. Though many of
is uncertain. This name features once more in Ptole- Dnyaneshwara’s early activities took place in Paithan,
my’s geography as that of a port on the western coast he himself does not mention the town in his literary
near ‘Semylla/Tiamulla’, or modern Chaul (McCrindle works. The association with Dnayneshwara, however,
1885: 359). In this instance, it can be safely identified meant that Paithan was noted in subsequent Varkari
with Kuda, a harbour-town south of Chaul, known literature, starting with Dnyaneshwara’s posthumous
for a complex of rock-cut caves. But based on Ptole- biography by his associate Namdeva (Morwanchikar
my’s description, it is plausible to identify the ‘Hip- 1985: 78–80).
pokoura’ where ‘Baleokouros’ reigned as located not In the late 15th to early 16th centuries, a noted
on the coast, but in the hinterland. Possibly, it was a Marathi ‘saint’ Eknath resided at Paithan. Eknath
Greek name of Brahmapuri near Kolhapur, which was was a Brahmin and a follower of the egalitarian Vit-
a prominent emporium on trade routes that traversed thala sect. He was also acquainted with heterodox
the Deccan plateau and where a large number of coins doctrines such as Sufism. He contributed immensely
of the Kura dynasty, to which two rulers named ‘Vili- to Medieval Marathi poetic literature on the theme of
vayakura’ belonged, have been found (Rapson 1908: Bhakti. Some events of his ‘saintly’ life, that live in
lxxxvi-lxxxvii). the popular memory of Maharashtrians, took place
at Paithan. Noteworthy amongst these is his embrace-
ment of a low-caste child along the banks of Godavari
Marathi sources and another where he took a long journey back to the
river so he could water a thirsty donkey (Morwanchikar
The town of Paithan is closely associated with the 1985: 82–86).
flourishing of Marathi literature in the late Yadava
period (c. 12th–13th centuries ad). The Yadava capital
at Devagiri (Daulatabad of present) was not situated Folklore
very far from Paithan and the fact that Paithan had
evolved as a prominent centre of Brahmanic religion in The folklore surrounding the origins of the Satavahana
the centuries prior to Yadava rule meant it attracted the dynasty is intricately linked with Paithan. According
attention of the Yadavas, who patronized Brahmanism to a folk story, a Brahmin lady lived in Paithan. When
in a major way. The chief Marathi source for informa- she went to fetch water at a place called ‘Nagađoha’
tion on Paithan in this period is, however, the litera- along the River Godavari, the Naga (serpent) king
ture associated with the heterodox Mahānubhāva sect, Śésha impregnated her. The child born out of this
which was founded by a seer named Śri Chakradhara encounter was named Sālivāhana. As a child, he would
in the mid-13th century ad (Morwanchikar 1985: 11). make terracotta horses and soldiers to play with. When
Chakradhara travelled widely across the Marath- Vikrama, the ruler of Avanti (Ujjain), attacked the city
wada and Vidarbha regions to preach his doctrines of Paithan, Sālivāhana brought his toy army to life by
and won many followers. During these travels, he also sprinkling the magical nectar Amṛta on it (according to
visited Paithan and lived there for a while. It was at mythical belief, the Nagas guarded repositories of Amṛta
Paithan that he took vows of renunciation and became in the Netherworld and Sālivāhana had access to it
a Sanyāsin (mendicant). Early Mahānubhāva literature, owing to his Naga parentage). He defeated Vikrama
such as Leelā-Charitra, Chakradhara’s spiritual biog- with the help of this army and subsequently became
raphy, Sthānpothī or the gazetteer of Mahānubhāva the first king of the Sālivāhana (or Satavahana)
pilgrimage and Smṛtisthala, a compilation of Chakrad- dynasty (Morwanchikar 1985: 9–10).
hara’s memoirs, mentions Paithan at various instances Another Medieval myth connects Vikrama with
(Raeside 1976: 587, 594). Sālivāhana of Paithan in a different way. According
 Chapter 2  Historical Context   15

to this story, astrologers at Vikrama’s court predicted of Paithan (Bhandarkar 1925: 32–33), the appellation
that Vikrama would be slain at the hands of a man deriving from a ‘Prakritised’ version of Paithan’s San-
born in the Southern Country or Dakshin.āpatha. They skrit name ‘Pratishthana’. Two other peoples, namely
also pronounced the tell-tale ‘signs’ of the killer. Dis- the ‘Rathikas’ and ‘Bhojakas’, find a mention along-
turbed by the oracle, Vikrama commanded Vetāla, the side the ‘Petenikas’ in the same edict. As these are
king of Bhūta spirits, to recognize and find him so he identified amongst the early ethnic inhabitants of the
could be killed. Vetāla came to Paithan in his search Deccan plateau, their mention alongside the Petenikas
and discovered that the ‘signs’ matched a boy named helps the latter to be located in the same geopolitical
Sālivāhana, who lived in the city quarters inhabited by region and thus adds support to the identification of
potters and mud-workers. He was playing with a ter- the ‘Petenikas’ as the residents of Paithan. This claim,
racotta army of horses and elephants when Vetāla however, has been debated, mainly based on different
spotted him. Vetāla reported his discovery to Vikrama, ways the word could be derived from Sanskrit/Prakrit
whereupon Vikrama launched an attack on Paithan. (Bhandarkar 1925: 32–33).
But Sālivāhana brought his toy army to life by magic Amongst the inscriptions from Buddhist sites
and with its help defeated and killed Vikrama, thus such as the cave temples of the Deccan, Paithan finds
fulfilling the prophecy (Ketkar 1963: 80–101). mention as ‘Patithana’, a Prakrit version of its Sanskrit
These myths contain a curious combination of legend name. Three inscriptions of a votive nature, one in
and reality. They surely refer, in an indirect manner, to Kanheri cave no. 3 and the other two in the Pitalkhora
the rivalry between Satavahanas and the Kshatrapas of Chaitya cave, refer to residents of Paithan. The Kanheri
Ujjain, but pose Vikrama, the mythical hero of Avanti, inscription mentions the establishment of a monastery
as the Satavahana enemy. The second version described at Paithan (Lueders 1909–1910: 102, no. 988). The first
above assimilates another well-known ‘Vikrama’ myth – of the Pitalkhora inscriptions makes note of a dona-
that of his association with the spirit-king Vetāla – into tion of a pillar by an unnamed ‘Son of Saghaka’, a resi-
the Satavahana thread. The reference to the toy army dent of Paithan, while the second mentions the gift of a
seems to reflect popular memory about artefacts such as pillar by Mitadeva, from a perfumer (Gandhika) family
terracotta horses, elephants and human figurines being of Paithan (Lueders 1909–1910: 137, nos. 1187–1188).
regularly unearthed at Paithan. It is also plausible that Although Paithan is widely regarded as a capital of
the Naga-Brahmana root of the Satavahanas was a histor- the Satavahana dynasty (see below), it is not mentioned
ical reality, Nagas being a well-known ancient clan who in any capacity in any Satavahana inscriptions. The
were, in the course of time, mythified as the ‘dwellers of names of ancient geopolitical regions Aśmaka and
the Netherworld’ (Fergusson 1868: 125). Mūlaka feature in the eulogy of Gotamiputa Siri Sata-
A popular belief that Paithan is a ‘city turned kani inscribed in Nasik cave no. 3 by his son and suc-
upside down’ (Ultī Nagarī to quote the Marathi term) cessor Vasithiputa Siri Pulumavi (Lueders 1909–1910:
reflects the archaeological reality of encountering 122, no. 1123). But we have already seen the fallacies
older anthropogenic remains such as bricks under involved in associating Paithan with such ancient
the ground during later-day construction and agri- regions.
cultural activities (Yusuf 1938: 3). Similar beliefs also
exist about other ancient cities like Ujjain and usually
a local myth flourishes around the causes of the city Copper-plates
being turned upside down.
An administrative division or Āhāra named Supratishṭha-
āhāra is mentioned in two Vakataka copper-plate char-

Material sources ters, namely the Jamb copper-plate and the Wade-
gaon copper-plate of Pravarasena II (c. ad 400–440)
(Mirashi 1963: 10–15, 53–56). The first copper-plate is
Stone inscriptions dated to his 2nd regnal year, and the second, to the 25th.
The language of these plates is Sanskrit. V. V. Mirashi,
The inscriptional sources on Paithan are rather limited. while discussing the place names mentioned in the
Asoka, the great Mauryan emperor, mentions a people charters, tried to identify them using a ‘structuralist
named ‘Petenikas’ or ‘Pitinikas’. This mention has linguistic’ method, i.e. tracing the ‘Sanskritisation’
been widely recognized as a reference to the people roots of modern place names. He therefore contended
16   Shailendra Bhandare

that the administrative division should be identified the ‘capital’ of these kings; the term he uses is trans-
with the area surrounding the village Pothra in Hin- lated by McCrindle as ‘royal seat’, which may not
ganghat Taluka of Nagpur District. H. S. Thosar chal- necessarily mean a ‘capital’ in its true sense. Indeed,
lenged this attribution and tried to show, following the the modern connotations of the word ‘capital’ were
same methodology as Mirashi, that the places men- unknown to classical writers such as Ptolemy.
tioned in the plates can be located in the Marathwada What prompted Paithan’s identification as a Sata-
region, and thus, the region around Paithan should be vahana ‘capital’ are the secondary corroborations to
identified as Supratishṭha-āhāra (for details of Thosar’s Ptolemy’s reference which come from Indian sources.
arguments against Mirashi, see Morwanchikar 1985: Chief amongst these are some of the Jaina sources
29–30). To further substantiate his claims, Thosar drew described above but some substantiation is also to
on secondary evidence from the Kathā-Saritsāgara (see be found in the myths surrounding the origins of the
above in the section on literary evidence), which refers Satavahana dynasty. The Prakrit secular literature
to Paithan as Supratishṭha Nagari. Thosar’s contention described above also makes similar connections. It is
that places mentioned in the copper-plates do exist in worth noting, however, that both these sources hardly
the Marathwada region is worthy of consideration. The make a distinction between ‘Satavahana’ (and ‘Saliva-
fact remains, however, that Supratishṭha-āhāra and hana’ to use the term widely employed in the texts) as
Supratishṭhita Nagari are two essentially different words, a dynasty and a person. As evident in these texts, ‘Sata-
and even if one is linked to Paithan, to what extent the vahana’ as a dynastic name is non-existent; to them, it
other can be linked in the same way is open to inquiry. means only a personal name. This is rather ambigu-
Thosar’s identification of Supratishṭha-āhāra is there- ous and undermines the reliability of such sources in
fore not conclusive. drawing firm historical conclusions. Furthermore, it
Two copper-plate charters of the Rashtrakutas are is also true that even these secondary sources do not
associated with Paithan. The first, given by king Govind explicitly mention Paithan as the ‘capital’.
III (c. ad 793–814) in ad 794 and known as the ‘Paithan Inscriptional evidence from the Satavahana period
copper-plates’, records the donation of a village to a is silent about Paithan and its association with the
group of Brahmins (Kielhorn 1894–1895: 103–110). dynasty. The name of the town does not feature in any
This village, named ‘Limbārāmikā’ in the plates, was of the known Satavahana inscriptions, let alone there
part of an administrative division named Pratishṭhāna being a direct reference to the seat of government being
Bhukti, conceivably comprising the area around Paithan. there. The ancient regions Aśmaka and Mūlaka feature
Limbārāmikā is noted to have been a part of a group of amongst those mentioned in the eulogy of Gotamiputa
12 villages, known as ‘Sārākachchha’. The second cop- Siri Satakani inscribed in Nasik cave no. 3 and could,
per-plate charter given under the Rashtrakutas dates to only indirectly and inferentially, be taken as a refer-
the reign of king Indra III (c. ad 914–929) and is known as ence to Paithan. But we have already seen the fallacy
the ‘Jambgaon copper-plates’ (Mirashi 1966: 223–238). of associating Paithan with these regions, owing to a
It records the gift of a village named ‘Khaërondhi’, in general lack of evidence.
celebration of the anniversary of Indra’s coronation in Archaeologically, there exists enough evidence
ad 915. Khaërondhi is identified with Khirai, located a to indicate that Paithan flourished under the Satava-
short distance to the north-west of Paithan. This cop- hana rule, but it does not offer any indications of the
per-plate charter refers to Paithan by its Apabhramsa/ town being a capital. Recently, a sealing of a ‘Royal
Prakrit name variant ‘Paët..thāṇa’. Minister’, or Rājāmātya, named Mahasenadata has
been reported from Paithan (Godbole 2002–2003: pls
II-III). Of the four sealings published by Godbole, two

The Satavahanas and Paithan are fakes, whilst the other two are genuine but have
been misread. Correct readings are ‘Sivadatasa’ (‘of
Sivadata’, the name of an individual) and ‘Rajama-
As shown above, Paithan is recognized widely as the
chasa Mahasenadatasa’ (of the ‘Royal Minister’ Maha-
capital of the Satavahanas, mainly on the basis of textual
senadata).
sources. Most convincing amongst these is Ptolemy, who
They can be dated on palaeographical grounds to
mentions ‘Baithana’ as the ‘royal seat’ of a king named
the mid- or late-Satavahana period (c. 2nd–3rd centu-
‘Siro Polemaios’ (Majumdar 1960: 376). It is to be noted
ries ad) and could be taken as evidence of individuals
that Ptolemy does not explicitly mention the town as
with governmental responsibilities having resided at
 Chapter 2  Historical Context   17

Paithan during the Satavahana period. This, however, While successive researchers accepted this cor-
still does not offer a full substantiation of the ‘capital’ rection, Bhandarkar’s assertion that Dhanyakataka
being situated at Paithan. In this respect, it is worth was a Satavahana ‘capital’ has survived in the lit-
noting that a Rājāmātya named Hala is mentioned in an erature. Added to this was the need to accommodate
inscription at Kuda (Lueders 1909–1910: 112, no. 1057), Ptolemy’s mention of Paithan. A curious assertion of
and other Amātyas, such as Sivakhadila and Vinhu- historical detail followed this need – that of the Sat-
palita, are known from Nasik inscriptions (Lueders avahana Empire having two capitals, one at Dhan-
1909–10: 123–124, no. 1125). These do not necessarily yakataka and the other at Paithan (Yazdani 1960:
indicate that both places were ‘capitals’. 132). But it is evident that this has been inferred from
Notwithstanding this, the identification of Paithan erroneous data, namely the wrong reading of a word
as the Satavahana capital is often accepted without by Bhandarkar, and we have no further evidence that
question. The reading and interpretations of certain epi- either city was Satavahana ‘capital’.
graphic sources in conjunction with textual references To conclude, Paithan may have been a Satavahana
(predominantly Ptolemy) have contributed to it, and in city par excellence, and its early flourishing might be
general to the debate. attributed to the Satavahana period. By the time of
R. G. Bhandarkar drew the inference that Dhan- Ptolemy, it certainly seems to have become an empo-
yakataka (present-day Dharanikota in the lower- rium and was evidently a ‘royal seat’ of the Satavahana
Godavari or ‘Amaravati’ region of Andhra Pradesh) king Siri Pulumavi. The folklore and myths that sur-
was a Satavahana ‘capital’ on the basis of a mention round the city may even have roots in the city’s Sata-
of a word which he read as Dhanakaṭakasāminehi in vahana past. It is important to emphasize, however,
an inscription of Gotamiputa Siri Satakani at Nasik. that there is no clear and unequivocal evidence that
R. G. Bhandarkar took the word as an adjective for the Paithan was the capital of the Satavahanas at any time.
king’s name and translated it ‘by the Lord of Dhan-
yakataka’, thereby inferring that the capital of Siri
Satakani was situated at Dhanyakataka (Bhandarkar
1895: 41). But Senart, who edited and published this
inscription in the late 19th century, read the word as
Benakataka and not Dhanakataka (Senart 1905–1906:
7). V. V. Mirashi subsequently proved that the word
was indeed Benakataka and corrected Bhandarkar’s
reading to Benākaṭakasamanehi, referring not to ‘the
Lord of Dhanyakataka’ but to the monks (Saman.a, Skt.
Śramaṇa) from the region of Benakataka, or the valley
of River Bena in Vidarbha (Mirashi 1979: [9]).
Derek Kennet

Chapter 3 
Previous Archaeological Work at Paithan

A considerable number of archaeological sites have ficult. Nonetheless, the excavated sites that have been
been excavated and published in the western Deccan published do act as a useful repository of compara-
and the region surrounding Paithan. Although none tive information which has been drawn on in different
was specifically aimed at investigating Early Medi- places in the present volume. Meanwhile, systematic
eval levels or the Early Historic/Early Medieval transi- field survey and mapping of rural settlement of the
tion, many quite significant Early Historic sites have type that has been standard practice in Europe, the
been explored in some detail (Fig. 1.1). In fact, it was Mediterranean and the Near East since the 1950s are
H. D. Sankalia and his colleagues from Deccan College still relatively rare in India, and unfortunately, no such
Pune who initiated this work with the excavations at projects have been undertaken and published in the
Brahmapuri in the mid–1940s (Sankalia and Dikshit vicinity of Paithan that might help to contextualise the
1952), continuing with Nasik, Maheshwar and Nevasa site within its broader landscape, with the exceptions
through the 1950s (Sankalia and Deo 1955; Sankalia of Shaw’s work in the vicinity of Sanchi, some distance
et al. 1958, 1960). At around the same time, excava- to the north (Shaw 2007) and recent work in Vidarbha
tions had been taken up at Baroda (Subbarao 1953) reported on in Chapter 15 of this volume.
and at Prakash in the Tapti Valley (Thapar 1967). From At Paithan, previous to the excavations reported
then until the early 1970s, there was an almost con- in this volume, there had already been at least three
stant stream of excavations at sites such as Kayatha campaigns of excavation, two of which have been pub-
(Wakankar 1967; Ansari and Dhavalikar 1975), Kaun- lished. The first was conducted by the Archaeological
danpur (Dikshit 1968), Nagara (Mehta and Shah 1968), Department, H. E. H. The Nizam’s Government under
Paunar (Deo and Dhavalikar 1968), Ter (Chapekar the directorship of Syed Yusuf, beginning in May 1937
1969), Pauli (Deo and Joshi 1972) and Bhokardan (Deo (Yusuf 1938, 1939). The second was undertaken in 1965
and Gupte 1974), with Pauni also having been reex- in advance of the construction of the nearby Jayakwadi
cavated in the 1990s (Nath 1998), highlighting the dam and was directed by Dr M. G. Dikshit of the Depart-
fact that the quantity of excavations appears to have ment of Archaeology, Government of Maharashtra.
trailed off in recent years. In addition to the Early His- Dikshit undertook a rescue survey in the area that was
toric levels that they revealed, many of the excavation to be submerged and also opened a small trench in the
projects mentioned above uncovered some evidence vicinity of the Narasimha Temple on top of the ancient
of Medieval occupation, although this is for the most mound in order to check the stratigraphic sequence
part 14th century and later as a lack of evidence is (Dikshit 1973). A third, limited excavation was carried
reported at many sites from the Early Medieval period. out at Paithan in 1995 by the Department of Archaeol-
This is a phenomenon which appears to be partly due ogy and Museums of the Government of Maharashtra
to problems on chronology and interpretation but may in collaboration with the Department of History and
also have some validity as an indicator of a general Ancient Indian Culture of the Dr Babasahib Ambedkar
lack of urban continuity through this period (Sharma Marathwada University, Aurangabad at the Son-Tekadi
1987; Kennet 2004b, 2013). mound, but the results of this excavation are not pub-
Given the number of Early Historic excavations lished.
that have been carried out in the region, the lack of any The two published excavations both revealed
synthetic regional analysis of the results is notable. deep occupation sequences of four to six phases and
This seems partly to be due to problems with chronol- throw some light on the history of occupation at the
ogy, which is still very poorly defined, making it dif- site. Their results are summarized below.
ficult to link phases of activity at different sites with
any reliability, and partly due to the way in which the
data are presented, making intersite comparisons dif-

https://doi.org/10.1515/9783110653540-003
 Chapter 3  Previous Archaeological Work at Paithan   19

Yusuf’s 1937 excavations –– Third Stratum. This layer contained antiquities and
silver and copper coins of the Mughal Period.
The first excavations at the site began in 1937 when four –– Second Stratum. This layer can be dated to the
trenches were opened by Yusuf on the northern bank 18th and 19th centuries as it yielded coins of the
of the Godavari (Yusuf 1938, 1939). The location of his Asif Jahi monarchs together with other objects. No
trenches is shown in Fig. 1.4. In some places, these were details of the nature of occupation are given.
excavated to a depth of 7.5 m. The excavations identified –– First Stratum. This is the topmost layer consist-
six strata or layers of occupation, which can be summa- ing of buildings constructed in stone and lime,
rized as follows according to Yusuf’s report: and brick and mud, found together with British
–– Sixth Stratum. This layer, dated to the Satavahana silver rupees, gold and silver jewellery, copper and
period on the basis of coins, was characterised pottery utensils.
by a monumental architectural phase with the
richest cultural assemblage of the sequence. Two Despite bringing to light the historical importance of
rectangular brick structures were uncovered side the site, excavations were not continued the following
by side, with a narrow passage between them; the year and the disruptions caused by the outbreak of
walls of these structures are said by Yusuf to have the Second World War meant that the project was never
survived to a height of 1.06 m. These structures resuscitated. Yusuf ’s results are further discussed in
are the two Early Hindu temples that became a key Chapter 14.
focus of the present excavations (see Chapter 4).
To the south, at a slightly greater depth, a length
of drain was discovered, constructed of very large
bricks (10 × 24 × 62 cm) and no lime mortar. Three
Dikshit’s 1965 excavations
cylindrical brick structures, each 1.37 m in diameter,
Dikshit’s 1965 excavations in the vicinity of the Nar-
were found in line with the drain. Copper-alloy coins
asimha Temple on top of the ancient mound revealed
of the Satavahana kings, including a conglomerated
a four-period sequence (Fig. 1.4; Dikshit 1973). These
lump of 10 square coins, with the emblem of the
periods are summarized from Dikshit’s report as
Bodhi tree stamped on one face, were found with
follows:
traces of a cloth wrapped around them. Numbers
–– Period I. In these layers, two supposedly early
of individual similar square coins came from the
Satavahana coins with a Gaja-Lakshmi motif were
same layers. Barrel-shaped, globular and discoid
found together with glass beads, crystal ear-reels,
terracotta beads are reported as having been found
legged querns, a fine bone or ivory lion capital,
in large numbers at a very low depth in the trench,
kaolin figurines of a nude goddess and an ivory
which is probably intended to mean this period.
kohl applicator. Small fragments of worn North-
–– Fifth Stratum. This layer appears to represent a gap
ern Black Polished ware (NBP) were also found in
in occupation or a major re-levelling of the site. It
the upper levels. This period is dated by Dikshit
is up to 2.4 m deep and contained only a few minor
roughly to 300 bc–ad 100, probably based on the
antiquities of the Satavahana period. Above this
coins and NBP, although this is not specified by
layer, the collapsed structures of the fourth stratum
him.
were found. At this point, there is some ambigu-
–– Period II. At this level, thick deposits of mud and
ity in Yusuf ’s description about the number of
silt came to light to a depth of 1.22 m. These are
structural layers and their relationship to the Fifth
interpreted by Dikshit as flood deposits, but it
Stratum.
is unclear whether this is correct. This period is
–– Fourth Stratum. This is a large destruction layer,
dated c. ad 100–500, but no evidence is presented
possibly associated with a flood according to
to support this date and it is not further discussed.
Yusuf. Building remains were found tilting, slanted
–– Period III. A portion of a brick temple in a ruined
or collapsed over the remains of the Fifth Stratum,
condition came to light in these levels. The plinth
which caused Yusuf to suggest that they had been
and jangha were built of very large bricks, but the
destroyed by flood water. Antiquities included
size is unfortunately not stated. Iron objects and
Tughluq and Bahmani coins, which suggest a date
slag were found in the foundations of the structure.
of late 14th or 15th century.
This phase is tentatively ascribed to the Rashtrakuta
20   Derek Kennet

period (ad 600–800), but no evidence is cited to to confirm Satavahana-period occupation at the site.
support this date. In fact, both excavations report Satavahana coins in
–– Period IV. A disturbed layer in which an under- the lowest excavated levels, which suggests that – in
ground chamber of a house, approached by a flight these two particular locations at least – this may have
of steps in the walls, was found. The finds included been the earliest period of occupation. However, in
crude and poorly fired pottery, generally of the neither case is the excavation report detailed enough
common red variety. The period is dated ad 1700– to be certain that traces of earlier occupation were not
1800 but again, no supporting evidence is given. also present. Both excavations also report a thick layer
of abandonment or ‘flood deposits’ immediately above
These two excavations throw some useful light on the these levels, which is then followed by Medieval and
history of occupation at Paithan, although it is not very post-Medieval occupation levels. The significance
precise and the evidence for the interpretations given of these results, along with a reinterpretation, is dis-
by the two authors is not always clear. The presence of cussed in more detail in Chapter 14.
Satavahana coins is of no surprise and would appear
Part II
The Excavations
Derek Kennet

Introductory Remarks on the Excavations

The Paithan excavation project has a slightly complex After the end of the 1996 season, the Institute of
history. It was begun under the directorship of P.  N. Archaeology in New Delhi undertook a training exca-
Kamble and J. Howell, who directed the project during vation at the site, opening a trench, 35 × 5 m, close to
its 1996 and 1997 seasons. In 1998, it was handed over to the area where excavation had already been started
J. V. P. Rao and D. Kennet, who formulated a new set of (Trench C). The results of this work are not included
research questions and aims (see Chapter 1) and directed in the present report, but after further work in the 1997
the 1998 and 1999 seasons. season, a rough occupation sequence was elucidated
Some limited survey of the site and its environs and is described under Trench C in Chapter 5.
was undertaken during the 1996 and 1997 seasons, and In the second (1997) season, one of the excavation
pottery and other minor antiquities were found at a trenches dug by Yusuf in 1937 was cleared to allow
number of localities. During this work it was noted that further investigation of the two brick structures uncov-
the low plain to the east of the main mound within the ered by him. The southern section of Yusuf ’s trench
area enclosed by the bend in the river yielded a notable was cut back in an attempt to link the stratigraphic
amount of scattered occupational material, suggesting position of the buildings with later material, and a
that this area may have been quite densely occupied at test pit was excavated to establish the dating of earlier
some periods of the site’s history. In addition to this, occupation in this area (Trench A).
it was noted that there are several smaller mounds During this season, the Institute of Archaeology
around the main mound, which have ancient archae- trench at the southern end of the site (Trench C) was
ological material associated with them. There was cleaned and excavated down to natural across a 2-m-
speculation that these sites might be early monastic wide section in each of the seven sectors; the baulks
or religious centres but they have never been explored were also cut back to provide a rough sequence which
archaeologically and they may also result from other is reported in Chapter 5. The fortification that sur-
types of structures on the site’s margins. The most sig- rounds the southern end of the site was also examined
nificant of them is a mound on the bank of the river in this season. The fortification wall stands as a large
to the north of the Ek Nath temple, which has now mound at least 4 m above the surrounding area and is
been partially levelled to create a sports field, and a capped with a post-Medieval wall.
second mound, also on the river bank but to the east of The main focus of work during the third (1998)
the town, upon which there now stands the Maulana and fourth (1999) seasons was Trench A focussing on
Sahib Durga temple (Fig. 1.4). the brick structures revealed by Yusuf that had by this
In the first (1996) season, an excavation trench was time been identified as Early Hindu temples. Trenches
positioned on the edge of the occupation mound at B, D, E and F were also excavated at various locations
the southern end of the ASI archaeological area. This around the site to gain insights into the occupation
area was chosen because it appeared to be free from history of the site (Chapters 4 and 5).
the later Medieval layers that cover the higher areas
of the protected area to the north. The depression in
which the excavation was positioned appeared to be
enclosed within some form of wall or rampart which
could be seen in section at the edge of the mound.
Some brick structures that were eroding out of the side
of the mound were also exposed.

https://doi.org/10.1515/9783110653540-004
Derek Kennet

Chapter 4 
The Brick Temples in Trench A

This chapter describes the excavation of the two Early temples are no longer in use, it was possible to exca-
Hindu brick temples in Trench A. These two temples vate soundings underneath the floor levels and up
were the focus of two key research objectives of the against the exterior walls, thereby gaining evidence for
second stage of the project. The first was to take advan- changing floor levels and the nature of the foundations
tage of this rare opportunity to gain an understanding that were put down during the temples’ construction.
of the way in which Early Hindu brick temples such as The archaeological techniques of meticulous,
these were laid out and constructed and to see whether brick-by-brick recording and detailed, single-context
they underwent any significant alterations to their excavation and planning used at Paithan provide
plans or superstructures throughout their period of crucial evidence which helps to elucidate the complex
use. The second was to investigate the context of the architectural development of the two temples.
construction and use of the two temples in relation Although there are inevitably still many points of
to the stratigraphic history of the site, including the uncertainty, it will be seen that the work has provided
environmental and economic developments recorded numerous insights into their architectural develop-
in the adjoining excavated sequence. It was hoped that ment. These insights are unique and important; they
the answers to these questions would provide new not only are relevant to the history of the Paithan
insights into the changing nature of the relationship temples but also provide information that will improve
between religion, society and economy through this our understanding of the development of the Early
crucial period in India’s history. Hindu temple more generally. This is because these two
Most of the known temples of this early period are structures, rather than being grand architectural pro-
presented in the literature as complete and finished jects and the focus of royal patronage, appear to have
buildings that were planned and constructed exactly been modest buildings, typical of many thousands of
as they now are (e.g. Meister et al. 1988: 36–37). In such temples that must have been in use right across
fact, it is quite possible – even probable – that in many northern India at this time. Most importantly, it will
cases, these buildings underwent quite complex archi- be seen that the two temples were not static, but that
tectural developments through a number of phases of they were dynamic structures that underwent constant
enlargement, alteration and adaptation. It is impos- and, in some cases, fairly major changes and altera-
sible to know, however, whether this is the case or not tions, changes that must have reflected the changing
as very few Early Hindu temples have been excavated, nature of temple practice as well as the requirements
recorded and analysed anywhere in India to the nec- of the community that used them. The excavations
essary level of detail. This is because relatively few have also served to illustrate the distinctive nature of
temples remain, and many that do are still largely com- the temples’ physical construction and have provided
plete buildings, meaning that it is impossible to gain a unique insight into the sophisticated nature of the
insights into the nature of their construction and into foundations that were laid down underneath the two
how their architectural layout might have changed structures.
through time. By contrast, the two Paithan temples had By necessity, this chapter goes into considerable
already been badly damaged by ancient brick robbing archaeological detail in describing the temples’ con-
and the digging of foundation trenches for later Medi- struction and the results of the excavations in words,
eval buildings. For all the damage they have caused, drawings and photographs. For ease of use, the chapter
these disturbances do provide some crucial ‘windows’ is divided into two sections: the first gives a detailed
into the core of the structures, and it is these which description and overview of the development of each
have allowed the elucidation of the details of the con- of the temples, and the second provides summaries of
struction and the architectural phases through which the excavated trenches that were written up by the indi-
each of the temples passed. In addition, because the vidual excavators at the end of the fieldwork seasons;
24   Derek Kennet

it is envisaged that these reports will be of interest only XVb, XVIa, XVII a–b) (Fig. 4.4). As they were still visible
to those with specific questions about the details of the in 1997, it was decided to clean, record and study what
excavation. In addition, Table 4.4 provides details of appeared to be two potentially very interesting and
each of the 36 principal walls that were recorded. Ref- important monuments. The information published
erence is also made to Appendix I, where all the layers by Yusuf on the two temples was restricted to the two
(‘contexts’) recorded by the excavation project are black-and-white photographs mentioned above, plus
listed and described. Meanwhile, Chapter 14 presents a brief paragraph containing a few details on brick
further discussion of the context and significance of sizes and construction (Yusuf 1938: 4; 1939: 41).
the two temples.

Fig. 4.1: Plan showing the central mound at Paithan and the
location of the main trenches. Note that the trenches are not
shown to scale in this plan.

Fig. 4.2: Satellite image of the ASI protected area showing the loca-
Section 1: Trench A, The Brick tion of the trenches.

Temples Yusuf’s trench originally measured about 30 metres

along each side, with supplementary trenches going off
Trench A was opened in the 1997 season in the remains
to the south-west and the north-east, whereas Trench A
of one of the larger of the four trenches that had been
of the present excavations was restricted to an irregularly
excavated by Syed Yusuf in 1937. It is located towards
shaped area with maximum dimensions of 23 m north-
the northern extent of the area now administered by
south and 15 m east-west centred around the two temples
the ASI, close to the top of the steep slopes that drop
(Fig. 4.5).
down to the Godavari (Figs 4.1–4.3). Plates I and II in
In the 1998 season, the two temples were carefully
Yusuf ’s 1938 report appear to show the early stages of
cleaned (Fig. 4.6). Brick-by-brick plans and elevations
the excavation of this trench.
were made of both of them at a scale of 1:20 using an
The photographs that Yusuf published of the same
archaeological drawing frame. The temples were pho-
trench at the end of excavation suggest that the two brick
tographed and studied in detail and a preliminary
structures that he uncovered at the lowest level of excava-
architectural phasing was worked out. In the 1999
tion were two early temples (Yusuf 1938: pl. V; 1939: pls
 Chapter 4 The Brick Temples in Trench A    25

Fig. 4.3: View of Trench A from the south-west at the end of the 1998 season.

Fig. 4.4: Yusuf’s photograph of the 1937 excavations showing the two brick temples from the south (from Yusuf 1938: pl. v; 1939: pl. XVIIb).
26   Derek Kennet

Fig. 4.5: Plan of Trench A showing the location of test pits and excavation areas.
 Chapter 4 The Brick Temples in Trench A    27

season, efforts were directed towards excavation of 10 buildings; the second was to locate the temples within
test pits in and around the structures together with a stratified occupation sequence; and the third was to
some limited area-excavations adjacent to them. These establish the nature and date of the pre-temple occu-
soundings had numerous aims: the first was to elu- pation in this part of the site. A brief report setting out
cidate details of the foundations and construction of the main findings of the excavations in Trench A has
the temples as well as to clarify any remaining uncer- already been published (Kennet and Rao 2003).
tainties relating to the structural history of the two

Table 4.1: List of test pits and areas excavated in Trench A (see Section 2 of this chapter).

Area Code/Test Pit Description

NWNT Layers related to the robbing of the north wall of the North Temple.

EEST Eastern end of the South Temple east of wall 344.

W Area West of the North Temple.

South or Southern Area An area 9 m × 6 to 8 m in the south of Trench A.

GG North The garbhagriha of the North Temple.

S Wall of N Limited excavation on the robbed-out south wall of the North Temple (later TP2).

S of S Temple The area immediately to the south of the South Temple (later TP6).

TP1 To west of the South Temple to investigate stratigraphic sequence against and below temple.

TP2 To south of North Temple to investigate robbed out southern wall and possible
stratigraphic relationship between temples (originally ‘S Wall of N’).

TP3 Between outer walls of garbhagrihas of both temples to investigate stratigraphic relationship between
temples.

TP4 Between remaining walls of both temples to investigate stratigraphic relationship.

TP5 Across the interior of the mandapa of the South Temple to investigate flooring, wall construction and founda-
tion deposits.

TP6 To south of South Temple to investigate deposition history in this area (originally ‘S of S Temple’).

TP7 Against the southern wall of the South Temple at east end to investigate the development of wall and founda-
tions.

TP8 Inside the east end of the South Temple to investigate the relationship between walls.

TP9 Across the interior of the mandapa of the North Temple to investigate the succession of shrines, flooring, wall
construction and foundation deposits.

TP10 To the western end of the North Temple to investigate stratigraphic accumulation against the temple wall.

Much of the work of the present project involved features that are a part of the temples’ ancient history
recording and removing layers that had been depos- and those resulting more from the recent excavations
ited since the temples went out of use. These can be and restoration. As far as the present authors are aware,
divided into three groups: those resulting from the no detailed documentation of either Yusuf ’s work or
ancient collapse and robbing of the temple structures; the 1967 restoration exists. Yusuf ’s work, particularly,
those resulting from Yusuf ’s 1937 excavations or from removed almost all information relating to the later
the restoration of the temples that was carried out by phases of use and abandonment of the temples. One
the ASI in 1967; and deposits that have accumulated especially significant problem is what at first appeared
naturally over the structures since 1967. to be drainage ditches or wall-chasing trenches that
One of the biggest problems faced by the present had been excavated along the base of the outer walls of
excavation was to distinguish between layers and other the temples, probably by Yusuf. These effectively sever
28   Derek Kennet

any stratigraphic connection between the foundation of the test pits in Trench A are shown in Fig. 4.5, whilst
cuts of the temples and the uppermost layers through the area-excavations and test-pits referred to in the
which they were cut, making it impossible to know pre- text are listed in Table 4.1.
cisely from which level the temples were constructed. Descriptions of the two temples and their phasing
The two temples are referred to in this report as the are presented first, followed by more detailed field
‘North Temple’ and the ‘South Temple’. The locations reports from the test pits and area-excavations in

Fig. 4.6: General view of the temples from the north at the end of the 1998 season.

Section 2 of this chapter. Detailed descriptions of the Full dimensions are given in Fig. 4.9. Although Yusuf
principal walls are presented in numerical order in states that a feature of both of these structures is that
Table 4.4. Within the text, reference is made to the they were placed ‘in strict regard to cardinal points’
context numbers allocated to each of the layers, walls, (Yusuf 1938: 4; 1939: 41), in fact each is on a slightly
cuts and features defined during excavation; a list of different orientation that, in the case of the North
these can be found in Appendix I. Temple, is just under five degrees south of magnetic
west (265° measured with a prismatic compass in Feb-
ruary 1998).

The North Temple The temple is built entirely of fired bricks and
mud mortar. In a few places, traces of lime mortar
are present, but these are almost certainly the result
The North Temple consists of a square sanctum with
of the 1967 restoration mentioned above. This is con-
a tapering mandapa attached to its eastern side (Figs
firmed by Yusuf ’s statement that no mortar was found
4.7 and 4.8). The mandapa is subdivided by an inter-
on either of the buildings in 1937 (Yusuf 1938: 4, 1939:
nal wall which abuts the eastern face of the sanctum
41). The state of preservation is variable; in some
and forms a narrow passageway between itself and the
places, the walls are preserved to a height of 1.6 m,
mandapa wall. The maximum dimensions of the whole
whilst in others, such as the north wall and parts of
structure are 11.30 m east-west and 7.64 m north-south.
 Chapter 4 The Brick Temples in Trench A    29

Fig. 4.7: Plan of the North Temple showing the locations of the main walls.

Fig. 4.8: The North Temple from the north-east.

30   Derek Kennet

Fig. 4.9: Dimensions of the North Temple.

Fig. 4.10: Phases of the North Temple.

 Chapter 4 The Brick Temples in Trench A    31

Fig. 4.11: The west end of the North Temple from the west.

the south wall of the mandapa, they have been com- excavated against the south wall of the platform (see
pletely removed by brick robbers and by pits and con- TP3 in Section 2 of this chapter). These insights provide
struction cuts from overlying occupation levels. enough information to compile the schematic section
Five distinct phases of development can be discerned through the shrine and platform that is shown in Fig.
in the temple: Phases I to V (Fig. 4.10), and these are each 4.12.
described below. The foundations of the shrine are of considerable
interest. Before construction began, a large and prob-
ably square hole was excavated, measuring just over 6
Phase I m wide at the base and about 30 cm less at the top due
to the fact that the sides slope inwards. The original
The earliest phase consists of a small square shrine (wall depth of the hole was probably around 1.5 m, although
554) measuring about 2 m along each side, which is situ- Yusuf ’s excavations have removed any precise indica-
ated in the centre of a square brick platform (walls 555, tion of the level from which the foundations were cut.
969) that was 0.55 m high and 4.2 m along each side The cut was deep enough to cut through all underlying
(Fig. 4.11). The original height of the shrine is unknown layers containing anthropogenic deposits and to reach
as only 20 courses of the wall are preserved and no undisturbed natural soil.
indications remain of the way in which it was roofed. The eastern and western limits of the foundations
In fact not very much of the structure is now visible as were not exposed, so it is impossible to be certain of
it is largely concealed by the thick Phase II wall (556) the precise shape and size of the foundation trench.
that was built on top of it. A later pit (341), however, On the north side, the cut was exposed by Test Pit 9,
has cut into the north-east corner of the Phase II wall where it has been shown to be 1.64 m from the corner of
and has exposed the construction of the Phase I shrine the platform (wall 969). This is more than twice the 80
and platform. In addition, excavation of the fill of the cm distance on the southern side as exposed in Test
garbhagriha and the levels below it has exposed the Pit 3. The reason for this difference is not clear, but it
foundations of the shrine, as has Test Pit 3 that was may be that the cut had a different shape on the north
32   Derek Kennet

Fig. 4.12: Schematic section through the garbhagriha of the North Temple from the north.

side. Another possibility is that the foundation cut It seems as if something heavy must once have stood
broadened out towards the east to incorporate a porch here, although this is more likely to have been during
or mandapa that was attached to the original shrine the construction of the temple than during its use, as
and platform. It will be noted that the section line of this feature it is overlain by a brick-jelly layer (923) that
Test Pit 9 is 1.5 m further east than the section in Test underlies the walls of the garbhagriha. The foundation
Pit 3. No evidence of any such structure has yet come levels were then covered with the layer of brick-jelly
to light, however. Unfortunately, time did not allow (923), 3 cm thick, and it is onto this that the platform
deeper excavation of Test Pit 9, which would certainly and shrine walls were built.
have clarified the shape of the Phase I foundation cut. Construction probably began with the lowest two
The foundation trench was then systematically courses of wall 554 (labelled 970) where the solid brick
filled with layers of subrounded cobbles and small wall is thinner than it is above and the lowest courses
boulders of igneous rock, on average 20 to 30 cm in of wall 555 (labelled 969) (Fig. 4.12). Once these walls
size, in a matrix of loose silt alternating with layers were in place, loose rubble made up of silt, stone
of more compact silt. Some of these layers contained and brick (941) was deposited in between them and
fragments of bricks and pottery and charcoal flecks, flattened. The upper part of wall 554 was then con-
but on the whole, the anthropogenic content was lower structed, against which a two-course brick surface
than would be expected in a redeposited layer from an (932) was put down on top of 941. It is possible that
archaeological site, suggesting that the earth was exca- construction of the platform stopped at this point, at
vated some distance away from any focus of occupa- which time it would have been about 30 cm high, and
tion. that the upper part of wall 555 and the related fill (911)
The surface of the topmost compact silty layer represent a later phase of construction. Alternatively,
(927) preserves a shallow rectangular depression about it is possible that the platform was simply built in two
50 cm × 40 cm and less than 1 cm deep that is situated stages. The top of 911 was destroyed by the cut of the pit
just towards the south-east corner of the garbhagriha.
 Chapter 4 The Brick Temples in Trench A    33

that exposed it, so it is not possible to say how regular outer surfaces of the bricks show no traces of weath-
its upper brick surface was. ering. This simple recess may represent an early stage
The bricks used in this phase measure, on average, in the development of adhisthana mouldings. It seems
7 × 25 × 40 cm and are soft-fired and an orange or red to be most closely paralleled at Ramtek, where, on
colour. the Kevala-Narasimha temple at least, it seems to be
datable to the 5th century (Meister et al. 1988: fig. 30).
This point is further discussed below.

Fig. 4.14: The garbhagriha of the North Temple after excavation to

the lowest foundation fill of boulders. Note the vertical incised lines
in the middle of each wall beneath the level of the blackboard.

Shrine 554 measures 2.11 m east-west and 2.04 m north-

south on the exterior whilst the interior space meas-
ures 1.08 m east-west and 1.10 m north-south. It still
stands to a height of about 20 courses, although it must
originally have stood much higher.
Once the shrine had been constructed, two thin
lines were incised into the brick-jelly floor (923) of the
garbhagriha interior, one crossing the centre of the
garbhagriha between the middle of the north and south
walls and the other, similarly, between the middle of
the east and west walls, the two lines crossing in the
Fig. 4.13: The Phase I structure of the North Temple showing wall
554 and the simple ‘Ramtek-like’ moulding exposed in a post-
centre of the chamber. The incised lines continue ver-
Medieval pit cut. tically up the middle of the walls to the eighth course
(Fig. 4.14). The function of these lines is not clear, but
they may have been related to the layout and definition
Wall 554, the wall of the central shrine or garbhagriha,
of the sacred geometry of the temple.
presents two notable features. Firstly, the wall is two-
The brick-jelly surface and incised lines were then
bricks wide until the eighth course, at which point it
covered with further fills within the garbhagriha con-
thins to a single brick width, leaving a 26-cm-wide ledge
sisting, once again, of alternating layers of stones and
on the interior that was later filled by wall 968, which,
compact and loose silt at least up to the level of the
to judge by the bricks used, is of a much later date. At
eighth course of the wall. The stones used in these levels
the same height, there is a simple two-brick-high recess
are somewhat smaller than those used in the lower foun-
running around the exterior of the shrine, which is
dations. These fills were probably intended as a base for
5 cm deep on the northern face and 2 cm deep on the
the original floor of the shrine, but they may have been
eastern face (Figs 4.13 and 4.15). A decorative feature
deposited at a much later date; it is impossible to be
such as this almost certainly indicates that this wall
certain. There is certainly no evidence of floor surfaces
was intended as a free-standing structure, although the
34   Derek Kennet

Fig. 4.15: Isometric reconstruction of Phases I, II and III of the North Temple from the south-west.

below this point – the brick-jelly surface 923 being

too friable to have served as a floor. Above the eighth
course of the wall, further stone/silt fills continue to
about the 12th or 13th course, but these abut wall 968
and are therefore of a much later date.

Fig. 4.16: Elevation of walls 555 and 556.

Having described the Phase I shrine, two key ques-

tions emerge. The first is from which direction was the
original shrine entered? It is impossible to answer
this question with any certainty because so little of
the upper part of wall 554 is visible. Clearly, there was
no opening in any direction below the eighth course of
the wall as all four walls are clearly intact and inter-
bonded up to this level. Above this, it seems that all
four faces of the wall stand to at least about the 16th
course, but it is impossible to be certain because, as
Fig. 4.17: The robbing on the north side of the North Temple from
has been mentioned, the inner face of the wall is the east.
concealed by later wall 968 above this point and the
outer faces are concealed by the later Phase II wall. garbhagriha that was subsequently sealed. Above this
These later walls may conceal an original opening in the level, the garbhagriha may have been open to the west,
 Chapter 4 The Brick Temples in Trench A    35

Fig. 4.18: Details of plinth mouldings from various walls from both the North and South Temples.

where the wall is now four of five courses lower than it Phase II
is on the other three sides. It may also have been open
to the east from a few courses higher, but the western Phase II represents the construction of a structure
wall of the Phase III mandapa stands to at least this (556) directly upon the Phase I platform, encasing the
height, demonstrating that there was clearly no com- free-standing garbhagriha. The resulting construction
munication between the mandapa and the garbhagriha is square, about 4.2 m along each side, with a triratha
at the time the mandapa was constructed. It is possi- pattern of 10-cm projections along the central two
ble, although rather unlikely, that this Phase III wall metres of each face (Figs 4.7, 4.9, 4.11, 4.15). The wall
may conceal an earlier opening to the east that was is one-brick thick and stands on the outer edge of the
closed during Phase III or before (see discussion below Phase I platform. The space between this wall and the
under Phase III). So, although it therefore seems likely wall of the Phase I garbhagriha is filled with a thick
that the entrance to the garbhagriha was from the west, hearting of roughly coursed brick bats and silt, result-
it is not possible to be certain of this. It is, however, ing in a wall with a total thickness of about 1.5 m. Up
possible to be certain that the floor level of the shrine to 13 courses of the wall are preserved on the south
would have been between 60 cm and 1 m higher than face, the lowest course being a levelling course resting
the ground surface surrounding the temple. directly upon the Phase I platform, some of the bricks
The second question regards whether or not there of which have been thinned in order to establish a
were two phases of construction in the Phase I plat- level base. There is evidence of a later rebuilding in
form (walls 969 and 555) or do these walls and their the upper courses of the wall, which is discussed below
associated fills simply represent two stages in the con- (see ‘Later use and robbing’).
struction of a single-phase platform. Unfortunately, it The jointing of wall 556 is a little finer than that of the
is impossible to answer this question without disman- Phase I walls 554 and 555, the courses of fine silt measur-
tling the whole shrine and it must therefore remain ing around 0.5 cm. The eastern face is well preserved,
unanswered. showing a pattern of predominantly stretchers (Fig.
36   Derek Kennet

4.16). There is a small offset of around 6 cm between actually quite separate and that Phase I was a free-
the top of wall 555 and the base of 556. There also standing shrine for a period of time before Phase II was
appears to be a slight difference in orientation of about constructed.
0º 43’ between the two structures.
The bricks used in this phase are very similar in
feel and firing to those used in Phase I, measuring, on Phase III
average, 7 × 25 × 40 cm.
Phase II effectively represents the encasing of the Phase III represents the addition of a substantial
Phase I garbhagriha and platform within a more sub- mandapa to the east of the Phase II shrine. The
stantial and elaborate mulaprasada, which contin- mandapa consists of a brick wall (557), 1.22 m thick,
ued to be free-standing and unattached to a mandapa in the shape of a tapered rectangle enclosing an inter-
so far as it is possible to tell. There is no indication nal area that measures 5.20 m north-south at its widest
of the nature of the roof, although the massive walls point and 6.06 m east-west.
were possibly intended to support a fairly substantial Wall 557 is preserved only in a few places. On the
superstructure such as a brick shikhara. Indeed, the north side, it was removed by robbing, leaving only two
completed temple may not have been dissimilar at stumps standing (502 and 514) (Fig. 4.17). The exact loca-
this time to the well-known brick temple at Bhitargaon tion of the original wall is, however, discernible due to
(Meister et al. 1988: 36–37), although, of course, on a the preservation of the lowest course of mud mortar in
much more modest scale. the bottom of the robber cut. At the east end, the wall
has not been robbed but it has been damaged and partly
concealed by later rebuilding and restoration. Along the
south side, most of the south face of the wall has also
been removed by robbing so that only two stretches of the
original outer face are preserved. At the west end of the
south side, where the wall abuts the Phase II structure,
the wall is better preserved and stands a maximum of 11
courses high.
In this same south-west corner of the mandapa,
a 24-cm-long section of adhisthana mouldings is pre-
served. The moulding is very similar to the better-pre-
served mouldings of Phase II/IV of the South Temple
Fig. 4.19: East-west section through wall 514 showing extra face
559/1213.
(Fig. 4.18A). A short section of the same moulding is
also preserved on the small surviving section of the
wall on the north side (514; Figs 4.18B and 4.19), sug-
The Phase II structure presents two questions, both
gesting that these are the remains of the original
quite similar to those posed by the Phase I shrine. The
mandapa mouldings.
first is the question of the entrance to the shrine. As has
As can be seen in Fig. 4.19, a crude later facing (559)
been stated above, there is absolutely no sign within
was added to wall 514 which included similar mould-
the fine brickwork of the eastern face of wall 556 that
ings at the same height, although the base of the facing
there could once have been an entrance in that wall.
is much higher than the base of the original wall. This
It must therefore be assumed that the entrance to the
represents a later thickening of the wall to the west,
Phase II shrine was from the west (e.g. towards the
the reason for which is not known. This part of the wall
river) in this phase.
has had its stratigraphic relationships removed by
The second question is whether the Phase I struc-
robbing and later excavation, making it impossible to
ture was ever actually used as a shrine or temple, or
understand the circumstances or date of this devel-
whether it simply represents a stage in the construction
opment. There is no evidence for a similar thickening
of the Phase II structure. It is difficult to be certain,
anywhere else on the wall.
but the absence of any inter-bonding between walls
of the two phases, the slight difference in alignment
and the difference in the quality of the brickwork and
jointing all strongly suggest that the two phases are
 Chapter 4 The Brick Temples in Trench A    37

Fig. 4.20: Sketch of the joint between walls 556 and 557 at the southwest corner of the North Temple from the south-west. Arrows indicate
where a brick abuts another.
1. This irregular ‘T’ or ‘L’ shaped brick with a curved corner protrudes from the west face of 557 and abuts the triratha offset of wall 556,
the curve of the brick protruding out above the right-angled corner made by the lower three courses of 557 that abut 556 and sit on
top of it. Is this simply poor finishing suggesting that these courses were originally below ground?
2. Although broken, this brick, in the middle of the curved moulding, is ‘L’ shaped, suggesting that it turned the corner from the south
face to the west face. It is too broken to be certain.
3. Here, three bricks seem to make up a short section of the western face of the wall, which the moulding courses now abut (e.g. (7)
and above). There is no suggestion of the mouldings turning to the west face of 557 here, but it is not clear how this would have been
organized. The west face (3) is 18 cm to the east of the lower face of wall 557.
4. As in Fig. 4.21, the lower levelling course of wall 556 has been cut into to provide a base for wall 557.
5. Here, as can be seen in Fig. 4.21, the upper courses of wall 557 seem to have been inter-bonded/threaded into the face of wall 556.
It is very difficult to understand which was the original face of wall 556. The question arises, was there an earlier mandapa-type
structure built onto or with wall 556 underneath wall 557? As the northern relationship is destroyed, it is only here that the evidence
remains but it is unclear.
6. The upper courses of purple bricks are clearly part of a later rebuilding (wall 971). The weathering suggests that these bricks have
stood exposed for some time. The upper courses of wall 557 appear to abut this, but it seems difficult to imagine that this could have
been the case.
7. This is a moulded brick with a ‘hammer head’ chamfer as on the plinth mouldings of the South Temple. It has been cut into to the
west.
8. These bricks, which appear to be the eastern end of the triratha offset in wall 556, actually abut the face of wall 556. This is not the
case on the west side of the triratha offset, where they are inter-bonded. This suggests a rebuilding or something more complicated
which is hidden.

The overall impression is that wall 557 has been inter-bonded into wall 556, which required some rebuilding of wall 556 itself. It seems
that wall 557 originally had mouldings on its outer face and there may be a more complicated sequence of phasing here that could only
be unravelled by dismantling the structure.
38   Derek Kennet

Fig. 4.21: The joint between the Phase II and Phase III brickwork of the North Temple, looking west in the southern half of the
mandapa, with wall 558 to the immediate right.

Fig. 4.22: Test Pit 9 sections. (A) Composite section across the whole test pit and the mandapa of the North Temple looking west. (B) East-
west section against wall 556 below the southern part of wall 558 where it crosses the test pit. (C) The western section of the test pit to the
north of the Phase I structure showing the edge of foundation cut 1171.
 Chapter 4 The Brick Temples in Trench A    39

In the south-west corner of the mandapa, the join

between the Phase II (556) and Phase III (557) walls is
partly preserved. This is a complicated relationship,
made more difficult to understand by later rebuild-
ings of the upper courses of both walls, which have
obscured or destroyed the original brickwork. The
three-dimensional sketch in Fig. 4.20 shows the visible
details as they are. It seems that wall 557 was keyed
into wall 556 and this process involved the removal
and replacement of bricks from some parts of the face
of 556. Much later, the upper seven or eight courses
of the eastern face of wall 557 appear to have been
rebuilt, leaving only a thin face of this wall sandwiched
between the rebuilding and a much later rebuilding of Fig. 4.23: Feature 1137, the earliest evidence for a possible shrine
the upper 10 courses of wall 556 (971). There are some in the west end of the mandapa of the North Temple.
problems with this interpretation, however, such as
the brick labelled ‘1’ in Fig. 4.20, which must be a part
of wall 557 but which projects beyond its western face.
Fig. 4.21 shows the same joint between the two
phases, this time looking towards the west from the
interior of the southern aisle of the mandapa. Here
it can be seen that the lower courses of the mandapa
clearly abut and overlie the Phase II wall and platform,
with some brick cutting and inter-keying that must
have been achieved by the removal and replacement
of bricks in wall 556. In Fig. 4.21, a slight decline in
the quality of the jointing of the Phase II brickwork is
evident above about the 10th or 11th course, which may
be indicative of a later rebuilding above at this point. Fig. 4.24: North Temple, showing the remains of a possible simple
Although a relationship between the two phases shrine 983 against the western wall of the Phase III mandapa.
has been proposed above, without dismantling the
walls it is impossible to be certain of having a full The addition of a mandapa to the east of the Phase II
understanding of this rather complex situation. shrine is problematic because it would appear to indi-
From the plan, it can be seen that the westerly face cate that the entrance to the Phase II shrine was from
of 557 has a slightly different orientation from that of the same direction. As has been noted above, however,
the original Phase I structure 554 (Fig. 4.7). This must there is absolutely no evidence that the Phase II shrine
be deliberate, but the reason for it is unknown. was ever open to the east; indeed, it is more likely
Once again, the bricks used in this phase are very that the entrance was from the west, towards the river.
similar in feel and firing to those used in Phases I and II, In this respect, the traces of possible cultic activity
measuring, on average, 7 × 25 × 40 cm. uncovered by Test Pit 9 from the earliest levels inside
The latest internal surface of the mandapa that was the west end of the mandapa are very important (Fig.
exposed by excavation was designated layer 550 along the 4.22) (see Section 2 of this chapter). Although no sculp-
north side of the temple. It appears to be the equivalent ture has come to light, a stratigraphic succession of at
of surfaces 1010, 1100 and 1080 in Test Pit 9 and layer least three rectangular impressions or arrangements of
319 in Test Pit 2. It consists of a compact grey silt with bricks against the eastern wall of the Phase II struc-
possible lime inclusions at a height of 449.42–449.47 in ture within the mandapa indicates that this place was
relation to the site datum. There may well have been most likely used for cultic activity, albeit consisting of
later surfaces that were removed by Yusuf, and there simple, crude constructions. The earliest shrine (1137),
certainly seem to have been earlier surfaces. a simple rectangular depression, is located directly on
top of the mandapa foundation deposits (Fig. 4.23).
Evidence for a second shrine was found immediately
40   Derek Kennet

above but higher up in the sequence (983). This indi- Phase IV

cates that the location was already being used for
simple cultic activity immediately after construction Phase IV represents the subdivision of the interior of
of the mandapa and that it continued to be used for the mandapa by the construction of wall 558, creating
some time (Fig. 4.24). The third shrine (551) is again in what was either a trapezoidal inner chamber or a raised
the same location but is a part of Phase IV and will be platform around the shrine and a narrow ambulatory
discussed in the section below. The question posed by between itself and the mandapa wall (Fig. 4.7). Wall
these shrines is, why should they have been located 558 encloses a space measuring 2 m north-south at the
against the back of the Phase II structure? There are widest point and 4.8 m east-west. The entrance was
two obvious possibilities. One is the cultic activity probably to the east, although no trace of it survives.
originally related to a subsidiary deity that was located
in a niche in the upper part of the outer walls of the
Phase II structure. If this were the case, it is possible
that this location at the back of the Phase II structure
had already become important for worship before
the construction of the mandapa. The mandapa may
then have been constructed to formalise the cult and to
enclose it within the temple. Unfortunately, it is impos-
sible to know whether this is the case because the foun-
dation cut for the mandapa would have destroyed any
evidence for earlier cultic activity in this location. A
second possibility is that, for some unknown reason,
construction of the mandapa was deliberately intended
to relocate worship to the eastern side of the Phase II
Fig. 4.25: Photo of the rectangular depression in layer 551 that
structure, although it is notable that no attempt was might indicate the location of a shrine.
made to open a connection between the garbhagriha
and the mandapa during the Phase III construction. In
the absence of any further evidence, this is the most
that can be said about the rationale behind the Phase
III mandapa.
The foundation cut for the mandapa was also
revealed by Test Pit 9. It is much shallower than the
foundation cut of the Phase I structure, although the
precise level from which it was dug is not known as the
relevant levels were destroyed by Yusuf ’s excavations.
The foundations consist of only one layer of igneous
rocks and silt, rather than the multiple sequence of
silt/rocks/silt found in the Phase I foundations. The
reason for this difference is not clear. It may be that
Fig. 4.26: Wall 383/544 of Phase IV in the North Temple before the
the importance of such foundations had diminished by excavation of Test Pit 9.
the time the mandapa was constructed. Alternatively,
because the mandapa did not include a garbhagriha,
The wall is preserved up to seven courses high,
it may have been felt that such elaborate foundations
although the entire south-east corner above the lowest
were unnecessary.
two courses is part of the 1967 reconstruction. The wall
Fuller details of the shrines and the sequence
is about 52 cm wide and is of a different type of con-
excavated within the mandapa, as well as the foun-
struction to any of the earlier walls in the temple. It
dations, are given in the Test Pit 9 report in Section 2 of
is a solid brick wall with no rubble fill, consisting of
this chapter.
a double row of bricks. The silt jointing is quite crude
– certainly much cruder than the Phase II wall. Most
significantly, the bricks of this wall are of a larger size
 Chapter 4 The Brick Temples in Trench A    41

than any used in the previous phases measuring 6.5 tion. It is also possible, however, that layers 639 and 551
× 26 × 42/43 cm. The introduction of a new construc- were deposited around the base of a shrine or idol that
tion technique and a larger brick size is significant for stood on bricks 983, and that the depression in 551 was
linking together the chronology and phasing of the caused by the removal of this object at a later date. To
North and South Temples. judge by the location of the stones in layer 639, the latter
The construction of wall 558 follows the deliber- scenario appears to be more likely, but in effect, there is
ate raising of the floor within the mandapa by about not much difference between the two interpretations; in
50 cm to the level of surface 1010, which was achieved both cases, the shrine continued to be used and the level
through the deposition of a further sequence of silt and of the interior of the mandapa continued to be raised.
stone foundation levels (1050, 1115). It is not certain,
however, that the raising of the floor level and the con-
struction of wall 558 occurred at the same time. The Phase V
detailed stratigraphic evidence discussed under Phase
III above, and also in the report on Test Pit 9 in Section Elucidation of the developments of Phase V is
2 of this chapter, suggests that shrine 983 may have extremely difficult, due partly to the nature of the
been in use on floor 1010 for a period of time before relevant black cotton soil deposits (1252 and 395) and
wall 558 was built. This suggests that wall 558 was also to later disturbance. At some point, a crude wall
built with the intention of enclosing the already exist- (383/544) appears to have been constructed of brick
ing shrine 983 or of raising a low platform on which the bats and broken tiles in order to surround the earlier
shrine could then be placed. shrine within a small rectangle, apparently ignoring
Shrine 983 was subsequently buried by the delib- the existence of the Phase IV wall 558, which was prob-
erate deposition of a further level of silt and stone ably therefore already destroyed. The area enclosed
foundation deposits (639) upon which a 20-cm-thick measures 3.40 m north-south and 1.70 m east-west.
layer of black cotton soil was laid down (551). On the The best record of this wall is the photograph taken
upper surface of 551, a large rectangular sunken area by Yusuf in 1937 shortly after the structure was first
was noted against the western end of the mandapa revealed (Fig. 4.4). In this photograph, it is clear that
(Figs 4.22 and 4.25). The sunken area measures 140 wall 383/544 still stood higher than the top of wall 556.
cm north-south by 90 cm east-west and is 15–20 cm in Since then, the wall has survived very badly, possibly
depth. It is thought that this represents the location of because it was not well made in the first place, and
a large rectangular object or structure that must have all that remained to be recorded in 1998 was a row of
been part of a third successive shrine in the same loca-

Fig. 4.27: Schematic section/view of the garbhagriha wall of the North Temple looking north showing the extent of later rebuild 971.
42   Derek Kennet

Fig. 4.28: The garbhagriha of the North Temple from the west showing possible evidence of late use.

Fig. 4.29: General view of the South Temple from the southeast during excavation.
 Chapter 4 The Brick Temples in Trench A    43

bricks and tiles standing on end in lime mortar within a 1248, which is a rebuilding of the upper three courses
shallow cut and a row of brick bats and tiles lying on the of the south side of the Phase III mandapa wall 557, and
soil retained behind them (Figs 4.22 and 4.26). The pres- wall 559, which is a later facing to mandapa wall 514.
ence of lime mortar demonstrates that these were placed The most significant rebuilding is wall 971, which is a
here during the 1967 restoration and probably represent rebuild of the upper parts of the Phase II structure 556
a rather desperate attempt to preserve an already very (Figs 4.20 and 4.27). From around the seventh course
badly damaged structure. They serve only to indicate the of wall 556, the bricks are notably thicker, measuring
approximate location of the wall, and perhaps the bricks 7/7.5 × 20 × 33/35 cm. They tend to have a purple colour-
and tiles from which it was built. ation and are fired to a much higher temperature than
Whether this wall formed a small earth-filled the bricks in earlier parts of the temple. They also often
platform, as seems most likely, or a small chamber is show traces of vitrification on the surface. These bricks
unknown. The platform would have effectively raised are similar to the bricks used in the rebuild of wall
the level of the shrine by creating a raised surface 561 of the South Temple and seem to represent a later
that has now been completely destroyed. As has been phase of reconstruction and repair of both temples, the
stated, the fact that wall 383/544 ran over the top of the date of which is not known but which probably took
Phase IV wall 558 demonstrates that this earlier wall place between the 9th and the 11th or 12th centuries,
had already been destroyed and its upper courses had although it could have been later. Robbing and erosion
been removed by this time, but the condition of the of the structure make it impossible to say very much
rest of the temple during Phase V is unknown. It is about the construction of which these bricks form part.
probably no coincidence that wall 558 is much better They are found on the north, south and west sides of
preserved underneath the Phase V structure than it is the Phase II structure but not on the east face. It is pos-
to the east, as can be clearly seen in Fig. 4.26. This may sible that the whole of the superstructure of the garb-
indicate that the western part of the wall was incorpo- hagriha needed replacing at some point in time. The
rated into the Phase V structure and that the rest of the bricks of the rebuilding are now badly weathered, sug-
wall was removed. gesting that they have been exposed to the elements
It is impossible to know with certainty the nature for much longer than the lower parts of the structure.
of the temple during this latest phase of its use. The Excavation along the course of the north wall of
very fragile remains of Phase V have been exposed to the mandapa has revealed that this wall was system-
60 years of weathering and erosion as well as undoc- atically robbed out after the temple fell into disuse.
umented excavation and restoration. All that can be The date of the robbing is unknown, as is the level
said with confidence is that the temple continued in from which it occurred, as all related levels were
use as a place of cultic activity through the construc- removed by Yusuf ’s excavations. The precise course
tion of what was probably a crude earth-filled platform of the original wall is shown by the lowest level of
located against the eastern wall of the garbhagriha mud mortar, which is still visible in the base of the
structure in a place that had already been in use as a robber cut (Fig. 4.17). The details of the robbing are
shrine, possibly for as long as several hundred years, discussed in Section 2 of this chapter (NWNT).
but that had certainly once seen distinctly better days. A series of rough brick-bat steps over the top of the
Wall 558 no longer existed, and it is impossible to be remains of wall 556 indicate that there has been access
certain that rest of the mandapa walls were still stand- to the ruined garbhagriha from the west or river side
ing – they may also have been demolished and robbed for some time (Figs 4.11 and 4.28). This may of course
by this time. The crude Phase V structure seems most relate only to the period since the 1937 excavations, but
likely to represent rather sporadic and unstructured the height of the walls surrounding the garbhagriha
use of an already old and established shrine that was indicates that this is the only direction from which it
now located within the half-standing remains of an could have been approached, at least during its final
old, ruined temple. form. It therefore seems most likely that the shrine
was also approached from this direction during its
early stages of use, but it would be impossible to verify
Later use and robbing this without dismantling some of the later walls.
The last phase of activity identified in this temple
There is evidence of later rebuilding of the upper is associated with the restoration of the buildings in
courses of some of the temple walls, for example wall 1967. At this time, a shallow scoop was dug against the
44   Derek Kennet

outside face of the eastern length of wall 557. It seems underwent a very different set of developments to its
that along the line of the robbed wall, a shallow ditch northern neighbour and each of the South Temple
was dug (337). The presence of lime-mortared bricks phases is certainly not contemporary with the equiva-
within this cut makes it clear that it results from the lent of the North Temple. The problem of the removal
1967 restoration. The surviving stumps of wall that by Yusuf of archaeological deposits relating to the
were discovered along the length of the mandapa wall later periods of use and abandonment applies equally
were also conserved with mortar pointing. After resto- to this temple. The five phases are described in turn
ration, it appears that the related trenches were back- below.
filled to some degree, after which silt had accumulated
across the area to a depth of about 30 cm before the
1998 excavations began. Phase I

As with the North Temple, the foundations of the struc-

The South Temple ture are of considerable interest. They consist of a large
– presumably rectangular – pit that is just under 8 m
wide and at least 9 m long, although the precise length
The South Temple consists of a rectangular garbhagriha
is unknown. The sides of this pit slope steeply down
with a rectangular mandapa attached to its eastern side
to a flat bottom that is 6.9 m wide and 20–25 cm below
(Figs 4.29 and 4.30). The mandapa was once subdi-
the level at which layers of archaeological deposits
vided by a north-south wall towards its eastern end,
of the surrounding area containing anthropogenic
and the fragmentary remains of further internal
deposits sit upon underlying natural soil (Fig. 4.33).
walls are also preserved. The maximum dimensions
The pit was then carefully back-filled with a repeat-
of the whole structure are 9.4 m east-west and 6.25 m
ing sequence of layers of small boulders of igneous
north-south (Fig. 4.31). As with the North Temple, this
rock packed into a very clean silt matrix followed by
temple is not oriented precisely east-west. The main
a thinner compact clayey layer. This sequence was
axis is oriented just under 4° south of magnetic west
repeated five times before a thin, even mud surface
(266° measured with a prismatic compass in February
was spread across the entire foundation trench and it
1998), which is about 1° different from the orientation
is directly onto this surface that the temple walls were
of the North Temple. The reason for this difference is
constructed (Figs 4.33–4.36). Once the walls had been
not clear. The South Temple is located immediately to
completed, a further sequence of stone and silt layers
the south of the North Temple, being separated by a
was deposited inside them to raise up the interior level
minimum distance of less than 1 m from the southern
to what is thought to have been the floor of the Phase
wall of the North Temple.
I temple. Outside the walls, the upper 50 cm of the
As with the North Temple, this temple is built almost
foundation cut flares outwards and the resulting space
entirely of fired bricks and mud mortar, although stone
between this and the outer face of the temple wall was
has been used in some later parts of the structure. The
filled with layers of silt or compact rubble.
state of preservation is also variable; in some places, the
Excavation below the base of the standing walls of
walls are preserved to a height of 1.35 m above the present
the temple in Test Pit 5 during the 1999 season revealed
ground surface, whilst in others, such as the eastern
the presence of an early phase (Phase I) that was not oth-
walls, they are preserved only in plan. In contrast to the
erwise visible (Fig. 4.32). All that remains of the Phase
North Temple, the outer walls of the mandapa are on the
I structure is a five or six course solid brick wall (1224)
whole better preserved and have been less damaged by
which sits directly on top of the mud surface on top of
robbing, but the internal walls that once subdivided the
the foundation fills described above. During Phase I,
interior of the mandapa are very badly preserved, having
the temple consisted of a large rectangular garbhagriha
been almost completely removed by the foundations of
(2.7 m wide and 1.42 m deep) that was attached to an
much later post-Medieval houses that once stood on this
almost square mandapa (4.49 m wide and 4.08 m deep)
spot.
which was open along most of its eastern face. The
Five distinct phases of development can be dis-
garbhagriha and the mandapa were connected by
cerned in the temple’s history: Phases I–V (Fig. 4.32).
an opening 1.47 m wide. This opening, which is now
This is the same number of phases as the North Temple
completely buried, can be seen in Figs 4.37 and 4.38
but this is purely coincidental because the temple
underneath the Phase II walls. In the middle of it, a
 Chapter 4 The Brick Temples in Trench A    45

Fig. 4.30: Plan of the South Temple showing the location of the main walls.

Fig. 4.31: Dimension plan of the South Temple.

46   Derek Kennet

Fig. 4.32: Phases of the South Temple.

large, heavily vitrified brick is visible which may have no evidence of the superstructure remains. This is prob-
been a step or part of a threshold between the two ably because the temple was completely rebuilt after this
chambers. The layer underneath was probably the time, but it is not absolutely certain that a temple of this
floor level related to this phase of the temple. phase ever actually existed. There is a possibility that
The brick masonry of the Phase I wall is similar to wall 1224 is simply the foundation course for the Phase
that of the later walls of the temple, and for the most II temple. This seems very unlikely, however, given the
part, the wall thickness is identical. In the area around differences in plan between the structures of Phase I and
the garbhagriha, the Phase I wall is, however, as much those of Phase II, especially around the garbhagriha,
as 35 cm thinner than the later walls, which means unless the layout of the temple was changed halfway
that the interior space of the garbhagriha would have through construction. It seems most likely that the super-
been up to 70 cm wider in Phase I than it was in Phase structure of the Phase I temple was completely disman-
II. A section through these walls is shown in Figs 4.36 tled before the Phase II temple was constructed.
and 4.39.
It is important to note that Phase I is represented
only by the foundation courses of the wall – absolutely
 Chapter 4 The Brick Temples in Trench A    47

Fig. 4.33: West-facing composite section across the South Temple showing foundations.

Fig. 4.34: View of the eastern section of Test Pit 5 across the interior Fig. 4.35: Interior of the South Temple during excavation of Test
of the South Temple looking south-east showing the successive Pit 5 showing a level of foundation fill consisting of small boulders.
foundation fills.
48   Derek Kennet

Fig. 4.36: Composite east-west section through the west end of the South Temple looking north.

the Phase I structure, but with some relatively minor

changes to the layout.
Upon excavation, it became clear that the first
four brick courses above foundation wall 1240 are built
of a different masonry to the higher superstructure of
the temple (Fig. 4.39). The lowest of these is a levelling
course, using split or wedge-shaped bricks to create
a level platform out of the Phase I wall. After the con-
struction of the Phase II walls, the interior was raised
using another sequence of silt/stone foundation layers
similar to those described above, which was then
capped by a floor, some 20 cm higher than the Phase
I floor.
Fig. 4.37: The wall between the mandapa and the garbhagriha in As has been mentioned above, the Phase II walls
the South Temple looking west during the excavation of Test Pit 5. around the garbhagriha are as much as 35 cm thicker
The Phase I gap in this wall can be clearly seen below the Phase II than the Phase I walls (Fig. 4.36). This caused the space
wall.
within the garbhagriha to be considerably reduced
during Phase II. In Phase I, there had also been a wide
opening between the garbhagriha and the mandapa,
Phase II but this opening may have been narrowed during
Phase II by the construction of wall 1251, which prob-
Phase II of the South Temple consists of the reconstruc- ably supported the walls on either side of a doorway,
tion of the same temple on the foundation walls of but there is no trace of the doorway so it is impossible
to know how wide the opening was during this phase.
 Chapter 4 The Brick Temples in Trench A    49

Fig. 4.38: Western section of Test Pit 5 across the South Temple in front of the wall between the garbhagriha and the mandapa.

Walls 1240/1247 and 1251 are crudely inter-bonded (Fig.

4.40), suggesting that the Phase II structure was built
in stages, but it is difficult to establish exactly what
the sequence was and why it should have been built
in this way.

Fig. 4.40: View of the corner formed by walls 1240/1251 and 1224
in the southwest corner of Test Pit 5 looking southwest. The arrows
show the boundaries between the upper and lower walls. The inter-
Fig. 4.39: Cross-section of walls 1247 and 1224 South Temple. bonding between walls 1240 and 1251 is clearly visible just to the
right of the corner and the difference in the quality of the brickwork
That the Phase II temple was adorned with adhisthana of the two walls is also clear.

mouldings is made clear by the fact that three courses

of them are preserved on the south-east corner of the east corner, three remaining courses of the mouldings
temple and for short sections on either side of the can be seen (Fig. 4.42). Here they are abutted and over-
opening in the eastern face (Fig. 4.41). In the south- lain by the mouldings of the Phase IV rebuilding of
50   Derek Kennet

the upper parts of the wall on this side of the temple. other arrangement for closing off the mandapa to
The Phase II mouldings ran around the corner onto the the outside. This may explain its cruder construction
short walls on either side of the eastern face for about style; alternatively, it may simply reflect the fact that it
111 cm (Fig. 4.43). There was then a 41-cm section of wall was built to less exacting standards.
with no mouldings on either side of the entrance to the It seems most likely that the wall was built before
mandapa, which may have been left free for column the eastern extension to the temple (Phase IV), and
bases to be placed against the wall. The mouldings this is certainly what is indicated by the wall’s strati-
appear to have had a very similar pattern to those of graphic relationship with compact silt surface 346 to
Phase IV, which are described in detail below. its east (Section 2 below: EEST). It therefore seems sen-
sible to include it as a separate phase in its own right,
as it does reflect a significant re-design of the temple’s
Phase III layout.
During this phase, three post-holes were cut into
Phase III represents a relatively minor alteration to the the same silt surface 346 that abuts wall 344 outside
Phase II temple. It involves the construction of a wall the eastern end of the temple. These are numbered 358,
(344) across the eastern opening of the mandapa. This 523 and 525. The two that have been excavated have
wall is of a much cruder construction than any that had revealed that they were the bases of posts measuring 20
been built up to this time in either of the two temples. It to 25 cm across. It seems most likely that these were
consists of two faces of brick bats rather than complete wooden posts, although there is no proof of this. Two
bricks, the space between them being filled with brick of the posts were situated against the eastern face of
rubble and silt. The wall now survives to a height of four the temple symmetrically on either side of the eastern
courses. opening to the mandapa and seem likely, therefore, to
have been part of a wooden entrance or façade struc-
ture immediately in front of the temple. The third post
is situated 70 cm or so in front of the temple to the
south and it is not matched by another in a similar
location to the north.

Fig. 4.41: Remains of mouldings in wall 1240 at the eastern end of

the South Temple looking west.

The reason for the construction of this wall is not clear

and, as will be noted below in the report on the eastern
end of the South Temple (Section 2 below: EEST), Fig. 4.42: View showing the eastern end of the south wall of
neither is it absolutely clear to which phase it belongs. the South Temple looking north (see Fig. 4.46 for a key to wall
It may have been built to close off the wide eastern numbers).

entrance to the mandapa completely, or as the foun-

dation for a wall which included a narrower entrance- By Phase III, the surface level of the building had
way. No trace of any such entrance remains in the wall, clearly risen considerably since the time of the origi-
however, so it is impossible to be certain of this or to nal construction, partly by natural accumulation and
have any idea how wide any such entrance may have partly by deliberate raising. Surface 346, which was an
been. The wall may not ever have been built beyond external surface to the temple until it was enclosed by
a few courses high and may have been intended only the Phase IV walls, is at a much higher level than the
as a base to support a wooden screen or some such
 Chapter 4 The Brick Temples in Trench A    51

surfaces outside the building that are assumed to be impossible to be certain, it seems most likely that this
the original floor surfaces. reconstruction happened at the same time as the Phase
IV extension, although it could in fact have happened
at any time thereafter. The most likely scenario seems
Phase IV to be that, at this time, the whole temple, or large parts
of it, was dismantled down to about the height of the
Of all the phases, Phase IV represents the most signifi- chamfered course of the mouldings. The three new
cant transformation of the South Temple’s original plan. extension walls were then built to the same height as
During this time, the temple was extended to the east by the standing bases of the Phase II walls, and then the
about 1.75 m, which involved the construction of exten- whole of the temple superstructure was rebuilt on top
sions to the northern and southern walls of the mandapa, of this base (Fig. 4.44).
as well as a new eastern wall. It also appears to have
involved reconstruction of substantial parts of the upper
walls of much of the temple.

Fig. 4.44: Elevation of the western wall of the South Temple

above the eastern section of Test Pit 1 showing the fills of the
temple foundations.

Fig. 4.43: South-facing section of Test Pit 8 showing the sequence

of floors that abut the eastern face of the Phase I – III walls of the Phase IV involved no discernible changes to the plan
South Temple. or size of the garbhagriha, but the shape and size of
the mandapa were quite different. It now became a
The three new walls to the east (345, 359 and 352) are all rectangle 4.49 m north-south by 5.87 m east-west.
of roughly similar construction (Fig. 4.30). Although There is no trace or indication of the location or size of
little is preserved of walls 345 and 352, it can be seen the entrance to the temple in the remains of wall 352,
that the lower five courses are faced with crudely but the absence of any entrance in the north and south
jointed brick bats above which the upper courses are of walls demonstrates that the entrance continued to be
a much better construction with finer joins. Only wall from the east, as may have been expected. The exten-
359 on the south side is preserved to any height; here, sion increased the internal space of the mandapa from
the lower courses are again crude and badly jointed, 18.3 m2 to 26.3 m2. Why this was done is unknown, but it
but those of the upper courses are finer and continue may be that much of the temple superstructure was in
the Phase II mouldings. As can be seen in Fig. 4.42, wall need of repair at this time and the extension was only a
359 abuts wall 1247 of Phase II. However, the abutment part of a bigger project of repair and rebuilding.
does not continue above the chamfered course of the By Phase IV, the floor levels of the temple had
mouldings, which indicates that the whole of the wall once again risen, partly through natural accumulation
was reconstructed from above this level after the con- and partly deliberately. This makes it likely that the
struction of wall 359. This upper wall is labelled 560 in remains of the Phase III wall 344 would already have
order to distinguish it stratigraphically. Although it is been buried during Phase IV and would no longer
52   Derek Kennet

have been visible within the temple. A small section western and southern walls of the temple are most
of brick-bat flooring (353) is preserved to the eastern likely to date to the time of the Phase-IV extension and
side of wall 352, and this may indicate the floor level rebuilding (Fig. 4.18C and D). This is confirmed by the
of Phase IV, although it may also be somewhat later. fact that the character of the brick masonry changes
Wall 352 was much later robbed out along its northern notably above the first inter-bonded course between
half, revealing details of construction and foundation the Phase II walls and the Phase IV extension. So far
(Fig. 4.45). as it is possible to be certain, the Phase II and Phase IV
mouldings were identical, with only one minor difference
that will be described below.
The first 10 courses below the Phase IV mould-
ings have a Header-Stretcher-Stretcher-Header pattern
which is varied at the corners, then, from the 11th
course, the moulding consists of two courses of bricks
with a single chamfer or curve on their upper edge,
followed by one slightly recessed course with a double
chamfer (one on the top and one on the bottom edge),
above which is a three-course curved moulding (Figs
4.18D, 4.42 and 4.46). Above this, the wall is badly
eroded but there are some indications that further
mouldings were present.
Fig. 4.45: The robbed-out wall 352 at the east end of the South It should be noted that the double chamfered
Temple looking south. bricks of the Phase II walls have a particular feature that
is not present in the Phase IV mouldings. At one end
of each brick, a short 9-cm section has been left un-
Adhisthana mouldings chamfered (Fig. 4.18D). This end is always laid towards
the south on the west face and towards the east on
The earliest extant adhisthana mouldings on the the south face. These bricks have the appearance of
exterior of the temple walls belong to Phase II. It is a simple fluted pilaster with a square base laid on its
unknown whether or not the Phase I structure was side. The equivalent bricks of the Phase IV mouldings
adorned with them as nothing survives of the super- are chamfered along their whole length.
structure of this phase. However, only the lowest The mouldings are very similar to the small section
three courses of the chamfered bricks of the Phase II of moulding preserved in Phase III of the North Temple
mouldings are preserved below the Phase IV rebuild- (wall 514), with only a few minor differences.
ing mentioned above, whereas the whole of the Phase
IV mouldings are still extant (Fig. 4.42). It is also clear,
as has been stated above, that the mouldings on the

Fig. 4.46: Elevation of the southern face of the southern wall of the South Temple (wall 560).
 Chapter 4 The Brick Temples in Trench A    53

Phase V no way of verifying the sequence and the order given

here is based on relative floor levels and comparative
There are four walls inside the mandapa of the South quality of construction.
Temple which all appear to belong to a final phase of
internal re-organization. These consist of two frag-
mentary parallel internal walls (510 and 511), which Later use and robbing
divided the mandapa into three aisles, and an interior
thickening of both the southern and the northern walls After Phase V, a number of minor changes and additions
(509/562 on the south side and 528 on the north). These were made, before the temple finally fell into disrepair
latter walls are each one-brick thick and the space and was robbed of its building materials, buried and
between them and the original temple walls is filled damaged by later construction activity on the site. As has
with silt. Wall 562 seems to have been inter-bonded already been stated, the latest phases of the temple’s
with the upper courses of the Phase IV wall 359, but use are the least well understood because the related
the two walls are unlikely to be contemporary. The stratigraphic layers were largely removed by Yusuf ’s
relationship between the Phase IV North Temple wall excavations in 1937.
and Phase V wall 528 is not clear as the upper parts of There is evidence that substantial parts of the temple
the north wall have been reconstructed. The Phase V superstructure were rebuilt once again after the Phase
walls are of very poor construction and preservation: IV rebuilding. The upper courses of the north wall (561)
only one or two courses of each of them have been pre- and the surviving upper parts the south wall (1241, 1242)
served. They are all built on rather shallow and crude are made of a very different type of brick to that used in
foundations of clay and stone. earlier phases. These are ‘purple’ highly fired bricks of
As can be seen in Fig. 4.33, all of these walls were a notably smaller size than those used elsewhere in the
constructed from around the same floor level and they building. They measure 6/7 × 17 × 27 cm and they are
appear to have been part of a crude re-organization often so highly fired that they show signs of vitrification
of the internal space of the mandapa during the last on the surface. Similar bricks, although of a slightly dif-
period of the temple’s use. It is not clear how far 510 ferent size, were used in the latest rebuilding of the North
and 511 originally extended to the east; in the recon- Temple garbhagriha (wall 971). This late rebuilding of the
struction (Fig. 4.32), they are shown as forming an north wall of the South Temple is interesting as it has a
internal unit similar to wall 558 in Phase IV of the very simple type of moulding consisting of one brick with
North Temple, but this is purely speculative based on a deeply chamfered upper edge inset by about 4.5 cm, five
comparison with the North Temple. courses above which there is a further 2 cm recess. This
It is possible that Phase V should also include the ‘moulding’ continues over the abutment between walls
later rebuilding of the upper courses of the external 560 and 345, thus proving that it occurred later than the
walls of the temple (561, 1241, 1242; see below), but Phase IV extension. The simple moulding seems to have
this seems unlikely because a very different type of been intended as a crude imitation of the original Phase
brick was used in these rebuildings, which have there- II/IV mouldings, which are found at the same height on
fore been included in the next section ‘Later use and the other walls of the temple. From a photograph pub-
robbing’. lished by Yusuf (1939: pl. XVIIa), it is clear that this wall
It should be stressed that there is no stratigraphic stood much higher at that time and had at least two more
evidence which proves that the Phase V walls are later offsets in upper courses, which have collapsed or been
than the Phase IV extension. Indeed, it should be noted removed since 1937.
that the internal thickening of the north and south It is impossible to suggest a date for this late
walls (509/562 and 528) does not, at present, extend rebuilding. So far as it is possible to tell, it did not
further east than the eastern end of the Phase III struc- involve any change in the plan of the temple, and it has
ture. This may be an accident of survival, as seems not therefore been allocated a separate phase number.
most likely, but it is nonetheless possible that Phase As was mentioned above, it is possible that the
V is actually an internal re-organization of the Phase rebuilding is contemporary with the internal changes
III structure and that it preceded the Phase IV exten- of Phase V, although this seems unlikely as the bricks
sion. If this were the case, then walls 510/511 obviously used in the Phase V walls are of a completely different
would not have extended so far to the east as they are type.
shown on the reconstruction. There is, unfortunately,
54   Derek Kennet

As is described below in more detail, excavation at that have been excavated clearly extend beyond the limits
the eastern end of the South Temple (Section 2 below: of the present excavation.
EEST) revealed a number of features that provide some
brief insights into further changes that took place
during the very latest phases of the temple’s life. They
include a crude wall built of large architectural stone
fragments on top of the Phase IV eastern wall 352.
In addition, wall 349 is abutted to the east by a now
fragmentary silt and brick-bat alignment (354) that
once continued to the north and which may indicate
that the South Temple was once incorporated into a
larger architectural complex. There is also a group of
large flat stones to the east of wall 352, which continue
beyond the edge of the trench. The use or re-use of
stone is important, as stone was not used as a building
material in any of the temple’s earlier phases. Where
the re-used architectural fragments noted above came
from is unknown, but it is possible that the two brick
temples described here were in fact part of a larger
temple complex, the rest of which still lies buried in
the vicinity.
Once the temple had fallen out of use and been
buried, occupation returned to the area in the form
of large post-Medieval town houses, the fragmentary
remains of which can be seen across the area today. The
construction of these substantial houses involved the
excavation of deep foundation trenches that caused a
lot of damage to underlying archaeological layers.

The possible boundary wall

In the area to the south and west of the temples some Fig. 4.47: Foundations of the Early Medieval perimeter wall 699 in
rather fragmentary evidence has come to light of a possi- the South Area of Trench A looking north. The stones of wall 691 are
ble stone boundary wall or walls that may have been built visible immediately to the west. At the top of the photograph, the
foundations of 699 turn to butt the south-west corner of the South
to surround the two temples at a late date (see Section
Temple.
2 below: South Area and W Area; and context numbers
691, 699/578, 688, 671, 616, 903, 617) (Fig. 4.47). The evi-
These walls were constructed at least partly of stone, or
dence for the presence of walls is strong, but given the
so it seems, and they are obviously quite crude – they
little that remains, the function of the walls, their date
are not straight or regular and they have clearly not
and the full plan of any complex that they may have
been carefully planned. They give the impression of
been a part of are purely speculative. Fig. 4.48 shows
a rather piecemeal construction. They were probably
these walls and gives some idea of their layout so far as
constructed very late in the temples’ history and, if
it is known, but it should be remembered that these
they did form a boundary to the complex, they may
are badly disturbed levels where preservation is poor.
reflect some important changes in the way that the
It should also be noted that excavation to the north
temples were used and considered.
and east was not possible due to the limits of Yusuf ’s
trench, and it is therefore quite possible that similar
walls also existed on the other side of the two-temple
complex but that they have not come to light. The walls
 Chapter 4 The Brick Temples in Trench A    55

Fig. 4.48: Trench A showing the remains of the possible late enclosure walls 578/699 and 691 and associated features to the south
of the South Temple and feature 617 to the west of the North Temple.
56   Derek Kennet

General discussion certainly occurred towards the end of the temples’ lives
in the later rebuildings and alterations, such as Phase V
Although no inscriptions or images were found directly of the South Temple and Phase V of the North Temple,
in association with the two temples, either by the present of which relatively little evidence remains. The size and
excavations or by Yusuf, it seems fairly certain from their type of bricks varied throughout the period that the
design and development that these are ‘Early Hindu’ temples were in use. The earliest bricks were smaller
temples – a term that is used with recognition of the asso- than those used in the middle period of the temples’
ciated terminological problems (see Flood 2003 for a dis- lives, whilst the latest bricks used for rebuilding were
cussion). not only the smallest used in the structures but also
Having said this, a fragmentary stone sculpture of the highest fired. This subject of brick sizes is dealt
a female was found in Trench B, some 70 m to the south with in more detail below as it is key to the linking of
of Trench A, and the Balasaheb Patil Government the phasing of the two structures.
Museum at Paithan contains a number of terracotta, There is no clear evidence to indicate how the
kaolin and stone figurines that apparently come from temples were roofed. Some fragments of tiles were
Paithan, although it is not known exactly whereabouts retrieved here and there during the excavation, but it
they were found. These are illustrated in Chapter 8 seems likely that these were used for flooring rather
(Figs 8.17 and 8.18) and Chapter 11 (Figs 11.6 and 11.7) than roofing as the quantity is so low. Had the temples
and further discussed in Chapter 14. been roofed by tile, a very large quantity of tile would
have been expected.

Construction
Foundations
Both of the temples are constructed entirely from fired
brick and silt. Silt was used for the pointing of the bricks One of the most interesting insights into the construction
and there is no evidence for the use of lime mortar of the temples that was discovered during the excavations
before the 1967 restoration. The only stone used was relates to the foundations. In both temples it seems that
in the foundation deposits and in what are certainly the foundations are much more substantial than would
later additions and alterations to the temples, such as be structurally required for brick buildings of this size.
some of the walls at the east end of the South Temple As has been stated, they consist of massive, vertically
(EEST) and the possible boundary walls. The quality of sided rectangular pits, in each case slightly larger than
the brickwork and jointing varied at different periods; the temple (Fig. 4.49). The foundation pits are 2 m deep
the finest jointing is probably to be seen in the middle and are cut down into virgin soil. The foundation pits have
phases of the temples’ lives, with the earlier and later been carefully back-filled with successive layers of boul-
work being of a lower quality. The crudest construction ders (up to about 30 cm) and silt, alternating with layers

Fig. 4.49: A composite section through the two temples and their foundations.
 Chapter 4 The Brick Temples in Trench A    57

of compact silt. This sequence of layers was repeated five the actual text itself. The text also suggests that there
times before the walls were constructed on the uppermost is significance in the fact the Paithan foundations are
level of compact silt. After the construction of the walls, a cut down to natural soil, thereby reaching pure earth
further sequence of slightly thinner layers of boulders and uncontaminated with human artefacts.
silt was deposited within the temple before the floors were This is not the first time that such elaborate, ritual
laid. Judging by the finds contained within the silt layers, foundations have been reported from the excavation of
the silt was dug from nearby on the site and incorporates a temple in India. I. K. Sarma recorded similar foun-
residual artefacts such as pottery, bone and other materi- dation deposits below the Chalukyan Sangameshvara
als that are commonplace in all soil at Paithan. temple and Papanasi group of temples in Alampur
There are slight differences between the nature of district, Andhra Pradesh (Sarma 1993: 356–361) and
the foundations of the North and South Temples. The similar foundations have also been reported, more
stones used in the North Temple are smaller and more recently, underneath a Rashtrakuta Jaina temple at
angular, and the silt surrounding them seems to be Pattadakal in Karnataka (Venkateshaiah et al. 2006:
less compact and less pure than that used in the South 337–340). A search through the archaeological litera-
Temple. ture, however, also reveals that a less elaborate but
It seems certain that the rationale behind these very otherwise very similar foundation technique had been
substantial foundations was sacred rather than struc- used for domestic structures in the north-western
tural. Indeed, it will be seen that the Paithan foun- Deccan since at least Early Historic times. For example
dations follow the Brahminical literature precisely in at Brahmagiri (Kolhapur), the following description is
all aspects. It is worth citing Kramrisch’s translation of given of the foundations of domestic structures of the
the relevant source to demonstrate this: Early Historic period.

in extent the foundation pit is coterminous with the bounda- Two or more sides of a house were raised on a foundation pre-
ries of the Prasada; in depth it is equal to the height of man pared by embedding large pebbles, in two layers at least, of
standing with raised arms, or it is dug to the rock-bottom or sticky clay. The latter was either regur or brownish river mud.
until gravel is reached or the water level according to the geo- It has not been possible to ascertain whether these founda-
graphical conditions of the site. After the pit is dug, it should tions were made by cutting a trench in the earth. Over these
be filled with pure earth, eight finger widths (angula) high; clay and pebble bases the brick walls were erected. In order
on this layer another one is placed, one cubit in height and to strengthen the building, and to have a hard level surface,
composed of layers of strong stones each embedded in wet the intervening space, between the four walls, was at times
earth and separated one from the other by sand and earth; covered with roughly dressed stone slabs, and the whole filled
when the foundation has been laid so far, it is moistened with up with layers of sticky clay, mixed with lumps of laterite up
water, trodden by elephants, and levelled with heavy wooden to a height of three to four feet above the foundation.
stampers. On top of this, it is firmly packed and when one It appears that this method of preparing the foundation
fourth of the pit remains the first bricks are laid. was purely a local feature, and probably confined to the riv-
(Kramrisch 1946, I: 105). erine tracts of the Deccan.
(Sankalia and Dikshit 1952: 135).

This passage is from the Isanasivagurudevapaddhati

(vol. III). This is a Paddhati, or a ritual compendium in Sankalia and Dikshit then go on to describe a few
which all sorts of rituals are systematically treated and similar types of foundations from other parts of India,
which is partly based on earlier texts such as Agamas. such as Bihar, where slightly different materials were
These Paddhatis served as handbooks for priests in used but the concept was the same (Sankalia and
the first place, but not so much for architects. They Dikshit 1952: 135–137).
became popular from the 11th century onwards and the At Bhokardan a similar type of foundation was
Paddhati at issue is probably dated to the 12th century described for a platform structure of the Early Historic
and was probably written in South India. period:
There is a remarkable correspondence between the
details given in this text and the foundations of the Here was noticed a big platform paved with rubble rammed
in black earth … It consisted of 3 layers of rubble interspaced
Paithan temples, even down to the proportions and
with black earth, the third layer of rubbles being compara-
heights given. Not only does this text therefore help tively of bigger size than those on the surface.
to explain the complex nature of the Paithan founda- (Deo and Gupte 1974: 13).
tions, but it also offers incontrovertible proof that the
traditions that it sets out are clearly much older than
58   Derek Kennet

And finally at Nevasa, ‘elaborate foundations’ are dation cuts widen out quite markedly and this suggests
mentioned, again for domestic structures of the Early that they may have been re-cut at a later date, perhaps
Historic period. These were much shallower than in order to repair and re-point the lower part of the
those brought to light at Paithan, but the concept is walls (Fig. 4.50). This is quite a common phenomenon,
identical: which unfortunately makes it impossible to establish
from which level the original cuts were made. It will
The foundation for these structures varied in depth between 1 ft. have been noted, however, that exactly the same form
7 in. and 1 ft. 2 in. The first course consisted of black sticky clay
and depth of re-cut are a uniform feature of all of the
of a thickness of 4 in. to 7 in. Over this came the second course
comprising undressed stones which were capped with black foundation cuts on all sides of both temples; this sug-
clay. The thickness of the latter was so adjusted as to present gests that these widenings might actually be part of the
a level surface for laying the bricks. original foundation cuts.
(Sankalia et al. 1960: 53).

The evidence from these three sites suggests very The dating of the temples
strongly that the method of foundations used in the
Paithan temples is a sacred elaboration of a domestic Relative chronology
tradition that has been prevalent in this part of the
Deccan since at least the Early Historic period. The Both temples underwent five separate phases of devel-
fact that the foundation technique is so accurately opment. Although the absolute dating evidence for
described by the Isanasivagurudevapaddhati suggests these developments is still rather limited, it is pos-
that the rituals of temple construction as set out in the sible to set out a relative chronology based on some
Brahminical literature are probably sacred formalisa- aspects of the construction.
tions of much older domestic construction techniques It has been noted above that the first three phases
that, in this case at least, may have had a quite specific of the North Temple are built of bricks measuring 7 ×
geographical origin. This is a potentially useful insight 25 × 40 cm, but Phase IV was constructed of larger
into how, and perhaps where, some of the Brahminical bricks measuring 6.5 × 26 × 42/43 cm. The larger sized
rules related to temple construction were developed. bricks were also used to construct Phase I of the South
At the same time, it should be noted that the more Temple and all later phases of that temple. This sug-
or less contemporary Gupta brick temples excavated gests that Phase I of the South Temple was constructed
at Bhitari appear to have quite different foundations. later than Phase III of the North Temple, after the intro-
Temple 1 at Bhitari is partly constructed on a raised duction of a larger brick size.
plinth created by a grid of brick walls, the spaces Brick sizes do not indicate whether Phase IV of the
between the walls are filled with 60 cm of compact North Temple was built before or after Phase I of the
clay, whilst other parts of the temple have a founda- South Temple. However, the walls dividing the interior
tion of eight courses of brick sunk into natural soil of the mandapa in Phase IV of the North Temple are
(Jayaswal 2001: 50, 73). The foundations of Structure very similar to those of Phase V of the South Temple
1 of Temple 2 at Bhitari, which shares close struc- and are therefore likely to have been roughly con-
tural parallels with Phase II of the North Temple at temporary. In addition, the adhisthana mouldings of
Paithan, were not fully explored, however. It can be Phase III of the North Temple are very similar to the
seen from Fig. 14 in the Bhitari publication (Jayaswal mouldings of Phase IV of the South Temple, but it is
2001: 84) that there is a foundation cut for the trira- possible that Phases I to III of the South Temple had
tha plan sanctum which cuts two layers, but unfortu- similar mouldings. Based on these observations, a rel-
nately, it seems that excavation was not continued to a ative chronology has been set out in Table 4.2.
sufficient depth to ascertain whether or not the founda-
tions are similar to those at Paithan.
Two further points need to be made in relation to
the temple foundations. Firstly, no evidence of ritual
offering was found beneath the garbhagriha of either
temple, despite the fact that they were both com-
pletely excavated. Secondly, an examination of the
section from Test Pit 3 shows that the top of the foun-
 Chapter 4 The Brick Temples in Trench A    59

Fig. 4.50: West-facing section of Test Pit 3 showing the foundation cuts for both buildings.

Table 4.2: The proposed relative and absolute chronology of the two temples.

Phase Development Proposed date

North Temple Phase I Simple shrine 5th century or earlier

North Temple Phase II Shikhara construction Mid–5th century

North Temple Phase III Mandapa and adhisthana mouldings Late 6th or early 7th century or later

South Temple Phase I Change in brick size 7th century or later

South Temple Phase II

South Temple Phase III

South Temple Phase IV

North Temple Phase IV and South Interior walls subdivide mandapa in both temples
Temple Phase V

North Temple Phase V

Abandonment Later than the 8th century

This sequence is supported by the admittedly uncer- Pit 3, to judge by the height of the Phase I wall and the
tain evidence from the test pits where it was shown to level from which it was constructed, layers above 1129
be most likely that the South Temple was built after or 1124 (probably including the latter) very possibly
the construction of Phase I of the North Temple. In Test accumulated between the construction of the North
60   Derek Kennet

Fig. 4.51: South-facing section of Test Pit 1, Trench A, showing the foundation cut for the South Temple.

and South Temples, but there is no way of being abso- foundation cut for the temples and the surrounding
lutely certain of this. Unfortunately, the stratigraphy pre-temple deposits.
is not well enough preserved to contribute very much At the very broadest level, the structures can be
more to our understanding of the relative chronology dated between the 4th/5th and the 8th centuries ad,
of the two temples. It can, however, provide something during which time the architectural form of the Hindu
of a broader context for the construction of the temples temple emerged, but a more accurate chronology of the
at this location. Test Pits 1 and 3 both revealed sections individual phases is required in order to situate their
through the pre-temple deposits in the area of Trench development within an historical context. This can
A (see Section 2 below). These deposits were about only be achieved through parallels between architec-
1.5 m deep (Fig. 4.51) and can be subdivided into two tural elements of the Paithan temples and dated build-
broad phases; the first deposits tend to be rich in brick ings elsewhere. There are four stages in the develop-
rubble and other debris from construction or demoli- ment of the two temples for which such parallels can
tion activity. Above this, the deposits become silty and be proposed:
contain pottery and bone but little building material. 1. The recess in the wall of Phase I of the North Temple
Above these deposits, the temples were built. This evi- appears to be a very early form of plinth moulding
dence suggests that, although there was other building that can be compared to the deep recesses in the
activity going on not far away, the temples are the earli- mouldings of the Rudra-Narasimha, Bhogarama
est substantial constructions at this precise location. and Kevala-Narasimha temples at Ramtek, the
The silty deposits suggest a period of time when the area latter of which is dated to the 5th century by an
was a weed-covered dump close to a larger settlement inscription (Fig. 4.52) (Meister et al. 1988: 66–70).
immediately before the temples were built. A similar recess is also present in the earliest
phase of the Kapoteshvara temple at Chejerla (Fig.
4.53), which I. K. Sarma has dated to between the
Absolute chronology 2nd and the 4th century ad, although he does not
specify the evidence for this (1982: 138–139). The
The absolute chronology of these structures and their extreme simplicity of the Paithan moulding might
sub-phases is problematic because almost all of the suggest that Phase I of the North Temple should be
associated stratigraphic levels were removed without dated to the early 5th century ad, or possibly some-
record by Yusuf ’s excavations, especially the strati- what earlier.
graphic connection between the excavation of the
 Chapter 4 The Brick Temples in Trench A    61

Fig. 4.52: Plinth mouldings from Ramtek: (a) Rudra-Narasimha temple; (b) Kevla-Narasimha temple; (c) Bhogarama temple (after
Meister et al. 1988: fig. 40).

Fig. 4.53: Plinth mouldings on the main temple at Chejerla (from Sarma 1982: fig. 18).

2. The thickened walls of Phase II of the North towards the end of the 5th century ad or later, such
Temple were clearly intended to support a shi- as Nachana, Bhumara, Sakor, Elephanta Cave
khara over the garbhagriha. It has been suggested 1 and the stupa in Ajanta Cave 26 (Meister et al.
that the construction of shikharas is a stage of 1988: 39–40, 52–54, 74–75, 90–91), whilst it does
Gupta temple development that can be dated to not appear to occur on structures built earlier than
the second half of the 5th century on the basis of about 475 AD such as Bhitargaon, Kunda, Tigawa,
parallels with Bhitargaon, Deogarh and the brick Bhitari and Darra (Meister et al. 1988: 30–32,
temple at Bodhgaya (Meister et al. 1988: 23). The 35–37). This, however, serves only as a terminus
triratha plan of Phase II of the North Temple is also post quem for the construction of the mandapa
closely comparable to Structure 1 of Temple 2 at walls – the Paithan mouldings could have been
Bhitari, which has been dated to the second half built considerably later, because similar rounded
of the 5th century ad on the basis of similarities kumudas continued to be built for some centu-
with the Parvati Temple of Nachana, as well as the ries. It is difficult to date the Paithan mouldings
re-use of broken bricks that, it is thought, were left more accurately than this as the detailed scheme
over from the construction of nearby Temple 1 in of development of adhisthana mouldings has not
the first half of the 5th century ad (Jayaswal 2001: yet been fully elucidated in North India. Similar
125–127). mouldings are known from the third stage of the
3. The adhisthana mouldings in the walls of North Kapotesvara temple at Chejerla in Andhra Pradesh,
Phase III and South Phases II to IV are quite where a date in the 7th century ad has been sug-
sophisticated. In North India, the rounded kumuda gested, although there is no independent evidence
(or kalasa) commonly occurs on structures dated for this (Longhurst 1917–1918). Similar mouldings
62   Derek Kennet

also can be seen on the pedestal in the sanctum Section 2: test pit and area-excava-
of the Trivikrama temple at Ter, the precise date of
which is also unknown (Mate 1957).
tion reports from Trench A
4. A date for the addition of a mandapa to the mul-
After an initial cleaning and recording in the 1998
aprasada in Phase III of the North Temple can also
season, the full extent of Trench A was not excavated
be suggested. It could be said that the earliest
because it was too large to have been properly investi-
known mandapas are the gudhamandapas on the
gated with the time and resources available. Instead,
Rudra-Narasimha and Kevala-Narasimha temples
targeted smaller-scale excavations were used to eluci-
at Ramtek, but these are quite particular, as they
date specific details of the temples’ construction and
are part of the same structure as the sandhara garb-
phasing and to answer specific questions or to clarify
hagriha, and can therefore be ignored as relevant
areas of uncertainty. To this end, 10 test pits were
parallels. The first mandapas that resemble Phase
excavated (Test Pits 1–10) (Fig. 4.5) and area-excava-
III of the North Temple at Paithan, in that the
tion was conducted in six different parts of the trench
mandapa is more clearly separate from the mul-
(see Table 4.1).
aprasada, first occurred at the end of the 6th or
The reports presented below are edited versions
the early 7th century on temples such as the Laks-
of the end-of-season field reports written by the indi-
mana and Rama temples at Sirpur, the Rajivalo-
vidual excavators after completion of excavation. They
cana temple at Rajim, and the Parasuramesvara,
present the details of the stratigraphy which forms the
Mohini, and Uttaresvara temples at Bhuvanesh-
basis for the more interpretive phasing and description
vara (Meister et al. 1988: 230–236, 256–260, 265–
of the temples that have been presented above. They
267) in North India, and began to develop in the
also present the thoughts, impressions and doubts of
later 7th and early 8th centuries in the Karnata-
the excavators. Numbers in brackets are excavation
Dravida tradition (Hardy 1995: 67, 71, fig. 46).
context numbers that are listed in Appendix I.

The presence of a degraded copy of the coins of the

Sarvva-Bhattaraka and Maitraka dynasties in the
Test Pit 1 (TP1): the western wall of the
foundation deposits of the South Temple (coin No. 34,
context 700; see Chapter 9), which can be tentatively South Temple (Figs 4.44, 4.51 and 4.54)
dated to the 7th century, provides a terminus post
The sequence in this test pit was initially divided into a
quem for the construction of this temple.
number of horizons, which are described below. These
The final abandonment of the Paithan temples
were then allocated to the site’s periods as is described.
cannot be dated accurately because the relevant levels
The key layers mentioned in the description below are
are so badly preserved. An impressionistic evaluation of
shown in Fig. 4.51.
the limited amount of pottery from the South Area, where
This test pit was begun in the 1998 season during
a few layers related to the later use of the temples have
which time the upper 0.80 cm was excavated. The
been excavated, might suggest a date of about the 8th
topmost layer consisted of fairly clean silt (300) that
century ad for these levels, but it should be stressed that
had probably been deposited since the 1937 excava-
the pottery chronology for the Early Medieval period in
tions. Beneath this part of a collapsed brick wall was
India is still very imprecise and badly understood.
encountered (304) that looked like it might have fallen
The proposed dating described above is incorporated
from the upper courses of the west wall of the South
into Table 4.2 to give a dated summary of the develop-
Temple. More horizontal layers were revealed below this
ment of the two buildings. Further discussion of the sig-
that were cut by 311/963, the foundation cut of the South
nificance of the Paithan temples can be found in Chapter
Temple. Deposits 300, 301, 304, 305 and 307 overlaid the
14.
foundation cut for the South Temple, whilst 306, 310
and 313 were fills within it. Layers 308, 312, 316 and
318 predate the foundation cut and are roughly equiva-
lent to layers 908, 912, 913 and 922 that were excavated
in 1999 (see Horizon 4 below).
Excavations were continued in Test Pit 1 in 1999
with the aim of investigating pre-temple deposits down
 Chapter 4 The Brick Temples in Trench A    63

to natural soil. The following four horizons of deposi- 984), post-holes and pot/tile pits (e.g. 1029) were cut.
tion were noted: Following this, more rubble layers were laid down
(e.g. 966, 944). The upper surface (944) showed evi-
dence of compaction and represents the only truly
consolidated surface horizon identified in this area. A
small ash pit (946) and ash dump (948) were recorded
on this surface. The deposits of this horizon have been
allocated to Period 2.
Horizon 2 consists of two clay-rich deposits with
frequent grit and small stones (1127 and 1120). Deposit
1127 was cut by a large pit (1145). After this pit had been
filled and covered, a second pit was cut (1131), which
was filled with a pot and tile dump 1126. These depos-
its have been assigned a separate horizon principally
due to the lack of brick in layers 1127 and 1120, as brick
rubble is such a notable feature of the layers of Horizon
3 above. A small patch of burning (1141) was also noted
next to pit 1131. Layers below 1127 in this horizon have
been allocated to Period 1, whilst 1127 and the layers
above have been allocated to Period 2.
Horizon 1, the earliest horizon, consisted of a black
cotton soil with occasional pottery fragments (1159) to
a depth of 0.44 m directly above natural soil (1160). It is
thought that this may represent a buried plough soil.
The only additional feature excavated in this area
was the foundation cut to the South Temple 963 (=982).
This cut is slightly offset from the foundation courses, the
southern end being closer to the building. It was filled
with a number of alternating silt/stone and compact clay
layers which formed a thick platform. This cut was not
quite bottomed as it disappeared under the east section. It
was, however, recorded as cutting a low ash layer 1159. Its
Fig. 4.54: Eastern section of Test Pit 1 showing the foundation fills
under the west end of the South Temple.
greatest observed distance from the foundation courses of
the wall was 0.14 m.

Horizon 4, the latest horizon, marks a period of

reduced human activity compared to the lower depos-
Test Pit 2 (TP2): the south wall of the North
its. It consists of a series of water-deposited silt washes
(908, 912, 913, 921 and 930) which contained occasional Temple
ash washes/dumps. Two top-soil-like deposits (933 and
A test pit was opened over the remains of the south
922) were also noted. All of these deposits have been allo-
mandapa wall of the North Temple in order to investigate
cated to Period 4.
a possible robber trench. A sequence of very late surfaces
Horizon 3 marks the most intensive activity within
and cuts was revealed.
this sequence. It is defined by a series of rubble layers
into which a large number of pits were cut. The ear-
liest of these was cut 1113 and sump 1114, which was
Test Pit 3 (TP3): the foundation cuts of the
filled with deposits rich in building debris (1110 and
1099). Two rubble deposits, 1077 and 1063/1052, accu- two temples (Fig. 4.50)
mulated prior to pits 1057 and 1075 being dug. Pit 1075
The sequence in Test Pit 3 consists of a series of pre-
was filled with brick and tile (1074). More rubble layers
temple deposits sandwiched between two cuts: the
(1040 and 1013) were dumped here before ash pits (e.g.
foundation cuts for Phase I of the North Temple (1171)
64   Derek Kennet

and the foundation cut of Phase I of the South Temple While we can be fairly certain that 982 is cut from 993, it
(982). The former is complicated by re-cut 683. is not possible to be certain of this with 1171. All that can
The top fill of the re-cut (682), a loose black cotton be said is that it can be no earlier than 1129, although
soil, seems to be cut by 339, which may be the foun- comparison with Test Pit 10 may resolve this question.
dation cut for the North Temple Phase III mandapa For what it is worth, it does seem that the cut curves
(see ‘W Area’ below). Below 682 is a brick-jelly fill (942) inwards at the level of 1129 and probably was cut from
sitting over what were at first thought to be more layers about there.
extending over the length of the trench. It became The layers in between the two cuts form two dis-
apparent that this hard compact layer was itself a fill tinct groups. The uppermost are quite disturbed by cuts
of a much deeper cut (1171), 683 being a re-cut more (1011/1012, 1004/1005 and 994/993), coming down to
or less exactly on the same line and cut from 959, the an occupation layer (1030) cut by a series of features
highest homogenous layer. There is, however, some (1016, 1018, 1020, 1022, 1024, 1026 and 1028). Below
uncertainty about cut 339; firstly, it is not certain that 1030, the situation is simpler, consisting of thicker
it is the cut for the mandapa wall, and secondly, it is homogenous layers, e.g. 1124, 1129, 1140 and 1148, only
not absolutely certain that it cuts layer 682. one of which (1134) has a post-hole cutting it (1139).
Similarly for the South Temple, after the removal These layers are sometimes ashy (e.g. 1036, 1134) and
of some superficial layers, a homogenous fill (981) was sometimes result from a build-up of soil with rubble
revealed. It was crammed full of pottery and the cut inclusions (e.g. 1148). Unfortunately, it was not possi-
went down to a hard layer, at first thought to be the ble to say much about them due to their being trun-
base of the feature. After allowing this to weather, it is cated by the two foundation cuts 982 and 1171. Excava-
fairly certain that feature 974/975 is actually the same as tion was stopped after the removal of layer 1148.
981/982. In any case, both are cut from 993, although it
was at first thought that 975 cut 981. A similar feature
was found in Test Pits 1 and 5. Removing the hard layer Test Pit 4 (TP4): the stratigraphic relation-
1039 revealed a repeating sequence of deposits famil- ship between the two temples
iar from other parts of the trench as temple founda-
tion deposits, namely a silty-clay matrix surrounding Test Pit 4 was dug in an abortive attempt to establish the
small boulders, often with gravel tip lines, lying over stratigraphic relationship between the foundation cuts of
a compacted clayey layer upon which the stones were the North and South Temples. A 50-cm-wide sondage
set. Excavation was stopped at the third layer of boul- was excavated between the two buildings to a depth
ders beneath the wall (1068/1069), revealing the next of about 50 cm. Work was abandoned when it became
clay layer (1192). Therefore, cut 982 was not bottomed clear that no stratigraphic relationship remains
in this test pit. between the two foundation cuts in this area, the
Foundation cut 1171 showed a similar pattern. After crucial information having been removed by Yusuf ’s
removing the hard layer (1073), a sequence of a silty excavations. During the excavation of Test Pit 4, no
matrix surrounding boulders set on a more compact context numbers were assigned or finds collected.
clayey layer followed. The third hard layer of boulders
(1112) lay at the base of the cut. The foundation cut
(1171) was originally dug 40 cm into natural soil. Test Pit 5 (TP5): the interior and foundations
The similarities between the two cuts are obvious; of the South Temple (Figs 4.33–4.38)
both contain the same broad sequence of deposits,
both cut down into natural soil and both have a very Two parallel foundation trenches belonging to post-
compact upper layer on which the temple walls are Medieval buildings (594, 595) had disturbed the temple
built. The boulder layers in 1171, however, are much from above and had been mostly removed by Yusuf,
more patchy than in 982, while the layers in 1171 are but the bottom of their fills still remained. Their align-
more compact than in 982 and contain more small ment and length were made clear by the cut into the
abraded brick fragments. Gravel tip lines in the silty- south wall of the South Temple. A standing brick wall
clay matrix are evident in the fills of 982, but not in forms the eastern limit of Trench A running parallel to
1171, and a break in the homogeneity of deposits in these cuts, and this wall was probably part of the same
1171 is shown in layer 1103/1104, which is a mixed layer building for which the foundation trenches were dug.
of black cotton soil and a greenish-brown silty clay. The relationship between westerly cut 594 and pit 335,
 Chapter 4 The Brick Temples in Trench A    65

which cuts the north-west corner of the South Temple, This suggests that 667 must have been the floor level
was removed by Yusuf. The pit is probably earlier than of Phase I of the temple.
the foundation cuts. It is similar to pit 341, which is cut Following this, the temple was reconstructed on
into the north-west corner of North Temple, but this is a slightly different plan. The Phase I temple appears
probably coincidental. to have been knocked down, leaving only the lowest
The foundation cut for Phase I of the South Temple six courses remaining as a foundation base. Nothing
has been numbered 982, 1250 and 1181 in different therefore remains of the upper courses of the Phase I
places. It is a flat-bottomed cut with its base 20–25 cm wall, and neither is there any concrete evidence that a
below the level of natural soil (as seen in Test Pit 1 to temple of this phase ever existed, aside from the differ-
the west of the temple). The first fill consisted of small ence in plan between the Phase I wall and the Phase II
boulders of igneous rock packed into a very clean silt walls (see discussion above).
matrix (Fig. 4.35). This was overlain by a thin (c. 10 Subsequently, the brick-jelly layer 661 must rep-
cm) compact clayey layer containing numerous small resent the rebuilding of the temple during Phase II,
weathered pottery and rubble fragments on to which when walls 1240 and 1247 were constructed using the
a further layer of stones has been laid. The silt sur- Phase I wall as a foundation. From the section, it is
rounding the stones is less compact and rammed level quite clear that the brick-jelly has been trampled into
to prepare for the following stone layer. A total of five the clay surface 667, probably during the construction
boulder/silt/compact clay sequences were laid down process.
before a thin mud surface was spread across the entire After the construction of the Phase II walls, the inte-
foundation trench (953/955). rior was raised up to the level of mud floor 634 using a
There were few notable differences in the boulder, similar pattern of alternating boulder stones/loose silt/
silt or compact clayey layers throughout the fill, which compact silt layers (654/642/638) that was capped by the
was remarkably uniform. The first two stone layers con- floor. Through this process, the height of the floor was
sisted of tightly packed small to medium-sized boul- raised by 20 cm.
ders. The third and fourth layers consisted of evenly As can be seen in Fig. 4.36, the Phase II walls around
spaced larger boulders, each of which needed two the garbhagriha are as much as 35 cm thicker than
people to lift. The fourth stone layer (1070/1071) was the Phase I walls below them. This means that the
contained in compact silt (978/979) rather than a loose interior width of the garbhagriha was reduced by 70
silt matrix. But the matrix of the fifth layer (1002/1003) cm in Phase II. The garbhagriha was then filled with
was very loose by comparison. Upon this, an 8-cm-thick another sequence of stone and silt foundation layers
compact clayey silt layer (964/965) and mud surface (979/666/653). As they are physically separated from
were laid (953/955). each other by the interior wall of the temple, it is not
The foundation wall of the building (1224) is repre- possible to say precisely how the layers in the garb-
sented by the first six courses of brick, which were laid hagriha relate to those in the mandapa. It is notable,
directly upon the mud surface covering the foundation however, that the fills within the garbhagriha are purer
fill. Above this within the mandapa walls, the nature when compared to the construction fills inside the
of the foundation fill changes, becoming dirtier with mandapa; that is to say they contain fewer pot sherds,
numerous inclusions of rubble and broken pottery stones and other inclusions (e.g. 642/638/626/622).
(952/918/700). The same is not true of the equivalent It proved to be much more difficult to isolate possi-
level within garbhagriha (919), which does not contain ble floor surfaces within the garbhagriha than within
much pottery. This suggests that the fills are not the the mandapa. There are three ‘stone and silt’ founda-
same and were deliberately kept separate from each tion sequences in the Phase II garbhagriha, but there
other. A dark grey layer, 667, was deposited above 700 appears to be no definite surface between any of them.
and runs underneath the Phase II walls (1240/1247), as Within the mandapa, there is evidence that the
can clearly be seen where there is a gap or entrance way level was raised once more to create a floor using the
in the Phase I wall (1224), which divided the mandapa same silt and stone foundation sequence (626/622).
from the garbhagriha (Fig. 4.37). Here, layer 667 abuts This later raising of the floor seems to relate to Phase
a large, heavily vitrified brick situated in the middle of V of the temple, a much later reorganization of the
the entrance, suggesting that the brick may have been mandapa interior when the walls were thickened on
a step or part of a threshold between the two spaces. the interior (509, 528). The south-side thickening is not
shown in the section in Fig. 4.33 but a possible foun-
66   Derek Kennet

dation cut and fill (615/620) are visible. The two Phase compact rubble packing (342). Below this, a succes-
V interior dividing walls (510, 511) are shown in Fig. sion of silt/boulder stone foundation deposits fills the
4.33. Other evidence relating to Phase V was removed foundation trench, closely matching those uncovered
by Yusuf ’s excavations. in Test Pits 1 and 5.
A number of minor cut features such as pits and
post-holes were cut into the pre-temple compact silt
Test Pit 6 (TP6): south of the South Temple surfaces next to the temple, but nothing significant
(Fig. 4.55) enough to warrant a fuller description was located. A
thick, compact grey clayey siltwash (deposit 300) that
Test Pit 6 was opened against the southern wall of the has accumulated since Yusuf ’s excavations covered all
South Temple in order to investigate the nature of the the contexts in this area.
surfaces related to the temple and to reveal the foun-
dation cut in this area.
The sequence revealed consists of a succession of Test Pit 7 (TP7): the south-east corner of the
hard clay/silt surfaces (650, 926, 373, 987) above some South Temple
looser clayey silt deposits (1035, 1051). The lower
of these are clearly cut by the foundation cut for the This small test pit was excavated to investigate the
South Temple (1246 in this area), but because the foundation cut of the eastern extension of the mandapa
upper portion of the foundations might have been dis- wall on the south side of the South Temple and also
turbed by a possible later recut (372), it is impossible the original foundation cut for the Phase I wall. The
to be absolutely certain from which level the founda- test pit showed the existence of cut 372 which runs
tions were originally cut. Cut 372 describes the rubble- along the south wall of the temple, cutting surface 373.
filled broadening out of the foundation cut in the top The stratigraphic relationship between this cut, the
20 cm. It is very similar to the form of temple founda- original foundation cut for the temple (1230) and the
tion cuts in other parts of the trench (e.g. Test Pit 3; Fig. Phase IV mandapa extension cut (1178) was impos-
4.50), and it is not clear whether this is a later re-cut sible to verify with certainty. It may be that cut 372 cuts
or whether this was the original form of the founda- 1230 and its later fills and is then cut by 1178. But it
tion cut. Because the uppermost foundation fill (1031) is impossible to be certain that cut 372 and the origi-
that is clearly part of the original cut also seems to fill nal temple foundation are not one and the same. Three
the lower part of 372, it seems likely that, in this case, hard surfaces (1190, 1191 and 1231) were noted under-
this was the form of the original cut. In this case, most neath surface 373. The test pit was excavated only to a
of the horizontal silt/clay deposits shown in Fig. 4.55 shallow depth due to lack of time and no further sig-
were deposited before the temple was constructed. nificant results were obtained.
The upper part of the foundation cut is filled with a

Fig. 4.55: South-facing section, Test Pit 6.

 Chapter 4 The Brick Temples in Trench A    67

Test Pit 8 (TP8): the EEST (Fig. 4.43) ficult to be certain which features originally related to
the Phase V temple and which related to the restoration.
Test Pit 8 is a very small sounding that was excavated in These problems were compounded by the fact that the
1999 in order to investigate the stratigraphic sequence fills behind wall 383/544 were of loose black cotton soil
against the eastern wall of the Phase I–III South Temple within which it is difficult to discern cuts. Nonetheless,
(Fig. 4.43). The uppermost level encountered below the upon excavation, a large rectangular sunken area was
level reached in the 1998 season was a hard floor (1182) located against the western end of the mandapa in the
that abuts the eastern face of Phase II wall just below surface of deposit 551 (Figs 4.22A and 4.25). Measur-
the mouldings. This suggests that it is related to quite ing 140 cm north-south by 90 cm east-west and 15 to
a late phase of use of the temple. It overlies packing 20 cm in depth, it is thought that this depression may
1216 that raised the level of the floor by about 70 cm. represent the location of a large rectangular object
Below this, another compact floor surface (1245) was or structure that might have been part of a shrine. It
encountered, this time abutting the top course of the is difficult to suggest another interpretation for it. It
Phase I wall 1224, but which could be related to Phase I certainly abuts the Phase IV wall (558), but its strati-
or to a later phase. It overlies packing 1219, which itself graphic relationship with wall 383/544 is much more
overlies an earlier floor (1220) that is also related to the difficult to ascertain for the reasons given above. It is
Phase I temple or later. Below this, a number of thin thought most likely that shrine 551 was constructed
silt and brick-jelly layers (dug as 1220) overlie another within the space defined by wall 558 but that in Phase
hard surface 1227, which must have originally abutted V wall 383/544 eventually enclosed this shrine or a
the Phase I wall 1224, but which has been cut by 1226. later shrine on the same spot, possibly after the partial
Surface 1227 is the earliest surface that was uncovered destruction of the Phase IV wall 558.
here and must relate to the initial use of the Phase Below 551, a single course of boulder stones in loose
I temple. Cut 1226, which was filled with rubble and silt was revealed (639). This is clearly a foundation
stone layers 1222 and 1223, is a re-cut against the wall deposit, similar to those that have been revealed in
of the temple from slightly later in time. The reason for the North and South Temples. It abuts wall 558 and
this re-cut is not clear. Underneath 1227, a flat compact appears to have been laid down as a foundation for the
layer, 1228, was revealed that underlies the Phase I 551 ‘shrine’ described above.
wall and must therefore be the top fill of the original Upon removal of 639 three bricks (983) were uncov-
foundation cut of the temple, the precise location of ered positioned close to the western wall 556 and
which was not revealed in this small test pit. associated with compact surface 940 (Fig. 4.24). They
were situated upon some very distinctive deposits of
compact, clean silt with green and red hues (976 and
Test Pit 9 (TP9): the mandapa of the North 977). It seems likely that these bricks and the associ-
Temple (Figs 4.22–4.25) ated surface and silts are the remains of an earlier
shrine. This overlies a succession of trodden silt sur-
In order to investigate the interior of the mandapa of the faces (1010) that accumulated over what appears to be
North Temple, a long, 1-m-wide slit trench was excavated another foundation sequence (1121, 1115, 1050) con-
across the building from the south wall, across the robber sisting of a layer of medium-sized, semi-rounded boul-
trench to the northern section of Trench A. The trench ders of local igneous rock in a fine silt matrix similar
was positioned against the eastern wall of the Phase to deposits found elsewhere in the foundations of both
II structure (556). The sections in Fig. 4.22 show the temples. It is possible that the upper level of 1050 was
stratigraphic contexts and their relationship to the used as a surface for a period of time, but the related
Phase III, IV and V walls (557, 558 and 383/544, respec- deposits (1049, 1032, etc.) do not offer conclusive proof
tively). of this. Before the bricks of the 983 shrine were laid
The uppermost remaining layers in the interior down, a small cut (1007) was made against the wall
of the mandapa were excavated in the 1998 season. At through the trodden surface 1010. It is not known why
that time, Phase V walls 383/544 and the thick loose this cut was made.
black cotton soil that it retained (395, 551, 1252) were Wall 558, the Phase IV wall of the temple, was built
removed. The presence of lime mortar in the crude from the level of layers 940 and 976. It was built directly
Phase V wall 383/544 demonstrates that the 1967 res- on top of these layers without any foundations. It is
toration had disturbed this area, and this made it dif- impossible to be absolutely certain whether this wall
68   Derek Kennet

was built before or after the shrine 983. The wall over- of the mandapa. They are deposited within cut 1243,
lies 940, which appears to abut the bricks of the shrine. which is the foundation cut for the mandapa. They
Also, as can be seen from the east-west section along directly overlie the upper foundation deposit for the
the face of the southern part of the wall (Fig. 4.22B), Phase I structure (927/1186), the surface of which was
two bricks of the lowest course of the wall have been cut by 1243. This demonstrates that the foundations of
omitted, where it abuts wall 556, apparently to avoid dis- the Phase III mandapa were not excavated to the same
turbing deposit 976, which is associated with the shrine. depth as the Phase I foundations. This may be because
This suggests that the wall was built after a shrine was the mandapa is later or because, as an addition to the
already in use at this level; the wall being built perhaps garbhagriha, the sacred element of the construction
to enclose the shrine. may not have been taken so seriously.
The foundation deposits (1121, 1115, 1050) abut The cut for the Phase III mandapa (1243) is shown
the mandapa wall 557 to the south and therefore can in the northern part of the section of the test pit. On
be regarded as packing for a major raising of the floor this side, the cut of the Phase I structure (1171) is 1.64
level of the mandapa (1010). Below these deposits, m from the corner of the Phase I wall, which is much
another surface came to light (1142), this time covered further than the distance on the south side of the
with a coarse brick-jelly (1132) within which a very Phase I structure as exposed in Test Pit 3. The Phase
clear 1-m-square impression (1137) is visible against III mandapa foundation cut was probably excavated
the middle of the eastern face of wall 556 (Fig. 4.23). from a higher level than the Phase I cut due to depos-
Again, this suggests that a large rectangular object was its that had accumulated in the intervening period.
placed here against the western end of the mandapa Unfortunately, these deposits were removed by Yusuf
and that surface 1142 accumulated against this object. so it is now impossible to know from what height this
The most obvious interpretation is that this results cut was made. As is shown in Fig. 4.22C, cut 1243 is
from the use of this location as a shrine. Surface 1142 thought to have shaved the top corner of cut 1171 and
is the first floor level of the mandapa that came to light then to have followed the uppermost deposit within
in this sequence. the Phase I foundations (927). Unfortunately, time did
Below 1142, another foundation sequence of boul- not allow deeper excavation of Test Pit 9, which would
ders and silt came to light (1151, 1157). This time these have clarified the shape of both cuts and the relation-
foundation deposits underlie mandapa wall 557, indi- ship between them.
cating that they are related to the original construction

Table 4.3: Stages in the development and use of the shrine within the mandapa of the North Temple.

Stage

1 Possible use of a shrine against the eastern wall of the Phase II structure before construction of the Phase III mandapa.

2 Phase III mandapa constructed.

3 Shrine 1137 was constructed on top of boulder/silt foundations within the mandapa.

4 Further stone/silt foundation sequence apparently in preparation for raising the interior floor to surface 1010. Possible use
from the level of 1049 and above.

5 Accumulation of trodden surface 1010. The mandapa was clearly in use at this time but there is no evidence for a related
shrine. Had it gone out of use or was there a shrine of which no trace remains?

6 Shrine 983 was constructed. This seems to have involved the cutting of a small pit 1007 against the eastern wall, the purpose
of which is not known.

7 The Phase IV wall 558 was constructed around shrine 983 in order to enclose it.

9 Shrine 551 was constructed on top of stone/silt foundations within the Phase IV inner structure.

10 Phase V wall was constructed around shrine 551 or a later shrine, the remains of which have disappeared. It is possible that
this occurred after the collapse and robbing of Phase IV wall 558.
 Chapter 4 The Brick Temples in Trench A    69

On the south side of Test Pit 9, the join between walls of the garbhagriha as well as cutting into the Phase
556 and 557 is exposed (Fig. 4.21). This demonstrates I wall, where it had already been damaged by later
very clearly that wall 557 abuts wall 556 but that some robbing. Some limited cleaning and recording were
of the bricks in the upper courses of wall 556 were also carried out on the western face of the structure.
removed in order to allow the much cruder brickwork The uppermost fill within the garbhagriha con-
of the mandapa wall 557 to be keyed into it. It can also sisted of brick bats and silt (646). This layer seems to
be seen that one of the bricks in the lower courses of 556 result from post-collapse, or possibly post-Yusuf use
has been cut into for the same reason. of the area. Beneath this layer begins a succession of
The most important information to come from the silt/boulder foundation deposits similar to those in the
excavation of Test Pit 9, apart from elucidating the foundations of the South Temple (see Test Pit 5 above).
construction of the Phase III mandapa and its physi- These consist of alternate layers of cobble-sized stones
cal relationship with the Phase II walls, is the evidence of igneous rock within a matrix of compact silt (Fig.
of at least three successive shrines within the western 4.12). The stones are notably smaller than those used in
end of the mandapa against the wall of the Phase II the fill of the South Temple, and they are also smaller
structure. No evidence came to light to indicate that than those used further down in the foundation depos-
the eastern wall of the Phase II building had been used its of this building.
in this way before the construction of the mandapa, Having removed five successive silt/stone sequences
but any such evidence would in any case have been (658/676, 677/685, 687/698/907, 910/917), a fine, compact
destroyed by the construction of the mandapa. clay surface 927 was exposed. This surface was covered
Table 4.3 sets out the probable sequence of devel- with a dusting of brick-jelly less than 1 cm thick (923).
opments related to this shrine. Both 923 and 927 extended underneath the earliest wall of
the Phase I shrine (970/554). Clearly incised into the four
walls of the garbhagriha and onto the surface of 923 were
Test Pit 10 (TP10): the west end of the North four lines, running vertically down the middle of the inte-
Temple rior face of each of the four walls and continuing across
the surface of 923, forming a cross at the exact centre of
Test Pit 10 was opened to the west of the North Temple the garbhagriha. These scratches are just visible on the
after the upper levels in the area had been subjected walls in Fig. 4.14, below the blackboard. It seems that the
to an area-excavation (see ‘W Area’ below). The test scratches relate to the laying out of the sacred geometry
pit was excavated to a maximum depth of 40 cm in of the garbhagriha and the marking of its precise centre.
order to explore a shallow cut (909) that runs along Below this level, the silt/stone foundation layers
the western face of the Phase I structure. Three fills extend underneath the walls of the shrine. A small
were removed from this cut (943, 928 and 920). The cut trench was therefore cut by the present excavators
was shown to become shallower in the middle of the from the base of the walls downwards for a further
building, where an entrance to the structure might be 1.2 m until natural soil was encountered. Within this
expected (see also 617 in ‘W Area’ below). Excavation 1.2-m sequence, three further successive silt/boulder
of the fills revealed what appears to be the foundation sequences were revealed. The boulder stones in these
cut for the North Temple garbhagriha only 10 cm from layers are slightly larger than those in the layers above.
the wall, which is markedly different from 1171 on the Some minor variations in the composition and texture
south side (see Test Pit 3 above). It was confirmed that of the silt deposits were noted. These were given sepa-
the fill of the foundation cut (an un-numbered brick- rate context numbers which are marked on the section.
rubble matrix) was sealed by the layer cut by 909 and As can be seen in Fig. 4.49, this succession of layers
that the foundation cut cuts an ashy matrix below it links closely to that excavated to the south of the
(un-numbered). shrine in Test Pit 3 (see above). It is absolutely clear
that these layers represent a deliberate filling of the
large foundation cut 1171, which underlies the entire
The garbhagriha of the North Temple (GG Phase I structure of the North Temple. The western,
North) (Figs 4.11, 4.12 and 4.15) northern and southern limits of this cut were located,
allowing the size to be calculated and to demonstrate
In the 1999 season, investigation of the western end that it may have been rather unevenly shaped and is at
of the North Temple (Fig. 4.11) involved the excavation varying distances from the Phase I walls: about 80 cm
70   Derek Kennet

to the south, where it is most reliably defined, possi- A wide robber trench (377) was later dug along the
bly 10 cm to the west and possibly 160 cm to the north. entire length of wall 557 (Fig. 4.17). Due to later interven-
Once cut 1171 had been filled by silt/stone foundation tions, it is not possible to ascertain whether the robbing
sequences, the compact, level clay surface 927 was put took out the upper courses of 557 in the east and it is
down. This was then dusted with brick-jelly and the also not known from what level the robber trench was
walls of the Phase I shrine (970, 554) were built in stages dug. It seems to have been dug from the north side, as
directly onto this. Once these had been completed, the the southern edge was almost flush along the inner
platform wall was built, possibly in two distinct stages brick line of the wall. The robbing was very thorough;
(969, 555) and the gap between the outer platform walls all but two or three bricks of the inner and outer wall
and the shrine walls was filled with brick-bat rubble face and the hearting material were robbed. A few of the
(941) and layers of bricks (932, 911), thus completing bricks that were left in situ in the middle of the wall’s
the construction of the Phase I structure. length by the robbers were actually consolidated in their
position with lime mortar during the 1967 restoration,
probably because the restorers mistakenly believed that
Robbing of the northern wall of the North they were a free-standing wall (see ‘wall’ 502, which is
Temple (NWNT) (Fig. 4.17) visible in the background of Fig. 4.17). The precise loca-
tion of the original mandapa wall is known because the
Excavation along the northern wall of the North lowest course of mud mortar is preserved in the bottom
Temple demonstrated that after the temple had fallen of the robber cut, within which the shapes of the bricks
into disuse the northern wall of the mandapa was com- can still be seen.
pletely robbed out. The infilling of the robber trench follows broadly the
The first activity after the abandonment of the same pattern along its length: firstly, a probably deliber-
North Temple was the cutting of two almost circular ate back-filling with a compact silt and rubble (fills: 549,
pits (533 and 541), each measuring about 1.50 m wide, 505, 535 and 529) was followed by a loosely compacted
both of which have been partly destroyed by the later natural silting (fills: 327, 376). Where the robber trench
robbing cut. The remains of one is visible to the left cut through the softer fill of pit 541 some mixing of fills
side of the main robber cut (Fig. 4.17), where it partly occurred.
cuts into adjacent layers. Because the upper levels had To the west, a smaller robber trench (386) was cut
been removed by Yusuf ’s excavations, it is not known into wall 557 just north-east of Phase II wall 556. The
from which height these were cut. Both had nearly ver- fill is very similar to the upper, looser silt fills in the
tical edges rounding to flat bases and were dug down main robber trench 377; it therefore seems likely that
to approximately the same level. Both had very similar both robber trenches are contemporary. This robbing
fills, consisting of very soft, light grey, ashy silt con- left a small section of the original mandapa wall in situ,
taining large pieces of broken pottery. Across the base and this has been numbered 514.
of both pits, a thin layer of white organic fibres was The robber trench had been completely back-
preserved. Both pits were severely truncated by later filled before the next phase of activity took place in
robbing of wall 557, and their shape was only evident this area. A thin east-west trench (329, 396) was cut
as they were dug deeper than the lowest course of through the earlier robber trench fill along the same
bricks in the wall. alignment. This cut ran from the north edge of the
In the size, shape and nature of their fills, these remaining stump of wall 557, eastwards to brick pedes-
pits are very similar to pit 369 located to the south of the tal 375. The interpretation of this cut is problematic. It
South Temple. Both pits cut into wall 557. The bricks of is most likely to have been cut by Yusuf or by the 1967
the wall were cut to maintain the circular shape of the restorers hoping to locate the missing northern wall of
pits, indicating that the pit diggers were not interested the mandapa.
in robbing bricks. One slightly problematic aspect is One feature that is not understood is the mysteri-
that these pits cut into wall 557, although it is certain ous square brick pedestal 375, which stands exactly
that the wall was still visible on the surface because it within the robber cut and which is visible in the fore-
was later robbed out. The pits must therefore have been ground of Fig. 4.17. It is properly faced on all four sides
deliberately cut into a wall, but it is not clear why this and is clearly therefore not simply a remaining frag-
should have been done. ment of the robbed mandapa wall. It may be related
to the foundations of much later post-Medieval build-
 Chapter 4 The Brick Temples in Trench A    71

ings in the overlying layers. There is lime mortar in the to the south. After removal, this feature was excavated
upper four courses of this feature, indicating that it in two segments in Test Pits 3 and 10 (see above). At
was partly consolidated by the 1967 restorers. the junction between the south wall of the mandapa
of the North Temple and the garbhagriha, cut 339 had
been partially excavated in 1998, removing fill 338
The area west of the temples (W Area) (Fig. and leaving further probable brick rubble matrix 640.
4.48) Although it is impossible to be certain, it seems pos-
sible that this was the foundation cut for the mandapa.
After removing some superficial silt layers, several
archaeological layers and features were revealed in
the area to the immediate west of the North Temple. The southern wall of the North Temple (S
Layer 612/618/644 is an irregular deposit stretching Wall of N)
along the west section. At one time in the past, it had
been a muddy surface as a footprint was preserved in In this area, two cuts were identified relating to the
it. Removal revealed a very hard surface 657 into which south wall of the mandapa: 585 and 569, the former
two post-holes (656, 660) were cut. The surface appears being on the south-east corner of the structure and
to have been contemporary with a probable robbed- the other, larger cut running along the central part of
out stone wall that was originally contained within cut the wall. This cut contained two fills; a silty-clay loam
617, but it also sealed part of the fill of the cut (616), (573), evidently backfill/in-wash material, and a rubble
though the remaining stones from the wall (903) jutted matrix (577), probably collapse from the wall’s heart-
up through it. ‘Wall’ is probably the wrong description ing. Cut 585 contained only one fill (584). Both were
for 616/903/617 as the original form is unknown, it may cut down to the bottom course of bricks in the wall and
have been the stone base of a fence or some similar had thereby erased any evidence of the original foun-
boundary. Slightly S-shaped, it ran in a north-south dation cut except for a thin yellowish brown silty clay
direction to the west of the end of the Phase I/II walls of lying directly beneath the lowest course of bricks. In
the North Temple and south towards the possible large the area between the two cuts, however, a narrow cut,
stone enclosure wall excavated in the South Area (see filled with brick rubble, was visible.
below) (Fig. 4.48). It is possible that the two structures
are contemporary and related. Interestingly, the cut
became markedly shallower at just the point in front The EEST (Figs 4.30 and 4.45)
of the west end of the garbhagriha where an entrance
way would have been expected. It is also notable that During the 1998 season, cleaning and excavation of a
the stones of 903 on either side of this shallow section number of features were undertaken at the eastern end
are very deeply set compared to the others in the pos- of the South Temple, such as the foundation trench
sible wall, which were flat-bottomed. This might rep- (347) of wall 352, and a few post-holes and cuts. This
resent a gateway or passageway through the boundary was followed in the 1999 season by the excavation of
wall, if that is what this is. The northernmost stone had Test Pit 8 in this area, which has already been reported
a V-shaped socket cut in it. above.
Cut 681, aligned east-west, cut through the fill of The sequence in this area is key to the interpreta-
the boundary wall. It is assumed that this was a later tion of Phases III and IV of the South Temple. Although
burial of some sort as a few fragments of human skull, the broader picture is fairly clear, there are still some
although no other bones, were found in the fill (680). unresolved stratigraphic relationships and interpre-
Around the west, north and south sides of the North tations. Essentially, over a period of time, four major
Temple garbhagriha, a peculiar square C-shaped cut walls were added to the eastern end of the temple: 344,
was observed, two sections of which were excavated 345, 352, and 359 adjoining the already existing Phase
(909, 939). Layer 904 may have been a fill of this cut, II walls 1240 and 1247 (Fig. 4.30).
but the first proper fill was 920, a brick rubble layer The uncertainty revolves around the precise order
concentrated particularly on the west side of the garb- in which these four walls were built. This is indicated,
hagriha. Very well structured, it was probably laid to to some degree, by their stratigraphic relationship with
provide a firm footing for access into the garbhagriha, a compact silt surface (346) that covers the interior of
although the number and quality of bricks peters out the eastern end of the temple. Walls 1240, 1247 and 344
72   Derek Kennet

are all abutted by surface 346 and are therefore obvi- alignment (354) that once continued to the north. This
ously earlier than it, whilst walls 345, 352 and 359 are may indicate that the South Temple was later incorpo-
later as their foundation trenches can be seen cutting rated into a larger architectural complex that extends
through the surface. The foundation trench for wall 352 beyond the limits of the trench. There is a small patch
was excavated and very clearly cuts through this floor. of a brick-bat and silt floor surface (353) abutting wall
Cuts 521 and 527 of walls 345 and 359 respectively were 352 to the east, indicating the level of at least one
not, however, excavated and it is not therefore pos- phase of later use. Finally, a series of large flat stones
sible to be absolutely certain that they are foundation was recorded to the east of wall 352, where they disap-
cuts. They are very narrow (4 cm wide) and it is pos- pear beyond the limits of the trench.
sible that they might have been cut to hold some sort All of these features are clearly much later than
of wooden facing against the walls, in which case the Phases I–IV and give a limited and unfortunately
walls might be earlier than the surface. Given the fact rather fragmentary insight into the changing style of
that walls 359 and 352 are inter-bonded, albeit crudely, construction and manner of use of the temple in its
and the three walls obviously form part of single exten- latest phases.
sion of the temple towards the east, this seems quite
unlikely.
The sequence is thus reasonably certain; walls The area to the south of the temples (South
345, 359 and 352 were added to the east of the temple to Area) (Fig. 4.48)
lengthen the structure. But wall 344 could have been
built before, at the same time as or later than these A large area was cleared to the south of the South
walls and it is impossible to establish the order with Temple as there were some suggestions that this area
certainty. It has been included as a separate phase, preserved information relating to the later use of the
Phase III, because it does not seem to make sense any- temples and their surrounding area which had not
where else. been disturbed by Yusuf ’s excavations. The area is an
Three post-holes were then dug in this area. Two irregular shape, measuring between 6 and 8 m north-
are almost certainly contemporary (523, 525) as they south and 9 m east-west. Unfortunately, on excavation,
are located symmetrically on either side of the wide it was found that the area had in fact been subject to
entrance to the mandapa. The posts they contained Yusuf ’s excavations and probably also the 1967 res-
must have been part of some sort of entrance structure, toration, one or both of which involved the cutting
again probably made of wood. The third post-hole (358) of ‘wall-chasing’ trenches, drains and other features.
is not matched on the other side of the temple and may Nonetheless, excavation was continued in the hope
be related to something quite different. It is cut by the that some useful information might be retrieved from
foundation cut for wall 352 (347). the heavily truncated stratigraphy.
In addition, a number of apparently later features The deposits in the south-east consisted of a series of
in the area, mostly ‘islands’ of stratigraphy, now stand laminated water-sorted silts and coarse sands (592, 605,
in isolation since they were revealed by Yusuf ’s exca- 635, 678 and 670) and a number of loose rubble tumble
vations. It is therefore impossible to place them in a deposits (613, 619 and 663). These all existed east of
sequence or to link them together. They do, however, a linear flat-bottomed cut, 692, which ran north-south
provide some useful insights into the latest phases of along the entire length of this area. The slope-wash
the temple’s life. They include wall 360, which consists and rubble tumble deposits mentioned above all post-
of some very large broken carved stone architectural date this feature. It should be noted, however, that
fragments that have been placed on a layer of silt on whilst this cut was clear in the north, its southern limit
top of wall 352 (Fig. 4.56), apparently as part of a crude is obscure and difficult to define and was therefore not
rebuilding of the wall. There is a small rectangle of a fully excavated. It is thought that this cut may repre-
solid brick wall (349) abutting the eastern end of wall sent an old section line or trench edge of Yusuf ’s exca-
345, the purpose of which is unclear. It is also unclear vations. It effectively divides the area into two sections
whether it was matched by a similar wall against the to the east and west, each with a distinct depositional
eastern face of wall 359 on the south side – although it history.
seems likely that it was – as this is outside the limits The deposits to the east were markedly different,
of the trench excavated by Yusuf. Wall 349 is likewise consisting of a number of architectural and negative
abutted to the east by a truncated silt and brickbat features. The principal features of this area were two
 Chapter 4 The Brick Temples in Trench A    73

parallel walls 691 and 699/578 (Fig. 4.47) and a ditch The abandonment of these features is marked
(688) with a re-cut (671) running parallel that may rep- by extensive pitting in this area. Two large round pits
resent the line of another wall (Fig. 4.48). These fea- 628 and 631, filled with dumps rich in pottery sherds
tures all semi-enclose an area to the south of the South (fills 629 and 632 respectively), were cut into the top
Temple, although they were not visible in the west of the ashy fill (672) of ditch 671. Two amorphous pits
of the area, possibly because they were removed by were also cut into this ditch (623 and 669). Two small
Yusuf ’s excavations. pottery-filled pits (588 and 587) were cut into the top of
The two walls are of a very different nature; 691 wall 578, as was a rectilinear pit 607. The fills of these
consists of a single line of large semi-dressed igneous pits contain Early Medieval grey wares (see Chapter
boulders set on their sides or ends. It starts 2.1 m west 7: ‘Thick Grey wares’) and provide some indication of
of the south-west corner of the South Temple and runs a possible terminus ante quem for the activity in this
in a discontinuous line for 8.7 m roughly south before area.
returning eastward for 4.6 m. It appears to have a foun- It should be noted that some difficulty was expe-
dation cut (673), although as this was not excavated, it rienced in understanding this complex and disturbed
remains unverified. area, particularly in the south. Many of the early fea-
Wall 699, which was severely robbed by cut 577, tures were not excavated or were only semi-excavated
runs approximately 0.1 m east of wall 691. It starts in and were interpreted through a section which was cut
the south-west corner of wall 691 and runs for 8.7 m, it in the south-west corner of the area. No relationship
then returns east to abut the south-west corner of the could be established between walls 691 and 699 and
South Temple. It consists of a single course of broken the interpretation of cut 692 as the edge of Yusuf ’s 1937
bricks and stones, which are presumably the founda- excavation trench edge remains tentative.
tions of the original wall. This wall was left unexca- The general sequence of events can, however, be
vated. No relationship could be observed between the given as follows with a fair degree of certainty:
two walls due to disturbance by robbing activity, which An enclosure wall, either 691 or 699, whichever was
appears to have removed the original foundation cut of earlier, was constructed to enclose an area to the south
wall (699), if one ever existed. of the South Temple. This was consolidated or replaced
Another possible wall (578) was recorded superim- by the addition of another wall (691 or 699). A ditch (688)
posed on the alignment of wall 699 above the robber- was subsequently added on the outside of this enclosure.
cut fills 591 and 608. This was crudely constructed This was later re-cut (671).
from sub-angular stones laid randomly and surviving Wall 699 was robbed by cut 577 sometime in this
to two courses in height. This is either a wall or the sequence, and possibly replaced by wall 578. The
uppermost fill of the robbing cut 577 of wall 699. area then went out of use as an enclosure and this is
To the west and south of the large boulder wall 691 marked by a sequence of pottery-filled pits. A large
was a shallow V-shaped ditch (688), 1 m wide by 0.2 m cut (692), possibly from the 1937 excavations, then
deep. Its primary fill was a thin calcareous wash (679), truncated most of the deposits in the east of the area,
possibly resulting from the erosion of a plastered wall which was then subsequently filled by a series of slope
somewhere close by to the south. The ditch’s other fills washes and rubble tumble deposits.
consisted of loose rubble (675 and 663). This ditch was Table 4.4 lists and describes the 36 principal walls
only partially excavated. It was re-cut by ditch 671. It that were defined in the description and excavation
is thought that these ditches are later than boulder of the two temples in Trench A. The location of these
wall 691, as ditch 688 clearly cuts the foundation cut walls can be seen in Fig. 4.7 and Fig. 4.30 and other
of 691. figures where stated.
All of the above features, with the possible exception
of wall 578, are presumed to relate to a late period of use
of the South Temple.
74   Derek Kennet

Fig. 4.56: View of wall 360, built from re-used stone architectural fragments above wall 352 at the east end of the South Temple from the
north-west.

Fig. 4.57: Drawing of the north face of the north wall of the South Temple with an interpretive overlay.
 Chapter 4 The Brick Temples in Trench A    75

Table 4.4: Detailed description of the principal walls in Trench A.

Wall Number Temple/Area Description Phase Brick size

344 S Temple Phase III east wall of the South Temple. South III 6/7.5 × ? × ? cm

This is a late addition blocking the opening in the east end of the Phase II South Temple. Only four courses survive; it is heavily restored
and re-mortared and there may have been some reconstruction – it is now impossible to tell. The construction is crude, similar to 352,
using mostly brick bats. The wall is 71 cm thick with c. 42 cm of rubble fill. Height as standing is 29 cm.

345 A:S Temple Phase IV extension to east end of north wall of South IV Unclear
South Temple.

This is an eastern extension to the north side of the South Temple, equivalent to 359 on the south side. This wall extends the line of
560/561 to the east by 183 cm at the lowest course. For most of the wall’s length, only the lowest three courses survive; at the west end,
11 courses are preserved, but above course five, the bricks are inter-bonded with wall 560. This demonstrates that the eastern extension
was complete before the upper part of wall 561 was rebuilt. The most westerly five bricks of the lowest course are offset by about 5 cm in
a very irregular fashion; the reason for this is not clear. The jointing is quite crude and the wall is made almost entirely of slightly weath-
ered brick bats. The wall is between 89 and 92 cm wide with 55 cm of densely packed brick rubble in its core.

349 A:S Temple Short wall abutting wall 345 to east of South – uncertain phase Unclear
South Temple (north side).

The width of this wall is 98 cm and the length 38 cm. It consists of four courses of brickwork which have been built on to the eastern end of
wall 345. It is built of solid brick bats with no rubble fill.

352 A:S Temple Phase IV and V east wall of South Temple. South IV South V Variable

Eastern wall of Phases IV and V of the South Temple. The wall is robbed out along about 1.4 m at the north end and only about 2.4 m
remain. It is overlain by a later stone alignment/wall 360. The wall is 70 cm in width with about 38 cm of brick bats and silt as a core.
The wall has a stone foundation with a compact silt matrix (317) (Fig. 4.45). Only seven courses of the wall remain; the lowest five are
foundation courses not intended for view and the bricks are variable in size and type and have crude jointing. The upper two courses are
offset by about 3 cm and the silt jointing is a little finer. Still the brick types are quite variable and most of them appear to be brick bats
rather than complete bricks. The largest are 24 cm in length.

354 A:S Temple Small, crude brick alignment abutting east end of South – uncertain phase Unclear
wall 349.

Small alignment of brick bats abutting wall 349 at the eastern end of the South Temple. It is not clear if this is a wall or the remnant of a
brick surface.

359 A:S Temple Phase IV extension to east of south wall of South South IV 7 × 24 × 36 cm
Temple.

Wall 560 was extended 1.56 m on the south side by the construction of this wall, which abuts 560 up to the 14th course, above which level
it is inter-bonded (Fig. 4.42). The bricks of this wall are orange fired and measure 7 × 24/25 × 35/37 cm. This is the same size as the bricks
of the inter-bonded courses above except that they are thinner (c. 6 cm). The jointing of these courses is also cruder than the inter-bonded
courses above. The wall is 84 cm wide with about 40 cm of brick-bat and silt fill. The wall is inter-bonded with wall 252, which is of a slightly
cruder construction.

360 A:S Temple Crude stone wall on top of wall 352, east of South South IV South V N/A
Temple.

This wall consists of re-used carved stone architectural fragments lying in a rough alignment along the top of wall 352 (Fig. 4.56). It appears
to be a late rebuild of wall 351 and may be related to the very latest phase of use of the temple. It is separated from wall 352 by a 5-cm
deposit of silt, suggesting that the construction techniques of the period when the wall was built were extremely crude.
76   Derek Kennet

Wall Number Temple/Area Description Phase Brick size

383 A: N Temple Thin Phase V partition wall in west of North V unclear

mandapa, North Temple (equals 544).

This wall consists of 10 or 11 brick bats crudely set on end into a thick, gravely mortar running across the mandapa of the North Temple,
1.64 m from the western end. The brick bats are of different heights but their bases are at roughly the same level. The wall continues to
the south of wall 558 in the southern aisle of the mandapa as wall 544 where it turns and abuts the interior face of wall 557. The brick
bats are 5.5/6.5 cm thick. The mortar must originate from the 1967 restoration of the temple because Yusuf clearly states that no
mortar was used in the temple construction (Yusuf 1938: 4). But it is clear from his photograph (Fig. 4.4) that this structure existed and
that it was considerably larger when it was first excavated in 1937. What remains now appears to be the 1967 restoration of a rather
fragile structure excavated by Yusuf that acted as a retaining wall for a raised platform in the western part of the mandapa.

509 A: TP5 Single course thickening of interior of south wall of South V 7 × 14 × 23 cm

South Temple.

Wall constructed against the interior face of the south wall of the South Temple as part of the Phase V alterations to the temple also
involving walls 562 and 528. This is a one-brick wide wall constructed in a very crude fashion. Only about 17 bricks remain, not all of
which are complete. There are four courses held together with a crude mud mortar. The face is irregular and badly laid. The wall is sepa-
rated from the main temple wall by 7 cm of silt fill. Equals 562.

510 A: TP5 Interior wall in South Temple mandapa. South V Unclear

Inside the mandapa of the South Temple, two badly preserved walls (510 and 511) appear to have originally subdivided the mandapa into
three sections as part of the Phase V alterations. This wall is of a very crude construction, being built of brick bats on a rough, shallow foun-
dation of small stones. The wall has been disturbed by later cutting from an unknown height and possibly also by Yusuf’s excavations. Only
one course of bricks now remains, none of which are complete. The wall is 50 cm wide.

511 A: TP5 Interior wall in South Temple mandapa. South V Varies

Inside the mandapa of the South Temple, two badly preserved walls (510 and 511) appear to have originally subdivided the area into
three sections as part of Phase V changes. Wall 511 is of a very crude construction, being built of brick bats on a rough, shallow founda-
tion of small stones. The wall has been disturbed by later cutting from an unknown height and possibly also by Yusuf’s excavations. Only
two courses of bricks now remain, none of which are complete. The bricks are mixed, some being orange fired (7 × 22 × 38 cm) and some
purple fired (7.5 × 20 × 33 cm). The wall is 50 cm wide.

514 A:N Temple Remaining portion of original northern wall North III Varies
abutting shrine, North Temple.

This appears to be the original Phase III wall of the mandapa of the North Temple (557, see below). It is the only remaining section of this
wall on the north side of the temple and is preserved between the edge of a robbing cut and a pit dug through the structure from above.
The base of the wall is 107 cm wide thinning to 96 cm wide above the third course and 75 cm wide above the mouldings (Fig. 4.19). The
wall consists of two brick faces with a 40-cm hearting of brick bats and silt. The northern face of the wall shows that it had adhisthana
mouldings beginning with an offset in the eighth course, but these were later hidden by the addition of an outer face (wall 559/1213).
Only the lowest three courses of the mouldings are preserved (Fig. 4.18B). The lowest courses of the wall are built of bricks 40/41 cm
long, whilst the upper courses use variable brick sizes. The interior face of the wall is offset at the third course by 3 cm and at the fourth
course by 9 cm. There is another very small offset of about 2 cm at the eighth course.

528 A: TP5 Thickening to interior face of north wall, South South V Unclear
Temple.

Wall constructed against the interior face of the north wall of the South Temple as part of the Phase V alterations to the temple also
involving walls 562 and 509. This is a one-brick wide wall constructed in a very crude fashion from brick bats held together with a crude
mud mortar on a thin foundation of small stones.

544 A: N Temple Thin Phase V partition wall in south aisle of North V Unclear
mandapa, North Temple (equals 383).

See description of 383 above.

 Chapter 4 The Brick Temples in Trench A    77

Wall Number Temple/Area Description Phase Brick size

554 A:N Temple Phase I wall, North Temple. North I 7 × 25 × 40/41 &
33 cm

This is the wall of the shrine of Phase I of the North Temple. It is a solid brick wall, two-bricks wide at its base and one in the upper
courses, with no hearting. The wall still stands to 19 or 20 courses or 160 cm high. The bricks are quite finely jointed with mud mortar,
although only very little of this is exposed. The structure is quite complex (Fig. 4.12). The lowest two courses of the wall are only visible
on the interior, on the exterior they cannot be seen because rubble fill 941 underlies the third course. This indicates that the construc-
tion of wall 554 and the deposition of 941 were simultaneous. The lowest two courses of 554 have therefore been labelled 970 (see
below) in order to differentiate them stratigraphically. The width of 970 is not known as the outer face is not visible. Wall 554 is two-
bricks, or 50 cm wide at its base, although the lowest (or 3rd on the interior) course is inset by about 3 cm on the exterior. Above the sixth
(or eighth) course, the wall thins on the interior to a single brick thickness (about 30 cm) for the rest of its height, creating a 25-cm-wide
ledge on the interior. Wall 968 (see below) was built into this space at a much later date to thicken the wall to 50 cm. The 7th and 8th (or
9th and 10th including 970) courses of the exterior are inset by 5 cm on the north face and 2 cm on the east face, creating what appears
to have been a simple rectangular decorative inset moulding that probably ran around the whole shrine, although it is visible only in
a limited area where it has been exposed by pit 341. The bricks are quite soft-fired and pinky-yellow or red in colour and vary in size,
most measuring 7 × 25 × 40/41 cm, although some smaller examples measuring 7 × 25 × 33 cm were also noted. The coursing consists
almost entirely of stretchers on the interior, which is the only face that is fully visible. The whole square structure measures 2.04 m N-S
and 2.08 m E-W on the exterior, and the space in the centre measures 1.10 m N-S and 1.06 m E-W at the base. There is no discernible sign
of weathering on the exterior wall surface that would demonstrate that the structure was exposed to the elements for any length of time.

555 A:N Temple Phase I platform, North Temple. North I 7 × 25 × 40.5 cm

Technically, this number refers only to the upper three brick courses of the platform of the Phase I shrine of the North Temple, although
the number has also sometimes been used in general descriptions to refer more generally to the whole Phase I square platform including
the lower part of the wall 969 (see below) and fills 941, 932 and 911 (see Fig. 4.12 and discussion). Wall 555 is exposed only in limited
places. The bricks are soft and have a reddish colour and are jointed with a fine mud mortar less than 1 cm thick (Fig. 4.16). The top
course is a levelling course in which the bricks are between 4.5 and 6 cm thick, suggesting that they have been deliberately thinned to
level the surface of the platform. The wall retains a rubble fill that abuts wall 554, although it is not clear where the upper level of this
fill originally was as it has been disturbed by the construction of wall 556 during Phase II. The platform measures 4.22 m N-S and 4.21 m
E-W and is about 56 cm high, the upper 26 cm of which is technically wall 555.

556 A:N Temple Phase II structure with offsets, North Temple. North II 6.5 × 25 × 40 cm

Wall 556 refers to the Phase II triratha structure in the North Temple that is constructed upon the Phase I brick platform 555. The struc-
ture is square and about 20 cm smaller than platform 555, but the middle of each face projects out 10 cm along a 2-m section to the edge
of platform 555 (Fig. 4.12). At least 13 courses of the outer structure are preserved, so that it stands 1.10 m above the top of platform
555. The wall consists of an outer face of brickwork, which is filled with a brick-bat and silt hearting. The lowest five courses of the wall
are very finely jointed (less than 5 mm) using good-quality orange or orange-red bricks measuring 6.5 × 25 × 40 cm, although in the fifth
or sixth course, the bricks are thicker (7 or 7.5 cm) but otherwise the same size. Above this course, 6 × 20 × 33 cm purple fired bricks
with surface traces of vitrification are used and are part of a much later rebuild numbered 971 (see below). The bricks above this level are
badly eroded, suggesting that they have been exposed for longer than the lower courses, perhaps since Yusuf’s excavations. The cours-
ing consists mostly of stretchers with occasional shorter bricks (Fig. 4.16). There are small offsets in the wall in some places, without
apparent symmetry: there is a 3.5-cm offset above the third course at the south end of the west side and a 4 cm offset above the ninth
course at the west end of the north side. The eastern side of the structure has been cut into above the sixth course for the addition of
wall 557 (Phase III). This is a complex inter-bonding, that is discussed in more detail above.

557 A:N Temple Phase III mandapa wall, North Temple. North III 6.5 × 25 × 40 cm

This is the Phase III mandapa wall of the North Temple. It forms a large tapered rectangular structure that abuts wall 556 to the west.
The wall is 120 cm thick along its south, east and north sides, but only 94 cm thick along the western side. The wall consists of two
brick faces filled with a hearting of brick bats and silt. The bricks are orange and are finely jointed with silt. The wall has been almost
completely robbed out along its northern side, the eastern side is preserved but has been built over during reconstruction and much of
the south side has been robbed out or has been built over by reconstruction; the upper courses of the north face of this part of the wall
consist of smaller-sized bricks jointed with mortar, almost certainly resulting from the 1967 restoration. It is not clear, however,
if there was a later rebuild on top of this part of the wall before that time. This wall is the same as wall 514 (see above), which has a
different number only because it has been isolated by later robbing. There is a 3-cm offset along the interior face above the seventh
course which appears to run around the whole building. Above this offset, the jointing is noticeably cruder, especially at the west end,
suggesting perhaps that the structure was rebuilt from this level. This is just above the later floor level (1010). Adhisthana mouldings are
preserved along a 24-cm section of the west face (and also on the west face of wall 514), but probably once extended around the entire
mandapa. The mouldings are of a very similar design to those on wall 560 of the South Temple (see Fig. 4.18A).
78   Derek Kennet

Wall Number Temple/Area Description Phase Brick size

558 A:N Temple Phase IV interior wall, North Temple. North IV 7 × 26 × 42 cm

This wall forms a sub-rectangular structure inside the mandapa of the North Temple and is part of the Phase IV alterations. The wall is
51–54 cm wide and consists of two rows of bricks with no hearting. The low-fired, crumbly orange bricks are quite crudely jointed with silt
mortar (much cruder than wall 556). The wall sits on a compact clay foundation on the floor, which abuts wall 557. Wall 558 abuts wall 556
and 557 at the west end. At the west end of the south side of the wall, a 40-cm-long grey stone is built into the wall.

559 A:N Temple Rebuilding of outer face of wall 514. North III? Unclear

This wall (also numbered 1213) consists of three courses of soft-fired red bricks, finely jointed with a mud mortar, that have been added
to the western face of wall 514. The bricks were clearly part of an adhisthana moulding, as they are offset and the brick in the middle
course is chamfered (Fig. 4.19). The wall appears to have no substantial foundations, and the lowest brick sits on a deposit of silt that
abuts the face of wall 519. The base of this wall is level with the eighth course of wall 514. This is a rather strange situation and it is
difficult to understand, partly because only about 50 cm of this wall is preserved, and partly because the stratigraphic relationships
and related deposits have been removed by previous excavations. This wall appears to represent a thickening of wall 514 by about 30 cm
towards the west that preserves the mouldings. The chamfered course in the mouldings is at exactly the same level as the same course
in wall 514. No similar thickening was observed on the opposite side of the temple.

560 A:S Temple Phase II wall of South Temple. South II 6/7 × 22/24 × 43 cm

This is the upper part of the main wall of Phase IV of the South Temple. It is partly constructed upon the Phase II walls 1240 and 1247
(see below and the section in Fig. 4.33), except in the eastern extension, where it would have been constructed on top of walls 345,
359 and 352. This wall was originally numbered 351 (south side) and 361 (north side), and in the site documentation before Test Pit 5
was excavated, walls 560, 1224 and 1240 were considered to be the same. The wall is about 101 cm thick and has a 50–55-cm thick
rubble hearting consisting of brick bats and silt. About five brick courses of wall 560 survive above wall 1247, although this may vary
across the structure, with adhisthana mouldings beginning with a chamfered brick in the second course (Fig. 4.18C–D). The bricks of the
moulding courses are slightly thicker than in the rest of the wall (7 × 22 × 40 cm), and appear to be slightly higher fired. See also eleva-
tions of the wall from different directions (Fig. 4.46). The jointing is cruder than 1240/1247 and the coursing consists mostly of headers.
It is assumed that the top of the vertical abutment at the eastern end of the south wall marks the distinction between walls 560 and
1240/1247 (Fig. 4.42), but this is far from certain and it is impossible to verify because not all faces of the wall are exposed, and the
difference between the two walls is not always apparent. It is also more than likely that the construction of wall 560 took place from slightly
different levels across the trench and from different levels on the interior and exterior face of wall 1240/1247. That is to say that the top of
1240/1247 may not have been level when 560 was built upon it.

561 A:S Temple Late rebuilding of north wall, South Temple. South V? 6 × 16 × 27 cm

This represents a late rebuild of wall 560 along the north side of the South Temple. The reconstruction took place from the 12th or
13th brick course above the lowest course of wall 1247 (Fig. 4.57). Above this level, the bricks are hard-fired and purple coloured with
frequent signs of vitrification on the surface. These bricks are smaller than those used in the earlier walls. The jointing is quite fine.
The lowest two courses are a little larger and these may represent an earlier rebuild of which no other evidence survives. The coursing
pattern is HSSSSSHSSSSSH or HSSSSHSSSSH. There is a 2-cm offset in the sixth course of 561 that runs along the whole wall. Wall 561
has removed the upper courses of the adhisthana mouldings of wall 560 below, leaving only the first chamfered course, but no attempt has
been made to re-create the mouldings. The type of bricks used in wall 561 is similar to those used in apparently later rebuilds of parts of the
garbhagriha structure of the North Temple (see wall 971 below).

562 A:S Temple Phase V thickening of inner face of south wall, South V 7 × 18 × 35 cm
South Temple.

This wall consists of four courses of bricks built against the inner face of wall 560 and 359 on the south side of the South Temple
mandapa. The wall is only one-brick wide and is separated from wall 560 by a 5-cm-thick silt deposit. One of the bricks is intact enough
to measure and is 7 × 18 × 35 cm. The upper two courses of this wall are inter-bonded with wall 359, suggesting that some reconstruction
of the temple superstructure occurred during the Phase V alterations. The base of the wall appears to rest on silt, although this was not
excavated and it is likely that there is a shallow foundation deposit below as there is for wall 509. This wall is almost certainly the same
as 509, but the two parts have been separated by the destruction of the middle part of the south wall.

578 A:South Area Wall foundation in cut 577. N/A

691 A:South Area Wall of large irregular igneous boulders. N/A

 Chapter 4 The Brick Temples in Trench A    79

Wall Number Temple/Area Description Phase Brick size

699 A:South Area Broken-brick wall foundation cut by 577. N/A

Together, these three deposits (578, 691, 699) represent a now completely robbed stone enclosure wall that appears to have abutted the
south-west corner of the South Temple and then turned sharply towards the south, where it disappears out of the excavated area (Fig. 4.48).
Deposit 578 consists of a very crude alignment of angular igneous rocks that were found within robber cut 577. It seems most likely that
they represent some of the stones from the original wall that were left behind by the robbing and have been buried within the fill of the
robber trench. What appears to have been the foundation of the original wall is represented by 699, a compact deposit of brick bats
and stones at the base of the robber trench beneath 578 (Fig. 4.47). To the immediate west of the robber trench is another crude and
discontinuous alignment of large stones (691) that may represent the course of a parallel wall or which may have been robbed from wall
578/699 but which were left at the site. There are two further cuts (671 and 688) close by and on the same alignment that might be the
foundation cuts or robber cuts of related walls that have now completely disappeared. That there was a wall (or walls) here is in no doubt as
the robbing and the foundations clearly show its course. Whether it was brick or stone is not certain, but the large amount of stone that is
found in deposits 578 and 691 suggests that the wall was constructed of large blocks of igneous rock.

968 A:GG North Internal thickening of wall 554 of North – uncertain phase 7.5/8 × 18/19 ×
garbhagriha, North Temple. 25/26 cm

This wall is an internal thickening of the Phase I garbhagriha wall of the North Temple (554). It is constructed on the internal inset in
the eighth course wall 554 and continues to the height to which the garbhagriha is preserved today (Fig. 4.12). Ten courses of this wall
are preserved standing to a height of 78 cm. The wall is built of bricks which are crudely jointed with silt. The construction is not of good
quality, mostly stretchers are visible on the interior and they are not neatly arranged. The small brick size suggests a fairly late date of
construction through comparison with other bricks in the two temples.

969 A:GG North Possible earlier phase of wall 555, North Temple. North I 6.5/7 × 24/25 ×
39/40 cm

This wall is the lower four courses of the outer edge of the Phase I platform of the North Temple 555 (Fig. 4.12). It forms a square exactly
the same dimensions as 555 (4.22 m N-S and 4.21 m E-W). The build is exactly the same as 555 and it may have been built at the same
time. It is built of bricks that are finely jointed with silt mortar. The top of the wall is level with the possible brick surface 932 with which
one brick is inter-bonded. It is not clear whether this wall is separate from 555, in which case it represents an earlier, lower, platform of
the Phase I shrine, or whether it is simply a stage in the construction of the Phase I platform.

970 A:GG North First two courses of wall 554, North Temple. North I Unclear

This wall is the lowest two-brick courses of wall 554. They are thinner than the upper courses of 554, although it is impossible to say
by how much as the outer face of 970 is not visible. The wall was defined on a stratigraphic technicality because these bricks are clearly
abutted by fill 941 upon which the upper courses of wall 554 rest. Whilst in reality the whole Phase I platform structure was probably
constructed together, technically these two courses are an earlier part of the construction (see Fig. 4.12). Equals 554.

971 A:GG North Rebuilds to wall 556, North Temple using purple North – uncertain phase 6.5 × 18/24 × 35/36 cm
bricks.

Above the seventh or eighth course of the Phase II wall of the North Temple (556), the bricks change from softer-fired, larger orange
bricks to smaller, harder-fired purple bricks similar in aspect but of a different size to those used in the construction of wall 561 on the
South Temple (Fig. 4.27). Robbing and erosion of the structure make it impossible to say anything about the nature of the construction of
which these bricks form a part. They are found on the north, south and west sides of the Phase II structure but not on the east face. The
bricks are badly weathered, suggesting that they have been exposed to the elements for longer than the lower part of the structure.

1213 See 559

1224 A:S Temple Foundation wall for Phase I of South Temple. South I 6.5/7 × 24 × 42/43
cm

Wall 1224 is the lowest five or six courses of the Phase I walls of the South Temple (Fig. 4.39). The wall is 70 cm wide and is made of solid
brick, three bricks wide, with no hearting. The bricks are orange or yellow in colour and are quite crude and soft-fired. They are quite
crudely jointed with 2-cm-thick mud mortar. Certainly, the jointing is cruder than walls 1240/1247 and 560. Mostly, the courses are
either completely headers or stretchers, but there is no apparent pattern in some places. In some places, the wall is 35 cm thinner than
walls 1240 and 1247 that are built on top of it.
80   Derek Kennet

Wall Number Temple/Area Description Phase Brick size

1240 A:S Temple South Temple Phase II wall – north side. South II 6.5 × 24 × 43/44 cm

Upon excavation, it became clear that the first four brick courses above foundation wall 1224 are built of a different masonry to the upper
part of the wall. These courses are also crudely inter-bonded with wall 1251 (Fig. 4.40). The bricks are soft-fired and orange yellow in
colour and measure 6.5 × 24 × 43/44 cm. The silt jointing is quite fine. The wall consists of two brick faces filled with brick-bat and silt
hearting. The interior face shows mostly stretchers. The wall is 80 cm wide and is constructed upon wall 1224 (see above and the section
in Fig. 4.39). The lowest course is a levelling course, using split or wedge-shaped bricks. It seems likely that the wall below the rebuild
evident on the outer face of the south wall must be 1240 (Fig. 4.42). The original adhisthana mouldings in this wall continue around the
corner of the eastern end of the Phase II temple, for a distance of about 111 cm, then there is a break of 41 cm before the entrance gap
that is now filled by wall 344. These mouldings have been partly concealed by the later construction of walls 345 and 359 that extended
the temple to the east during Phase IV. Equals 1247 on the south side.

1241/ A:S Temple Late rebuilding of south wall, South Temple. South V? 6.5 × 24 × 33/34 cm
1242

Wall 1241/1242 is the uppermost four to seven courses of the southern wall of the South Temple (Fig. 4.30). The two walls were clearly
once part of the same rebuild but they are now separated by cut 595. The bricks on these walls are small and higher fired than those
used in wall 560. The colour is slightly purple. The two walls appear to be on a slightly different alignment to wall 560 below, being a few
degrees to the west. There are no associated floor levels, but it seems quite likely that the upper rebuild of the north wall (561) is part of
the same rebuilding as these two walls.

1247 A:S Temple South Temple Phase II wall, south side. South II 6.5 × 24 × 43/44 cm

This is the same wall as 1240 but on the south side of the temple. Although no visible join between the two walls has been exposed, it is
clear that they must be part of the same wall. On this side of the temple the wall is directly overlain by wall 560.

1248 A:N Temple Late rebuild of mandapa wall 557, south side North – uncertain phase 7 × 25 × 42/43 cm
North Temple.

This is a late rebuild of the upper courses of wall 557 on the south side. Very little of this rebuild remains. The jointing is cruder than in the
lower levels, but the bricks are not the vitrified purple variety used in very later rebuilds. Up to three courses remain.

1251 A:S Temple Phase II wall of garbhagriha of South Temple. South II 6.5 × 24 × 43/44 cm

This wall was initially numbered 1240 until the join with 1240 was seen during excavation of TP5 (Figs 4.38 and 4.40). The wall is shown
to have better quality masonry than 1240 and to be inter-bonded with it, 1251 having been constructed first, leaving open teeth at the
ends into which bricks have been crudely slotted to bond 1240 to this wall. The jointing is very fine (<5 mm) and the bricks are the same
size as wall 1240 being laid as stretchers with only occasional headers. As can be seen in Fig. 4.38, the wall incorporates a few vitrified
bricks and some fragments of stone, most notably a large piece in the middle of the older entrance to the garbhagriha in wall 1240 that
this wall effectively closes. A number of cut bricks are also evident where this wall was fitted onto wall 1240 below it. It is assumed that
the wall encloses the whole of the garbhagriha, but each of the sections of wall abuts each other in the corners of the garbhagriha, at
least on the outer face. This probably reflects the method of construction.
J. Varaprasada Rao and Derek Kennet

Chapter 5 
The Excavation of Trenches B, C, D, E and F

Although the excavations around the two temples in From deposit 440, where the Period 2 levels begin, a
Trench A were the main focus of the project, a series lot of ash and evidence for fire and burning came to light
of smaller trenches was opened across the mound in in contexts 441, 442, 443, 444, 445, 446 and 447. A large
order to investigate the depth and nature of the stratig- pot was found covered with a thick, burnt mud coating
raphy and the chronology of occupation on different (441) together with a smaller pot (445), both of which
parts of the mound. The aim was to delimit the extent were associated with a black burnt deposit (443). A
of the Early Historic urban centre and to understand number of fragments of burnt clay were also recovered
the extent of Early Medieval occupation. These aims from the floor (450). All of these layers together suggest
were only partly fulfilled due to the difficulty of exca- that, for a period of time, this area was given over to
vating trenches in the centre of the modern town. In some sort of industrial process involving fire. Unfortu-
the event, five further trenches were excavated, with nately, no evidence was uncovered that gives any more
only one (Trench F) being located outside the ASI pro- precise indication of what this process might have
tected area at the south end of the site. Trench loca- been. At this point, the size of the trench was reduced
tions are shown in Fig. 4.1. This chapter consists of to 2 × 2 m in the south-east corner in order to speed up
reports of each of the excavations in each of these five the excavation.
trenches. Reference is made throughout to the context Below this ‘industrial’ activity, the nature of the
numbers, which are listed and described in Appendix deposits in this trench changes significantly. A sequence
I. For period dating and phasing, see Chapter 6. of 23 more or less horizontal floor levels was revealed,
beginning with floor 450 (Figs 5.3 and 5.4). Apart from
the uppermost two floors, 450 and 451, the rest of this

Trench B sequence is allocated to Period 1. All floor levels had a

gentle slope downwards towards the west. Most of the
(Figs 5.1–5.4)
floors are made of thick, semi-compact silt and clay, but
some of the lower floors are thinner and have traces of
Trench B was excavated during the 1998 and 1999
lime on them. Some are associated with pits and post-
fieldwork seasons. It is located to the south of Trench
holes, and some with hearths, but the limited size of
A (Fig. 4.1). The trench was originally 4 m × 4 m but was
the trench did not allow the investigation of build-
reduced in size during the excavation to 2 × 2 m.
ing styles or techniques or the nature of occupation
The corner of a stone structure (401) came to light
throughout this period. A few coins were recovered
immediately below the surface in the south-east corner
from these levels.
of the trench (Figs 5.2 and 5.4). Sixty cm below this, six
Below the sequence of Period 1 floors, undisturbed
stone pillar bases were uncovered cut into floor 413
natural soil (743) was reached at a depth of about 3.2
underneath the upper layers of dumping and pitting
m from the surface. Natural soil seems to slope down
activity (403, 404, 408, 409, 411, 425). Two of these
towards the south-west. The archaeological deposits
bases were outside the limits of the trench, the other
in Trench B are about 3.2 m in depth, with the Early
two are covered by wall 401.
Historic levels making up the lower 1.3 to 1.5 m of this.
Below this, a succession of uneven floors or surfaces
(417, 427, 429, 439) came to light associated with pits
filled with pottery and other rubbish as well as an ash
deposit (428) and thick dumping deposits such as 437,
438. A pit (432/433) associated with these levels had a
white plaster base. These deposits, and those above
them, can be allocated to Period 4. No Period 3 remains
were found in this trench.
https://doi.org/10.1515/9783110653540-005
82   J. Varaprasada Rao and Derek Kennet

Fig. 5.1: Trench B, view of the south and east sections of the trench at the end of excavation.

Fig. 5.2: Trench B showing wall 401 and related pillar bases to the west.
 Chapter 5  The Excavation of Trenches B, C, D, E and F   83

Fig. 5.3: Trench B, north-facing section.

Fig. 5.4: Trench B, west-facing section.

84   J. Varaprasada Rao and Derek Kennet

Trench C total amount of pottery collected. The most significant

(Fig. 5.5) find from this phase is a lead bulla.
Phase CIII: The thick soil layer of this phase is cut
In the 1996 season, a number of 5 × 5 m trenches were into by numerous pits and shallow troughs, which do
opened on the edge of the archaeological mound of not appear to be structural. The pottery is mixed, with
Paithan towards the extreme south of the ASI protected Thick Grey ware and other wares occurring in almost
area (Figs 4.1 and 4.2). This trench was excavated by J. equal amounts. The finds include beads, bangles and
Howell and A. Kumble in the 1996 and 1997 seasons. A kohl applicators, as well as ceramic pendant loom
summary report of the small finds and antiquities from weights and cylindrical counters.
this season is included in Chapter 10 of the present Phase CIV: This phase is the topsoil and recent
volume. It is impossible to incorporate the evidence intrusive activity on the site. The only structural activ-
from this trench into the periodisation proposed for ity in this phase is the ring-well found in trenches
the rest of the site due to the different methodology 130/450 and 135/450, which is the latest of the four ring-
that was used. The following summary is intended only wells found in the excavations. The predominant type
to give a sense of the nature of deposits in this part of of pottery is Thick Grey ware, which represents more
the site and to suggest how these may relate to the site than 30% of the sample. The finds from these deposits
periodisation. are probably mixed and contaminated, but the most
This area was chosen because it appeared to be free significant are a ceramic seal and one of the lead-
from the later Medieval structures and deposits that scrolled objects.
overlie the area to the north. At the edge of the mound, As can be seen, all of these layers would be allo-
two brick ring-wells were visible in the cliff above cated to Period 3 or later according to the present peri-
the river bank and it was decided to expose these odisation due to the presence of Thick Grey wares, but
by cutting back the section and getting a preview of the material from this trench was not included in the
the entire archaeological deposit down to the natural final analysis of pottery.
soil. The trench was located in a low depression A significant feature of this part of the site of
that appears to be enclosed within a low earthwork Paithan is the large number of circular brick or ter-
marking the location of a wall or rampart, probably of racotta structures, commonly referred to as ring-wells.
the Mughal period. At least eight of these features can be identified in the
The coordinates of the five trenches according to the eroding sections of the mound along the river bank.
site grid are as follows: 130/450, 135/450, 140/450, 145/450 Most are cut from relatively high levels on the mound
and 135/445. It is by these coordinates that they are and do not appear to be deep enough to reach down to
referred to in the description below. Context numbers the water table. This brings into question their inter-
1–41 were allocated during the 1996 season, and pretation as a device for drawing up drinking water.
context numbers 42–203, during the 1997 season (see Other interpretations that have been offered are that
below). these structures are either soak pits for dispersing
The sequence elucidated by the trench is summa- excess water, or silos for underground storage. The
rized below from the available excavation records: large number of ring-wells at Paithan have various
Phase CI: The earliest phase is sealed by a large soil designs, and the bricks or terracotta rings used in their
deposit, which includes a ditch and the material into construction are not of a uniform form or size.
which it is cut. In this phase, Thick Grey ware (Chapter In total, four ring-wells were discovered within the
7) is present, but in comparatively small quantities area excavated, two of which were exposed to their
compared to the wares that might be associated with entire depth, which was in excess of 2.1 m. They were
the Early Historical period, including Red Polished constructed in brick and bottomed out just below the
ware. The principal types of finds were ceramic ‘areca top of the natural soil. A third brick ring-well was found
nut’ beads, pendant loom weights, skin rubbers and within trenches 130/450 and 135/450 and was not fully
copper-alloy coins (Chapter 10). excavated. The fourth ring-well is constructed from ter-
Phase CII: This includes a phase of structural activ- racotta rings, only two of which survived. It was in a
ity overlying the deposits of the previous phase with a very fragmentary condition on the edge of the mound
drain and earthen floors. The proportion of Thick Grey and was not excavated. With the exception of these
ware begins to drop in this phase, down to 12% of the ring-wells, no structural evidence was found in these
trenches. In trench 135/445, a series of inter-cutting
 Chapter 5 The Excavation of Trenches B, C, D, E and F    85

Fig. 5.5: Trench C section (155/450, south-facing).

features and later disturbances such as tree roots and as resulting from dumping and extra-mural activities
animal burrows truncated all horizontal stratigraphy such as gardening, herding and well-digging probably
to a depth of 1.25 m. close to the city of Paithan but certainly outside it.
During the 1997 season, a second area was located An example of one of the trench sections drawn
at the extreme south end of the site next to the 1996 and excavated is shown in Fig. 5.5 (from Trench C,
excavations. Here, after the end of the 1996 season, 155/450). Within this sequence, which is fairly typical
students from the ASI Institute of Archaeology had of the whole area, layers 60 and above contained Thick
undertaken a training excavation, opening a trench Grey ware and would therefore be allocated to Period 3
35 × 5 m. In 1997, the same trench was cleaned and or later under the present periodisation system. Below
excavated down to natural across a 2-m-wide section this, the layers can be allocated to Period 2, with the
in each of the seven sectors opened. This gave a 35-m- exception of the lowest layer (93), which is probably
long section across this area of the site in which the dated to Period 1.
sequence could be examined. In order to gain more
information about the cultural assemblage, the baulks
between the sectors were subsequently excavated. A
series of terracotta and brick ring-wells and several pits
Trench D
were found cutting into the deposits, but no substantial
Trench D is situated at the northernmost limit of the
structural activity was identified. The cultural material
ASI reserve area, 50 m north-east of Trench A close to
within the deposits can be dated to the Medieval and
the modern-day Muslim quarter of the town (Fig. 4.1).
Early Historic periods. The deposits are interpreted
A small L-shaped trench was excavated here, either by
86   J. Varaprasada Rao and Derek Kennet

Dikshit in 1965 or by Yusuf in 1937, which seems then to most levels of the trench, two fairly deep deposits of
have been enlarged in the more recent past by the clan- compact black cotton soil and a loose silt layer with
destine robbing of building materials which has now a large number of pottery sherds (753, 754) covered a
been halted. Three metres of Medieval and post-Medie- further close sequence of four consecutive compact
val overburden had been removed by these activities, mud floor levels (759, 766, 767, 770) immediately above
providing an ideal location, at the very northern limit natural soil (777). Each of these floors was cut by a few
of the archaeological area, for investigating the Early post-holes or pits (e.g. 772), but none was associated
Historic layers of the site. with any brick structures. A large water storage jar
(760/761) was found cut into floor 766, and evidence
of burning and an ash filled pit were found associated
Trench D1 with floor 759.
(Fig. 5.6) All of the deposits of Trench D1 have been allocated
to Period 1 on the basis of the pottery that they con-
Trench D1 was excavated down to a similar depth to tained, with the exception of floor 752, which is allo-
Trench D2 and the results were broadly similar. D1 was cated to Period 2, and the loose wash on the surface
opened as a 4 × 4 m trench but was reduced down to 2 × 4 (751) that is allocated to Period 3.
m almost immediately in order to save time.
The sequence revealed by this trench is quite unre-
markable for its structural remains or evidence of
human activity. Below a lime floor (752) in the upper-

Fig. 5.6: Trench D1, east-facing section.

 Chapter 5 The Excavation of Trenches B, C, D, E and F    87

Trench D2 The uppermost levels in this trench, down to floor

(Fig. 5.7) 712, have been allocated to Period 2 on the basis of the
pottery that they contained. Layers below this down to
The majority of the upper layers are dumps of pottery natural have been allocated to Period 1.
apparently deposited at different times when this area
was inhabited during the Mughal and late Mughal times.
Large town houses with cellars were built during this
period by digging foundations into the Early Historic
Trench E
levels, thus explaining some reverse stratigraphy and
Trench E is located just to the south-east of Trench A.
a high degree of residuality in the finds from this area.
Excavation was taken up to investigate the nature of
Undisturbed archaeological layers occurred from around
the mounding here, which is considerably higher than
450.00 m above site datum (711), whereas the height of
the level of the surrounding area. A small 2.5 × 2.5 m
the ground surface was 453.31 m.
trench was excavated to a depth of over 3 m, but natural
soil was not reached. Excavation was stopped when it
became clear that the mounding in this area is pre-
dominantly post-Medieval in date and consists largely
of heavily disturbed or redeposited material. Fifty-five
contexts were recorded, including pits, a few floors,
one stone and one mud wall and a few hearths. Some
Early Historic material came to light, but it seems
mostly to be residual material from much later layers.
No in situ Early Historic occupation was excavated
here, although some of the lower levels encountered
may be datable to the Early Medieval period. The trench
produced some interesting and useful small finds that
throw some light on the date of occupation of Paithan
more generally. These are reported on in Chapter 8.

Trench F
(Figs 5.8 and 5.9)

Fig. 5.7: Trench D2, south-facing section, as dug.

Trench F (2 × 2 m) was opened up in the middle of the
old town of Paithan on the top of the mound in an area
The earliest occupants settled down on the natural
known as the ‘Nath Galli’ (Fig. 4.1). The aim of the
black cotton soil (730) in huts made of wooden posts
trench was to investigate the nature and chronology
with open fire places and with querns and mill-stones
of the deposits in this area.
as part of the inventory of their daily lives. Bones were
The trench was excavated to a depth of 3.5 m. The
found associated with fire places where butchering
top 1.2 m of deposits consisted of later dumping (778).
and roasting perhaps took place. Eight or nine intact
Below this, the deposits consisted of a series of clay
floor levels were encountered amongst the Early His-
floors with occasional post-holes, gravel spreads and
toric period levels (711, 712, 715, 718, 721, 722, 725, 726,
hearths (784, 787, 788, 789). At a depth of 2.2 m (451.6
729). A good number of bone points, arrowheads and
m above datum), a brick floor was encountered con-
beads of different material were also recovered from
structed of brick bats (790) and cut by two post-holes.
these layers.
This feature was by far the most substantial evidence of
The beginning of occupation seems to have been
occupation encountered in the sequence here. Below
marked by simple living after which the economic stand-
the brick floor, further clay surfaces were found (846,
ards gradually improved as signified by increased
851, 852) before natural soil (853) was reached at a
coinage, good-quality pottery, permanent hearths and
depth of 3.2 m (450.4 m above datum). The excavation
rammed-earth floors.
continued into the black cotton soil that constitutes
88   J. Varaprasada Rao and Derek Kennet

the natural deposits of the river terrace to a depth of Summary

15 cm.
The trench demonstrates that there does not seem The trenches reported on above have thrown some
to be a great deal of very substantial architecture of the limited light on the history of occupation in certain
Early Historic period in this particular part of the site, parts of the mound, most notably in the ASI protected
with the possible exception of the brick floor (790). The area, which is located towards the southern extent of
picture that is given is one of occupation on the river what appears to be the main archaeological mound-
terrace dating back to perhaps as early as the 1st/2nd ing of the site. Looking to the broader perspective,
century bc, and from that point onwards, there is evi- the results are unsatisfactory in the sense that they
dence of continual but not very substantial occupa- are too limited in size and number and do not give a
tion. clear and unambiguous picture of the development of
The very lowest levels in this trench, pit 850 and floor occupation over the whole site. This is because of the
851 are allocated to Period 1 on the basis of the pottery difficulty of finding suitable excavation sites in the
that they contained. The rest of the sequence has been modern town, which today covers the highest part of the
allocated to Period 2 with the exception of the surface Early Historic mound.
layer (778), which contains material of Period 3.

Fig. 5.8: Trench F, east-facing section. Fig. 5.9: Trench F, view of the west-facing section.

The structural evidence from the Early Historic period

has been shown to be very limited in all trenches, and
in none of them can it be claimed that the archaeo-
logical deposits are consistent with a densely occu-
pied urban centre of the Satavahana or Early Historic
 Chapter 5 The Excavation of Trenches B, C, D, E and F    89

period. Whether this reflects the nature of Satavahana/ The results from these trenches, earlier work at
Early Historic urbanism in this region or simply the the site, Trench A and the antiquities and coins from
fact that it was not possible to excavate trenches in the the site are brought together in Chapter 14, where the
right locations on the mound remains unclear. history of occupation at Paithan is considered.
Derek Kennet

Chapter 6 
Periodisation and Dating of the Sequence

Before outlining the periodisation of the stratigraphic There are deeper problems with the periodisation
sequences from Paithan, it is necessary to say a few words of Nevasa. This is shown in Table 6.1, which is based
about how it was achieved and to put this into the context on a detailed reading of the published report. It shows
of the way in which periodisation has been carried out on the occurrence of glazed wares that are not themselves
other excavated Early Historic sites in India. specifically discussed by the excavators in the publi-
cation. It can be seen that the glazed wares and ‘cela-
dons’, all of which date to the 11th/12th century and

Problems with the periodisation later to judge by the published photographs, occur
in layers that have been attributed by the excavators
of Early Historic sites in India to much earlier periods. For example, there are two
sherds of glazed ware from layers attributed to Period IV
At Paithan, as at many excavated Early Historic sites, (150–50 bc), which, although anomalous, may simply
several trenches were excavated in different locations be intrusive. More worrying is the much larger quan-
across the site. It should not be expected that the occu- tity that was found in layers attributed to Period V,
pational history of each trench will be the same. As the so-called ‘Indo-Roman’ period, that is dated from
can be seen by observing any modern town or village 50 bc to ad 200. This is indicative of some funda-
anywhere in the world, some areas of a settlement may mental problems with the periodisation of the Nevasa
be left open and abandoned whilst adjacent areas are sequence and demonstrates the importance of taking a
densely occupied. This situation may change in suc- holistic and quantified view of the pottery, small find
ceeding periods so that areas once built up turn to and coin assemblages and of applying rigorously the
wasteland whilst areas once unused are built up. This rule of terminus post quem. These problems are further
seems obvious, but nonetheless, the way in which the compounded by the fact that the pottery from Nevasa
stratigraphic sequences of many Early Historic excava- is published by period rather than by layer. This makes
tions in India have been periodised implies a belief that it impossible to disentangle the pottery sequence from
all parts of each site will share a similar occupational the flawed periodisation.
and depositional history and that all periods of occupa-
tion will be present in all trenches. Table 6.1: Occurrence of glazed wares in the Nevasa sequence.
Nevasa is a good illustration of this. An examina- Note that only one sherd of glazed ware is listed on page 339 of the
report from Period IV, but two are listed on page 345.
tion of the Nevasa report, published in 1960, shows that
a four-period occupational history has been imposed
Nevasa Period Glaze a Glaze b Glaze c Celadon Total
on all of the trenches excavated at the site, including
trenches I, X and Y, which are located at a consider- VI (ad 1400–1700) 45 10 5 60
able distance from the main trenches (A–H), despite V (50 bc–ad 200) 14 5 4 7 30
the fact that a Qutubshahi coin and sherds of glazed
IV (150–50 bc) 1 1 2
ware were found in the levels of trench I that are other-
wise ascribed to the ‘Indo-Roman’ period (Sankalia et
Similar problems can be shown to exist in the periodi-
al. 1960: 65, 204, 344–345). This is because the perio-
sation of many of the published Early Historic sites in
disation has failed to take into consideration that there
India, although there is no space here to go into these
may be ‘voids’ in the occupation history of some areas
in further detail. Nonetheless, much of the archaeo-
of the site and this has caused layers to be forced into
logical chronology of Early Historic India is still based
periods where they do not belong, grouping material
on the periodisation of such sites. This chronology may
of different dates into the same period.
have many serious flaws and is in need of a full-scale
review.
https://doi.org/10.1515/9783110653540-006
 Chapter 6 Periodisation and Dating of the Sequence    91

A further problem is that quite often, the occur- D1, D2 and F were, however, fully incorporated into the
rence of a single coin, for example a Satavahana coin, is site periodisation because the excavations in Trench C
enough for a layer or for an entire period to be attrib- were not conducted by the present authors or according
uted to the Satavahana period. The rule of terminus to the same system, and those in Trench E were explora-
post quem is thereby ignored. This rule states that any tory and did not achieve great stratigraphic depth (see
datable find, such as a coin, indicates that the layer Chapter 5).
within which it is contained was deposited at the time Taking into consideration the points that have
of the date of the coin or at any time since that date. been made above about the problems with the perio-
Thus, a Satavahana coin in a layer indicates only that disation of Early Historic sites in India, careful thought
the layer was deposited in the Satavahana period, or at was given to the best way of periodising the sequence
any time since. Indeed, it is well established that Sata- from these five trenches. A number of datable objects
vahana coins continued in circulation into much later or ‘chronological markers’ such as coins and sherds
periods (Sircar 1974: 18; Shastri 1992: 291–292). The of known wares such as NBP and RPW are present in
same concept applies equally to pottery; a single sherd these sequences, but generally in very small quanti-
of Northern Black Polished ware (NBP) or Red Polished ties. The small quantity makes it very difficult to know
ware (RPW) does not necessarily indicate that the layer whether or not any particular occurrence of such
within which it is contained is datable to the period objects is residual, or indeed when they are absent,
of circulation of that ware. It may be that the sherd is that the absence is not simply due to chance. It was
residual and the layer may have been deposited at a therefore decided that such finds provide a very poor
much later date, often several centuries later. Quanti- basis on which to correlate the five different trench
fication of the pottery assemblage can go some way sequences.
towards resolving this issue by highlighting ‘outliers’ Instead, the periodisation was achieved through
in the sequence that may be intrusive, or the tail-off of a combination of the quantified occurrence of known
residual pottery beyond its main period of circulation. chronological markers: both the rarer examples such
Quantification of pottery sequences is, however, rarely as NBP, RPW and glazed wares, as well as the more
available in published reports. common examples such as Thick Grey ware (GREY) and
Even when these points are taken into considera- Black and Red ware (B&R) Type 38 – but this was linked
tion, periodisation of a stratigraphic sequence is not to a quantified analysis of the occurrence through the
an exact science; it is always an approximation, an sequence of the key pottery types defined at Paithan
attempt to impose an orderly chronological structure on such as Types 7, 19, 32 and 37, the absolute dating of
what are in effect the results of sporadic and unstruc- which is not known but whose occurrence through the
tured human activity. Periodisations should always, stratigraphic sequences can be more precisely charted.
therefore, be open to criticism, review and revision as The reason for this is that, in being very common, their
new evidence comes to light. For this reason, the mate- absence in any particular layer can more reliably be
rial from a site should be published in enough detail taken as an indication they were not in circulation at
for the reader to be able to re-assess the periodisation the time that the layer was deposited. Consideration
if required. was also taken of the ‘percentage occurrence profiles’
The above problems and considerations were of some of the more common types and wares through
taken into account when deciding the way in which the the sequence (see below). All of this relied on a much
Paithan sequence was to be analysed and periodised. more precise definition of pottery wares and types than
is normally the case at Early Historic sites in India (see
Chapter 7). On this basis, the following four periods

Periodisation of the Paithan were defined. The absolute dating of each period is dis-
cussed in the following section.
sequence The stratigraphic matrices for Trenches A, A:TP1,
B, D1, D2 and F have been lodged with the Archaeologi-
A total of seven trenches were excavated at Paithan cal Data Service (http://dx.doi. org/10.5284/1017460).
during the 1996–1999 seasons (Trenches A, B, C, D1, D2, The matrix for the smaller trenches shows the occur-
E and F). Each of these provided a stratigraphic sequence rence of Types 7, 32 and 37, GREY and glazed wares.
reflecting the particular history of occupation in the part
of the site where it was located. Only Trenches A:TP1, B,
92   Derek Kennet

Period 1 ing that it was introduced into circulation at about the

same time as these types.
In these layers, which occur in the lowest levels of Type 32 is the rim of a large storage jar. It occurs in
all five trenches, a number of known chronological the sequences from trenches A:TP 1, B, D1 and D2, in three
markers were found: 22 sherds of NBP occurred in cases (Trench A:TP1, B and D2) very close to the same
eight out of 116 layers; 25 sherds of the very distinctive level at which Type 7 first occurs.
B&R Type 38 were present in 14 layers and occasional Type 37 is a jar with an everted rim. It occurs in the
microliths also came to light in some of the lowest sequences from trenches A:TP1 and D2 at very close to
layers of trenches D1 and D2, together with some rolled the same levels where Type 7 and the other types dis-
sherds of Jorwe ware. This period was also defined by cussed here first occur.
the complete absence of the key pottery markers of There are a few later, intrusive objects in layers attrib-
Period 2, namely RPW and Types 7, 19, 32 and 37 (see uted to Period 2, including a porcelain peacock’s head
below). (No: 678) datable to the 16th century or later, and eight
These same layers also contained a relatively high GREY sherds (see below).
percentage of B&R generally. Although there is con-
siderable variation, the percentage of B&R Type 10 is
consistently above 2% of the total sherd assemblage – in Period 3
many cases much higher. The proportion of Type 1 can
also be shown to have been increasing steadily through- Period 3 is defined quite simply by the occurrence of
out this period in each of the sequences (see below). the abundant GREY, Types 14, 15, 24, 26, 44, 45 and 46.
The lower part of the Trench F sequence has been This development is accompanied by a decline in the
allocated to this period on the basis of a single sherd of amount of Type 1 to below 5% of the total sherd assem-
NBP and a high percentage of B&R Type 10. blage in the trenches where the sequence continues
There are three intrusive GREY sherds from layers this late (trenches A:TP1 and B). In fact, as mentioned
attributed to Period 1 (see below). above, three sherds of GREY are recorded from layers
allocated to Period 1 and eight from layers allocated to
Period 2. These are considered either mis-attributions
Period 2 or intrusive, that is to say, they were either erroneously
recorded or they have made their way into early layers
Of the 86 layers allocated to Period 2, only two con- through soil disturbance of some kind. They have to be
tained sherds of RPW, both of which are from the Trench balanced against the 191 sherds of the same class that
A:TP1 sequence. One sherd of NBP was also present were found in the layers allocated to Period 3, and the
in the same sequence, but this sherd was judged to be 144 sherds in Period 4.
residual based on the presence of Types 19 and 32 in
the same or in stratigraphically lower contexts.
The precise lower limits of the period are defined Period ‘3-temple’
principally by the first occurrence of pottery Types 7,
19, 32 and 37. These types are all reasonably common Period ‘3-temple’ is used to differentiate those layers
in the levels where they occur but none of them occur within Trench A that are datable to Period 3 but that
below approximately the same level in each of the are part of the construction and foundations of the
trench sequences. two brick temples. These deposits do not result from
Type 7 is a very distinctive form. Its precise abso- domestic occupation and neither do they contain very
lute date is unknown, apart from being broadly attrib- much material. The artefactual and ecofactual assem-
utable to the later Early Historic period. It occurs in the blages from these layers are therefore likely to be of a
sequences of trenches A:TP1, B, D2 and F, but it does very particular nature. Because they make up such a
not occur in the lowest levels of any of these trenches large proportion of the excavated layers presented in
except F, where it only occurs in layers stratigraphi- this report, it was thought to be sensible to keep them
cally later than the single sherd of NBP. separate from Period 3 so that reliable comparisons
Type 19 occurs only in the Trench D2 and Trench can be made between the Period 3 assemblages and
A:TP1 sequences where it first occurs in levels very those from Periods 1, 2 and 4.
close to those where Types 7 and 32 first occur, suggest-
 Chapter 6 Periodisation and Dating of the Sequence    93

Period 4 One way to confront these problems is to calcu-

late the percentage of the most common pottery types
Period 4 is defined by the presence of late glaze wares and wares as a part of the pottery assemblage and to
such as Blue and White Frit (PERS), Slip-Painted Ware use these figures as a comparative tool for aligning and
(SLIP) and Chinese Blue and White porcelain (CBW). periodising sequences. This was attempted at Paithan
These wares are not very common, and no other pottery using Type 1, the most common single type from the
types have been defined that can be used in associa- site. Table 6.2 and Fig. 6.1 show the percentage occur-
tion with them, thus making this period distinction rence of this type through the lower parts of the Trench
perhaps the least reliable of those presented here. B, D1 and D2 sequences – all of which are allocated
to Period 1 as outlined above. There is a remarkable
degree of correlation between the three sequences.
Period 4? Each of them contains percentages that are just under
10% in the lowest levels but which increase steadily to
‘Period 4?’ represents layers that were excavated in 25% and more towards the later Period 1 layers.
Trench A:NWNT which are clearly later than the aban- Of course, the percentage of a type in any particu-
donment and robbing of the temples and are most lar layer or sequence is likely to be as much due to the
likely therefore to belong to Period 4, but there are no activities that took place in the vicinity as it is to the
finds of glaze wares to demonstrate this unequivocally. chronological occurrence of the type, and these data
Further descriptions and discussions of the pottery must therefore be treated with some caution. Nonethe-
types and wares can be found in Chapter 7, along with less, the pattern shown in Table 6.2 adds support to the
the relevant external dating evidence. periodisation that has been proposed.

Trench A: phases of temple construction and

Periodisation based on occurrence
use
profiles of key types
Within Trench A, the stratigraphy related to the construc-
As has been stated above, there are various problems tion and use of the temples was subdivided into architec-
with using the presence of relatively rare datable tural phases that are specific to each of the two temples
finds such as coins and sherds of NBP or RPW as ‘type and which have been described in detail in Chapter 4.
fossils’ with which to establish the periodisation of a These are distinct from the periods that are defined
sequence. One of these is that any one of these type and described in the present chapter.
fossils might be residual and the relevant layer might
therefore be much later in date than some of the finds
contained within it. A second problem is that the
absence of a relatively rare find from any particular
Absolute dating
layer might be due to chance: it does not necessarily
Chronology is still the most significant problem faced by
mean that the find was not in circulation when the
archaeologists of the Early Historic and Early Medieval
layer was deposited. A third is what to do with those
periods in India. The archaeological chronology of the
layers – often the majority – that do not contain any
Early Historic period of the western Deccan region is based
of the type fossils at all: which period should these
largely on 1950s excavations by the esteemed H. D. Sanka-
be allocated to? These problems make it difficult to
lia and his colleagues at sites such as Brahmapuri, Nasik,
establish a reliable periodisation around what is often
Maheshwar and Nevasa (Sankalia and Dikshit 1952; Sanka-
such a limited framework of datable evidence. The use
lia and Deo 1955; Sankalia et al. 1958, 1960). The dating
of more common pottery types, such as those used
of these sites is, in turn, based on earlier excavations at
above in the definition of Period 2, can help to reduce
sites such as Arikamedu, Sisupalgarh, Brahmagiri, Chan-
these problems, but they cannot resolve them entirely
dravali and Taxila. The chronology that was established
because of the paradox that the more common a type
during this period has never been fully set out, critiqued
is, the more meaningful its absence from a layer is
and argued, but it has nonetheless become the received
likely to be, but at the same time the more likely it is to
wisdom amongst archaeologists working in the region.
occur residually in later levels.
 94   Derek Kennet

Table 6.2: Percentage occurrence of Type 1 in the lower parts of the sequences from Trenches B, D1 and D2 (the top three rows show the
context numbers as grouped for this calculation, and the lower three show the amount of Type 1 as a % of the total sherd assemblage
from those context groups).

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21

B 743 736 733 494 497 493 492 487 486 485 484 483 478 479 477 474 473 472 471 469 468
735 731 498 480 470
734 495 481
732 496 482

D1 777 776 770 771 767 766 760 759 752

775 772 761
773 762
774 763
764
765
768
769

D2 729 727 726 725 722 723 721 719 718 716 715 713 712 711 710 709
728 724 720 717 714 708

B 3.75 13.24 0.00 3.99 12.09 9.22 16.35 6.02 13.10 13.08 9.40 13.60 10.65 18.32 18.59 15.20 15.15 19.27 24.68 19.18 32.39

D1 10.34 2.30 6.52 18.75 11.15 22.02 0.00 30.56 33.33

D2 20.29 11.41 6.12 7.22 9.15 4.26 11.94 13.76 14.06 14.29 20.68 10.53 22.84 19.67

Fig. 6.1: Graph showing the percentage of Type 1 through the lower parts of the sequences from Trenches B, D1 and D2 (sherd count as % of
total sherd assemblage).
 Chapter 6 Periodisation and Dating of the Sequence    95

It is, however, highly problematic and imprecise and is its occurrence on archaeological sites in the Arabian/
certainly now in need of thorough revision. The ‘fixed Persian Gulf (Kennet 2004a: 65–66).
points’ are based almost entirely on identifications of Finally, at some sites such as Nevasa, the presence
coins, NBP, RPW and – at one or two sites – the pres- of Roman amphorae was used as dating evidence.
ence of Roman amphorae. Other pottery wares, such Unfortunately, it has recently been shown that many of
as B&R, Painted Andhra Ware and GREY, are also used, the Roman amphorae reported from India are, in fact,
but these wares are not themselves independently misidentifications of Mesopotamian ‘torpedo jars’.
dated other than through stratigraphic associations These jars, although they are themselves quite impre-
with the ‘fixed points’ listed above. cisely dated, are known to have continued in use and
On examination, it becomes clear that the absolute circulation beyond the 9th century ad (Kennet 2004a:
dating of all of the ‘fixed points’ noted above can now be 63; Tomber 2007).
called into question for one reason or another. The factors described above all demonstrate that
Firstly, the dating of pre-Satavahana coins is, for the the existing chronological framework for the Early
most part, still relatively imprecise and provides only Historic period in this region of India now needs to be
the broadest chronological framework, whilst the clas- carefully reviewed and the chronology proposed by
sification and dating of Satavahana coins, although the excavators of the sites mentioned above cannot be
they have improved and changed since the reports used as a reliable framework. This means that there is
mentioned above were published, are still subject to very little external evidence with which to establish
debate and disagreement (see for example Bhandare the absolute dating of the Paithan sequence, which
1998; Gupta 1999). But the two most significant prob- has therefore been based almost entirely on the 39
lems related to coins are residuality and the failure identifiable coins that were retrieved. The coins, which
to take the concept of terminus post quem into consid- are presented in Chapter 9, were identified and dated
eration. These points are particularly important in a according to the schema proposed by Bhandare (1998).
region where low-denomination Satavahana coins are The coins have been counted and tabulated so that pat-
known to have continued in circulation in large quan- terns of occurrence and residuality are absolutely clear
tities until much later periods, whilst coins of later (see Chapter 9, Table 9.1).
dynasties (e.g. the Vakatakas) appear to be very rare In addition to the coins, there are two further
on most sites (see Chapters 9 and 14). This means that elements of absolute dating, neither of which is very
in some cases, post-Satavahana layers can contain precise. The first is that provided by the temples them-
nothing other than low-denomination Satavahana selves, which suggests that the North Temple was first
coins, which has led archaeologists to assign a date to constructed in the 4th or 5th century ad and the South
such layers which is far too early, often by as much as Temple in the 6th or 7th (Chapter 4, Table 4.2). The
300 or 400 years. This fact may help to explain the second are two C14 dates taken from fragments of car-
apparent absence of Vakataka layers at many sites in bonized wood that were retrieved from the foundation
the region. deposits of each of the two temples (Table 6.3).
Secondly, the chronological span of NBP has been Both of the charcoal fragments, being datable to
established for the Gangetic region of India where, the 4th and 3rd centuries bc, are much older than the
on the basis of C14 dates, it can be shown to have in temple foundations, as is demonstrated by the coins
production for as long as 500 years (see Chapter 7). from those layers and by the architectural dating of
Nonetheless, the date when NBP first began to occur the temples, which has been mentioned above. At first
in significant quantities in the western Deccan region glance, it might seem rather coincidental that the two
is not known, neither is it known how long NBP con- charcoal fragments should give such similar dates, but
tinued to be used in this part of India or when its use it is obvious that the earth in the foundation deposits
finally died out. is likely to have been brought from the same source not
Thirdly, there are also problems with RPW. The far from the temples and will have contained artefacts
introduction of this ware is generally placed in about and environmental material from earlier archaeologi-
the 1st century ad and the ware is thought to have cal deposits. This issue is further discussed in relation
gone out of use between the 4th and the 6th centu- to the coins in the foundation deposits in the Appendix
ries ad (Chapter 7). Although neither date is firmly to Chapter 9.
established, however, production can now be shown
to have continued until the 8th century ad through
96   Derek Kennet

Table 6.3: C14 dates from the temple foundations.

Sample Context Measured C14 13C/12C ratio Conventional C14 age 2 sigma calibration
age

Beta 227163 North Temple 2240 ± 40BP –24.8o/oo 2240 ± 40 BP 390–200 bc

(927)

Beta 227164 South Temple 2280 ± 40BP –25.5o/oo 2270 ± 40BP 400–340 and 320–210 bc
(1002)

Broadly speaking, the occurrence of coins through the Period 2: Satavahana (late 1st century bc to
sequence suggests the dates for the Paithan periodisa- late 1st century ad although possibly as late
tion that are set out in the paragraphs below. as the 4th century ad)
All of the 18 identifiable Satavahana coins found at
Pre-Period 1: (1400–4th century bc) Paithan can be dated to the immediate pre- and post-
Gotamiputa Satakani period suggesting activity at
The rolled Jorwe sherds mentioned in Chapter 7 prob- the site between the late 1st century bc until the late
ably indicate a presence at the site roughly between 1st century ad. The absence of coins actually attrib-
1400 and 700 bc at the broadest level. Following this, uted to Gotamiputa Satakani’s reign (e.g. ad 55–79 or
no silver punch-marked coins of the Magadha-Maurya 66–90; Cribb 1992: 134; Bhandare 1998) is, however,
‘Karshapana’ series were found in excavated layers, notable. This absence might indicate that all or part
although seven are present in the Balasaheb Patil Gov- of the site was abandoned through his reign, but it
ernment Museum at Paithan (Chapter 11, 1–7). These seems more likely that it results from a change in the
might indicate activity at Paithan as early as 370 bc – a pattern of coin circulation or deposition during the
date which tallies with the earlier end of the date range same period. At the same time, it is important to note
of the C14 dates reported above – but, as the precise that only one Satavahana coin (Chapter 9, coin 17) was
find spots of these coins are not known, it is impossible actually found in the excavated layers of this period,
to be certain that they were originally found at Paithan. along with two earlier ‘hollow cross’ type coins. The
latter are clearly residual, but it should nonetheless be
noted that, although the layers of Period 2 are likely to
Period 1: Pre-Satavahana (200–100 bc) have been deposited during the late 1st century bc to
late 1st century ad period (a period during which activ-
The eight pre-Satavahana ‘hollow cross’ type coins ity at the site is suggested by the coins that have been
retrieved from levels of this period can be dated found in later layers), it is of course quite possible that
between 200 and 100 BC (Chapter 9). These are the coin 17 is residual and that the layers of Period 2 were
earliest coins that came to light from the excavations. actually deposited at a somewhat later date. This must
Technically, of course, they provide only a terminus have been before the 4th/5th century, as is shown by
post quem but, because eight such coins were found in the stratigraphic relationship with Period 3 levels, but
these levels and no later coins were found with them, it could just as easily have been in the 2nd to 4th cen-
despite the fact that later coins can be shown to have turies ad.
been deposited in later layers at the site in quite large
quantities, this dating can probably be taken as being
reasonably reliable (see Chapter 9, Table 9.1). Period 3 and Period 3-temple: post-Satava-
hana (4th/5th to 8th century ad)

The dating of Period 3-temple is provided by three

pieces of evidence:
–– the 4th/5th century ad architectural dating of the
earliest construction of the North Temple;
 Chapter 6 Periodisation and Dating of the Sequence    97

–– the 6th/7th ad architectural dating of the construc- It will be noted from the discussion above that
tion of South Temple (Chapter 4) and there is no clear evidence from the excavations for
–– the latest of the two post-Satavahana coins in activity between the end of the 1st century and the
the South Temple foundation deposits (Chapter 9, 4th/5th centuries ad. Again, this apparent void may be
coins 33 and 34), which is probably also datable to due to a decline in activity over all or part of the site,
the 7th century ad (whilst the earlier of these two but it may also be due simply to a change in the pattern
coins is obviously residual). The end of the period of coin circulation or deposition during this period.
has been dated to about the 8th century ad but this
is based largely on an estimation of how long the
temples may have remained in use and is not reli- Period 4: 14th/15th to 17th century and later
able.
On the basis of the coins, Period 4 can be dated to
The dating of Period 3 is slightly more complicated. the Late Medieval and British periods. Some of the
Analysis of the pottery indicates that it is contemporary pottery found in Period 4 levels, e.g. Blue and White
with Period 3-temple, but post-Satavahana coins are Frit (PERS), might suggest that this rather loosely
only present in Period 3-temple levels. This is almost defined period encompasses activity as early as the
certainly due to the fact that, because post-Satavahana 14th century, but the majority of datable finds such as
coins are much less common than Satavahana coins, coins and imported Chinese porcelain can be dated to
they are generally found only in larger assemblages, the 15th to 18th centuries.
but it does leave open the possibility, as does the strati- The apparent lack of activity between the aban-
graphic sequence, that some Period 3 deposits were donment of the temples and the 14th/15th centuries is
deposited somewhat earlier. something that has been noted at many sites. It is possi-
There appears to be no reliable evidence for the ble that occupation moved or contracted to a core that
dating of the introduction of the so-called ‘Early Medi- is now buried underneath the modern town of Paithan
eval’ grey wares (GREY) from the publication of any during this time. The relatively limited investigation of
other excavated site. These wares are often suggested layers of this period does not allow further discussion
to have begun to circulate in the 8th century ad, but, of this point.
if the dating of the Northern Temple here at Paithan
is reliable, they can be shown to have been in circu-
lation at Paithan by the 4th or 5th century or shortly
thereafter.
Derek Kennet

Summary Remarks on the Excavations

The three chapters in Part II have set out the results to lack of personnel. The results were not on the whole
of past and present excavations at Paithan. These are instructive or promising given the heavy overburden
brought together in Chapter 14 with the results of the of modern occupation, soil movement, dumping, veg-
studies of the artefacts and environmental evidence etation and general build-up of material. In addition,
retrieved by the present excavations. an attempt was made to excavate trenches across the
The research aims for the 1998 and 1999 seasons whole site during the 1998 and 1999 seasons with the
have already been set out in Chapter 1. For the most same aim. Despite the importance of this aspect of the
part, these were successfully dealt with and the results research, the difficulty of finding suitable space and
have provided a number of important insights into the of negotiating access and permissions with private
chronological development of the site and particularly landowners proved extremely difficult. Thus, with
into the development of the two Early Hindu temples the exception of Trench F and the trench excavated by
and the area of occupation immediately surrounding Dikshit in 1965, the insights into the spatial develop-
them. Indeed, Chapter 4 presents the most detailed ment of Paithan that are provided in the present report
stratigraphic analysis of the development of an Early are based entirely on the area within the present ASI
Hindu brick temple that has yet been published from protected area at the southern extent of the main
anywhere in India. The fact that it has been possible mound. This is certainly not broad enough and may
to link this to a systematic, stratigraphic and quanti- not present a fully representative picture, meaning
fied study of artefacts and environmental samples that it has only been possible to sketch out in Chapter
presents us with the first archaeological study of the 14 a hypothetical outline of Paithan’s development
foundation and development of an Early Hindu temple based on the information that is presently available.
within its environmental and economic context. The The present author has made the point elsewhere that
results of this work and its broader implications are until the practice of ‘urban archaeology’ becomes
further discussed in Chapter 14. more widely established across India, allowing a more
Most of the 1998 research aims were successfully holistic approach to be taken to the study of archaeo-
dealt with, but, given the size and scope of the present logical sites such as Paithan that have continued to be
research project, it was not possible to achieve com- occupied until the present day, our archaeological
pletely satisfactory progress with the investigation of understanding of the development of ancient urban-
the geographic extent and development of the whole ism – especially Medieval urbanism – will not be able
site. Some limited survey had been carried out during to achieve very much progress (Kennet 2013).
the 1996 season and this was continued during the 1998
and 1999 seasons, but only at a very limited scale due
Part III
The Finds
Derek Kennet

Brief Remarks on the Finds

Part III of this volume presents five specialist reports on studies, but it is hoped that the value of a more precise
the artefactual evidence from the excavations (pottery, definition of wares and types and a full presentation
coins, glass, small finds) along with notes on related of the quantified occurrence of each through the exca-
material from the Balasaheb Patil Government Museum vated sequence will be clear. It is also hoped that such
at Paithan. an approach, if more widely adopted, will encourage
A key aim of the present project was to use a system- and facilitate inter-site comparisons and regional syn-
atic approach to collection and a quantified approach thesis of archaeological evidence, which is an aspect
to the study of all materials. This included sieving of archaeological analysis that is still relatively under-
all excavated earth and the collection and counting developed in Early Historic India.
of all finds. Although such approaches are generally The finds from the first two (1996 and 1997)
regarded as having become standard practice with seasons were not fully studied due to the way in which
the advent of the ‘new’ or ‘processual’ archaeology the project had to be organized at that time. Chapter
of the 1960s, they are still rarely applied to the study 10 deals with the most important finds from these
of Early Historic and Medieval material from India. seasons, whilst Chapter 11 presents material from the
Most published excavation reports, such as those of private antiquity collection of the late Sri. Balasaheb
Early Historic sites listed in Chapter 3, simply present Patil, who was a prominent local dignitary and col-
a selection of material that is regarded by the excava- lector of archaeological objects. Although none of this
tor as being representative of the periods and phases material can be attributed to Paithan with absolute
that were defined at the site. Whilst this approach has certainty because the circumstances of its acquisition
a certain validity, it fails to make the most of the infor- are not recorded, it seems likely that most of it was
mation contained within the material and it also makes found at or close to Paithan and thereby provides a
inter-site comparisons difficult and less meaningful useful illustration of the type of material that is turned
than they might be, thereby discouraging regional syn- up in this locality by farming and construction work.
thesis. Some may find the approach taken here, espe- The key results of the studies presented in this
cially that taken in the pottery study (Chapter 7), quite part of the volume are further discussed and contextu-
a marked step away from the more traditional pottery alised in the conclusions (Chapter 14).

https://doi.org/10.1515/9783110653540-007
Derek Kennet

Chapter 7 
The Pottery

Introduction through the stratigraphic sequence. It was hoped that

such an approach would provide a means with which
to link the excavated sequences from the different
A total assemblage of 205,267 sherds was retrieved
trenches as well as providing a tool that might be of
from excavated contexts at Paithan during the 1998
use to researchers working at other sites in the region.
and 1999 seasons. From this assemblage, the pottery
from five trenches or test pits (‘Main sequences’), along
with a selection of assemblages from other trenches
Traditional methodology
(‘Secondary assemblages’), was selected as the study
sample, comprising 86,787 sherds or about 42% of the
The system of pottery description and analysis that
total excavated assemblage. The remaining 58% was
is traditionally used in the publication of excavation
inspected for imported wares and other key diagnos-
reports in this region of India is to divide the pottery
tics, but was not further studied. This includes most of
assemblage into what are loosely called ‘wares’, of
the material from Trenches C and E plus material from
which there are many, for example ‘Red ware’, ‘Coarse
Trench A other than TP1.
Red ware’, ‘Black and Red ware’ and ‘Burnished and
The ‘Main sequences’ were selected as providing
Slipped Red ware’. The problem with this approach
the most reliable and complete overview of the Paithan
is that, although most experienced archaeologists
sequence, whilst the ‘Secondary assemblages’ were
working in the region believe they know roughly what
selected in order to supplement these by covering
these ware names refer to, the terms have never been
aspects of the excavations that were not included in
precisely and formally defined and neither has their
the ‘Main sequences’.
chronology. Because of the lack of clear definitions, the
The study of the Paithan pottery was undertaken
ware names often appear to be used very loosely and
during 3 weeks in February 2000. It included 75,990
in many cases refer to different wares at different sites.
sherds from the five ‘Main sequences’ and 10,797
For example at Bhokardan, the term ‘Drab black’ ware
sherds from nine ‘Secondary assemblages’, giving the
is used, whilst at Maheshwar and Prakash, ‘Coarse
total of 86,787 sherds mentioned above. The vast major-
Black’ or ‘Darkish-grey’ ware is reported (Sankalia et
ity (72,201) of these were body sherds and were counted
al. 1958: 140, 149, 154, 163; Thapar 1967: 93; Deo and
but not further classified (with the exception of a few
Gupte 1974: 91). Similarly, at Nevasa, ‘Tan slipped’
wares) because they proved impossible to classify reli-
ware is reported, whilst at Bhokardan, the name ‘Mica-
ably using the naked eye and a ×10 hand lens (see the
slipped tan ware’ is used, and at Maheshwar, ‘Tan’
discussion on methodology below). The quantitative
ware is reported (Sankalia et al. 1958: 87, 116, 146,
analysis was therefore carried out only on rim-sherds,
158, 164; Sankalia et al. 1960: 264, 287; Deo and Gupte
a total of 14,586 of which were recorded from the
1974: 108). To give a further example of the problem, at
studied assemblages. These are distributed across the
Bhokardan, we hear of ‘Mica-slipped red ware’, which
periods in the following way: Period 1: 5,176; Period 2:
is not reported at Nevasa, Maheshwar or Prakash (Deo
5,424, Period 3: 1,495, Period 3-temple: 407 and Period
and Gupte 1974: 89, 99–100). By contrast, ‘Thick coarse
4: 2,084. These are all reasonably good sized assem-
red’ ware was found at Maheshwar but not at Bhok-
blages, with the exception of Period 3-temple, which
ardan (Sankalia et al. 1958: 164). The question faced by
is too small to allow reliable quantified comparisons
the pottery analyst is: are the wares described by these
in some cases. Details of the assemblage are set out in
names the same, sharing the same date and possibly
Table 7.1.
even manufacturing centre? Alternatively, are they
The main aims of the pottery study were to set out
attempts by archaeologists to impose the same poorly
a clearly defined classification of types and fabrics and
defined classification system on to localised pottery
to provide a quantified analysis of their occurrence
traditions that are in reality quite different? Or are
102   Derek Kennet

they simply descriptive terms that are not suitable for provide us with a mechanism for identifying chrono-
making inter-site comparisons? Are these apparent dif- logical or regional variation between assemblages. This
ferences indicative of regional or chronological varia- lack is clearly reflected in the lack of progress over
tions between the sites in question, or are they simply the last 45 years or so towards a more precise ceramic
the result of the use of imprecise and poorly defined chronology for the Early Historic period. The 300- to
terminology? The general consensus seems to be that 400-year-long archaeological periods that are in use
there are broad patterns of technological and stylistic today are the same as those that were set out in the late
change in pottery manufacture that encompass large 1950s and early 1960s by Sankalia and his colleagues
areas of Early Historic and Early Medieval India, by and there has been no perceptible progress towards refin-
which ‘Black and Red ware’ precedes a variety of Red ing them.
wares and then Grey wares. This is potentially a very Having said this, it should also be noted that local
important issue and one that certainly requires further course wares of the Early Historic and Early Medieval
investigation if we are to improve our understanding of period in India are particularly challenging to the
the chronology and of the production and distribution ceramic analyst. Clear definition of fabrics is particu-
systems of these periods. But how can the question be larly difficult given the generally low degree of stand-
further investigated if a precise and coherent ware ter- ardisation and the high levels of continuity between
minology does not exist? periods. These problems are especially obvious when
compared to other regions of the world such as the
Table 7.1: Sherd numbers for the main sequences and secondary Mediterranean or the Near East. Such differences may
assemblages included in the final pottery study.
be related to different structures in the organization
of production and distribution but further research is
Trench Rims Other sherds Totals
needed into this question.
Main sequences

A:TP1 2,874 18,780 21,654

B 4,230 21,679 25,909 Methodology at Paithan

D1 810 3,287 4,097
The traditional classification system discussed above
D2 3,318 8,177 11,495
was initially tried at Paithan, but it was soon aban-
F 2,251 10,584 12,835 doned because it quickly became obvious that, as has
Secondary assemblages been stated above, these terms are so vague and poorly
defined as to prevent the identification of patterns of
A:South Area 117 1,604 1,721
change within the sequence. The fabrics of traditional
A:South Area, TP6 0 359 359 ‘wares’ such as ‘Coarse red ware’, ‘Burnished and
A:TP3 867 6,351 7,218 slipped red ware’ and ‘Slipped red ware’ were com-
A:TP3, TP5 116 1,277 1,393 pared and found to be indistinguishable with the use
of a ×10 hand lens or the naked eye. Instead, it was
Other (E, GG North, NWNT, 3 103 106
clear that a number of coherent ‘types’ based on rim
TP9)
shapes were present in the sequence, some of which
Totals 14,586 72,201 86,787
had discreet chronological patterns of occurrence. It
was therefore decided that a formal typology based on
The fact is that almost all of the ‘wares’ mentioned rim forms linked to a more precise description of clay
above and many others like them that are presented fabrics was the most promising method by which to
in excavation reports are loosely descriptive terms classify the Paithan assemblage.
based entirely on observation of surface treatments In developing and using this system, great care
and body colour rather than on clay fabrics, inclusions was taken to quantify the assemblage from each layer
and tempers. In fact, the latter are not described to an accurately. Only in this way has it been possible to gain a
acceptable standard in any of the excavation reports clear impression of the chronological life span of indi-
that have been examined by the present author. vidual types. Quantification has also helped, at least
It can therefore be said that the traditional system to some degree, to overcome the perennial problem of
of pottery classification used in the region does not multi-period sites, that of residuality. By quantifying
 Chapter 7 The Pottery    103

the assemblages, it is often possible to get a reasonable Fabric 1 (local buff fabric)
sense of when a type ceased to circulate from the point
at which there is a notable decline in its overall propor- This is a coarse fabric that is quite variable in terms
tion of the assemblage. of the proportion of various inclusions, the firing tem-
So, whilst it would be wrong to suggest that the perature and colour, although other basic features do
system adopted at Paithan has provided us with the not change. The variability makes it difficult to define
final key to the pottery chronology of the Early Historic coherent sub-fabrics within the pottery using only a
period in this part of India, it certainly does offer a ×10 hand lens. It is possible that more sub-fabrics may
more precise and reliable method with greater poten- exist than those that have been defined below.
tial for the identification of chronological and regional As a whole, this fabric makes up between about
patterning within and between pottery assemblages. 50% and 80% of the total pottery assemblage from
In this study, a distinction is made between ‘fabric’, each period (see Table 7.2).
‘class’ and ‘ware’. The ‘fabric’ is the clay of which the
pottery is made, including the temper and other mineral
inclusions (natural and deliberately added). Fabrics were Fabric 1a
identified and described with the assistance of a ×10 hand
lens and a Munsell soil colour chart. A ‘class’ is any group Types: 1, 1a, 1b, 2, 3, 4, 5, 8, 12, 16, 17, 18, 22, 25, 27, 30,
of pottery with consistently similar variables that allow it 31, 32, 33, 34, 35, 36, 37, 40, 41, 42, 47, 50.
to be meaningfully grouped or classified together. ‘Ware’ –– Traditional classification: Depending on the
has a the same meaning as ‘class’, although it should be surface treatment, pottery made in this fabric
noted that, in more general parlance, the use of the word might traditionally have been classified as ‘Red’,
‘ware’ might be taken to indicate that the pottery was ‘Buff’, ‘Burnished and slipped red’ or ‘Tan’ ware.
manufactured in the same locality or at the same kiln,
whereas ‘class’ suggests only a working subdivision of Table 7.2: The occurrence of the main fabrics by period, by rim-sherd
count above, and by rim-sherd count as a percentage of the total
material for the purposes of archaeological study.
period rim-sherd count below.
It was found that the majority of the pottery at
Paithan is made of a single, almost certainly local
Fabric 1 2 3 3-temple 4
fabric (Fabric 1), or a variant thereof. It is quite vari-
able in quality, firing and the nature of the inclusions, 1a 3,790 4,043 855 323 835
which probably reflects the loose organization of the 1b 129 137 34 5 73
manufacturing process rather than necessarily indi- 1c 3 53 1 0 161
cating the existence of distinct industries. Many of the
2 3 11 195 5 145
other fabrics (e.g. Fabric 2 and 3) appear to be closely
3 897 581 44 24 73
related to Fabric 1 but seem to have undergone differ-
ent methods of levigation, tempering and firing. The 4 4 20 13 0 72

highly variable nature of these fabrics makes it diffi- 5 0 0 0 0 67

cult, if not impossible, to identify the products of indi- Other 350 579 353 50 658
vidual potters or industries consistently without the Total 5,176 5,424 1,495 407 2,084
use of scientific techniques.

1a 73.2 74.5 57.2 79.4 40.1

1b 2.5 2.5 2.3 1.2 3.5
Pottery fabrics 1c 0.1 1.0 0.1 0.0 7.7

Fabrics 1, 2 and 3 described below make up between 93% 2 0.1 0.2 13.0 1.2 7.0

at the highest (Period 1) and 62% at the lowest (Period 3 17.3 10.7 2.9 5.9 3.5
4) of the pottery assemblage from each period, or 85% 4 0.1 0.4 0.9 0.0 3.5
of the total pottery assemblage from the site. Fabrics 4 5 0.0 0.0 0.0 0.0 3.2
and 5 are much less common and tend to be specific to
Other 6.8 10.7 23.6 12.3 31.6
certain types and wares.
104   Derek Kennet

–– Occurrence: This is the most common fabric in Fabric 1c

all periods at Paithan, making up over 70% of
the pottery assemblage in Periods 1 and 2 (and Types: 19, 41, 48, 55.
3-temple), dropping to 57% by Period 3 and 40% –– Description: This is again the same fabric as Fabric
by Period 4. 1a, but in this case, there is an absence of large
–– Description: The fabric colour varies from red inclusions and of vegetable temper, giving the frac-
through reddish yellow to very pale brown (2.5YR ture a slightly smoother and purer look, although
6/6; 5YR 7/6; 7.5YR 6/5–7/6; 10 YR 6/4–8/4). It is the fine grainy structure is still readily visible with
normally poorly fired, very weak and breaks as a ×10 hand lens.
easily as a stale biscuit, although there is a light
snap in the better-fired examples. The fracture is
very rough and the structure of the clay is notably Fabric 2
blocky and grainy. The limited range of inclusions
encompasses the following in varying proportions Types: 14, 15, 24, 26, 44, 45, 46.
and amounts: (1) curved white laminar inclusions –– Traditional classification: Pottery made in this fabric
up to about 1 mm which appear to be the remains would traditionally have been classified as ‘Grey’,
of seed husks; (2) small (1–2 mm) grey/white sub- ‘Thick grey’ or ‘Medieval grey’ ware.
angular or rounded inclusions; and (3) very fine –– Description: This fabric is finer and better levigated
sand inclusions that give the clay a grainy struc- than Fabric 1a and tends to have a smoother fracture
ture. There is considerable variation in the pres- as a result. The clay has a less blocky structure to it,
ence and especially proportions of these inclu- and there are fewer large inclusions. On the whole,
sions, but these variations do not appear to show it is also denser, stronger and better fired. Vegeta-
any clear pattern in relation to the types that have ble or chaff tempering is present, but it is difficult
been defined. They probably reflect an arbitrary to see in section and is normally only detectable
range of clay treatments that were used by potters. as voids on the clay surface. Some sherds contain
An attempt was made to subdivide this fabric on occasional large (<0.75 mm) rounded grains, that
the basis of the proportion of the inclusions. This are probably quartz. This fabric is fired to between
might have yielded useful results with a greater a very dark grey to an olive grey (5Y 3/1–4/1; 5Y4/2),
investment of time than it was possible to give, but normally being darker on the edges than at the
the apparently random variation that came to light core. There is some variation in the fabric, but this
suggested that this approach would not necessar- is less than for Fabric 1a. It seems quite likely that
ily yield useful results and it was abandoned at a this clay came from the same source as Fabric 1 but
relatively early stage. that it was mixed and fired according to a different
–– Surface treatment: The core of the fabric is often system.
a greyer colour than the surface. The surface
treatments vary with type, the most common are
reddish brown slips and washes that are some- Fabric 3
times burnished.
Types: 10, 11, 29, 38, 39, 58.
–– Traditional classification: Pottery made in this fabric
Fabric 1b would traditionally have been classified as ‘Black
and red’ ware or as ‘Burnished black’ ware depend-
Types: 6, 7, 9, 43, 49. ing on the colouring of the exterior.
–– Description: This is the same fabric as Fabric 1a –– Description: This is a finer fabric than Fabric 1a.
except that it has a high percentage of sand grains It breaks with a light snap and has a rough angular
that are large enough to be visible to the naked eye fracture and a very fine grainy structure due to a
(up to 0.3 mm). Chaff temper is also less evident. moderate amount of fine, badly sorted, sub-angu-
The fabric has a different feel to Fabric 1a, and lar grey grits, the smallest of which are almost
once familiarity has been established, the two can too small to see even with a ×10 hand lens. There
easily be distinguished with the naked eye. are occasional signs of chaff or other vegetable
temper, but these are not very common. The fabric
 Chapter 7 The Pottery    105

is fired to between a greyish brown and a very dark common in Periods 1 and 2 and may have gone out
grey (2.5Y 5/2–3/1). The external surface of Types of use by Period 3, the few remaining sherds being
10 and 38 is black, with the exception of the lower residual. Likewise, Fabric 2 seems to be datable pre-
part of the exterior, from about 2 cm below the rim, dominantly to Period 3; the few examples from earlier
where it is fired to between a yellowish red and a layers may be misidentifications or intrusive sherds.
reddish yellow (5YR 5/6 to 7.5YR 7/6). These are the The amount of Fabric 1a declines in Period 3 and Fabric
vessels that are traditionally referred to as ‘Black 5 occurs only in Period 4.
and Red ware’ (see below). Other vessels (Types
11, 29, 39, 58) are entirely black on the interior and
exterior surfaces; these are the vessels that are
traditionally referred to as ‘Burnished black ware’
Pottery wares
(see below).
In addition to the typological and fabric classification,
which formed the basis of the Paithan pottery study, a
number of ceramic wares or classes were also identi-
Fabric 4
fied or defined and these are described below. Some
are already well known from the published literature
Types: 20, 21.
and can be approximately dated by evidence from
–– Traditional classification: Pottery made in this fabric
other sites, whilst others are defined here for the first
would traditionally probably have been classified as
time.
‘Fine red’ ware.
–– Description: This fabric is weakly fired and breaks
easily with hardly a snap. It is well levigated and
Jorwe ware
has a smooth fracture. There are very few visible
inclusions. The fabric has a fine grainy structure
At Paithan, eight very small and heavily eroded sherds
that is visible with a ×10 lens; it is caused by very
of a red, finely levigated, well-fired pottery ware came
dense, tiny, well-sorted grains that are almost too
to light in the natural soil layer (777) that was excavated
small to see with the ×10 lens. The fabric is fired to
in the lowest levels of Trench D. These sherds seem to
an even reddish yellow (5YR 6/6–7.5YR 6/6). All of
be related to a type of pottery noted in Proto Historic
the sherds are small jars or bowls with quite thin
levels at other sites in the region, which is wheel-made
walls (<5 mm). They are all covered with a thin mat
and has a fine, well-levigated fabric with an orange-
slip/paint (on the exterior only in the case of jars),
red core, thin walls and black-painted designs. This
which is red to weak red in colour (2.5YR 5/4–5/6).
ware is variously called ‘Jorwe ware’, ‘Jorwe Nevasa
red ware’, ‘painted red ware b’, ‘painted black on red
ware’, ‘black-painted red pottery of the Jorwe ware’ or
Fabric 5
‘Jorwe Nevasa painted red ware’ and has been found
at Nasik, at Nevasa in Period III, at Navdatoli in Period
Types: 51, 52 (and White and Red Painted [WARP] class).
III and at Prakash in Period Ib, amongst other sites
–– Traditional classification: Unknown.
(Sankalia and Deo 1955: 161–162; Sankalia et al. 1958:
–– Description: This is a fine, well-levigated fabric
85–86; Sankalia et al. 1960: 205–207; Thapar 1967: 36,
with a smooth fracture. It is soft to medium fired
49–54). At Nevasa, which is close to Paithan, it made
and can break with a slight snap. There are no
up 65% of the pottery of Period III, whilst at more
large inclusions, but there are occasional small
distant Prakash, it made up only between 5% and 15%
(0.1 mm) unidentified white flecks in the clay. The
of the roughly contemporary Period 1b assemblage
clay also has a very fine grainy structure caused
(Sankalia et al. 1960: 205; Thapar 1967: 26). The ware
by fine grits that are almost too small to see with a
is normally dated to between about 1400 and 1000 bc,
×10 lens. The fabric fires to an even reddish yellow
with a later ‘degenerate’ form said to continue as late
(7.5YR 6/6–6/8).
as 700 bc, although some of this dating seems to be
based on uncalibrated C14 dates (Ghosh 1989, I: 239).
Table 7.2 shows the occurrence of the main five fabrics
As the sherds found at Paithan are very small and
through the sequence where chronological change
eroded and none of their shapes or original surface
can be noted. For example, Fabric 3 is clearly more
106   Derek Kennet

decorations are preserved, it is impossible to be abso- The presence of NBP is often used as a chronological
lutely certain of this attribution, although it seems ‘marker’ in excavation reports. It is found over the whole
likely to be correct given the lack of alternatives and of the Ganges area and into the western Deccan as far
the predominance of this ware in Proto Historic levels south as Brahmapuri (e.g. Thapar 1967: fig. 21). There is
at nearby Nevasa. A jar that is on display in the Bala- some discussion over the precise chronology of its pro-
saheb Patil Government Museum at Paithan is another duction, but this has mostly centred on the date of its
example of this ware that is thought to have been found introduction. Erdosy has reviewed the C14 evidence for
at Paithan (see Fig. 11.6, 1). NBP production in the Ganges area around Kausambi,
Sravasti, Rajghat and Vaisali. He places its introduction in
the second half of the 6th century bc – although this has
Northern Black Polished ware been arrived at by averaging dates from a number of dif-
ferent sites and contexts (Erdosy 1995: 104) – and the end
Northern Black Polished ware (NBP) is a well-known of its production at about 100 bc, although this is based
fine, black-slipped ware that was probably manufac- on unspecificied associated coins, seals and sealings
tured in the Ganges region (Sankalia et al. 1958: 132– (Erdosy 1995: 105). It is important to note two things in
134; Thapar 1967: 71–75; Ghosh 1989, I: 251–257). The relation to Erdosy’s chronology: firstly that the C14 dates
sherds found at Paithan have a very fine, well-levigated on which it is based are not very precise due to the nature
grey to light grey body (2.5YR 6/1–7/1) that is very strong of the calibration curve at this period, and secondly, that
and well fired and breaks with a snap. The fracture is the dating of the end of production is largely speculative
conchoidal and there are not usually any visible inclu- and is not well supported by evidence (Erdosy 1995: 100–
sions, although in some cases, rare rounded red inclu- 105). Even if Erdosy’s dates are correct, they apply only to
sions (grog?) are visible alongside occasional small the core area of NBP production in the Ganges valley and
angular air holes. The surface is covered with a high- the area immediately surrounding. It is possible – or even
quality, thin black slip that appears almost to have highly likely – that it came into use at different times in
vitrified in some cases to a glaze, giving off a lustrous different regions (e.g. Sankalia et al. 1960: 69; Ray 1986:
sheen (Figs 7.1–7.2). 47). The date at which it first began to circulate widely

Fig. 7.1: Northern Black Polished ware sherds (NBP) (scale in cm).
 Chapter 7 The Pottery    107

Fig. 7.2: Classes NBP, RPW, SLIP and CBW.

in the western Deccan and surrounding areas has never unknown. Therefore, although the circulation of NBP in
been accurately established, and neither has the date at the western Deccan region may one day be more accu-
which it ceased to circulate. To give some idea of the way rately dated, until further evidence has come to light, any
in which NBP has been dated in this region, at Nevasa, attempt to use it for more precise dating is purely specu-
it was dated to the ‘late Mauryan period’; at Prakash, to lative (Thapar 1967: 25).
the third/second or possibly 1st century bc; at Nasik, to Twenty-three sherds of NBP were recorded at Paithan,
400–200 bc; at Maheshwar, 400 to 100 bc; and at Ter, 22 of them in layers allocated to Period 1; the remaining
between 200 bc and the 1st century bc (Sankalia and Deo sherd was found in a layer allocated to Period 2 and must
1955: 28; Sankalia et al. 1958: 19, 22; Sankalia et al. 1960: therefore be residual (Tables 7.3 and 7.4).
69; Thapar 1967: 74–75; Chapekar 1969: 17). However,
close inspection of these reports, as well as reports on
other sites in the region where NBP has occurred, such Red Polished ware
as Brahmapuri, Mansar, Pauni and Nagara, reveals that
none of these sequences contains any reliable dating This is another well-known ware about which much has
evidence with which to establish the first or latest occur- been written (Pinto Orton 1991). It has a very fine, well-lev-
rence of NBP. Bearing in mind the lack of reliable evi- igated, brick-red body covered by a thin orange-red slip,
dence for the end of production in the Ganges area, this which is often burnished. The fracture is smooth with
means that, at present, in the western Deccan NBP can, no visible inclusions. Mica is visible on the surface. The
strictly speaking, only be used as a terminus post quem most common forms are a carinated pot with an almost
to date the layers in which it occurs to the mid–6th bc or horizontal out-turned rim that is notched on its outer
later. Of course it is almost certain that it came into use face and a sprinkler (Figs 7.2–7.3).
somewhat later in the western Deccan, and, to judge from Red Polished ware (RPW) is found over much of
excavation reports, there was probably a period of time central and northern India, although some of it was
during which it circulated in reasonably large quantities, probably produced in Gujarat, perhaps in the region
but, at present, the chronology of these developments is around Amreli, the site which has yielded the greatest
108   Derek Kennet

number of types (Rao 1966: 51–59; Pinto Orton 1991). glaze and decorated with under-glaze cobalt. It is pos-
It was first defined at Baroda and is normally dated to sible to subdivide CBW into different styles based on
between the 1st and anywhere between the 3rd and the the decorative scheme (e.g. Swatow and Kraak) (Figs
5th centuries ad, although sometimes as late as the late 7.2 and 7.4).
sixth (Subbarao 1953: 56–64; Sankalia et al. 1958: 161; Large-scale production and export of CBW from China
Rao 1966: 52–53; Whitehouse and Williamson 1973: did not begin until the first or second quarter of the 14th
39; Ghosh 1989, I: 259; Pinto Orton 1991: 46). The evi- century at the earliest (Medley 1976: 176; Guy 1986: 76;
dence for the first appearance of RPW in the western 1990: 26). It started to become common in elite circles
Deccan between the 1st bc and 3rd ad is demonstrated in the Near East at the end of the 14th century (Lane
in Period V at Nevasa, where it is associated with rea- 1947: 27–28; Gray 1948–49: 30; Krahl 1986, II: 482), but
sonable numbers of confirmed Roman amphorae of a it did not begin to be traded widely until the mid to late
known date (Sankalia et al. 1960: 69, 280–281, 307; 15th century and probably began to arrive in India in
Gupta et al. 2001; Tomber 2007: table 1, 979). There are quantity at around the same time (Krahl 1986, II: 533;
very few cases, however, where the dating evidence 1997: 154). It continued to be exported to the western
for its disappearance stands up to detailed critical Indian Ocean and Europe until the late 18th century,
scrutiny. For example, Rao proposes a ‘lower limit’ of when imitations began to be manufactured in Europe
the beginning of the 5th century based upon the fact (Lunsingh Scheurleer 1974: 37–38).
that a coin datable to ad 380 was found in one of the Thirty sherds of CBW occurred at Paithan, and their
layers containing RPW at Amreli (Rao 1966: 53), but presence was used to define layers of Period 4 (Tables
there is clearly no reason, based on this evidence, that 7.3 and 7.4). Of these, 16 sherds, including all but two of
RPW could not have continued in use much later. A those mentioned below, come from Trench B, between
more convincing case is Period III at Paunar, which is layers 402 and 435. One of the youngest sherds in this
dated to the 6th century and later by coins of the Kala- sequence (P1612) comes from the earliest layer (435),
churis and Vishnukundins and Vakataka-style sculp- indicating high levels of re-deposition. Of the Trench B
ture (Deo and Dhavalikar 1968: 7). No RPW is reported sherds, there are two that might be as early as the 15th
from layers 4 and 5 that make up Period III, whereas 52 century (P1614, P2165), and there are also two sherds of
sherds are reported from the preceding Period IIb 16th century Kraak porcelain (P2167, P2168). The rest
(Deo and Dhavalikar 1968: 47–69). Another is Prakash, of the material is datable to the 16th/17th centuries,
where RPW occurs only in levels 28–25 (Thapar 1967: with the exception of two sherds (P1612, P2164) of the
24), below levels containing figurines and coins dated Kangxi period (ad 1662–1722).
to the 4th to 8th centuries, suggesting that it was out In addition, there is one Chinese sherd that is not
of circulation by the 6th/7th centuries. On the other CBW, but that might conceivably be significantly older:
hand, the evidence from the excavations at Kush in P1613. This might possibly be a Southern Song Jingdez-
the United Arab Emirates (UAE) has suggested that hen product of the 12th to 13th century, but it is prob-
RPW may have continued to circulate as late as the 8th ably a Ming or Qing period Fujian white ware. It was
century ad (Kennet 2004a: 65–66). It is possible that found in context 435 in Trench B associated with sherd
there are regional differences in the distribution of P1612, which is 17th to 18th century in date, along with
RPW and that Gujarati RPW continued to be manufac- other sherds mentioned above that are datable to the
tured and traded by sea later than RPW circulated in 15th to 17th centuries. Sherds P2173 and P2174 (both
the western Deccan. datable to the Wanli period, ad 1573–1620) come from
Only 10 sherds of RPW were recorded at Paithan, Trench A South Area.
three occurred in Period 2, and the others occur in
layers allocated to Period 3-temple, Period 4 or in layers
that were not allocated a period (Tables 7.3 and 7.4). Blue and White Frit (PERS)

This ware is made of a good-quality thick frit or stone-

Chinese wares (CBW and other) paste body; the interior and the exterior of the vessels
are coated in a transparent glaze that covers under-
All but one of the sherds of Chinese pottery are Chinese glaze painting in cobalt blue. It is often mistaken for
Blue and White porcelain (CBW), which is a well- Chinese porcelain, although the body is notably differ-
known class of porcelain, covered with a transparent ent, as are the quality of the glaze and decoration. Frit
 Chapter 7 The Pottery    109

Table 7.3: Summary of selected ceramic classes from Paithan by period (rim-sherd counts above; percentages of total
period rim-sherd assemblages below).

Ware Period 1 2 3 3-temple 4 Total

Black and Red 878 569 44 21 72 1,584

Black Burnished 19 12 3 1 35

NBP 5 5

RPW 1 1 2

GREY 3 11 195 5 144 358

PERS 1 1

WARP 69 69

CHIN 12 12

Other 4,271 4,831 1,256 378 1,784 12,520

Total 5,176 5,424 1,495 407 2,084 14,586

% % % % %

Black and Red 16.96 10.49 2.94 5.16 3.45

Black Burnished 0.37 0.22 0.74 0.05

NBP 0.10

RPW 0.02 0.05

GREY 0.06 0.20 13.04 1.23 6.91

PERS 0.05

WARP 3.31

CHIN 0.58

Table 7.4: Summary of body sherd counts of by period for those wares for which body sherds were counted separately.

1 2 3 3-temple 4 No period Total

Jorwe 8 8

NBP 17 1 18

RPW 2 2 3 1 8

CHIN 18 18

PERS 3 1 4

SLIP 7 7

WARP 6 6

Other 21,518 25,290 9,260 3,702 12,261 101 72,132

Total 72,201
110   Derek Kennet

wares were produced in the Near East from about the 12th
century onwards, but this style of frit ware, which closely
imitates CBW porcelain, was produced in the Near East
from the 14th until about the 19th century (Soustiel 1985:
214), and this is supported by excavation data from the
UAE (Kennet 2004a: FRIT.BW). Only five sherds of PERS
occurred at Paithan, all in layers allocated either to Period 4
or for which it was not possible to allocate a period (Tables
7.3 and 7.4, Fig. 7.5).

Fig. 7.3: Red Polished ware sherds (RPW).

Fig. 7.4: Chinese Blue and White porcelain sherds (CBW).

Slip-painted ware (SLIP)

White and Red Painted (WARP)
Types: 51, 52.
This is a glazed ware with a coarse red earthenware
body that has a fine sandy texture. It has a smooth
This is an unglazed painted earthenware that was
fracture and the body is fired an even red (2.5YR 5/8).
defined by this study. It consists of small closed
The seven sherds retrieved all come from small closed
vessels such as jars that are made of Fabric 5 and are
vessels such as jars. They are all covered on the outside
decorated on the outside and over the rim with a thin,
with a transparent lead glaze that contains some
white mat paint that is then decorated crudely with
splashes of green. This covers decoration painted in
thin horizontal red bands of paint (Fig. 7.7).
a thick, cream yellow slip. The motives are lines and
This class includes types 51 and 52 as well as a
arches and foliage. There is no glazing or decoration
number of unclassified body sherds. Seventy-five
on the interior surface or on the top of the flattened rim
sherds were found at Paithan, these occurred only in
(Figs 7.2 and 7.6).
layers allocated to Period 4 where it was quite common
The date of this class is not known. Slip-painted
(Tables 7.3 and 7.4). The high number of rim-sherds rel-
ware first appeared in the Near East in around the 11th
ative to body sherds is caused by an anomalous find in
or 12th century ad, but these sherds are likely to be con-
context 435 of Trench B.
siderably later, probably datable to the 14th century or
somewhat later. At Paithan, only seven sherds were
recorded, all from layers allocated to Period 4 (Table
7.4).
 Chapter 7 The Pottery    111

Thick Grey wares (GREY) ‘Darkish-Grey ware’ (Prakash). They tend to appear in
Types 14, 15, 24, 26, 44, 45, 46. large quantities in the later Early Historic or Early Medi-
eval layers, for example at Nevasa in Period VI (14th–
These types make up are very distinctive class of 18th century ad), at Maheshwar in Period VI (ad 100–
pottery that is known from excavations elsewhere in 500) and at Prakash in Period IV (6th–11th century ad),
the region. They are all made of Fabric 2 and are large, although there is no clear evidence to show precisely
grey, sturdy, mostly thick-walled carinated jars with at which date they first began to appear; the dating
everted rims of a variety of shapes. Most are burnished of some of the periods mentioned is also in need of
and many are decorated with ribbing on the outside. review (Sankalia et al. 1958: 137, 140, 150–151, 154–155, 163;
The types defined in this ware are all highly variable, Sankalia et al. 1960: 333; Thapar 1967: 15, 89, 93; see also
suggesting a low level of standardisation. This might the discussion below related to types 14 and 24). Given the
reflect the relatively informal organization of the obvious significance of this ware to the archaeology
industry that produced these wares. of the Early Medieval period, it is surprising that it is
not more widely discussed in the literature (e.g. Mehta
1979: 42–49; Ghosh 1989, I: 260).

Fig. 7.7: White and Red Painted ware sherds (WARP).

Fig. 7.5: Blue and White Frit sherds (PERS).

At Paithan, Thick Grey ware was a very significant

and obvious component of the Period 3 assemblage
of which it made up over 13% (Table 7.3). Indeed, its
appearance in the sequence was used to mark Period
3. This holds true with the exception of 14 sherds, three
from Period 1 and 11 from Period 2, that are considered
to be intrusive, although it should be noted that at
Maheshwar, it occurred in smaller quantities through
the Early Historic levels (Sankalia et al. 1958: 150). This
suggests that these wares may have been in existence
from earlier but have become more abundant towards
the end of the Early Historic period. They are poten-
tially extremely important wares for our understand-
ing of the Early Historic/Early Medieval transition.
If the dating of the North Temple here at Paithan
Fig. 7.6: Slip-painted ware sherds (SLIP). is accepted, Thick Grey wares can be shown to have
been in circulation by the 4th or 5th century or shortly
At other sites in the region where they have been thereafter.
encountered, these types are referred to as ‘Coarse
Black ware’, ‘Black ware’ (Nevasa, Maheshwar) or
112   Derek Kennet

Black Burnished ware (BBW) it occurs as late as the Early Historic period in a series of
Types 11, 29, 39, 58. forms similar to those found at Paithan (Ghosh 1989, I:
94–95, 234–238).
This ware consists of sherds made of Fabric 3 that are At Paithan, it consists entirely of wheel-made,
entirely black on the interior and exterior surfaces. slightly closed bowls that are made of Fabric 3 and are
The surface treatment never varies, being heavily bur- fired black with the exception of the lower part of the
nished in all cases. This ware is made of exactly the exterior, from about 3 cm below the rim, where it is
same fabric as Black and Red ware and, aside from the fired to between a yellowish red and a reddish yellow
absence of ‘inverted firing’, shares a very similar firing (5YR 5/6 to 7.5YR 7/6) due to having been fired upside
technique. The forms, however, are quite different from down. At Paithan, the surface treatment never varies,
Black and Red ware, consisting of small, closed vessels being heavily burnished in all cases, although at other
such as jugs, jars and one or two bowls, although no sites, for example Nevasa and Maheshwar, a few coarse
complete profiles came to light. and unburnished examples are also reported (Sankalia
Pottery described as ‘burnished black’, ‘black et al. 1958: 134; Sankalia et al. 1960: 259, 276).
slipped’ or ‘black-slipped burnished ware’ has been At sites where it is recorded, Black and Red ware
found at other sites in the region, for example Mahesh- seems to decline in abundance from a peak in Early
war and Nevasa (Sankalia et al. 1958: 153–154; Sanka- Historic levels, dwindling to much lower quantities by
lia et al. 1960: 278, 301–306, 323). At Maheshwar, it Medieval times, which is exactly the pattern observed
is first mentioned in Period VI (ad 100–500) and is at Paithan (see below). For example, at Maheshwar,
said to have been most common in Period VII, the so- Black and Red ware made up around c. 14.7%+ of the
called ‘Muslim-Maratha phase’ (Sankalia et al. 1958: Period IV assemblage (400–100 bc), 11.4% of the Period
151), whilst at Nevasa, it first appears in Period V (50 V assemblage (100 bc–ad 200), 8.4% of the Period VI
bc–ad 200), where it is represented by ‘a few sherds’, assemblage (ad 200–500) and only a few sherds of
and then is said to have become the ‘standard ware’ the Period VII assemblage (Medieval) (Sankalia et al.
of Period VI (‘Muslim-Maratha’), although the report 1958: 137, 143, 150, 162). At Nevasa, it was present in
contradicts itself further on by stating that red ware Period IV (150 bc–ad 150), made up a lower quantity in
was more common in Period VI (Sankalia et al. 1960: Period V (ad 50–200) and was not mentioned in Period
278, 323). VI (ad 1400+) (Sankalia et al. 1960: 259, 276, 322–324).
At both sites, the forms present in these wares and At Prakash, in Period II (600–100 bc), Black and Red
their occurrence through the sites’ sequences seem to made up a total of 34.83%, declining from 55.48% to
associate them with the ‘Coarse Black ware’ and ‘Black 9.09% through that period, whilst in Periods III (ad
ware’ mentioned above under Thick Grey wares, whilst 50–200) and IV, it was not mentioned (Thapar 1967:
at Paithan, their fabric seems to associate them with 27, 84–93). At Bhokardan, in Period IA (early Satava-
Black and Red ware (below), as does the fact that they hana), it made up ‘much less than other wares’, whilst
occur predominantly in Periods 1 and 2 together with a in Period IB (late Satavahana), it made up a ‘large
few sherds from later layers that are probably residual number of sherds’, and in Period II (post-Satavahana),
(Table 7.3). ‘less bulk’ (Deo and Gupte 1974: 91–92, 103–104, 110).
These figures can be compared with Paithan, where
Black and Red ware made up the following proportions
Black and Red ware of the period-assemblages: 16.96% in Period 1, 10.49%
Types 10, 38. in Period 2, 2.9% in Period 3 and 3.45% in Period 4. It
is possible to see a broadly similar decline in the circu-
This is a well-known ware, or perhaps ‘manufacturing lation of this ware at all sites in the region, although
technique’ would be a better term, that has been recorded there are some, for example Bhokardhan, where the
from most Early Historic sites that have been excavated pattern does not seem to hold up well, possibly due to
in the region as well as more widely across India (e.g. imprecise quantification.
Sankalia et al. 1958: 134–135, 141–143, 144–146, 150, 152–
153, 162, 164; Sankalia et al. 1960: 259–263, 276, 282–286;
Thapar 1967: 68–71; Deo and Gupte 1974: 91–92, 103–104,
110). The ware has a long history, possibly stretching
back as early as the Harappan period in some places, but
 Chapter 7 The Pottery    113

Other wares reality, there are marked variations in colour, fabric,

surface treatment, form and manufacture amongst
Table 7.3 shows the occurrence of each ware by period these materials, and for this reason, ‘red ware’ was not
as rim-sherd counts and percentages of total period rim- further used as a category in this study. This presented
sherd counts. The percentages are based on the number the problem of how to classify the remaining 80% to
of rims as the majority of body sherds were not classified 90% of the ‘red’, ‘buff ’ and ‘tan’ coloured rim-sherd
according to ware. It should also be noted that many assemblage that remained after the wares that have
rim-sherds were not allocated a ware type, as has been been defined above had been removed. As explained
noted above. These sherds are included in the group above, after a considerable amount of trial and error, it
called ‘Other’ in the table. These sherds were, however, was decided that it was not possible, using the human
classified by type and their occurrence is shown in Table eye and a ×10 hand lens, to define wares within this
7.5. Whilst most body sherds were not classified by ware, material with characteristics that were consistent,
those belonging to particular diagnostic wares such coherent and usable. At the same time, it was noted
as NBP, RPW, etc., were also recorded. These are not that it was possible to subdivide the forms and shapes
included in the percentage analysis in Table 7.3 as they of the vessels into consistent and easily recognizable
would obviously bias the trends elucidated there, but types. It was therefore decided that a form- and type-
these sherds are listed in Table 7.4 so that their occur- based approach was the most effective method of clas-
rence through the sequence can be seen. sifying the rim-sherd assemblage from the excava-
Table 7.3 makes clear the relative significance of tions. The resulting typology is presented below.
different wares through the sequence. Particularly
notable are the decline in Black and Red ware from
Period 1 onwards and the increase in Thick Grey ware
in Period 3. One very obvious absence from Table 7.3
Typology
and the discussion above is the group of wares that
Fifty-six separate types were defined, all of which are
are normally described as ‘red’, or ‘coarse red’ or ‘red
based on rim forms. Each type is described below and ref-
slipped’ wares. Such wares normally make up a large
erence is given to illustrations in Figs 7.8 to 7.18. A small
proportion of the Early Historic assemblages from sites
graph is included at the end of the chapter showing the
in the region; for example in Period IV at Mahesh-
‘occurrence profile’, or the percentage each type makes up
war, ‘coarse red’ and ‘red slipped’ made up 47.4% and
of the period rim assemblages through the sequence. The
29.3%, respectively, of the assemblage, whilst in Period
data upon which these graphs are based are presented
IV at Nevasa, a ‘wide range of shapes’ are recorded in
in Table 7.5. Tables 7.6 to 7.11 also show the occurrence of
coarse and slipped red pottery, whilst ‘coarse red ware’
types through the sequences of each of the main trenches,
is reported as being the most common ware in Period V
context by context.
(Sankalia et al. 1958: 137; Sankalia et al. 1960: 256, 277).
In order to avoid repetition, where types are made
The same is true at Prakash, where red ware is reported
of Fabrics 1 to 5, which have been described above, this
as making up most of the ‘other wares’ that themselves
is simply indicated. Where the fabric is unique to the
make up almost 65% of the Period II assemblage, whilst
type, then a full description is given below, with the
the Period III assemblage is described as being ‘wholly
type description. It should be noted that all vessels are
of monochrome red ware’ (Thapar 1967: 27, 75, 84). As
wheel-made unless otherwise stated.
can be seen in Table 7.2, at Paithan, rim-sherds that tra-
It is acknowledged that the approach adopted here
ditionally would have been defined as ‘red wares’ (e.g.
is at variance with that generally taken in the study
Fabrics 1a, 1b and 1c – which include at least five differ-
of Early Historic pottery in India. For this reason, an
ent combinations of slipping and burnishing) made up
attempt has been made to suggest what nomenclature
roughly 76% of Period 1, 78% of Period 2, 60% of Period
might have been used for each type in the more com-
3 and 51% of Period 4 rim assemblages. Although the
monly used classification system, for example ‘coarse
‘Red ware’ category was not used at Paithan because
red ware’, ‘slipped red ware’, etc. This is intended
it is too broad and poorly defined to be reliable, these
to provide a link if it should be needed; it should be
approximate figures are given here to allow rough com-
stressed that such terms are only for rough guidance
parisons to be made with other sites.
and are not precisely defined.
However, in most of these cases, ‘red ware’ is
little more than a crude description of clay colour. In
114   Derek Kennet

Table 7.5: Types by period based on rim-sherd count (left) and percentages of total rim-sherd count (right). ‘UNIQ’ means
unique sherd; the most important of these are described in Table 7.15.

Type 1 2 3 3-temple 4 Total Type 1 2 3 3-temple 4

1 2,005 708 440 299 516 3,968 1 38.74 13.05 29.43 73.46 24.76
10 853 569 44 21 72 1,559 10 16.48 10.49 2.94 5.16 3.45
1A 625 1,503 82 4 2,214 1A 12.07 27.71 5.48 0.98
1B 314 1,002 160 3 1,479 1B 6.07 18.47 10.70 0.74
2 212 21 5 30 268 2 4.10 0.39 0.33 1.44
4 175 174 15 3 57 424 4 3.38 3.21 1.00 0.74 2.74
8 115 84 5 21 225 8 2.22 1.55 0.33 1.01
3 110 159 18 30 317 3 2.13 2.93 1.20 1.44
16 105 171 36 9 73 394 16 2.03 3.15 2.41 2.21 3.50
9 63 75 9 147 9 1.22 1.38 0.60
5 46 52 10 15 123 5 0.89 0.96 0.67 0.72
43 45 20 19 3 11 98 43 0.87 0.37 1.27 0.74 0.53
30 27 13 2 42 30 0.52 0.24 0.10
38 25 25 38 0.48
6 21 16 1 23 61 6 0.41 0.29 0.25 1.10
33 16 39 8 1 11 75 33 0.31 0.72 0.54 0.25 0.53
36 12 6 18 36 0.23 0.11
40 12 10 3 25 40 0.23 0.18 0.14
42 9 4 13 42 0.17 0.07
58 9 1 10 58 0.17 0.02
29 7 11 3 1 22 29 0.14 0.20 0.74 0.05
NBP 5 5 NBP 0.10
20 4 20 13 72 109 20 0.08 0.37 0.87 3.45
25 4 10 19 13 46 25 0.08 0.18 1.27 0.62
24 3 8 153 1 4 169 24 0.06 0.15 10.23 0.25 0.19
39 3 3 39 0.06
41 3 39 36 78 41 0.06 0.72 1.73
35 2 5 3 10 35 0.04 0.09 0.20
18 1 4 2 1 8 18 0.02 0.07 0.13 0.05
55 1 24 25 55 0.02 1.15
57 1 15 16 57 0.02 0.72
32 41 8 2 9 60 32 0.76 0.54 0.49 0.43
7 26 6 1 23 56 7 0.48 0.40 0.25 1.10
31 15 4 19 31 0.28 0.27
19 14 14 19 0.26
37 13 13 37 0.24
34 5 3 8 34 0.09 0.20
12 2 4 10 16 12 0.04 0.27 0.48
26 2 27 4 33 26 0.04 1.81 0.98
14 1 10 19 30 14 0.02 0.67 0.91
22 1 2 1 2 6 22 0.02 0.13 0.25 0.10
27 1 19 1 3 24 27 0.02 1.27 0.25 0.14
RPW 1 1 2 RPW 0.02 0.05
50 7 5 12 50 0.47 0.24
44 5 65 70 44 0.33 3.12
47 5 34 39 47 0.33 1.63
48 1 125 126 48 0.07 6.00
51 62 62 51 2.98
46 31 31 46 1.49
13 30 30 13 1.44
45 19 19 45 0.91
49 16 16 49 0.77
53 15 15 53 0.72
 Chapter 7 The Pottery    115

Type 1 2 3 3-temple 4 Total Type 1 2 3 3-temple 4

CHIN 12 12 CHIN 0.58

54 11 11 54 0.53
23 8 8 23 0.38
15 7 7 15 0.34
56 7 7 56 0.34
52 5 5 52 0.24
PERS 1 1 PERS 0.05
UNIQ 8 21 23 30 82 UNIQ 0.15 0.39 1.54 1.44
NON 335 557 330 50 504 1,776 NON 6.47 10.27 22.07 12.29 24.18
Total 5,176 5,424 1,495 407 2,084 14,586

Throughout the description of types below, reference is probably anomalous due to the small size of
is occasionally made to whether the absence of a par- the assemblage in that period. There is a marked
ticular type is ‘significant’ or not. In this context, ‘sig- decline in Period 4, many of the sherds from which
nificant’ means that three sherds or more of the type may be residual, although it seems unlikely that the
would have been expected in the period assemblage type ever went completely out of use. The smaller
concerned, given the size of that assemblage and the version (Type 1a) is more common in Periods 1 and
proportion that the type makes up of other period- 2, but this changes in Period 3, where the larger
assemblages where it is found. version (Type 1b) becomes more common.
–– External parallels: Given the frequency of Type 1
at Paithan, it is surprising that it does not receive
Type: 1 greater mention in the published pottery reports
Traditional ware name: Coarse red ware from other sites in the region. This may be because
it is a ubiquitous, rather simple, shape with a broad
–– Description: This is a very common, coarse- chronological range, but it may also be the case
ware bowl with nearly straight sides and a simple that this type is particularly common at Paithan.
pointed rim that is sometimes slightly thickened It is frequently found at Brahmapuri (Sankalia and
and upturned. The diameter is normally 10–12 cm, Dikshit 1952: fig. 18, type 52), whereas at other sites,
although up to 18 cm has been observed. The bowl there is usually no indication of how common this
has a flat, string-cut base. Type 1 was subdivided type is, and some of the illustrated types may have
into two subtypes, Types 1a and 1b, based on size been single examples. For example in Period V at
(see below). Nevasa, there are two types (76 and 97) that are
–– Ware: Fabric 1a. broadly comparable in terms of shape, although
–– Illustration: Fig. 7.8. the size of 76 is different, as is the fact that it is red
–– Subtypes: Type 1a is a small version with a maxi­ slipped (Sankalia et al. 1960: figs 129 and 138). At
mum wall thickness of 5.5 mm and a rim diam- Nasik, which is considerably further distant than
eter of 7–12 cm. Type 1b is a large version with a Nevasa, better parallels can be found in Period IIA
minimum wall thickness of 5.5 mm and a rim diam- in types 13, 13a and 13b, and also types 62 and 64 in
eter of 12–20 cm, averaging about 17 cm. Period IV (Sankalia and Deo 1955: figs 16 and 31).
–– Internal dating evidence: Fig. 7.23 shows the occur- At Bhokardan, parallels can be found in Periods IA
rence of Types 1a and 1b as well as the total for all and IB (Deo and Gupte 1974: fig. 5 type 1 and 1a, fig.
Type 1. Type 1 is the most common type recorded in 11 type 63), at Ter in Periods I and II (Chapekar 1969:
the assemblage, making up between 50% and 60% fig. 7 type 11b, fig. 15 type 33), at Prakash in Periods
of the rim assemblage in Periods 1 and 2, with a II and IV (Thapar 1967: fig. 23 type 1 and 1a, fig. 29
decline to 45% and 25% in Periods 3 and 4, respec- type 1) and at Maheshwar in Period IV (Sankalia et
tively. The very high percentage in Period 3-temple al. 1958: fig. 60 type 82).
116   Derek Kennet

Type: 2 and c, fig. 10 type 4c, fig. 11 type 5 and fig. 12. type
Traditional ware name: Coarse red ware 6c; Sankalia et al. 1958: fig. 74 type 103; Sankalia et
al. 1960: fig. 118 types 46 and 47; Thapar 1967: fig.
–– Description: This is a small jar with an everted, 28 type 19a; Chapekar 1969: fig. 4 types 2d and 3a,
slightly thickened rim above a short neck. There is fig. 15 type 30f; Deo and Gupte 1974: fig. 5 type 7b,
some variation in the shape of the rim. The diam- fig. 6 type 17b, fig. 7 types 24, 24a and 24b, fig. 12
eter is about 10 cm. type 71b). The fact that this type was found only in
–– Ware: Fabric 1a. the later periods at Maheshwar and Prakash seems
–– Illustration: Fig. 7.8. to go against the evidence for it being a predomi-
–– Internal dating evidence: This type was very nantly early type at Paithan.
common in Period 1 but shows a marked decline
after that time. It seems likely that it only occurred
in Period 1 and that the few sherds from later Type: 4
periods are residual (Fig. 7.23). Traditional ware name: Slipped red ware
–– External parallels: This is a reasonably common
type in Period 1 at Paithan, but it is not possible –– Description: This is a small jar with a thickened,
to find reliable parallels in other published assem- outward-curving rim marked by a distinctive notch
blages in the region. This may be because this type is on the interior. The diameter is normally about
not particularly distinctive and is difficult to identify 12–15 cm. The surface is covered with a thin red
with certainty in published drawings. slip/wash, which is more evident on the exterior.
There are marks of wet clay on the surface. This
type is often very similar to Type 16.
Type: 3 –– Ware: Fabric 1a.
Traditional ware name: Burnished and slipped red ware –– Illustration: Fig. 7.8.
–– Internal dating evidence: This type has a strange
–– Description: This is a large jar with a thick, everted occurrence profile: it was abundant in Periods 1
rim (diameter 14–16 cm), normally rounded but and 2 but appears to have ceased circulating after
also occasionally squared or triangular. There is that time, although a significant number of sherds
always a distinctive notch on the inside just below was also found in Period 4 layers. These are prob-
the top of the rim. In this respect, it is similar to ably residual and reflect the high levels of residu-
Type 4. All examples have a thick, polished, red ality noted in that period (Fig. 7.23).
slip/paint on the interior and exterior surfaces. –– Comments: One fragment of Type 4 comes from
There is also occasionally some light ribbing on context 738, low down in the Trench D2 sequence.
the outside. The size seems to vary considerably. This is unusual, being thinner than normal and
There are also similarities with Types 7, 13 and 25. having a polished red slip, which Type 4 never nor-
–– Ware: Fabric 1a. mally has meaning that it is probably a mis-iden-
–– Illustration: Fig. 7.8. tification.
–– Internal dating evidence: This type makes up
between 2% and 3% of the assemblage in Periods 1
and 2, but there is a sharp decline in the quantity Type: 5
of this type after Period 2, suggesting that the type Traditional ware name: Slipped red ware
may have gone out of circulation at that time. The
sherds from Periods 3 and 4 are probably residual –– Description: This is a lid with a pinched edge and
(Fig. 7.23). high ridge. The diameter of the widest part is about
–– External parallels: This is a distinctive and 14 cm. It is coated with a fine, matt red slip/wash on
common type at Paithan, so it is no surprise that both the interior and exterior.
parallels can be found at other sites in the region, –– Ware: Fabric 1a.
for example in Period IIa at Nasik, Period IV at –– Illustration: Fig. 7.8.
Nevasa, Periods 1A and 1B at Bhokardan, Periods –– Internal dating evidence: The percentage of this
I and II at Ter, Period III at Prakash and Period V at type declined after Period 2, but it is not possible to
Maheshwar (Sankalia and Deo 1955: fig. 9 type 3a be certain whether sherds from Periods 3 and 4 are
 Chapter 7 The Pottery    117

Fig. 7.8: Types 1 to 7.

118   Derek Kennet

residual or not. They are quite abundant, probably Type: 7

indicating that the type had a very long life span Traditional ware name: Burnished and slipped red ware
extending right into Period 4. The type’s absence
from Period 3-temple is to be explained by the –– Description: This is a relatively rare but very dis-
small size of the assemblage in that period. tinctive jar with an everted, thickened rim, the
–– External parallels: Parallels can be found at most external vertical face of which is marked with two
sites in the region, for example at Brahmapuri in or three deep horizontal notches. The diameter is
‘late Satavahana’ and ‘Brahmani’ levels, at Bhok- 12–20 cm. The vessel is covered with a thick pol-
ardan in Periods IA and 1B, at Nasik in Periods ished red paint/slip on both surfaces.
IIA and IV, at Ter in Periods I and II, at Prakash in –– Ware: Fabric 1b.
Periods II and III and at Maheshwar in Period V –– Illustration: Fig. 7.8.
(Sankalia and Dikshit 1952: fig. 18 type 56; Sankalia –– Internal dating evidence: This form only occurs
et al. 1958: fig. 71 type 98; Sankalia and Deo 1955: from Period 2 onwards and is very abundant in
fig. 20 types 21–24, fig. 33 types 73–80; Thapar 1967: Period 4, which may be linked to the high levels
fig. 23 type 8, fig. 27 type 1; Chapekar 1969: fig. 6 type of residuality in that period. This type’s clearly
5, fig. 13 type 26, fig. 17 type 34; Deo and Gupte 1974: significant absence from Period 1 makes it a useful
fig. 6 type 10, fig. 12 type 73). chronological marker at Paithan (Fig. 7.23).
–– Comments: The presence of this form was used to
make the distinction between levels of Period 1 and
Type: 6 Period 2 when the sequence was periodised (see
Traditional ware name: Burnished and slipped red ware Chapter 6).
–– External parallels: This type also occurs at Nevasa
–– Description: This is a large jar or vase with a thick in Period IV, at Bhokardan in period 1A, at Nasik in
wall and a very heavy, thick everted rim with a Periods IIA and IV, at Ter in Period I and at Prakash
diameter of about 24 cm. Internal and external sur- in Period II (Sankalia and Deo 1955: fig. 9 type 3e–h,
faces are both coated with a thick polished red slip/ fig. 21 types 26a and b, fig. 35 type 94; Sankalia et
paint, which is often quite well preserved. al. 1960: fig. 118 type 46c; Deo and Gupte 1974: fig.
–– Ware: Fabric 1b. 6 type 18, fig. 8 type 41a; Thapar 1967: fig. 25 type
–– Illustration: Fig. 7.8. 30; Chapekar 1969: fig. 6 type 6a–c).
–– Internal dating evidence: This type has a prob-
lematic occurrence profile. It certainly circulated
during Periods 1 and 2 and then seems to have dis- Type: 8
appeared in Period 3. Period 4, however, yielded a Traditional ware name: Slipped red ware
higher percentage of this type than either Period
1 or 2. This is difficult to account for; it may be –– Description: This is a wide bowl with an almost ver-
explained by the generally high levels of residu- tical, in-turned rim that is carinated and slightly
ality in Period 4 or it may be related to problems thickened on the carination, sometimes enough to
identifying this type in some cases. make the profile almost triangular. The rim diam-
–– Comments: The distinction between this type and eter is around 28 cm. The vessel is often covered
Type 9 is not always clear. with a thin, matt slip similar to Type 5, but this is
–– External parallels: Parallels for this type are not never polished.
common; roughly similar types have been pub- –– Ware: Fabric 1a.
lished from Ter in Period II and at Prakash in –– Illustration: Fig. 7.9.
Period II (Thapar 1967: fig. 25 type 25; Chapekar –– Internal dating evidence: The occurrence profile
1969: fig. 10 type 20e). of this type shows a clear and marked decline in
abundance from Period 1 onwards, with a slight
increase in Period 4 that is probably due to the
high levels of residuality in that period. Fig. 7.23
shows a histogram of the occurrence of this type
through the sequence.
 Chapter 7 The Pottery    119

–– External parallels: It is not always possible to be –– Internal dating evidence: This type declines very
certain of identifications of this type made from markedly after Period 1. It probably went out of use
published drawings. Probable parallels have been during the course of Period 2 and the sherds from
found at Nevasa in Period IV, Nasik in Periods IIA Periods 3 and 4 are probably residual (Fig. 7.24).
and IV, Ter in Period II, Prakash in Period III and –– External parallels: This is an extremely common
Maheshwar in Period V (Sankalia and Deo 1955: type and has been found at all sites in the region
figs 17 and 17a types 17d–j, figs 18–19 types 18 and where Black and Red ware has been encountered,
19; Sankalia et al. 1958: fig. 72 type 99, fig. 73 type for example Nevasa Periods IV and V, Bhokardan
100; Sankalia et al. 1960: fig. 119 type 49; Thapar Periods IA, IB and II, Nasik Period IIA, Ter Period
1967: fig. 27 type 4; Chapekar 1969: fig. 12 type 23d). I, Prakash Period II and Maheshwar Period IV
(Sankalia and Deo 1955: figs 26 and 37; Thapar 1967:
fig. 20; Sankalia et al. 1958: fig. 63 type 89; Sankalia
Type: 9 et al. 1960: fig. 115 type 28e–i, fig. 124 fig. 59a and
Traditional ware name: Burnished and slipped red ware b; Chapekar 1969: fig. 6 type 15; Deo and Gupte 1974:
fig. 9 types 50–51, fig. 15 types 115–116, fig. 17 type
–– Description: This is a bowl with tightly curved sides 138).
and a thick, broad, everted rim that is slightly thick- –– Note: Two of the illustrated sherds (1766 and
ened and rounded at the end and with a diameter of 863) have crude incised symbols on their exteri-
16 cm or more in some cases. The surface is some- ors just below the rim. This arrow-like symbol is
times covered with a polished red slip/wash. known from other sites in the region; for example
–– Ware: Fabric 1b. at Prakash, it was found in Period I on a sherd of
–– Illustration: Fig. 7.9. Jorwe ware (Thapar 1967: 67, fig. 19 8), whilst at
–– Internal dating evidence: Type 9 almost certainly Nasik, it was found on a sherd of red ware whose
ceased to circulate after Period 2 or early in Period period is not reported (Sankalia and Deo 1955: 84,
3. The few sherds found in Period 3 levels are prob- fig. 8a 1).
ably residual (Fig. 7.24).
–– Comments: During the course of the pottery study, it
became clear that Types 6 and 9 are probably differ- Type: 11
ent sizes of the same type, Type 9 being the smaller (Black Burnished ware) Traditional ware name:
of the two. Burnished black ware
–– External parallels: Similar types are reported from
Nevasa Period V, Nasik Period IIA, Ter Period II –– Description: This is a small jar with a small everted
and Maheshwar Period IV (Sankalia and Deo 1955: triangular rim, the diameter of which is about 12
fig. 15 type 10c; Sankalia et al. 1958: fig. 60 type 83; cm. The surface is heavily burnished. The type is
Sankalia et al. 1960: fig. 129 type 74, fig. 135 type 86; represented only by three small sherds.
Chapekar 1969: fig. 16 type 32g). –– Ware: Fabric 3.
–– Illustration: Fig. 7.9.
–– Internal dating evidence: No examples of this type
Type: 10 were found in the stratified sequence (the illus-
(Black and Red ware) Traditional ware name: Black and trated example is from an unstratified layer).
Red ware

–– Description: This is the more common of the two

Black and Red ware forms. It is a simple, almost
vertically sided bowl/cooking pot. It is thin-walled
(up to 5 mm) with a heavily burnished gloss over
the slip both inside and out. The rim diameter is
normally between 17 and 20 cm but can be smaller.
–– Ware: Fabric 3.
–– Illustration: Fig. 7.9.
120   Derek Kennet

Fig. 7.9: Types 8 to 11.

 Chapter 7 The Pottery    121

Fig. 7.10: Types 12 to 15.

122   Derek Kennet

Type: 12 –– Illustration: Fig. 7.10.

Traditional ware name: Coarse red ware –– Internal dating evidence: Apart from one sherd in
Period 2, this type occurs only in Periods 3 and 4. Its
–– Description: This is a large carinated bowl with a absence from Period 3-temple is probably due to the
distinctive rounded and thickened rim that flares small size of that assemblage. Given that ten sherds
outwards above the carination. It is a coarse, thick- were found in Period 3, it is likely that its absence
walled vessel. The diameter is 25–30 cm. It is fired from Period 1 and limited presence in Period 2 are an
to a dark grey/black red. Traces of a surface wash accurate reflection of its dating (Fig. 7.24); see the
are only occasionally seen. This is quite a variable discussion under ‘Thick Grey ware’ above.
type. –– External parallels: As with Thick Grey wares, this
–– Ware: Fabric 1a. type tends to be more common in the later levels at
–– Illustration: Fig. 7.10. excavated sites, for example at Nevasa in Period IV,
–– Internal dating evidence: This is not a common at Nasik in Period IV and at Ter in Period II (Sanka-
type. There are two sherds from Period 2 but it is lia et al. 1960: fig. 118 types 48 and 48a; Sankalia
mostly restricted to Periods 3 and 4, particularly and Deo 1955: fig. 39 types 118–119; Sankalia et al.
the latter. Because the type is so rare, it is possible 1960: fig. 118 types 48 and 48a; Chapekar 1969: fig.
that its absence from Period 1 is due to chance. It 21 types 41–42).
may have been in circulation at that time (Fig. 7.24).

Type: 15
Type: 13 (Thick Grey ware) Traditional ware name: Grey ware
Traditional ware name: none
–– Description: This is a Thick Grey ware jar with an
–– Description: This is a large, broad-mouthed jar everted rim and ribbing below the rim on the exte-
with a squared, thickened, vertical rim. There is rior. The walls of the vessel around the rim tend to
a notch or indent on the interior and incised lines be rather thinner than in Type 14. The diameter is
on some examples below the rim. The diameter is about 18 cm. This type does also occur occasionally
about 17–18 cm. The burnished surface is preserved in red-washed ware.
on a number of examples, one or two are also red –– Ware: Fabric 2.
slipped/washed. –– Illustration: Fig. 7.10.
–– Ware: This is a type-specific fabric. It is a coarse –– Internal dating evidence: This type occurs only in
grey ware with occasional very small grit inclu- Period 4.
sions. The dense fabric is medium to soft-fired with
a smooth fracture and black body (7.5YR 2.5/1). One
Type: 16
example contains vegetable temper, and the frac-
Traditional ware name: Coarse red ware
ture shows a more laminar structure. This example
is partly fired to a reddish yellow (7.5YR 6/6). This
–– Description: This is a simple jar, the rim form is
fabric appears to be a slightly better levigated
similar to Type 4, the rim being smaller, more
and fired version of Fabric 1a.
everted and lacking the distinctive notch on the
–– Illustration: Fig. 7.10.
inside. The rim is quite variable, it is thickened,
–– Internal dating evidence: This type occurs only in
rounded and slightly triangular, even slightly
Period 4.
overhanging. The diameter is about 13 cm. It is not
slipped or washed in most cases.
Type: 14 –– Ware: Fabric 1a.
(Thick Grey ware) Traditional ware name: Grey ware –– Illustration: Fig. 7.11.
–– Internal dating evidence: This type occurs in all
–– Description: This is a Thick Grey ware jar or cooking periods and there is no obvious chronological pattern
pot with a heavy, everted rim and thick walls. The (Fig. 7.24).
diameter is about 25–30 cm. The exterior is normally –– Comments: It was noted during the study that this
burnished and decorated with incised lines. form appears to become finer in the later levels
–– Ware: Fabric 2.
 Chapter 7 The Pottery    123

Fig. 7.11: Types 16 to 22.

124   Derek Kennet

of Trench B, where the surface is more commonly is wholly absent from Trench B, which is puzzling,
either burnished or slipped. but must simply be due to chance.
–– External dating evidence: A roughly similar type
is published from Brahmapuri in levels reported to
Type: 17 date to the ‘late Satavahana’ period (Sankalia and
Traditional ware name: Coarse red ware Dikshit 1952: fig. 15 types 9 and 10).

–– Description: This is an almost vertically sided

beaker or small jar with a slightly everted rim and Type: 20
slightly curved sides. It is possibly a variation on Traditional ware name: none
Type 16. The diameter is about 12 to 14 cm.
–– Ware: Fabric 1a. –– Description: This is a small, thin-walled jar with a
–– Illustration: Fig. 7.11. finely shaped, everted rim. There is a carination on
–– Internal dating evidence: The rarity of this type the body in some examples. The diameter is nor-
makes it impossible to gain an accurate idea of its mally 8–10 cm. It is fine-walled with a red wash on
occurrence through the sequence. the smoothed exterior.
–– Ware: Fabric 4.
–– Illustration: Fig. 7.11.
Type: 18 –– Internal dating evidence: This type occurs in all
Traditional ware name: Burnished and slipped red ware periods, but it is very rare in Period 1 and only
becomes common in Period 4 (Fig. 7.24).
–– Description: This is a red or black ware jar with
a bifoliate everted rim. There is often an incised
notch on the interior. The diameter is about 14 cm. Type: 21
It is covered with a thick polished red slip/wash. Traditional ware name: none
–– Ware: Fabric 1a. This type is also occasionally fired
black. –– Description: This is a small beaker with a fine
–– Illustration: Fig. 7.11. everted rim and curved sides. The diameter is 6–8
–– Internal dating evidence: Only a few examples of cm. It has a thin body and a distinctive red slip/
this type were found; they become increasingly wash.
common from Period 1 to 3, after which time the –– Ware: Fabric 4.
type ceased to circulate. The few sherds found in –– Illustration: Fig. 7.11.
Period 4 are almost certainly residual. –– Internal dating evidence: No examples of this type
–– External parallels: A similar type is reported from occurred in the quantified sequence (the illus-
Nevasa in Period V (Sankalia et al. 1960: fig. 130 trated example is from an unstratified layer).
type 75, fig. 131 type 77).

Type: 22
Type: 19 Traditional ware name: Coarse red ware
Traditional ware name: Slipped red ware
–– Description: This is a small jar with an almost hori-
–– Description: This is a large plate with a slightly zontally everted, bifoliate rim. The diameter is
raised, flattened rim. The diameter is about 24 cm. about 12–13 cm.
There is a very thin matt red wash on the interior –– Ware: Fabric 1a. There is a slight snap to the break
which gives a distinctive appearance. with a smooth to hackly fracture and no vegetable
–– Ware: Fabric 1c. temper.
–– Illustration: Fig. 7.11. –– Illustration: Fig. 7.11.
–– Internal dating evidence: The presence of this form –– Internal dating evidence: Very few sherds of this
was used to make the distinction between Period type came to light. There is only one sherd in
1 and Period 2 when the sequence was periodised Period 2 and two in Periods 3 and 4. Because this
(see Chapter 6). It occurs only in Period 2. The type type is so rare, its absence from Period 1 cannot be
 Chapter 7 The Pottery    125

taken as an indication that it was not in circula- –– External parallels: Parallels to this type are most
tion at that time. commonly found in the later levels at excavated
–– External parallels: A similar type is reported from sites, for example at Nevasa in Periods V and VI,
Bhokardan Period IB (Deo and Gupte 1974: fig. 13 at Brahmapuri in levels dated ‘late Satavahana’
types 81a–c). or ‘Brahmani’, at Bhokardan in Periods IA, IB
and II, at Nasik in Periods IIA and IV, at Ter in
Period II and at Maheshwar in Period VI (Sankalia
Type: 23 and Dikshit 1952: fig. 17 type 37a, fig. 19 types 65–66;
Traditional ware name: none Sankalia and Deo 1955: fig. 27 type 40, fig. 39 types
120–121; Sankalia et al. 1958: fig. 78 type 119, fig. 79
–– Description: This is a broad-mouthed jar with an types 120–122; Sankalia et al. 1960: fig. 140 type
everted, rounded rim with a slight notch on the 104, fig. 157 type 127, fig. 158 type 129; Chapekar
interior. The diameter is about 18 cm. The surface is 1969: fig. 17 type 40; Deo and Gupte 1974: fig. 9 types
smoothed with a thin wash. 46–48, fig. 15 types 112–114, fig. 17 type 133).
–– Ware: This is a type-specific fabric. It is a dense,
well-levigated fabric with a smooth, almost sub-
conchoidal fracture. The body breaks with a slight Type: 25
snap and appears to be only medium fired. The clay Traditional ware name: Burnished and slipped red ware
has few inclusions apart from occasional laminar
particles of chaff or other vegetable temper and –– Description: This is very similar to Type 3 in almost
a very fine gritty structure that is caused by fine all respects, but it does not have an interior niche on
grains that are almost invisible even with a ×10 the rim. The diameter is around 14 cm.
hand lens. The body is an even reddish yellow –– Ware: Fabric 1a.
(7.5YR 7/8). –– Illustration: Fig. 7.12.
–– Illustration: Fig. 7.12. –– Internal dating evidence: This type occurs in all
–– Internal dating evidence: Eight sherds of this type periods but is most common in Period 3. Some of the
occur in Period 4. Period 4 sherds might be residual (Fig. 7.24).

Type: 24 Type: 26
(Thick Grey ware) Traditional ware name: Grey ware (Thick Grey ware) Traditional ware name: Grey ware

–– Description: This is an extremely variable type. –– Description: This type was used as a catch-all for
It is a jar with an everted rim that is rounded and very fragmentary rim sherds of thick grey polished
slightly thickened at the end. The outward turn of ware with a fairly heavy rolled rim. Some of these
the rim is marked by a sharp carination on the inte- may in fact be Type 24, but they are too fragmen-
rior. The rim diameter is between 16 and 25 cm, tary to be certain. The diameter is normally about
and the wall thickness is between 6 and 7 mm. The 20 cm.
jar was also sharply carinated around the middle of –– Ware: Fabric 2.
the body. The surface is burnished to a high polish –– Illustration: Fig. 7.12.
which looks like a slip. Turning marks are visible on –– Internal dating evidence: Apart from two sherds in
the surface. Although the basic concept of the jar is Period 2, which may be intrusive, this type is found
the same, there is a high degree of variability in the almost entirely in Period 3 and Period 3-temple.
details of the form and the shape of the rim. As the type is represented by 27 sherds in Period 3,
–– Ware: Fabric 2. it seems certain that its absence from Period 1 is an
–– Illustration: Fig. 7.12. indication that it was not in circulation at that time
–– Internal dating evidence: Small quantities of this (Fig. 7.25; see also the discussion under ‘Thick Grey
type occur in all periods, but it is very abundant in ware’ above).
Period 3 (Fig. 7.24; see the discussion under ‘Thick
Grey ware’ above).
126   Derek Kennet

Fig. 7.12: Types 23 to 27.

 Chapter 7 The Pottery    127

Type: 27 nally and externally. The diameter is about 15 cm.

Traditional ware name: Burnished and slipped red ware The surface is covered with an unpolished red slip.
–– Ware: Fabric 1a.
–– Description: This type has a slightly complex –– Illustration: Fig. 7.13.
everted rim with a niche on the upper surface. The –– Internal dating evidence: This type was very abun-
width of the everted part of the rim is somewhat dant in Period 1, but it declined in Period 2 and
variable, but the concept is the same. The diameter had completely disappeared from circulation by
is 12–14 cm. The surface of these sherds has a fine Period 3. The two sherds in Period 4 are probably
red polished slip on the interior and exterior. residual (Fig. 7.25).
–– Ware: Fabric 1a.
–– Illustration: Fig. 7.12.
–– Internal dating evidence: Apart from a single sherd Type: 31
in Period 2, and a few residual sherds in Period 4, Traditional ware name: Slipped red ware
this type is found exclusively in Period 3. As the
type is represented by 19 sherds in Period 3, it –– Description: This is a somewhat enigmatic type,
seems certain that its absence from Period 1 is an consisting of a bowl with flaring sides and a
indication that it was not in circulation at that time slightly incurving, almost vertical rim. It com-
(Fig. 7.25). monly has an unpolished red wash or slip on the
surface. The diameter is between 15 and 20 cm.
–– Ware: Fabric 1a.
Type: 28 –– Illustration: Fig. 7.13.
Traditional ware name: none –– Internal dating evidence: This type is found only
in Periods 2 and 3. It is abundant enough for its
–– Description: This type number was not used. absence from Period 1 to be significant.

Type: 29 Type: 32
(Black Burnished ware) Traditional ware name: Traditional ware name: Slipped red ware
Burnished black ware
–– Description: This is a rolled rim of a large storage
–– Description: This is a simple, almost vertically vessel. The width of the rolled rim itself is between
sided bowl/cooking pot with a widely flaring rim. 3.5 and 5.5 cm and the diameter is a minimum of
It is thin walled (up to 5 mm) with a heavy burnish 35–40 cm. The vessel walls are a minimum of 1.75
both inside and out. The rim diameter is 10–20 cm. cm in thickness. This is a really large storage vessel
The body is black and the surfaces are heavily bur- of the type that would not have been moved around
nished. once installed. The surface is lighter than the core
–– Ware: Fabric 3. of the fabric, and it is smoothed and covered with
–– Illustration: Fig. 7.13. a mat red slip.
–– Internal dating evidence: The percentage of this –– Ware: Fabric 1a. This type tends to be rather weakly
type appears to increase from Period 1 to Period 2, fired and breaks easily with no snap.
after which time it must have died out, the sherds –– Illustration: Fig. 7.13.
in Period 3-temple and Period 4 are probably –– Internal dating evidence: This type does not occur
residual. in Period 1 and its presence was used to make the
distinction between Periods 1 and 2. It was most
abundant in Period 2, after which time it slowly
Type: 30 declined. It is abundant enough for its absence
Traditional ware name: Slipped red ware from Period 1 to be significant (Fig. 7.25).
–– Comments: The presence of this form was used
–– Description: This is a small, gently carinated bowl to make the distinction between levels of Periods
with a rounded rim and a distinct mat red wash inter- 1 and 2 when the sequence was periodised (see
Chapter 6).
128   Derek Kennet

Fig. 7.13: Types 29 to 33.

 Chapter 7 The Pottery    129

–– External parallels: The introduction of these very Type: 35

distinctive large storage jars appears to have hap- Traditional ware name: Slipped red ware
pened relatively late at some sites, such as Nevasa
where they first appear in Period IV and at Mahesh- –– Description: This is a large, slightly closed basin,
war where they occur in Periods V and VI. They do with a thickened, rolled rim with an almost triangu-
occur, however, in earlier levels at other sites in lar profile. The diameter is between 30 and 35 cm.
the region such as Bhokardan Period IA and Ter The surface of the vessel is covered with a thick matt
Period II (Sankalia et al. 1958: fig. 76 type 111, fig. red slip which has fired grey in some cases.
91 type 162; Sankalia et al. 1960: fig. 120 type 55; –– Ware: Fabric 1a.
Chapekar 1969: fig. 19 type 35; Deo and Gupte 1974: –– Illustration: Fig. 7.14.
fig. 9 types 53 and 53a). –– Internal dating evidence: This type occurs in
Periods 1 to 3, in increasing quantities, although
numbers are very small. There are no residual
Type: 33 sherds from Period 4. The reason for the absence
Traditional ware name: Slipped red ware of this type from Period 3-temple is probably due to
the small size of that assemblage.
–– Description: This is a smaller version of Type 32,
a large storage vessel with a large rounded or
everted rim, of variable shape. The diameter is Type: 36
24–30 cm. The type is most often smoothed and Traditional ware name: Coarse red ware
covered with a matt red wash. One or two examples
are fired grey. –– Description: This is an angular form of Type 2
–– Ware: Fabric 1a. that is found in only a few layers in Trench D2. It
–– Illustration: Fig. 7.13. is unslipped and unburnished. The diameter is
–– Internal dating evidence: This type occurs in all about 15 cm.
periods but is most common in Period 2 (Fig. 7.25). –– Ware: Fabric 1a.
–– External parallels: A similar type was found at Bhok- –– Illustration: Fig. 7.14.
ardan in Period IA (Deo and Gupte 1974: fig. 7 types –– Internal dating evidence: This type occurs only in
27 and 27a). Periods 1 and 2, but its absence from later assem-
blages may be due to chance, given the small
numbers concerned.
Type: 34
Traditional ware name: Coarse red ware
Type: 37
–– Description: This is a bowl with flaring sides and Traditional ware name: Burnished and slipped red ware
an everted, thickened rim with a deep notch just
below the rim on the inside where it has been –– Description: This is a medium-sized jar with an
rolled over. The diameter is 15–20 cm. The surface everted, slightly thickened rim. It is very similar to
of the vessel is smoothed and a lighter colour than Type 3 except that it lacks the distinctive notch on
the body internally and externally. the interior. The diameter is around 15–17 cm. The
–– Ware: Fabric 1a. This form is slightly higher fired than type is normally coated in a polished red slip.
others in this fabric. –– Ware: Fabric 1a.
–– Illustration: Fig. 7.14. –– Illustration: Fig. 7.14.
–– Internal dating evidence: Only eight examples –– Internal dating evidence: Only 13 examples of this
of this form were recorded, so it is difficult to be type were recorded, all from Period 2. Its presence
certain of the reliability of its occurrence through was used to help define Period 2 from Period 1. It
the sequence. It occurs only in Periods 2 and occurs in the lower part of Trench A:TP1 and once
3, most commonly in Period 2. As it is not very in the upper part of D2. Its absence from Trench B
common, it is possible that its absence from the is perhaps to be explained by chance.
Period 1 assemblage is due to chance. –– Comments: The presence of this form was used to
make the distinction between levels of Period 1 and
130   Derek Kennet

Fig. 7.14: Types 34 to 39.

 Chapter 7 The Pottery    131

Period 2 when the sequence was periodised (see eter is around 20 cm. The surface is coated with a
Chapter 6). polished red slip on the interior and exterior.
–– Ware: Fabric 1a.
–– Illustration: Fig. 7.15.
Type: 38 –– Internal dating evidence: This type occurs in
(Black and Red ware) Traditional ware name: Black and Periods 1 and 2. There are no examples from Period
Red ware 3 or 3-temple, suggesting that it had ceased to be
used by that time. The sherds from Period 4 are
–– Description: This is a simple, straight-sided bowl probably residual.
with a vertical pointed rim that is marked by having
the top 8 mm or so of its inner edge shaved steeply
to a triangular point. The diameter is about 12 cm. Type: 41
The surface is slipped and heavily burnished. Traditional ware name: Slipped red ware
–– Ware: Fabric 3.
–– Illustration: Fig. 7.14. –– Description: This is a simple bowl with gently
–– Internal dating evidence: This type occurs only in curving sides and a straight, pointed rim. The diam-
Period 1 (Fig. 7.25). eter is up to about 26 cm. In some cases, the surface
–– Comments: Only 25 sherds of this type were is smoothed and it is coated with a thin mat wash
recorded; it was used as a marker of the Period 1 on the interior and over the rim.
assemblage. –– Ware: Fabric 1a and occasionally Fabric 1c.
–– External parallels: Roughly similar types are –– Illustration: Fig. 7.15.
reported from Nevasa Period IV and Maheshwar –– Internal dating evidence: This type has a strange
Period V, although no precise parallels to this occurrence profile; it is present in Period 1 in very
type could be found amongst the Black and Red small quantities, it is more common in Period
ware from early layers at the published sites in the 2, totally absent from Period 3 and 3-temple and
region (Sankalia et al. 1958: fig. 69 type 92; Sanka- appears to have been quite abundant in Period 4
lia et al. 1960: fig. 116 type 29). (Fig. 7.25).

Type: 39 Type: 42
(Black Burnished ware) Traditional ware name: Traditional ware name: Burnished and slipped red ware
Burnished black ware
–– Description: This is a small jar with a thin body, a
–– Description: This is a plate or a very shallow bowl short neck and a slightly flaring rim with a small
with flat sides and a carinated rim of about 15 mm rounded thickening on the end. The diameter is
height. The diameter is 13 or 14 cm. The vessels about 12 cm. The interior and exterior surfaces
are black-bodied and are heavily burnished on the are covered with a thin mat or polished red or red-
exterior. brown slip. This type is quite similar to Type 20,
–– Ware: Fabric 3. but it has a more pronounced neck.
–– Illustration: Fig. 7.14. –– Ware: Fabric 1a. This type is harder fired than
–– Internal dating evidence: Only three sherds of this others in this fabric.
type were recorded, and they are all from Period 1. –– Illustration: Fig. 7.15.
–– Internal dating evidence: This type was present in
Period 1, it then declined in Period 2 and does not
Type: 40 occur in any later levels.
Traditional ware name: Burnished and slipped red ware

–– Description: This is a large ‘nailhead’ jar with an

everted thickened rim with a triangular profile and
a deep indent behind it on the interior. The diam-
132   Derek Kennet

Fig. 7.15: Types 40 to 44.

 Chapter 7 The Pottery    133

Type: 43 Type: 46
Traditional ware name: Burnished and slipped red ware (Thick Grey ware) Traditional ware name: Grey ware

–– Description: This is a jar with a short neck and a –– Description: This is a very crude grey-ware plate or
simple out-turned flaring rim similar but less dis- very shallow bowl with a slightly raised, carinated
tinctive than Type 9. The diameter is about 18 cm. rim. The diameter is about 28 cm. The base has
The surface is burnished and may have had a thin been deliberately impregnated with coarse grits,
brown slip. but the outside above the carination and the inte-
–– Ware: Fabric 1b. rior are both smoothed, although not burnished.
–– Illustration: Fig. 7.15. –– Ware: Fabric 2.
–– Internal dating evidence: This type occurs in all –– Illustration: Fig. 7.16.
periods in reasonable quantities with no obvious –– Internal dating evidence: This type occurs only in
chronological pattern. Period 4 but it is common enough for its absence
from Period 3 and earlier assemblages to be signifi-
cant.
Type: 44
(Thick Grey ware) Traditional ware name: Grey ware
Type: 47
–– Description: This is a jar with a carinated body, a Traditional ware name: Slipped red ware
short neck and an everted rim that is partly rolled
back on itself. There is a good deal of variability in –– Description: This is a small jar with a short neck
the precise form. There are often shallow incised and a slightly thickened rim with a shallow notch
decorative bands around the exterior of the jar on the outermost face. The diameter is about 10
below the rim together with rows of incised slits. cm. The surface is coated with a thin brown slip or
The diameter is normally about 14 cm. The surface wash that is not burnished.
is often smoothed and burnished and is a lighter –– Ware: Fabric 1a.
colour than the body. The body walls are notably –– Illustration: Fig. 7.16.
thinner than Type 15; in this case, they are about 4 –– Internal dating evidence: A small quantity of this
mm thick in the middle of the vessel. type occurs in Period 3, but it was most abundant
–– Ware: Fabric 2. in Period 4. It is common enough to suggest that
–– Illustration: Fig. 7.15. its absence from Period 2 and earlier assemblages
–– Internal dating evidence: This type occurs from is significant.
Period 3 in small quantities but was most abundant –– External parallels: Rough parallels to this form can
in Period 4. It is common enough for its absence be found at Nasik in Period IIA and at Prakash in
from Periods 1 and 2 to be significant. Period II (Sankalia and Deo 1955: fig. 21 type 26;
Thapar 1967: fig. 25 type 37), but it is difficult to be
certain of identifying this form in published draw-
Type: 45 ings.
(Thick Grey ware) Traditional ware name: Grey ware

–– Description: This is a low, flat grey ware bowl or Type: 48

dish with a flat base that is deliberately impreg- Traditional ware name: Coarse red ware
nated with coarse grit. It has a low, steeply sloping
side and a slightly everted rim. The diameter is –– Description: This is a small, crude, hand-made
about 25 cm. The exterior surface is untreated but bowl with very low sides and a flat base. The diam-
the interior is burnished. eter is between 6 and 12 cm. There is no evidence of
–– Ware: Fabric 2. deliberate surface treatment.
–– Illustration: Fig. 7.16. –– Ware: Fabric 1c.
–– Internal dating evidence: This type occurs only in –– Illustration: Fig. 7.16.
Period 4. It is common enough for its absence from –– Internal dating evidence: This type has a very
Period 3 to be significant. similar occurrence profile to Type 47; it occurs
134   Derek Kennet

Fig. 7.16: Types 45 to 48.

 Chapter 7 The Pottery    135

Fig. 7.17: Types 49 to 51.

136   Derek Kennet

in Period 3 in small numbers, but only became exterior is covered in white paint decorated with
common in Period 4. It is just about common horizontal red painted bands around the body.
enough to suggest that its absence from the Period 1 –– Ware: Fabric 5, WARP.
and 2 assemblages is significant. –– Illustration: Fig. 7.17.
–– Internal dating evidence: This type occurs only in
Period 4 and its absence from earlier assemblages
Type: 49 does seem to be significant (Fig. 7.25).
Traditional ware name: none

–– Description: This is a large bowl or basin with a flat Type: 52

base and straight flaring sides with a very slightly (White and Red Painted ware) Traditional ware name:
incurved rim with a finger-wide grove on the outside. none
The walls are very thick (9–10 mm) and the diameter
is between 30 and 45 cm. There is no evidence of any –– Description: This is a small globular jar with no
surface treatment beyond a smoothing. neck and a slightly thickened rim that has been
–– Ware: Fabric 1b. This type tends to be made of a folded back on itself. The walls are very thin (3–4
denser than average Fabric 1b. mm) and the diameter is about 6 cm. The exterior and
–– Illustration: Fig. 7.17. at least a part of the interior are painted with a thin
–– Internal dating evidence: This type occurs only white paint.
in Period 4. Its absence from earlier assemblages –– Ware: Fabric 5, WARP.
does seem to be significant. –– Illustration: Fig. 7.18.
–– External parallels: Parallels to this type can be –– Internal dating evidence: Only five sherds of this
found at Maheshwar in Periods V and VI (Sankalia type occur, all in Period 4.
et al. 1958: fig. 73 type 100, fig. 84 type 140).

Type: 53
Type: 50 Traditional ware name: none
Traditional ware name: Coarse red ware
–– Description: This is a small, fine, carinated jar
–– Description: This is a crude, thick-walled, cari- with a slightly everted geometrically shaped rim in
nated bowl with a simple, almost vertical, rounded a very distinctive fine grey ware that is unique to
rim. The walls are about 10 mm in thickness, and this type. The walls are very thin (3–4 mm) and the
the diameter is between 16 and 20 cm. diameter is between 6 and 8 cm. The surface was
–– Ware: Fabric 1a. This type has a sandier fabric than burnished on the exterior whilst the vessel was still
usual. on the wheel.
–– Illustration: Fig. 7.17. –– Ware: This is a type-specific fabric. It is well levi-
–– Internal dating evidence: This type occurs in gated with a fine, grainy structure and an almost
Period 3 and 4. The reason for its absence from smooth fracture. It is medium to soft fired and
Period 3-temple is probably due to the small size breaks easily. The grainy structure is caused by
of that assemblage, but its absence from the Period fine grits that are almost too small to see with a
1 and 2 assemblages does seem to be significant. ×10 lens. There are no other inclusions. The fabric
fires to an even olive grey to light olive brown (5Y
5/2–2.5Y 5/3).
Type: 51 –– Illustration: Fig. 7.18.
(White and Red Painted ware) Traditional ware name: –– Internal dating evidence: This type occurs only in
none Period 4 and does not seem to have been in circula-
tion any earlier.
–– Description: This is a small, very fine-walled, glob-
ular jar with an everted rim. The walls are 3 or 4 mm
thick and the diameter is between 5 and 6 cm. The
 Chapter 7 The Pottery    137

Type: 54 cm. The surface is normally covered with a slip on

Traditional ware name: none the interior and exterior that is either mat red or
sometimes purple-brown.
–– Description: This is a short-necked jar with an –– Ware: This is a type-specific fabric. It is a well-
everted rim that is flattened and slightly grooved levigated fabric with a smooth fracture and a
on top in order to hold a lid. One sherd has evi- very fine grainy structure. There are no visible
dence of a spout in the upper part of the body. The inclusions. The body is fired to an even reddish
rim diameter is 9 cm. It has a mat red slip on the yellow (5YR 7/6).
exterior and over the rim; the interior has a light –– Illustration: Fig. 7.18.
coloured wash. –– Internal dating evidence: This type occurs only in
–– Ware: This is a type-specific fabric. This is like a Period 4. It is common enough for its absence from
harder-fired version of Fabric 1a. Despite the better earlier periods to be significant.
firing, it still breaks easily. It has a very rough
fracture, with a coarse blocky structure. There
are frequent, large (2 mm) air holes caused Type: 57
by burning vegetable temper and numerous Traditional ware name: none
rounded white inclusions up to 3 mm. The core is
a reddish yellow (5YR 6/6). –– Description: This appears to have been the tobacco
–– Illustration: Fig. 7.18. container from a hookah or water pipe. It has a
–– Internal dating evidence: This type occurs only in broader crucible than Type 56 above, and no grill of
Period 4 and almost certainly did not circulate any holes is visible on any of the preserved examples. It is
earlier. a small object (diameter 8 cm) with a wall thickness
of 3 mm. In some cases, the surface is burnished. The
quality of the turning and the clay is variable.
Type: 55 –– Ware: This is a type-specific fabric. It is a very fine,
Traditional ware name: none grey ware, well fired with a smooth fracture. It has
a very grainy structure and there are no visible
–– Description: This is a small jar with a short neck inclusions. The body is a dark grey (5YR).
and an everted rim. The walls can be as thin as 3 –– Illustration: Fig. 7.18.
mm in some cases. The diameter is between 10 and –– Internal dating evidence: Apart from a single
14 cm. The exterior of the jar is sometimes lightly sherd in Period 1, which must be intrusive, this
burnished, probably on a wheel. type occurs only in Period 4, and its absence from
–– Ware: Fabric 1c, but fired to an even brown to yellow- earlier periods is likely to be significant.
ish brown (10YR 5/3–10YR 5/4).
–– Illustration: Fig. 7.18.
–– Internal dating evidence: With the exception of Type: 58
one sherd in Period 1 that is probably intrusive, (Black Burnished ware) Traditional ware name:
this type occurs only in Period 4 and almost cer- Burnished black ware
tainly did not circulate earlier.
–– Description: This is a small jar with a slightly flaring
neck and a thickened rim with an almost angular
Type: 56 profile. The diameter is about 11 cm. The body is black
Traditional ware name: none and the surface is heavily burnished inside and out.
–– Ware: Fabric 3.
–– Description: This is a small tobacco container from –– Illustration: Fig. 7.18.
a hookah or water-pipe. It has quite an elaborate –– Internal dating evidence: This type occurs only
shape, the exact number of ribs and the precise form in Periods 1 and 2. Nine sherds were recorded from
varying somewhat, but always being defined by the Period 1 and a single sherd from Period 2, which is
ash-stained crucible connected by small holes to probably residual.
the hollow tube below it. The height is about 5.5
cm and the diameter of the widest part is about 3.5
138   Derek Kennet

Fig. 7.18: Types 52 to 58.

 Chapter 7 The Pottery    139

Distribution on site Analysis of the assemblage and

The quantified pottery sequence of each of the five
sequence
main trenches is presented below in a tabular format by
The typology-based approach to the study of the
context, type and sherd count.
Paithan assemblage that has been set out above has
proved to be successful in that it has been possible
to propose a number of well-defined types that have
Trench A: Test Pit 1
clear and demonstrable chronological profiles that
make them useful chronological markers. Quantified
See Table 7.6 for distribution of types.
analysis of the occurrence of the types through the
sequence has helped to refine this through the iden-
tification of obviously residual and intrusive types,
Trench B
despite very high levels of residuality and, more impor-
tantly, apparently high levels of continuity in pottery
See Table 7.7 for distribution of types.
tradition and manufacture.
Period 1 is marked by Types 2 and 38 and, to some
degree, Type 10, each of which either declined mark-
Trench C
edly or went out of use completely by Period 2. Almost
all other types present in the Period 1 assemblage con-
Trench C was excavated during the 1996 and 1997
tinued to be used in Period 2 in roughly similar quanti-
seasons. The pottery was only briefly studied and the
ties, and some of them also continued into Period 3,
assemblage was not included in the main analysis that
although normally in much lower quantities. In Period
is presented in this chapter. Nonetheless, the results
2, Types 7, 19, 31, 32, 34 and 37 appeared; these were not
are of some interest and give an impression of how the
in circulation in Period 1, and their absences from that
sequence in this part of the site relates to that in the
period appear to be significant. Of these, only Types 19
main trenches.
and 37 appear to have ceased circulation by Period 3,
Table 7.8 shows an obvious change in the nature of
whilst the others continued. When it comes to Period
the assemblage between layers 83 and 69. In layer 69, a
3, only Types 44, 47, 48 and 50 occur that were not
number of new types such as 13, 16, 17 and 18 occur for
present in earlier periods, together with Types 14, 24,
the first time, followed shortly by GREY forms 14 and
26 and 27 that were previously in circulation in Period
15. In the following deposits, most of the forms which
2, but in such low quantities that they were largely
were common in the earliest levels appear to become
insignificant. Some of these types seem to go out of cir-
increasingly less frequent (i.e. Types 6, 7, 8, 9). Layers
culation by Period 4, for instance Types 24, 26 and 27,
60 and above probably correspond roughly to Period 3,
whilst others continued to circulate, some in increased
but it is more difficult to allocate a period to the layers
amounts. In Period 4, we see the introduction of a large
below this.
number of new types, for example Types 13, 15, 23, 45,
46, 49, 51, 52, 53 and 56, whilst types such as Types 44,
47 and 48 continued to circulate, but in notably larger
Trench D1
amounts than in Period 3.
Having set out this preliminary framework, it
See Table 7.9 for distribution of types.
would have been useful to have been able to compare
it to published assemblages from other excavated sites
in the region as has been done, where possible, in the
Trench D2
description of types above. In order to attempt this,
seven regional sites were looked at in detail, namely
See Table 7.10 for distribution of types.
Bhokardan, Brahmapuri, Maheshwar, Nasik, Nevasa,
Prakash and Ter, each yielding some parallels to the
types defined at Paithan. It did not, however, prove
Trench F
to be the case that the closer sites, such as Nevasa (50
km) and Bhokardan (95 km), yielded more parallels or
See Table 7.11 for distribution of types.
140   Derek Kennet

Table 7.6: Types from Trench A: Test Pit 1 by rim-sherd count and layer in stratigraphic sequence.

306 310 308 312 912 913 922 921 930 933 945 944 966 1013 1040 1058 1052 1077 1099 1120 1127 1158 1159
1195 908 316 318 946 997 1074 1063 1110
948 985 1105 1111
986
988
1009
1014
Type 1043
1 1
1A 3 1 4 4 2 1 17 9 13 4 18 70 96 71 22 74 64 54 60 21 4 3
1B 3 7 9 1 11 32 22 24 3 41 258 84 115 66 172 54 103 30 4 1
2 1 1 4 3 1 4 2 2 2 2
3 3 9 3 4 1 6 6 10 5
4 1 1 2 1 1 7 8 8 2 12 8 2 4
5 1 1 3 1 2 1 1 4 6 1
6 1 1 1
7 1 2 4 1 1 2 4
8 1 4 8 2 4 2 5
9 2 2 1 3 3 4 4
10 1 5 20 5 19 1 27 12 19 13 9 1
12 1 1
16 1 4 1 3 5 1 2 3 13 5 8 1 7 3 10 7
18 2 1 1
19 2 2 3 1 1 2 1 1
20 1 5 1 2 1 2 1 1 2 1 2 1
22 1 1
24 1 8 12 6 9 11 2 2
25 3 4 2 1 5 1 3 1 1 1 1
26 4 8 8 3 2 6 1 1
27 1 6 5 2 2 1 1 1
29 2
31 1 1 1 1 1 2 1 3 4 2 1
32 1 1 1 1 2 2 3 2 1 1 1
33 1 1 2 1 1 1 8 2 7 1 3 1
34 1 2 2
35 1 1 1 1 1 1 2
37 1 1 2 1 3
RPW 1
NON 3 3 36 26 1 8 5 33 34 28 45 97 41 45 14 44 21 31 7 6
UNIQ 6 2 2 3 3 1 1 1 1
Rim total 17 7 81 78 9 43 8 118 85 81 11 120 511 272 293 117 375 193 267 136 43 5 4

NBP body 1
RPW body 1 1
Body total 108 621 807 90 533 69 963 687 847 105 861 3276 1069 1638 527 2140 1149 2112 686 451 38
 Chapter 7 The Pottery    141

a wider range of parallels than the more distant sites in a particular trench (a period-sequence). To give an
such as Maheshwar (300 km), Prakash (250 km), Nasik example, ‘TrB-1’ indicates the Period 1 material from
(177 km) or Ter (148 km). The parallels that have been Trench B. The 15 period-assemblages were selected
found are intriguing, and they suggest that major devel- from across the site in order to give a representative
opments in pottery technology and style followed sample.
regional trends, whilst at the same time there is evi- Some of the same data are used in the appendix to
dence, in the types and wares that are restricted only to Chapter 9 to calculate coin deposition rates in the temple
Paithan, that local pottery manufacture was the norm foundation deposits.
and each site is also likely to have had a range of types In order to investigate sherd density, the amount of
and styles that were unique to it. excavated earth was calculated for each of the period-
More detailed comparisons with the published assemblages. By combining this with the sherd count
assemblages from the region are rendered impossi- from the same levels, it is possible for the density of
ble due to problems with the way in which the mate- sherds per cubic metre to be calculated and compared
rial has been studied and published, most notably (Tables 7.12 and 7.13).
the lack of precision in the definition of wares and The data and analysis presented above and the
fabrics and the lack of systematic quantification. Were ranked analysis in Fig. 7.19 makes clear that, in general,
it not for these limitations, it would by now be pos- the Period 1 assemblages have amongst the higher
sible to attempt a much-needed synthetic regional densities of sherds at the site and that this density
review of pottery wares and types that would be an appears to decline gradually through the history of
important step towards the creation of a reliable the site, with the lowest densities occurring in Period
archaeological chronology of the Early Historic and 3 and 4 deposits.
Early Medieval periods in this region of India. Unfor- When averages are taken by period, the figures
tunately, such advances are still a considerable way suggest that the amount of pottery deposited at the
in the future and are dependent on a fundamental site declined by around 20% between Periods 1 and 2,
change in the way that pottery is collected, studied and 32% between Periods 2 and 3 and 54% between Periods 3
published on archaeological excavations in India. and 4 (Table 7.14). This represents an overall decline of
75% between Periods 1 and 4.
This very consistent pattern must be indicative of

Period-based analysis of pottery much higher levels of pottery deposition during the
earlier part of the site’s history. This, in turn, is prob-
deposition rates ably indicative of higher levels of manufacture and use
of pottery, but this is more difficult to determine with
The following analysis uses 15 ‘period-assemblages’ certainty as other factors related to discard and to
from Trenches A, B, D and F to evaluate the relative localised conditions can be involved.
density of sherds per cubic metre of excavated earth. The significance and implications of these results are
The aim is to investigate possible changes in the rate of further discussed in Chapter 14.
pottery deposition through the stratigraphic sequence
at the site. A ‘period-assemblage’ is the assemblage
from the contexts or layers from a particular period
142   Derek Kennet

Table 7.7: Types from Trench B by rim-sherd count and layer in stratigraphic sequence.

Type 402 408 413 415 417 418 427 428 430 429 438 440 439 437 441 443 444 450 451 452 453 456 457
406 426 420 432 448
414 422 435

1 65 15 4 17 5 4 2 188 12 56 50 75 23 14 2 54 5 36 90 48 19
2 1 1 21 1 2 1 3 1 2
3 6 1 2 11 4 3 3 4 3 1 1 2
4 9 3 1 1 21 1 5 4 12 1 4 12 1 1
5 2 6 1 1 1 4 2 3 1 3 1
6 6 1 7 1 2 1 5 2 3 4
7 2 1 3 8 9 1 1
8 8 1 1 4 1 5 1 6 2 1 1 4 1
9 3 1 6
10 4 2 1 1 10 1 9 10 34 2 1 14 1 7 18 16 1 3
12 1 1 8
13 16 1 1 10 2
14 1 1 6 2 3 6
15 2 1 1 3
16 6 5 4 26 5 10 3 14 4 2
18 1
20 9 1 1 3 51 1 5 1 1
22 1 1
23 5 3
24 4
25 1 1 1 9 1
27 3
29 1 5 1
30 1 1
32 1 1 3 2 1 1 13 7
33 2 1 3 2 1 2 2 3
35
38
40 1 1 1 8 1
41 4 3 1 6 2 1 1 10 5 2 1
43 1 7 2 1 1
44 3 2 2 3 52 3
45 16 1 1 1
46 10 8 10 1 2
47 10 22 1 1
48 17 1 11 1 91 3 1
49 2 3 1 10
50 3 1 1
51 3 54 1 4
52 4 1
 Chapter 7 The Pottery    143

Type 458 465 467 468 469 471 472 473 474 477 481 478 483 484 485 486 487 492 493 495 494 731 499 735
470 497 736

1 47 58 99 23 442 38 21 25 57 58 70 33 37 11 17 19 8 17 20 26 12 9 42
2 2 4 8 27 2 3 1 3 11 6 9 2 2 3 1 1 2 8 1
3 2 3 1 3 1 1 1 5
4 4 5 3 2 1 2 2 1
5 1 1 2 1 1 2 1
6
7
8 4 2 2 1 1 2 2
9 2 2 1 1 3 1 2
10 9 6 20 4 5 2 3 2 6 1 4 2 3 2 2 2 1 1 3 2 9 2 29
12
13
14
15
16 2 4 4 1 1 1 1
18
20 2 1
22
23
24 1
25 3
27
29 2 2 1 1
30
32
33 3 1 1 1
35 1
38 2 1 1 1 1 1 8
40 2 1
41
43 4 1
44
45
46
47
48
49
50
51
52
144   Derek Kennet

Table 7.7 (continued): Types from Trench B by rim-sherd count and layer in stratigraphic sequence.

Type 402 408 413 415 417 418 427 428 430 429 438 440 439 437 441 443 444 450 451 452 453 456 457
406 426 420 432 448
414 422 435

53 15
54 11
55 3 4 3 1 12 1
56 1 2 3 1
57 2 2 11
RPW 1
CHIN 2 1 9
PERS 1
NON 102 19 14 22 25 19 1 186 21 25 23 44 3 24 2 7 1 14 13 9 1 4
UNIQ 5 5 18 1 1 1
Rim Total 319 2 58 32 67 60 57 4 898 61 145 115 228 38 70 9 115 14 61 132 99 4 30

NBP Body
RPW Body 1 2
CHIN Body 5 1 1 9
PERS Body 1 1 1
SLIP Body 2 1 1 2 1
WARP Body 4 2
NON Body 3458 32 567 142 402 224 318 28 3879 446 1 614 1140 163 1248 173 292 128 121 513 229 65 166
Body Total 3465 32 567 142 403 224 321 28 3896 448 3 614 1142 164 1248 173 292 128 121 513 229 65 166
 Chapter 7 The Pottery    145

Type 458 465 467 468 469 471 472 473 474 477 481 478 483 484 485 486 487 492 493 495 494 731 499 735
470 497 736

53
54
55 1
56
57 1
RPW
CHIN
PERS
NON 6 4 13 5 8 2 4 1 1 1 1 1 4 2 1 2 1 4 15
UNIQ
Rim Total 74 83 156 36 498 48 31 33 73 73 89 47 45 19 23 22 16 18 27 36 41 1 17 106

NBP Body 1
RPW Body
CHIN Body
PERS Body
SLIP Body
WARP Body
NON Body 395 418 774 35 1807 106 78 132 302 239 293 263 227 98 107 123 117 86 190 179 260 0 51 1014
Body Total 396 418 774 35 1807 106 78 132 302 239 293 263 227 98 107 123 117 86 190 179 260 0 51 1014
146   Derek Kennet

Table 7.8: Types from Trench C by rim-sherd count and layer in stratigraphic sequence.

Type 48 52 60 69 83 84 87 93
1 4 5 6 16 6 14 7
3 2 1 7 7 1 9 8 2
4 1 4 6 7 17
6 3 2 3 11 1
7 1 2
8 3 2 7 4 5 5
9 1 1 3 1 6
10 2 2 5 15 8 23 1 5
11 1 1 1 2
12 4 3 2
13 3 4 1 1
14 4
15 4 3 1
16 4 9 1
17 2 2
18 1 1
Rim total 32 19 37 71 31 88 22 7

Table 7.9: Types from Trench D1 by rim-sherd count and layer in stratigraphic sequence.

Type 751 752 759 760 766 767 773 770 775 777
1 75 6 55 109 255 21 9 2 6
2 2 7 1 1
3 3 2 4 5 2
4 2 3 6 7 1 1
5 1 1 4
8 3 1 1 2 2
9 1 2
10 4 1 8 43 36 2 8 3 2
14 2
16 8 3 4
20 1
24 4 2
32 5 1
33 2 1 1 3
38 1 1
43 5 1
58 1 1
NBP 1
NON 13 3 5 5 23 2 1
Rim total 127 12 78 0 174 352 24 25 8 10

NBP body 3 3 1
Body total 229 6 102 365 321 1936 88 113 79 48
 Chapter 7 The Pottery    147

Table 7.10: Types from Trench D2 by rim-sherd count and layer in stratigraphic sequence.

Type 701 702 703 705 706 707 709 710 711 712 713 715 716 718 719 721 723 722 725 726 728 729

1A 2 24 74 11 10 8 188 88 243 301 4 244 2 211 12 36 4 46 13 13 21 12

1B 2 10 5 4 6 8 18 19 10 150 3 46 35 1 49 17 2
2 1 3 17 19 2 20 3 24 10
3 1 2 13 2 4 6 7 6 9 1 4 16 1 5 4
4 1 2 4 3 4 11 5 23 44 53 33 3 3 1 1
5 2 2 3 2 1 4 1 3 4 1 5 1
6 2 1 1 1 16 1 4
7 1
8 1 2 1 6 4 3 6 6 5 2 4 2
9 1 3 3 1 1 2 4 3 3 2 2 4 2 2
10 1 3 27 19 21 12 22 8 30 23 1 28 1 68 5 39 2 61 19 80 9 30
14 1
16 6 13 4 1 1 6 5 5 21 1 15 7 1 1 1
18 1 1 1
19 1
24 1 1 1
25 1 1 1
29 1
30 1 7 10 3 7
32 1 1 4 1
33 1 1 1 2 4 1 1
35 1
36 3 3 8 4
37 3 2
38 1 1 4 2
39 1 2
40 1 2 2
41 2 11 2 2 1
42 2 2 3 4 1 1
43 1 2 1 1 2 2 1
NBP 2 2
NON 2 2 4 4 4 3 5 2 6 12 23 1 9 2 8 6 11 4
UNIQ 1 2 1 2 1 1 1 7
Rim Total 8 22 75 50 43 24 58 47 81 131 3 135 1 217 10 95 8 101 31 147 24 55

NBP Body 4 3 1
Body Total 6 151 498 238 270 115 30 345 101 950 31 839 11 1989 84 510 82 703 149 804 271 0
148   Derek Kennet

Table 7.11: Types from Trench F by rim-sherd count and layer in stratigraphic sequence.

Type 778 785 787 788 789 797 798 846 851 852 853

1 273 3 81 106 29 27 2 182 186 16

2 4
3 9 7 17 1 1 1 22 46
4 1 1 9 1 8 28 1
5 8 4 1 2 3 12
7 2 3 5 1 2
8 4 5 14 70 1
9 7 6 22 8 26 1
10 15 11 97 8 5 104 243 6
12 4
14 8
16 11 9 17 2 7 52 4
20 11
24 3
27 1
29 1 2
30 12
33 2
34 3
40 5
41 1 23
43 9 9 33
44 5
48 1
50 7
58 1 7
NON 40 11 23 3 7 1 38 111 5 4
Rim total 404 3 138 314 45 43 6 437 823 34 4

NBP body 1
Body total 882 31 594 1322 183 203 86 2571 4446 216 50
 Chapter 7 The Pottery    149

Conclusion Period 4 is also notable for a slightly greater variety

of classes than are found in earlier periods, especially of
In addition to the points set out above, a number of higher-quality wares (Table 7.4). In Periods 1 and 2, the
further points can be made on the basis of the com- presence of small quantities of NBP and RPW is noted,
bined discussion of the types, wares and fabrics in rela- but by Period 4, four high-quality classes are present
tion to the site’s sequence. The first is the high level of in the assemblage (CHIN, PERS, SLIP and WARP), at
residuality that is present in Period 4. This can be seen, least three of which are imported to Paithan. It is also
for example, in the occurrence of Types 2, 6, 7, 12, 14, notable that Period 4 has a higher number of new types
16, 20, 25, 41 and a few others, which make up a similar than other periods (see above) and a greater number of
or higher percentage of the Period 4 assemblage than ‘unique’ sherds. Both of these points suggest, again,
they do the assemblages of earlier periods when the that there was a greater variety of types and wares in
types were actually in circulation. This suggests that circulation at this time. A similar picture is given by
the layers of Period 4 contain a lot of redeposited earth the pottery fabrics. In Periods 1 and 2, Fabric 1 and its
from earlier levels. This is probably due to the deep variants make up over 75% of the pottery. This drops to
foundations that were dug when the large brick town 60% in Period 3 and 51% in Period 4 (as noted above,
houses were built in the later Medieval period, which the Period 3-temple assemblage is too small to be sta-
resulted in earth from 2- or 3-m depth being brought to tistically reliable in such a comparison). These obser-
the surface along with the archaeological material that vations suggest a much greater diversity of pottery
it contained. This material then entered the deposits of manufacture and trade in the later Medieval period,
the later Medieval period and it is from such deposits when it seems likely that Paithan might have been
that it was recovered by the excavation. Only full quan- somewhat better integrated into a broader regional
tification of the pottery assemblage allows this sort distributive economy than perhaps it had been during
of detail to become clear. If it is true for the pottery, the Early Historic period. It is therefore strange to note
then it will also be true for small finds, coins and other that the late period, which is sometimes referred to
antiquities. as the ‘Muslim-Maratha’ period in regional excava-
Period 3-temple presents something of an enigma. tion reports, is often said to show a declining quality
It is made up of layers that are largely foundation of pottery production, although no such decline was
deposits and construction layers related to the two noted at Paithan.
temples and might therefore be expected to contain a Conversely, the predominance of a single fabric
largely re-deposited assemblage from earlier periods. (Fabric 1a) in Periods 1 and 2, where it makes up over
This is in fact likely to be the case as is suggested by 70% of the total assemblage, might suggest exactly
the fact that Types 1 and 10 are over-represented in the opposite, that is to say, a production and distri-
Period 3-temple compared to Period 3, whilst other bution system with very little diversity, focussed, at
types, most notably Type 24 but also Types 3, 25 and least as far as pottery is concerned, on a single produc-
27, appear to be markedly under-represented. These tion location and possibly a single industry. But this
anomalies suggest that, although the Period 3-temple is somewhat speculative. More work needs to be done
and Period 3 assemblages are broadly contemporary, on these issues before any more can be said with any
it is likely that they have quite different depositional degree of certainty. They do, however, point the way
histories, with Period 3-temple deposits containing to some potentially interesting and useful avenues for
much higher amounts of re-deposited material that is further research.
much earlier and relatively few examples of contem- In summary, it seems that the use of a formal typol-
porary types. Having said this, it should be remem- ogy linked to detailed definitions of fabrics and wares
bered that the Period 3-temple assemblage is quite and a system of full quantification of the assemblage is
small – between one third and one thirteenth the size likely to yield more useful results than the traditional
of the other assemblages – meaning that it is less likely system of loosely defined wares has done to date. It
to contain rarer examples of types and wares and also is difficult to see, otherwise, how progress might be
that the figures from it are more likely to be affected by made towards a better understanding of the Early His-
freak occurrences of one or two sherds that would not toric and Early Medieval pottery chronology.
affect a larger assemblage.
150   Derek Kennet

Fig. 7.19: Ranked analysis of the density of sherds per metre cubed from 15 period-assemblages from across the site.

Table 7.12: The number of sherds and the amount of earth excavated Table 7.13: The density of sherds per cubic metre from 15 period-
from a sample of period-assemblages from across the site (Trench A assemblages from across the site.
Test Pit 1 (TP1) Periods 1–4; Trench B (TrB) Periods 1, 2 and 4; Trench
D1 (TrD1) Period 1; Trench D2 (TrD2) Periods 1 and 2; Trench F (TrF)
Period Period-assemblage Sherd density
Periods 1, 2 and 3; Trench A Garbha Griha of the North Temple (GGN),
Period 3-Temple and Trench A Test Pit 5 (TP5) Period 3-Temple). 1 TP1-1 10,938
1 TrF-1 3,884
Period Sherds Meters³ 1 TrD2-1 3,015
TrB-1 1 10,602 5.15 1 TrB-1 2,059
TrB-2 2 2,646 4.11 1 TrD1-1 1,918
TrB-4 4 15,832 16.64 2 TP1-2 4,180
TP5-3T 3T 10,184 8.60 2 TrF-2 1,440
TrF-1 1 5,787 1.49 2 TrD2-2 891
TrF-2 2 6,423 4.46 2 TrB-2 644
TrF-3 3 1,286 3.53 3 TP1–3 2,588
TrD1-1 1 6,772 3.53 3 TrF–3 364
TrD2-1 1 7,658 2.54 3T GGN–3T 1,260
TrD2-2 2 4,382 4.92 3T TP5–3T 1,184
TP1-1 1 3,500 0.32 4 TrB–4 951
TP1-2 2 17,516 4.19 4 TP1–4 414
TP1-3 3 6,263 2.42
TP1-4 4 837 2.02
GGN-3T 3T 3,088 2.45
 Chapter 7 The Pottery    151

Table 7.14: Average density of sherds per meter cubed based on the Catalogue of ‘unique’ sherds
data from Table 7.13 (not including the anomalously high figure for
TP1–1 of 10,938 per m³). A total of 82 ‘unique’ sherds (UNIQ) was recorded that
either could not be classified according to the typology
Period Average sherds per m³ and classes set out above or which have specific traits
1 2,719 that are of particular interest. The most significant of
2 2,170 these sherds are illustrated in Figs 7.20–7.22 and are
3 1,476
described below (Table 7.15).
3-temple 1,222
4 683

Fig. 7.20: Sherds with incised marks: P863, P1766, P2552 and sherd P2553, Roman amphora.
152   Derek Kennet

Table 7.15: Catalogue of unique sherds.

P. No. Context Period Description Figure

863 729 1 Black burnished ware (Fabric 3). Symbol or letter incised onto Fig. 7.9,
sherd after firing. Fig. 7.20
1173 729 1 Coarse red ware, thick red slip, no trace of burnish. Fig. 7.21
1254 715 1 Black burnished ware, no slip or burnish on the interior. Fig. 7.21
168 966 2 Light orange-red ware, soft-fired, no slip or burnish. Fig. 7.21
216 988 2 Red earthenware, partial slip and burnish on the exterior but none on the interior. Fig. 7.21
Stamped or moulded decoration exterior.
1257 712 2 Very unusual fabric, very light, pink (7.5YR 7/4), lots of Fig. 7.21
organic temper and very coarse. No slip or burnish.
2263 586 3 Coarse red ware, breaks easily, slightly rough fracture, fine sandy texture, grey Fig. 7.21
corer and reddish exterior (pink 7.5YR 8/4) but grey core very thin (c. 4 mm). Exte-
rior and rim painted white with red blotches and lines on rim.
2264 586 3 Coarse reddish sandy ware, rough fracture. No visible inclusions. Body reddish brown Fig. 7.21
(5YR 5/3), surface smoothed, small mica grains visible.
2267 632 3 Grey ware with incised decoration, Fabric 2. Fig. 7.21
2277 632 3 Grey ware with incised decoration, Fabric 2. Fig. 7.21
2278 632 3 Fabric 2 but fired to a variable reddish colour. Fig. 7.21
2279 632 3 Red ware, slipped but no polish. Fig. 7.21
2280 632 3 Grey ware, Fabric 2. Fig. 7.21
2281 632 3 Grey ware, Fabric 2. Fig. 7.21
2282 632 3 Light grey, buff ware with burnish but apparently no slip. Lots of organic temper. Fig. 7.22
Fabric is dark grey with a reddening on the exterior surface 3 mm deep. Sandy texture,
occasional large (2 mm) white rounded inclusions.
2285 632 3 Grey ware, Fabric 2. Fig. 7.22
1388 402 4 Spout, slipped but not polished. Fig. 7.22
1390 402 4 Fabric 1c. Small thickened rim of a narrow-necked vase or jar. Traces of mat red slip on Fig. 7.22
the exterior and interior.
1556 430 4 Heavy, well-fired grey ware with a dark core and lighter grey edges. Surface is Fig. 7.22
smoothed. Incised decoration.
1608 435 4 Red earthenware, slip and burnish on the exterior only, stamped/moulded decoration. Fig. 7.22
1610 435 4 Grey ware with traces of pink or red mat paint on the exterior only. Fig. 7.22
1611 435 4 Thin body up to 3.5 mm, soft-fired, fine sandy texture. Body greyish brown (10YR 5/2). Fig. 7.22
No slip on the interior (closed vessel). The exterior has a fine red burgundy slip. Body
moulded with petal decoration.
1660 432 4 Heavy, well-fired grey ware with a dark core and lighter grey edges. Surface is Fig. 7.22
smoothed. Incised decoration.
1662 432 4 Coarse red earthenware, no slip or burnish, fired grey and red. Fig. 7.22
1766 438 4 Black burnished ware (Fabric 3). Symbol or letter incised onto Fig. 7.9,
sherd after firing. Fig. 7.20
2553 334 4 Roman Dressel 2–4 amphora handle. Dense hard fabric, very strong. Breaks with a Fig. 7.20,
sub-conchoidal fracture. Dense badly sorted, sub-angular black inclusions up to c. 0.5 Fig. 7.22
mm. Fabric reddish yellow (5YR 6/6).
2552 532 4? Black burnished ware (Fabric 3). Symbol or letter incised onto Fig. 7.20
sherd after firing.
 Chapter 7 The Pottery    153

Fig. 7.21: Unique sherds, Periods 1 to 3.

154   Derek Kennet

Fig. 7.22: Unique sherds, Periods 3 and 4.

 Chapter 7 The Pottery    155

Fig. 7.23: Occurrence profiles of Types 1, 2, 3, 4, 7 and 8. The histograms show occurrence as a percentage of the total period rim-
sherd assemblages. Note that vertical scales differ on the histograms.
156   Derek Kennet

Fig. 7.24: Occurrence profiles of Types 9, 10, 12, 14, 16, 20, 24 and 25. The histograms show occurrence as a percentage of the total period
rim-sherd assemblages. Note that vertical scales differ on the histograms.
 Chapter 7 The Pottery    157

Fig. 7.25: Occurrence profiles of Types 26, 27, 30, 32, 33, 38, 41 and 51. The histograms show occurrence as a percentage of the total period
rim-sherd assemblages. Note that vertical scales differ on the histograms.
M. Kasturi Bai

Chapter 8 
The Small Finds and Glass

Introduction No: 802 Trench: B Context: 741 Period: 1

Object: Core Material: Jasper.
Description: Amorphous core of green jasper. Fig. 8.1.
A total of 869 small finds were found during the exca- Date: Chalcolithic.
vations at Paithan (Table 8.1). The finds are described No: 704 Trench: B Context: 485 Period: 1
and discussed below, accompanied by a full catalogue. Object: Core Material: Chalcedony.
The excavations also provided 56 coins, which are Description: Fluted core of chalcedony. Fig. 8.1.
described in Chapter 9. Date: Chalcolithic.
Finds were collected in the field during excavation, No: 730 Trench: B Context: 493 Period: 1
either by hand or in 5-mm mesh sieves through which Object: Flake Material: Carnelian.
all excavated earth was passed. The objects were then Description: Amorphous flake of carnelian. Fig. 8.1.
Date: Chalcolithic.
cleaned and registered before being studied.
No: 728 Trench: B Context: 493 Period: 1
Table 8.1: The main categories of small finds.
Object: Blade Material: Chalcedony.
Description: Broken, backed blade of chalcedony(?). Fig. 8.1.
Category Total Date: Chalcolithic.

Coins 56 No: 801 Trench: B Context: 736 Period: 1

Microliths 7 Object: Blade Material: Chalcedony.
Jewellery 375 Description: Fragment of thin blade of chalcedony with use-
Metal objects 203 marks. Fig. 8.1.
Household objects 67 Date: Chalcolithic.
Play objects 98 No: 823 Trench: A:TP9 Context: 1168 Period: 3-temple
Figurines 32 Object: Blade Material: Chert.
Miscellaneous 4 Description: Backed blade of chert with slightly broken distal end,
Unidentified 27 use-marks on one margin and deliberate blunting on the other
Total 869 margin. Fig. 8.1.
Date: Chalcolithic.

Microliths Jewellery
Altogether, seven microliths were recovered from the
A total of 375 items of jewellery came to light during
site, all of which belong to the Chalcolithic period. Of
the excavations (Table 8.2), including beads, bangles,
these, six microliths were obtained from Period 1 and
finger-rings, toe-rings, ear-studs and pendants. The
one from Period-3 temple, which is a backed blade.
miscellaneous jewellery items are mainly categorized
Among the six pieces from Period 1, there were a flake
under kohlsticks, hair-pin and inlay objects. Of the
core, a fluted core, two blades and one flake. The mate-
jewellery objects, beads outnumber the other catego-
rials used for the preparation of these microliths are
ries, followed by bangles, finger-rings and pendants
crypto crystalline materials such as jasper, carnelian
etc. The occurrence of beads decreases from Period
and chalcedony.
1 to Period 3 but in Period 4 it is equal with Period 1.
No: 481 Trench: D1 Context: 777 Period: 1 Bangles, however, increase from Period 1 to 4.
Object: Core Material: Jasper.
Description: Residual core(?) of jasper.
Date: Chalcolithic.

https://doi.org/10.1515/9783110653540-008
 Chapter 8 The Small Finds and Glass    159

Fig. 8.1: Microliths.

160   M. Kasturi Bai

Table 8.2: Jewellery objects by period.

Period 1 2 3 3-temple 4 4? No period Total

Type

Bead 43 22 35 22 44 6 13 185
Bead(?) 1 1
Bangle 4 4 33 12 44 6 40 143
Finger ring 2 2 3 1 3 1 3 15
Ring-stone 1 1 2
Toe ring 2 2
Ear-stud 1 1 2 1 5
Pendant 2 2 2 1 3 10
Hairpin 1 1
Kohlstick 1 2 3
Inlay object 2 1 3
Foil 1 1 1 3
Band 1 1
Wire 1 1
Total 52 34 78 35 101 15 60 375

Beads from Period 1, terracotta beads are predominant, and

other materials are only found in quantities below
A total of 185 beads came to light during the excavations five. The occurrence of terracotta beads gradually
(Table 8.3). These are illustrated in Figs 8.2–8.4. decreased from Period 1 to Period 4 as it was replaced
Period 1 shows the usage of different materials like by glass. The glass beads, by contrast, increase from
carnelian, coral, crystal, glass, jasper, shell, topaz Period 1 to Period 4.
and terracotta for bead making. Among the bead finds

Table 8.3: Summary of beads by material and period.

Period 1 2 3 3-temple 4 4? No period Total

Material

Carnelian 5 2 2 1 2 12
Chalcedony 1 2 3
Coral 1 1 1 3
Crystal 1 1 2
Garnet 1 1
Glass 2 4 17 11 24 4 5 67
Jasper 4 2 3 3 12
Mother of pearl 1 1
Pearl 1 1
Shell 1 1 4 1 2 1 10
Stone (Unidentified) 2 1 2 5
Terracotta 26 15 7 4 8 1 6 67
Topaz 1 1
Total 43 22 35 22 44 6 13 185
 Chapter 8 The Small Finds and Glass    161

Fig. 8.2: Beads of different materials.

162   M. Kasturi Bai

Fig. 8.3: Terracotta beads.

 Chapter 8 The Small Finds and Glass    163

Table 8.4: Summary of non-terracotta beads by shape and period.

Period 1 2 3 3-temple 4 4? No period Total

Shape

Almond 1 1
Barrel 1 2 2 1 1 7
Barrel (Faceted) 1 1
Short barrel 1 1
Long barrel 1 1
Bicone 2 1 1 2 6
Short bicone 1 1 2
Biconvex 1 1
Cylinder 1 1 2
Long cylinder 1 1 1 3
Short cylinder 2 2 11 5 11 1 1 33
Square cylinder 1 2 1 1 5
Disc 1 1 1 1 4
Elongated oval 1 1
Lenticular 1 1 1 3
Spherical 4 6 5 11 2 2 30
Squat globular 2 1 1 4
Tiny 1 3 2 1 7
No ID 1 2 1 2 6
Total 17 7 28 18 36 5 7 118

Different shapes of beads were made during Period 1, Description: Spherical bead of carnelian. Fig. 8.2.
although there are few made from materials other than No: 507 Trench: B Context: 453 Period: 1
terracotta (Tables 8.4 and 8.5). Generally, the shapes Object: Bead Material: Carnelian.
preferred in Period 1 are barrel, bicone, cylinder and Description: Small, disc-shaped bead of carnelian. Fig. 8.4.
spherical, with slight variation among them. In Period No: 508 Trench: B Context: 453 Period: 1
2, the beads found are fewer in number and show cylin- Object: Bead Material: Carnelian.
drical and barrel shapes. In Period 3, the occurrence Description: Truncated, bicone bead of carnelian with hexagonal
of beads increased from Periods 1 and 2, and as far as section. Figs 8.2 and 8.4.
shapes are concerned, it is similar to Period 1. In Period No: 625 Trench: B Context: 467 Period: 1
4, a large number of beads were found and the shapes Object: Bead Material: Carnelian.
are the same as Period 1. Description: Spherical bead of carnelian.

No: 555 Trench: B Context: 450 Period: 2

Object: Bead Material: Carnelian.
Carnelian beads Description: Small, cylinder bead of carnelian with square section.
Fig. 8.4.

No: 414 Trench: D2 Context: 718 Period: 1 No: 556 Trench: B Context: 450 Period: 2
Object: Bead Material: Carnelian. Object: Bead Material: Carnelian.
Description: Lenticular diamond-shaped bead of carnelian, etched Description: Small, short cylinder bead of carnelian(?) with a circu-
lines faintly visible. Figs 8.2 and 8.4. lar section, white etched circle on the body.
Ref: A similar bead was recovered from Period IIB of Nasik without
No: 261 Trench: D1 Context: 751 Period: 3
etched lines (Sankalia and Deo 1955: 88, fig. 45, 20).
Object: Bead Material: Carnelian.
No: 415 Trench: D2 Context: 718 Period: 1 Description: Square cylinder bead of carnelian with rectangular
Object: Bead Material: Carnelian. section. Fig. 8.2.
164   M. Kasturi Bai

Table 8.5: Summary of terracotta beads by shape and period.

Period 1 2 3 3-temple 4 4? No period Total

Shape

Amlaka-seed 1 1
Arecanut 5 10 3 1 1 1 4 25
Bicone 2 1 3
Biconvex 1 1
Convex 1 1
Cylinder 1 1
Disc 1 1
Lenticular 1 1
Long Barrel 1 3 4
No ID 1 1 2
Plano-convex 1 1
Pot shaped 2 2 4
Pumpkin shaped 1 1
Short Barrel 1 1
Spherical 8 2 2 1 13
Squat globular 1 1
Top shaped 3 1 1 1 6
Total 26 15 7 4 8 1 6 67

Ref: A similar bead was found in the Satavahana layers of Brahma- No: 58 Trench: A:S of S Temple Context: 303 Period: 4
puri (Sankalia and Dikshit 1952: fig. 30, 4). Object: Bead Material: Chalcedony.
Description: Small, spherical bead of chalcedony.
No: 4 Trench: A:TP1 Context: 305 Period: 3
Object: Bead Material: Carnelian. No: 102 Trench: A:NWNT Context: 376 Period: 4
Description: Spherical bead of carnelian. Object: Bead Material: Chalcedony.
Description: Small, spherical bead of chalcedony. Fig. 8.4.
No: 486 Trench: A:TP5 Context: 700 Period: 3-temple
Object: Bead Material: Carnelian.
Description: Spherical bead of carnelian.

No: 28 Trench: A:TP9 Context: 320 Period: 4 Coral beads

Object: Bead Material: Carnelian.
Description: Small, spherical bead of carnelian. No: 510 Trench: B Context: 453 Period: 1
Object: Bead Material: Coral.
No: 37 Trench: A:TP9 Context: 320 Period: 4
Description: Short cylinder bead of coral with circular section.
Object: Bead Material: Carnelian.
Description: Spherical bead of carnelian. No: 557 Trench: A:TP1 Context: 908 Period: 3
Object: Bead Material: Coral.
Description: Small, squat globular bead of coral with circular
section.
Chalcedony beads
No: 70 Trench: A:TP1 Context: 300 Period: 4
Object: Bead Material: Coral.
No: 17 Trench: A:TP1 Context: 308 Period: 3
Description: Deformed, long cylinder bead of coral with circular
Object: Bead Material: Chalcedony.
section.
Description: Spherical bead of chalcedony. Fig. 8.4.
 Chapter 8 The Small Finds and Glass    165

No: 498 Trench: B Context: 448 Period: 2

Object: Bead Material: Glass.
Description: Fragments of a bead of pale grey translucent glass.

No: 613 Trench: A:TP1 Context: 966 Period: 2

Object: Bead Material: Glass.
Description: Lug collared barrel bead of pale blue glass. Fig. 8.2.

No: 791 Trench: A:TP1 Context: 1126 Period: 2

Object: Bead Material: Glass.
Description: Short cylinder bead of turquoise glass over pale yellow
base.

No: 797 Trench: A:TP1 Context: 1126 Period: 2

Object: Bead Material: Glass.
Description: Barrel and eared bead of brownish yellow opaque
glass.

No: 5 Trench: A:TP1 Context: 305 Period: 3

Object: Bead Material: Glass.
Description: Standard square cylinder, faceted bead of green trans-
lucent glass(?). Fig. 8.2.

No: 255 Trench: A:South Area Context: 589 Period: 3

Object: Bead Material: Glass.
Description: Small, spherical, truncated bead of navy blue opaque
glass.

Fig. 8.4: Beads of different materials. No: 256 Trench: A:South Area Context: 589 Period: 3
Object: Bead Material: Glass.
Description: Small, spherical bead of pale yellow opaque glass.

Crystal beads No: 274 Trench: A:South Area Context: 591 Period: 3
Object: Bead Material: Glass.
Description: Small, short cylinder bead of blue translucent glass
No: 707 Trench: B Context: 487 Period: 1
with circular section.
Object: Bead Material: Crystal.
Description: Irregular-shaped bead of crystal. No: 310 Trench: A:South Area Context: 608 Period: 3
Object: Bead Material: Glass.
No: 824 Trench: A:TP9 Context: 1168 Period: 3-temple
Description: Almond-shaped bead of violet translucent glass. Fig.
Object: Bead Material: Crystal.
8.2.
Description: Fragment of a circular bead of crystal.
No: 311 Trench: A:South Area Context: 608 Period: 3
Object: Bead Material: Glass.
Description: Faceted, long cylinder bead of bluish-green opaque
Garnet bead glass with hexagonal section. Fig. 8.2.

No: 313 Trench: A:South Area Context: 608 Period: 3

No: 612 Trench: A:TP5 Context: 964 Period: 3-temple Object: Bead Material: Glass.
Object: Bead Material: Garnet. Description: Cylinder bead of blue opaque glass with circular
Description: Lenticular diamond-shaped bead of garnet. Fig. 8.4. section.

No: 405 Trench: A:South Area Context: 647 Period: 3

Object: Bead Material: Glass.
Description: Standard bicone bead of black opaque glass with cir-
Glass beads
cular section.

No: 432 Trench: D2 Context: 721 Period: 1 No: 409 Trench: A:South Area, W Area Context: 651 Period: 3
Object: Bead Material: Glass. Object: Bead Material: Glass.
Description: Tiny, short cylinder bead of cream coloured opaque Description: Small, short cylinder bead of turquoise glass over
glass(?). yellow base with circular section.

No: 436 Trench: D1 Context: 767 Period: 1

Object: Bead Material: Glass.
Description: Short bicone bead of cream coloured opaque soft
glass, slightly broken. Fig. 8.2.
166   M. Kasturi Bai

No: 459 Trench: A:South Area Context: 675 Period: 3 Description: Tiny, spherical bead of grass-green translucent glass.
Object: Bead Material: Glass.
No: 726 Trench: A:TP5 Context: 1092 Period: 3-temple
Description: Tiny, short cylinder bead of pale yellow opaque glass
Object: Bead Material: Glass.
with circular section.
Description: Tiny, spherical bead of pale yellow opaque glass.
No: 525 Trench: A:TP1 Context: 921 Period: 3
No: 741 Trench: A:TP9 Context: 1096 Period: 3-temple
Object: Bead Material: Glass.
Object: Bead Material: Glass.
Description: Tiny, spherical bead of yellow opaque glass.
Description: Short cylinder bead of turquoise glass over pale yellow
No: 585 Trench: F Context: 778 Period: 3 base with circular section.
Object: Bead Material: Glass.
No: 821 Trench: A:GG North Context: 1164 Period: 3-temple
Description: Small, short cylinder bead of pale yellow opaque
Object: Bead Material: Glass.
glass.
Description: Segmented (three) spherical bead of dark blue
No: 631 Trench: A:TP6 Context: 373 Period: 3 opaque glass with circular section, wire-wound. Fig. 8.2.
Object: Bead Material: Glass.
No: 822 Trench: A:TP9 Context: 1168 Period: 3-temple
Description: Spherical and truncated bead of cream opaque glass,
Object: Bead Material: Glass.
broken into two pieces.
Description: Broken bead of pale green glass.
No: 635 Trench: A:TP3 Context: 973 Period: 3
No: 51 Trench: A:S of S Temple Context: 303 Period: 4
Object: Bead Material: Glass.
Object: Bead Material: Glass.
Description: Short cylinder bead of yellow opaque glass with cir-
Description: Short cylinder bead of turquoise glass over yellow
cular section.
base with circular section. Fig. 8.2.
No: 636 Trench: A:TP3 Context: 973 Period: 3
No: 63 Trench: A:TP5 Context: 330 Period: 4
Object: Bead Material: Glass.
Object: Bead Material: Glass.
Description: Small, short cylinder bead of yellow opaque glass
Description: Small, short cylinder bead of pale yellow opaque
with circular section.
glass.
No: 733 Trench: A:TP6 Context: 1051 Period: 3
No: 64 Trench: A:TP5 Context: 330 Period: 4
Object: Bead Material: Glass.
Object: Bead Material: Glass.
Description: Short cylinder bead of pale green glass with circular
Description: Small, short cylinder bead of turquoise glass over
section.
yellow base with circular section.
No: 856 Trench: A:TP6 Context: 1035 Period: 3
No: 72 Trench: A:TP9 Context: 320 Period: 4
Object: Bead Material: Glass.
Object: Bead Material: Glass.
Description: Short cylinder bead of sky-blue translucent glass
Description: Small bicone bead of yellow opaque glass with circu-
with circular section.
lar section.
No: 449 Trench: A:TP5 Context: 666 Period: 3-temple
No: 101 Trench: A:NWNT Context: 376 Period: 4
Object: Bead Material: Glass.
Object: Bead Material: Glass.
Description: Tiny bead of black glass.
Description: Small, short cylinder bead of yellow opaque glass with
No: 451 Trench: A:TP5 Context: 667 Period: 3-temple circular section.
Object: Bead Material: Glass.
No: 103 Trench: A:NWNT Context: 376 Period: 4
Description: Truncated bicone bead of blue translucent glass.
Object: Bead Material: Glass.
No: 515 Trench: A:TP5 Context: 918 Period: 3-temple Description: Square cylinder bead of black glass with slight
Object: Bead Material: Glass. depression on each side. Fig. 8.2.
Description: Small and short cylinder bead of blue opaque glass.
No: 122 Trench: A:TP9 Context: 519 Period: 3
No: 542 Trench: A:TP5 Context: 919 Period: 3-temple Object: Bead Material: Glass.
Object: Bead Material: Glass. Description: Small, short cylinder bead of black glass with circu-
Description: Tiny, short cylinder bead of pale yellow opaque glass. lar section.

No: 554 Trench: A:TP5 Context: 902 Period: 3-temple No: 135 Trench: A:TP9 Context: 543 Period: 4
Object: Bead Material: Glass. Object: Bead Material: Glass.
Description: Tiny bead of blue glass. Description: Tiny, short cylinder bead of violet opaque glass.

No: 610 Trench: A:TP5 Context: 964 Period: 3-temple No: 136 Trench: A:TP9 Context: 543 Period: 4
Object: Bead Material: Glass. Object: Bead Material: Glass.
Description: Small, short cylinder bead of black glass with circular Description: Spherical bead of opaque blue glass.
section.
No: 154 Trench: B Context: 406 Period: 4
No: 611 Trench: A:TP5 Context: 964 Period: 3-temple Object: Bead Material: Glass.
Object: Bead Material: Glass.
 Chapter 8 The Small Finds and Glass    167

Description: Small, irregular-shaped bead of pale yellow opaque Description: Tiny, spherical bead of pale green opaque glass.
glass.
No: 105 Trench: A:NWNT Context: 379 Period: 4?
No: 165 Trench: B Context: 414 Period: 4 Object: Bead Material: Glass.
Object: Bead Material: Glass. Description: Tiny bead of translucent turquoise glass.
Description: Small, short cylinder bead of blue glass.
No: 108 Trench: A:NWNT Context: 386 Period: 4?
No: 189 Trench: B Context: 435 Period: 4 Object: Bead Material: Glass.
Object: Bead Material: Glass. Description: Barrel bead of black glass. Fig. 8.2.
Description: Six short barrel beads of grey translucent glass(?).
No: 139 Trench: A:NWNT Context: 546 Period: 4?
No: 226 Trench: B Context: 438 Period: 4 Object: Bead Material: Glass.
Object: Bead Material: Glass. Description: Small, truncated spherical bead of black glass with
Description: Short cylinder bead of blue glass. circular section.

No: 245 Trench: A:S Wall of N Context: 571 Period: 4 No: 92 Trench: A:S Temple Context: 366 Period: No period
Object: Bead Material: Glass. Object: Bead Material: Glass.
Description: Tiny bead of turquoise translucent glass. Description: Short cylinder bead of opaque, turquoise glass over
yellow base with circular section.
No: 319 Trench: A:TP6, South Area Context: 613 Period: 4
Object: Bead Material: Glass. No: 130 Trench: A:N Temple Context: 536 Period: No period
Description: Two short cylinder beads of, turquoise, translucent Object: Bead Material: Glass.
glass with circular section. Description: Small, barrel bead of pale yellow opaque glass.

No: 331 Trench: A:South Area Context: 619 Period: 4 No: 562 Trench: E Context: 821 Period: No period
Object: Bead Material: Glass. Object: Bead Material: Glass.
Description: Short bicone barrel bead of black glass with circular Description: Tiny, spherical bead of black glass with circular
section. section.

No: 345 Trench: A:South Area, TP6 Context: 635 Period: 4 No: 700 Trench: E Context: 831 Period: No period
Object: Bead Material: Glass. Object: Bead Material: Glass.
Description: Small spherical bead of blue glass. Description: Flattened, cylinder bead of turquoise glass over yellow
base with lenticular section.
No: 351 Trench: A:South Area, TP6 Context: 635 Period: 4
Object: Bead Material: Glass. No: 840 Trench: A:TP7 Context: 1225 Period: No period
Description: Tiny bead of black glass with circular section. Object: Bead Material: Glass.
Description: Disc-shaped bead of pale yellow glass, broken.
No: 356 Trench: A:TP5 Context: 633 Period: 4
Object: Bead Material: Glass. No: 33 Trench: A:NWNT, TP9 Context: 328 Period: 4
Description: Tiny, short cylinder bead of pale green opaque glass Object: Bead(?) Material: Glass.
with circular section. Description: Rectangular cylinder bead of bluish-green translucent
glass(?), under preparation.
No: 357 Trench: A:TP5 Context: 633 Period: 4
Object: Bead Material: Glass.
Description: Small, spherical bead of green opaque glass with cir-
cular section.
Jasper beads
No: 400 Trench: A:South Area, TP6 Context: 635 Period: 4
Object: Bead Material: Glass. No: 378 Trench: D1 Context: 766 Period: 1
Description: Small, spherical bead of yellow opaque glass. Object: Bead Material: Jasper.
No: 425 Trench: A:TP6 Context: 649 Period: 4 Description: Tiny bead of red jasper.
Object: Bead Material: Glass. No: 465 Trench: D2 Context: 722 Period: 1
Description: Bicone bead of yellow opaque glass with circular Object: Bead Material: Jasper.
section. Fig. 8.2. Description: Spherical bead of jasper. Fig. 8.2.
No: 484 Trench: A:South Area Context: 645 Period: 4 No: 509 Trench: B Context: 453 Period: 1
Object: Bead Material: Glass. Object: Bead Material: Jasper.
Description: Tiny, short cylinder bead of green translucent glass. Description: Twisted, pentagonal bicone bead of green jasper. Fig.
No: 518 Trench: A:W Area Context: 920 Period: 4 8.2.
Object: Bead Material: Glass. Ref: A similar bead was found at Nasik but in quartz material (see
Description: Biconvex bead of grass-green opaque glass with cir- Sankalia and Deo 1955: 90, fig. 45, 16) and also the same type jasper
cular section. Fig. 8.2. bead was reported from Period IV of Pauni (see Nath 1998: 69, fig.
28, 35).
No: 73 Trench: A:NWNT Context: 327 Period: 4?
Object: Bead Material: Glass.
168   M. Kasturi Bai

No: 776 Trench: F Context: 851 Period: 1 No: 614 Trench: A:TP1 Context: 966 Period: 2
Object: Bead Material: Jasper. Object: Bead Material: Shell.
Description: Gadrooned, barrel bead of jasper with collars. Fig. 8.2. Description: Short cylinder bead of shell.

No: 312 Trench: A:South Area Context: 608 Period: 3 No: 6 Trench: A:TP1 Context: 305 Period: 3
Object: Bead Material: Jasper. Object: Bead Material: Shell.
Description: Small, squat globular bead of red jasper(?). Description: Small, short cylinder bead of shell with circular
section.
No: 732 Trench: A:TP6 Context: 1051 Period: 3
Object: Bead Material: Jasper. No: 11 Trench: A:TP1 Context: 308 Period: 3
Description: Small, short cylinder bead of mustard yellow colour Object: Bead Material: Shell.
jasper(?) with circular section. Description: Barrel bead of shell(?) with circular section. Fig. 8.4.

No: 371 Trench: A:TP5 Context: 642 Period: 3-temple No: 558 Trench: A:TP1 Context: 912 Period: 2
Object: Bead Material: Jasper. Object: Bead Material: Shell.
Description: Tiny bead of mustard yellow coloured jasper(?). Description: Short cylinder bead of shell. Fig. 8.4.

No: 634 Trench: A:GG North Context: 938 Period: 3-temple No: 864 Trench: A:TP5, TP3 Context: 1124 Period: 3
Object: Bead Material: Jasper. Object: Bead Material: Shell.
Description: Tiny, squat globular bead of red jasper. Description: Thick, disc-shaped bead of shell, under process.

No: 849 Trench: A:TP5 Context: 979 Period: 3-temple No: 541 Trench: A:TP5 Context: 918 Period: 3-temple
Object: Bead Material: Jasper. Object: Bead Material: Shell.
Description: Short cylinder bead of maroon red jasper with circular Description: Small, disc-shaped bead of shell.
section.
No: 52 Trench: A:S of S Temple Context: 303 Period: 4
No: 155 Trench: B Context: 406 Period: 4 Object: Bead Material: Shell.
Object: Bead Material: Jasper. Description: Small, spherical bead of shell(?).
Description: Elongated oval-shaped bead of green jasper. Fig. 8.2.
No: 100 Trench: A:NWNT Context: 376 Period: 4
No: 355 Trench: A:TP5 Context: 633 Period: 4 Object: Bead Material: Shell.
Object: Bead Material: Jasper. Description: Short cylinder bead of shell with circular section.
Description: Tiny, squat globular bead of mustard yellow col-
No: 133 Trench: A:NWNT Context: 540 Period: 4?
oured jasper.
Object: Bead Material: Shell.
No: 401 Trench: A:South Area, TP6 Context: 635 Period: 4 Description: Small, short cylinder bead of shell with circular
Object: Bead Material: Jasper. section.
Description: Small, spherical bead of red jasper.

Unidentified stone beads

Mother of pearl beads
No: 759 Trench: F Context: 851 Period: 1
No: 589 Trench: A:South Area Context: 931 Period: 3 Object: Bead Material: Stone.
Object: Bead Material: Mother of pearl. Description: Barrel-shaped dull white stone bead, under prepara-
Description: Barrel-eared bead of mother of pearl with circular tion(?).
section. Figs 8.2 and 8.4. No: 800 Trench: B Context: 736 Period: 1
Object: Bead Material: Stone.
Description: Slightly broken spherical bead of grey stone(?).
Pearl beads No: 167 Trench: B Context: 418 Period: 4
Object: Bead Material: Stone.
No: 190 Trench: B Context: 435 Period: 4 Description: Big, irregular bead of grey stone, perhaps used for
Object: Bead Material: Pearl. animals.
Description: Two spherical beads of pearl with copper string.
No: 780 Trench: E Context: 843 Period: No period
Object: Bead Material: Stone.
Description: Square cylinder bead of green translucent stone with
squarish section.
Shell beads
No: 826 Trench: E Context: 1206 Period: No period
No: 502 Trench: B Context: 452 Period: 1 Object: Bead Material: Stone.
Object: Bead Material: Shell. Description: Spherical and truncated bead of pale green stone(?).
Description: Slightly broken, long cylinder bead of shell(?).
 Chapter 8 The Small Finds and Glass    169

Terracotta beads No: 553 Trench: D2 Context: 726 Period: 1

Object: Bead Material: Terracotta.
There are more terracotta beads in Period 1 and Period Description: Vertically broken cylinder bead of terracotta, flattened
on either end. Fig. 8.3.
2 compared to later periods (Table 8.3). Among these,
the most common shape is arecanut. The other shapes No: 620 Trench: B Context: 458 Period: 1
Object: Bead Material: Terracotta.
found are spherical, long and short barrel, bicone,
Description: Arecanut bead of terracotta with convex base and
top-shaped and pot-shaped. Arecanut beads increased truncated top.
from Period 1 to Period 2. Spherical beads are more in
No: 623 Trench: B Context: 467 Period: 1
Period 1 and decreased from Period 1 to Period 4 and
Object: Bead Material: Terracotta.
are completely absent from Period 2 (Table 8.5). Description: Broken, spherical bead of terracotta.
No: 288 Trench: D1 Context: 754 Period: 1
No: 655 Trench: B Context: 469 Period: 1
Object: Bead Material: Terracotta.
Object: Bead Material: Terracotta.
Description: Top-shaped bead of terracotta. Fig. 8.3.
Description: Slightly broken, truncated bicone bead of terracotta.
No: 342 Trench: D1 Context: 759 Period: 1
No: 659 Trench: B Context: 474 Period: 1
Object: Bead Material: Terracotta.
Object: Bead Material: Terracotta.
Description: Amlaka-seed-shaped bead of terracotta with black
Description: Top-shaped bead of terracotta, slightly truncated on
polish. Fig. 8.2.
either end. Fig. 8.3.
No: 366 Trench: D2 Context: 715 Period: 1
No: 708 Trench: B Context: 487 Period: 1
Object: Bead Material: Terracotta.
Object: Bead Material: Terracotta.
Description: Pot-shaped bead of terracotta, slightly broken at one
Description: Disc-shaped bead(?) of terracotta(?).
end, wheel marks visible. Fig. 8.3.
No: 709 Trench: B Context: 492 Period: 1
No: 377 Trench: D1 Context: 766 Period: 1
Object: Bead Material: Terracotta.
Object: Bead Material: Terracotta.
Description: Spheroid bead of terracotta. Fig. 8.3.
Description: Arecanut bead of terracotta with flat base.
No: 753 Trench: B Context: 493 Period: 1
No: 381 Trench: D2 Context: 718 Period: 1
Object: Bead Material: Terracotta.
Object: Bead Material: Terracotta.
Description: Spherical bead of terracotta, truncated on either end
Description: Biconical bead of terracotta, flattened on either end.
with black polish.
Fig. 8.3.
No: 755 Trench: F Context: 851 Period: 1
No: 416 Trench: D2 Context: 718 Period: 1
Object: Bead Material: Terracotta.
Object: Bead Material: Terracotta.
Description: Arecanut bead of terracotta, flat at the base and
Description: Spheroid bead of terracotta with black polish on body.
concave at the top. Fig. 8.3.
No: 434 Trench: D1 Context: 767 Period: 1
No: 764 Trench: B Context: 499 Period: 1
Object: Bead Material: Terracotta.
Object: Bead Material: Terracotta.
Description: Short barrel bead of terracotta, truncated on either
Description: Truncated, biconvex bead of terracotta.
end. Fig. 8.3.
No: 773 Trench: F Context: 851 Period: 1
No: 435 Trench: D1 Context: 767 Period: 1
Object: Bead Material: Terracotta.
Object: Bead Material: Terracotta.
Description: Top-shaped bead of terracotta, slightly broken at the
Description: Pot-shaped bead of terracotta with truncated top. Fig.
base, flattened on top. Fig. 8.3.
8.3.
No: 775 Trench: F Context: 851 Period: 1
No: 464 Trench: D2 Context: 722 Period: 1
Object: Bead Material: Terracotta.
Object: Bead Material: Terracotta.
Description: Slightly broken arecanut bead of terracotta with flat
Description: Spherical bead of terracotta with cleavage. Fig. 8.3.
base and flattened top.
No: 466 Trench: D2 Context: 722 Period: 1
No: 807 Trench: A:TP1 Context: 1144 Period: 1
Object: Bead Material: Terracotta.
Object: Bead Material: Terracotta.
Description: Big and half-broken spherical bead of terracotta.
Description: Plano-convex-shaped bead of terracotta.
No: 490 Trench: D2 Context: 726 Period: 1
No: 811 Trench: A:TP1 Context: 1150 Period: 1
Object: Bead Material: Terracotta.
Object: Bead Material: Terracotta.
Description: Broken arecanut bead of terracotta.
Description: Spheroid bead of terracotta with black polish.
No: 499 Trench: D2 Context: 729 Period: 1
No: 258 Trench: D2 Context: 702 Period: 2
Object: Bead Material: Terracotta.
Object: Bead Material: Terracotta.
Description: Half-broken spherical bead of terracotta with black
Description: Arecanut bead of terracotta with flat base. Fig. 8.3.
polish.
170   M. Kasturi Bai

No: 335 Trench: D2 Context: 706 Period: 2 No: 15 Trench: A:TP1 Context: 312 Period: 3
Object: Bead Material: Terracotta. Object: Bead Material: Terracotta.
Description: Arecanut bead of terracotta, flattened on either end. Description: Small, spheroid bead of terracotta.
Fig. 8.3.
No: 61 Trench: A:TP1 Context: 312 Period: 3
No: 339 Trench: D2 Context: 711 Period: 2 Object: Bead Material: Terracotta.
Object: Bead Material: Terracotta. Description: Slightly broken spacer bead of terracotta, concave at
Description: Arecanut bead of terracotta with convex base. Fig. 8.3. the bottom and convex on the top.

No: 341 Trench: D2 Context: 711 Period: 2 No: 260 Trench: D1 Context: 751 Period: 3
Object: Bead Material: Terracotta. Object: Bead Material: Terracotta.
Description: Top-shaped bead of terracotta with flattened top. Description: Arecanut bead of terracotta with flat base. Fig. 8.3.
Fig. 8.3.
No: 265 Trench: D1 Context: 751 Period: 3
No: 595 Trench: A:TP1 Context: 944 Period: 1 Object: Bead Material: Terracotta.
Object: Bead Material: Terracotta. Description: Small, arecanut bead of terracotta with flat base. Fig.
Description: Broken, pot-shaped bead of terracotta with black 8.3.
polish and banded decoration.
No: 267 Trench: D1 Context: 751 Period: 3
No: 596 Trench: A:TP1 Context: 944 Period: 2 Object: Bead Material: Terracotta.
Object: Bead Material: Terracotta. Description: Spheroid bead of terracotta in black colour.
Description: Pot-shaped bead of terracotta with black polish.
No: 668 Trench: A:TP3 Context: 1012 Period: 2
No: 652 Trench: A:TP1 Context: 997 Period: 2 Object: Bead Material: Terracotta.
Object: Bead Material: Terracotta. Description: Arecanut bead of terracotta with concave base, flat-
Description: Pumpkin-shaped(?) bead of terracotta, slightly tened top.
broken.
No: 805 Trench: A:TP3 Context: 1140 Period: 3
No: 676 Trench: F Context: 788 Period: 2 Object: Bead Material: Terracotta.
Object: Bead Material: Terracotta. Description: Top-shaped bead of terracotta, truncated on top.
Description: Arecanut bead of terracotta with convex base, trun-
No: 564 Trench: A:TP9, GG North Context: 927 Period: 3-temple
cated on top.
Object: Bead Material: Terracotta.
No: 684 Trench: F Context: 789 Period: 2 Description: Lenticular bead of terracotta, slightly concave at the
Object: Bead Material: Terracotta. top.
Description: Arecanut bead of terracotta with convex base and
No: 666 Trench: A:TP9 Context: 1010 Period: 3-temple
truncated top.
Object: Bead Material: Terracotta.
No: 720 Trench: F Context: 846 Period: 2 Description: Spherical bead of terracotta, truncated on either end.
Object: Bead Material: Terracotta.
No: 795 Trench: A:TP5 Context: 1130 Period: 3-temple
Description: Broken, arecanut bead of terracotta with slightly
Object: Bead Material: Terracotta.
convex base.
Description: Small, arecanut bead of terracotta with convex base
No: 723 Trench: F Context: 846 Period: 2 and truncated top.
Object: Bead Material: Terracotta.
No: 858 Trench: A:TP3 Context: 1095 Period: 3-temple
Description: Arecanut bead of terracotta, slightly concave at the
Object: Bead Material: Terracotta.
base. Fig. 8.3.
Description: Spherical bead of terracotta, truncated on either end
No: 725 Trench: A:TP1 Context: 1077 Period: 2 with black polish. Fig. 8.2.
Object: Bead Material: Terracotta.
No: 59 Trench: A:S of S Temple Context: 303 Period: 4
Description: Arecanut bead of terracotta with convex base, wheel
Object: Bead Material: Terracotta.
marks visible.
Description: Small spherical bead of terracotta.
No: 744 Trench: A:TP1 Context: 1099 Period: 2
No: 172 Trench: B Context: 427 Period: 4
Object: Bead Material: Terracotta.
Object: Bead Material: Terracotta.
Description: Arecanut bead of terracotta with slightly convex base.
Description: Crude and bicone bead of terracotta.
No: 788 Trench: A:TP1 Context: 1120 Period: 2
No: 173 Trench: B Context: 429 Period: 4
Object: Bead Material: Terracotta.
Object: Bead Material: Terracotta.
Description: Long barrel bead of terracotta, tapering on either
Description: Long barrel bead of terracotta, broken on either end.
end, broken at one end.
No: 175 Trench: B Context: 429 Period: 4
No: 793 Trench: A:TP1 Context: 1126 Period: 2
Object: Bead Material: Terracotta.
Object: Bead Material: Terracotta.
Description: Slightly broken long barrel bead of terracotta with cir-
Description: Arecanut bead of terracotta with concave base, flat-
cular section.
tened top.
 Chapter 8 The Small Finds and Glass    171

No: 195 Trench: B Context: 435 Period: 4 No: 196 Trench: B Context: 435 Period: 4
Object: Bead Material: Terracotta. Object: Bead Material: Terracotta.
Description: Six long barrel beads of terracotta with circular Description: Big, irregular bead(?) of terracotta used for animals
section, different sizes, perhaps also used as fish net weights. Fig. or net sinker(?).
8.2.

No: 280 Trench: A:TP6 Context: 596 Period: 2

Object: Bead Material: Terracotta.
Description: Top-shaped bead of terracotta with truncated top,
Topaz(?) beads
vertically broken.
No: 653 Trench: B Context: 467 Period: 1
No: 689 Trench: A:TP9 Context: 684 Period: 4 Object: Bead Material: Topaz.
Object: Bead Material: Terracotta. Description: Big, hexagonal, faceted barrel bead of topaz(?). Figs
Description: Arecanut bead of terracotta with convex base and 8.2 and 8.4.
truncated top. Ref: A similar type of bead made of crystal was found in Period II of
No: 76 Trench: A:NWNT Context: 332 Period: 4? Pauni (Nath 1998: fig. 27, 14).
Object: Bead Material: Terracotta.
Description: Arecanut bead of terracotta with flat base and flat-
tened at the top.
Bangles
No: 89 Trench: A:S Temple Context: 366 Period: No period
Object: Bead Material: Terracotta.
Description: Arecanut bead of terracotta with convex base.
A total of 143 bangles came to light during the excava-
tions (see Table 8.6), although they are very fragmen-
No: 285 Trench: A:South Area Context: 513 Period: No period
tary (Fig. 8.5). Of these, glass bangles are the most
Object: Bead Material: Terracotta.
Description: Arecanut bead of terracotta with concave base and common, followed by shell and then ivory. The bangles
flattened top. found in Period 1 and Period 2 are few in number, but
the number increases in Period 3 and Period 4.
No: 387 Trench: E Context: 801 Period: No period
Object: Bead Material: Terracotta.
Description: Arecanut bead of terracotta with flat base and concave
top. Copper bangles
No: 607 Trench: E Context: 824 Period: No period
Object: Bead Material: Terracotta. Five copper bangles were recorded, one from Period 3,
Description: Squat globular bead of terracotta. three from Period 4 and one from a layer that has not
No: 866 Trench: A:TP7 Context: 1191 Period: No period been allocated to a period. They are completely absent
Object: Bead Material: Terracotta. from other periods.
Description: Arecanut bead of terracotta with flat base.
No: 10 Trench: A:TP1 Context: 308 Period: 3
No: 528 Trench: E Context: 816 Period: No period Object: Bangle Material: Copper.
Object: Bead Material: Terracotta. Description: Fragment of a thin and broad bangle of copper,
Description: Irregular bead of terracotta with convex body and col- concave at the bottom and convex on top. Fig. 8.6.
lared concave base, slightly damaged on top.

Table 8.6: Summary of bangles by material and period.

1 2 3 3-temple 4 4? No period Total

Copper 1 3 1 5
Glass 2 1 13 26 23 65
Ivory 3 2 6 4 7 22
Shell 1 2 16 10 9 2 9 49
Stone 1 1
Terracotta 1 1
Total 4 4 33 12 44 6 40 143
172   M. Kasturi Bai

Fig. 8.5: Bangles of different materials.

 Chapter 8 The Small Finds and Glass    173

Fig. 8.6: Copper objects.

174   M. Kasturi Bai

No: 171 Trench: B Context: 427 Period: 4

Description: Fragment of a bangle of black glass with plano-convex
Object: Bangle Material: Copper.
section.
Description: Fragment of a thin and slender bangle of copper
alloy with circular section. Fig. 8.6. No: 334 Trench: A:South Area Context: 624 Period: 3
Object: Bangle Material: Glass.
No: 191 Trench: B Context: 435 Period: 4
Description: Fragment of a thin bangle of dark green translucent
Object: Bangle Material: Copper.
glass with rectangular section.
Description: Broken, thin bangle of copper alloy with a circular
section. Fig. 8.6. No: 453 Trench: A:South Area Context: 668 Period: 3
Object: Bangle Material: Glass.
No: 192 Trench: B Context: 435 Period: 4
Description: Fragment of a thin bangle of black glass with plano-
Object: Bangle Material: Copper.
convex section.
Description: Curved bangle of copper with one end slightly broad
and squarish, the other end narrow, round and broken. Fig. 8.6. No: 454 Trench: A:South Area Context: 670 Period: 3
Object: Bangle Material: Glass.
No: 661 Trench: E Context: 827 Period: No period
Description: Fragment of a thick bangle of black opaque glass with
Object: Bangle Material: Copper.
plano-convex section.
Description: Two fragments of a thin bangle of copper alloy with
grooves on either margins. No: 455 Trench: A:South Area Context: 670 Period: 3
Object: Bangle Material: Glass.
Description: Fragment of a thin bangle of green translucent glass
with grey coating.
Glass bangles
No: 457 Trench: A:South Area Context: 675 Period: 3
Object: Bangle Material: Glass.
The glass bangles recovered from Period 1 and 2 are Description: Fragment of a thick bangle of translucent, turquoise
fewer in number than in Period 3 and 4. These are glass with rectangular section, decorated by a shallow diamond cut
mostly monochrome. However, a very few polychrome design on top with thick silvery white paint coating.
glass bangles were also found, which exhibit different No: 586 Trench: F Context: 778 Period: 3
colours like black, green, blue, yellow and red in dif- Object: Bangle Material: Glass.
ferent hues. These are mostly plain but a few exhibit Description: Fragment of a bangle of black opaque glass with trap-
ezoidal section.
some decoration. Both thin and slender and broad and
thick varieties of bangles were found. They exhibit rec- No: 669 Trench: A:TP3 Context: 1012 Period: 3
tangular, squarish, biconvex, plano-convex and trian- Object: Bangle Material: Glass.
Description: Fragment of a bangle of dark blue translucent glass
gular sections.
with plano-convex section.
No: 419 Trench: D1 Context: 767 Period: 1
No: 670 Trench: A:TP3 Context: 1012 Period: 3
Object: Bangle Material: Glass.
Object: Bangle Material: Glass.
Description: Fragment of a broad bangle(?) of violet opaque glass.
Description: Fragment of a broad bangle of dark blue opaque glass
Fig. 8.5.
with two incised grooves design on the top.
No: 420 Trench: D1 Context: 767 Period: 1
No: 735 Trench: A:TP6 Context: 1051 Period: 3
Object: Bangle Material: Glass.
Object: Bangle Material: Glass.
Description: Fragment of a bangle of black opaque glass with tri-
Description: Fragment of a perforated bangle of dark blue opaque
angular section.
glass, twisted rope like and with a squarish section.
No: 743 Trench: A:TP1 Context: 1099 Period: 2
No: 239 Trench: A:South Area Context: 568 Period: 4
Object: Bangle Material: Glass.
Object: Bangle Material: Glass.
Description: Fragment of a broad bangle of black opaque glass
Description: Fragment of parrot green bangle of opaque glass
with roughly rectangular section.
with a triangular section and decoration of incised lines on either
No: 250 Trench: A:TP6 Context: 370 Period: 3 margin on the top.
Object: Bangle Material: Glass.
No: 320 Trench: A:TP6, South Area Context: 613 Period: 4
Description: Fragment of a pale yellowish green bangle of opaque
Object: Bangle Material: Glass.
glass over brown base.
Description: Fragment of a thin bangle of blue translucent glass
No: 276 Trench: A:South Area Context: 591 Period: 3 with plano-convex section, with a silvery-shine coating over the
Object: Bangle Material: Glass. body.
Description: Fragment of a bangle of pale blue translucent glass
No: 321 Trench: A:TP6, South Area Context: 613 Period: 4
with ovalish section, coated with silver paint.
Object: Bangle Material: Glass.
No: 333 Trench: A:South Area Context: 624 Period: 3 Description: Fragment of a broad bangle of yellow opaque glass
Object: Bangle Material: Glass. over a pale grey base with a roughly rectangular section.
 Chapter 8 The Small Finds and Glass    175

No: 322 Trench: A:TP6, South Area Context: 613 Period: 4 Description: Fragment of a broad and thin bangle of black opaque
Object: Bangle Material: Glass. glass with a rectangular section, decorated by a series of grooves.
Description: Fragment of a bangle of black opaque glass with a
No: 372 Trench: A:W Area Context: 643 Period: 4
triangular section.
Object: Bangle Material: Glass.
No: 323 Trench: A:TP6, South Area Context: 613 Period: 4 Description: Fragment of a thick and broad bangle(?) of dark
Object: Bangle Material: Glass. honey-coloured opaque glass with a plano-convex section.
Description: Fragment of a bangle of red glass with a plano-con-
No: 376 Trench: A:W Area Context: 644 Period: 4
vex section; it exhibits burnt sienna, green and chrome orange
Object: Bangle Material: Glass.
shades, decorated by incised design.
Description: Fragment of a thin bangle of dark green translucent
No: 324 Trench: A:W Area Context: 618 Period: 4 glass with a plano-convex section.
Object: Bangle Material: Glass.
No: 460 Trench: A:South Area Context: 678 Period: 4
Description: Fragment of a bangle of yellow opaque glass over
Object: Bangle Material: Glass.
pale grey base with a roughly rectangular section.
Description: Fragment of a thin bangle of pale golden yellow trans-
No: 325 Trench: A:W Area Context: 618 Period: 4 lucent glass with a plano-convex section of uneven thickness.
Object: Bangle Material: Glass.
No: 461 Trench: A:South Area Context: 678 Period: 4
Description: Fragment of a thin bangle of black glass with a tri-
Object: Bangle Material: Glass.
angular section.
Description: Fragment of a thin bangle of green opaque glass over
No: 327 Trench: A:South Area Context: 619 Period: 4 a brown base with a plano-convex section.
Object: Bangle Material: Glass.
No: 479 Trench: A:TP9 Context: 684 Period: 4
Description: Fragment of a thin bangle of green opaque glass over
Object: Bangle Material: Glass.
pale blue base with a squarish section.
Description: Fragments of a bangle of yellow opaque glass on a
No: 328 Trench: A:South Area Context: 619 Period: 4 cream base with a trapezoidal section.
Object: Bangle Material: Glass.
No: 480 Trench: A:TP9 Context: 684 Period: 4
Description: Fragment of a thin bangle of yellow opaque glass
Object: Bangle Material: Glass.
over a pale grey base with a rectangular section.
Description: Fragment of a thin bangle of black opaque glass with
No: 329 Trench: A:South Area Context: 619 Period: 4 a triangular section.
Object: Bangle Material: Glass.
No: 520 Trench: A:W Area Context: 920 Period: 4
Description: Fragment of a thin bangle of green opaque glass.
Object: Bangle Material: Glass.
No: 332 Trench: A:South Area Context: 619 Period: 4 Description: Fragment of a bangle of dark blue translucent glass
Object: Bangle Material: Polychrome glass. with a plano-convex section.
Description: Fragment of a bangle of polychrome glass with pale
No: 521 Trench: A:W Area Context: 920 Period: 4
grey at the base followed by yellow and red with white dots at the
Object: Bangle Material: Glass.
apex. Fig. 8.5.
Description: Fragment of a slightly broad bangle of bluish-green
No: 343 Trench: A:South Area, TP6 Context: 635 Period: 4 translucent glass with a plano-convex section.
Object: Bangle Material: Glass.
No: 527 Trench: A:TP10 Context: 928 Period: 4
Description: Fragment of a thin bangle of bluish-green translu-
Object: Bangle Material: Glass.
cent glass with a roughly triangular section.
Description: Fragment of a bangle of yellow opaque glass over a
No: 344 Trench: A:South Area, TP6 Context: 635 Period: 4 grey base with a plano-convex section.
Object: Bangle Material: Glass.
No: 690 Trench: A:TP9 Context: 684 Period: 4
Description: Fragment of a thin bangle of grass-green translucent
Object: Bangle Material: Glass.
glass with a plano-convex section.
Description: Fragment of a thin bangle of black glass with a trian-
No: 348 Trench: A:South Area, TP6 Context: 635 Period: 4 gular section.
Object: Bangle Material: Polychrome glass.
No: 392 Trench: E Context: 801 Period: No period
Description: Fragment of a bangle of stratified glass with green at
Object: Bangle Material: Glass.
the base followed by yellow, red and blue at the top with cut design
Description: Broken bangle of honey-colour opaque glass with a
and white dots in the depressions. Fig. 8.5.
squarish section and a silvery white paint coating on the body.
No: 350 Trench: A:South Area, TP6 Context: 635 Period: 4
No: 393 Trench: E Context: 801 Period: No period
Object: Bangle Material: Glass.
Object: Bangle Material: Glass.
Description: Fragment of a bangle of yellow opaque glass with
Description: Broken bangle of bluish-green translucent glass with a
a triangular section, and small dots of yellow glass affixed at the
silvery white paint coat on the body.
apex.
No: 397 Trench: E Context: 809 Period: No period
No: 359 Trench: A:South Area, TP6 Context: 635 Period: 4
Object: Bangle Material: Glass.
Object: Bangle Material: Glass.
Description: Broken bangle of black glass with a triangular section.
176   M. Kasturi Bai

No: 398 Trench: E Context: 809 Period: No period No: 712 Trench: E Context: 816 Period: No period
Object: Bangle Material: Glass. Object: Bangle Material: Glass.
Description: Broken bangle of blue translucent glass with a plano- Description: Broken bangle of dark green glass with a roughly
convex section. squarish section.

No: 438 Trench: E Context: 811 Period: No period No: 713 Trench: E Context: 816 Period: No period
Object: Bangle Material: Glass. Object: Bangle Material: Glass.
Description: Broken, broad and thick bangle of black opaque Description: Broken bangle of blue translucent glass with a circular
glass with a triangular section, decorated by white dots at inter- section.
vals on the top.
No: 765 Trench: E Context: 816 Period: No period
No: 439 Trench: E Context: 811 Period: No period Object: Bangle Material: Glass.
Object: Bangle Material: Glass. Description: Fragment of a broad bangle of black glass with a
Description: Fragment of a thin bangle of pale brown opaque glass plano-convex section.
with a rectangular section, decorated by bluish-green dots at inter-
No: 766 Trench: E Context: 816 Period: No period
vals on top.
Object: Bangle Material: Glass.
No: 530 Trench: E Context: 816 Period: No period Description: Fragment of a bangle of grass-green opaque glass over
Object: Bangle Material: Glass. a pale grey base with a triangular section.
Description: Broken bangle of blue translucent glass with a
No: 767 Trench: E Context: 816 Period: No period
roughly triangular section. Fig. 8.5.
Object: Bangle Material: Glass.
No: 534 Trench: E Context: 816 Period: No period Description: Fragment of a bangle of blue glass with incised deco-
Object: Bangle Material: Glass. ration on top.
Description: Fragment of a bangle of dark green translucent glass
No: 838 Trench: A:TP7 Context: 1191 Period: No period
with a plano-convex section.
Object: Bangle Material: Glass.
No: 550 Trench: E Context: 812 Period: No period Description: Fragment of a bangle of black glass with a round
Object: Bangle Material: Glass. section.
Description: Fragment of a bangle of pale brick-red opaque glass
No: 869 Trench: A:TP8 Context: 1216 Period: No period
over a pale grey base with a triangular section.
Object: Bangle Material: Glass.
No: 551 Trench: E Context: 812 Period: No period Description: Fragment of a bangle of yellow opaque glass over a
Object: Bangle Material: Glass. pale grey base with a roughly rectangular section.
Description: Fragment of a bangle of yellow opaque glass over a
No: 60 Trench: A:TP1 Context: 312 Period: 3
pale grey base with a roughly rectangular section.
Object: Bangle Material: Glass.
No: 552 Trench: E Context: 812 Period: No period Description: Fragment of a bangle of pale yellow opaque soft glass
Object: Bangle Material: Glass. with a circular section, flaky.
Description: Fragment of a broad and thin bangle of black glass
with a plano-convex section.

No: 561 Trench: E Context: 818 Period: No period Ivory bangles

Object: Bangle Material: Glass.
Description: Broken bangle of black glass with a triangular section.
A total of 22 ivory bangles were found. These are com-
No: 632 Trench: E Context: 826 Period: No period pletely absent in Periods 1 and 2. They were found in thin
Object: Bangle Material: Glass.
and slender as well as broad and thin varieties. Many of
Description: Broken bangle of black glass with a triangular section.
the bangles show excellent incised and embossed deco-
No: 633 Trench: E Context: 826 Period: No period rative motifs on their top surface.
Object: Bangle Material: Glass.
Description: Fragment of a broad bangle of black glass with a No: 251 Trench: A:TP6 Context: 370 Period: 3
plano-convex section. Object: Bangle Material: Ivory.
Description: Thin and slender bangle of ivory(?) with rectangular
No: 665 Trench: E Context: 829 Period: No period section.
Object: Bangle Material: Glass.
Description: Fragment of a bangle of black glass with thin incised No: 403 Trench: A:South Area Context: 647 Period: 3
grooves on top. Object: Bangle Material: Ivory.
Description: Fragment of a narrow and thin bangle of ivory with
No: 687 Trench: E Context: 829 Period: No period a rectangular section, decorated on top by triangular pattern, bor-
Object: Bangle Material: Glass. dered by incised grooves on either margin. Fig. 8.5.
Description: Fragment of a thick bangle of black glass with a cir-
cular section. No: 584 Trench: F Context: 778 Period: 3
Object: Bangle Material: Ivory.
 Chapter 8 The Small Finds and Glass    177

Description: Fragments of a bangle of ivory with a rectangular No: 138 Trench: A:NWNT Context: 546 Period: 4?
section; one piece is decorated on top by two rows of circles in Object: Bangle Material: Ivory.
shallow relief, and the second piece is a narrow and thin bangle Description: Fragment of a bangle of ivory(?) with a triangular
decorated on top by two rows of circles in shallow relief with traces section.
of pink paint. Fig. 8.5.
No: 23 Trench: A:TP2 Context: 317 Period: No period
Ref: Parallels can be drawn with a similar specimen from Period V of
Object: Bangle Material: Ivory.
Nevasa (Sankalia et al. 1960: 468, fig. 198, 4).
Description: Fragment of a bangle of ivory(?) with a biconvex section.
No: 489 Trench: A:TP5 Context: 700 Period: 3-temple
No: 563 Trench: E Context: 821 Period: No period
Object: Bangle Material: Ivory.
Object: Bangle Material: Ivory.
Description: Fragment of a bangle of ivory(?) with a roughly rectan-
Description: Broken, thin and slender bangle of ivory(?) with a
gular section.
rectangular section.
No: 608 Trench: A:TP5 Context: 955 Period: 3-temple
No: 662 Trench: E Context: 827 Period: No period
Object: Bangle Material: Ivory.
Object: Bangle Material: Ivory.
Description: Fragment of a bangle of ivory(?) with a roughly rectan-
Description: Fragment of a thin and slender flat bangle of ivory(?)
gular section.
with a rectangular section.
No: 62 Trench: A:TP5 Context: 330 Period: 4
No: 686 Trench: E Context: 829 Period: No period
Object: Bangle Material: Ivory.
Object: Bangle Material: Ivory.
Description: Fragment of an ivory(?) bangle with a roughly bicon-
Description: Five broken, thin, slender and flat bangles of ivory(?)
vex section.
with a rectangular section and occasional holes on body.
No: 97 Trench: A:NWNT Context: 376 Period: 4
No: 768 Trench: E Context: 816 Period: No period
Object: Bangle Material: Ivory.
Object: Bangle Material: Ivory.
Description: Fragment of a thick bangle of ivory(?) with a rectan-
Description: Broad, thin and flat bangle pieces of ivory with a rec-
gular section.
tangular section and one piece having a hole in the centre.
No: 98 Trench: A:NWNT Context: 376 Period: 4
No: 833 Trench: E Context: 1207 Period: No period
Object: Bangle Material: Ivory.
Object: Bangle Material: Ivory.
Description: Fragment of a bangle of ivory(?) with a roughly rectan-
Description: Broken bangle of ivory(?) which is under prepara-
gular section.
tion.
No: 121 Trench: A:TP9 Context: 519 Period: 4
No: 834 Trench: E Context: 1208 Period: No period
Object: Bangle Material: Ivory.
Object: Bangle Material: Ivory.
Description: Fragment of a slender and thin bangle of ivory with a
Description: Broken, thin and slender bangle of ivory(?) with a
rectangular section.
squarish section.
No: 183 Trench: B Context: 430 Period: 4
Object: Bangle Material: Ivory.
Description: Fragment of a broad bangle of ivory with a rectangular
section, decorated on top by a row of fishes in the centre in a low
Shell bangles
relief, traces of pink paint on the body. Fig. 8.5.
A total of 49 shell bangles came to light. Their number
No: 519 Trench: A:W Area Context: 920 Period: 4
Object: Bangle Material: Ivory. gradually increases from Period 1 to Period 3 and
Description: Fragment of a slightly broad bangle of ivory(?) with decreases in Period 4. Except for a few decorated
plano-convex section, decorated on top by circles in low relief in examples, all are plain bangles. Both thick and thin
the centre, bordered by incised grooves. varieties were found.
No: 79 Trench: A:NWNT Context: 336 Period: 4? No: 626 Trench: B Context: 467 Period: 1
Object: Bangle Material: Ivory. Object: Bangle Material: Shell.
Description: Fragment of a broad ivory bangle decorated on top Description: Fragment of a faceted bangle of shell(?) with a pen-
by two rows of circles in shallow relief, a horizontal line separating tagonal section.
the rows.
No: 594 Trench: A:TP1 Context: 944 Period: 2
No: 116 Trench: A:NWNT Context: 507 Period: 4? Object: Bangle Material: Shell.
Object: Bangle Material: Ivory. Description: Fragment of a bangle of shell with a rectangular
Description: Fragment of a bangle of ivory(?) with a rectangular section and with a patina.
section.
No: 772 Trench: A:TP1 Context: 1111 Period: 2
No: 134 Trench: A:NWNT Context: 540 Period: 4? Object: Bangle Material: Shell.
Object: Bangle Material: Ivory. Description: Fragment of a thin and slender bangle of shell with
Description: Fragment of a slender and thin bangle of ivory with a a rectangular section.
rectangular section.
178   M. Kasturi Bai

No: 9 Trench: A:TP1 Context: 307 Period: 3 Description: Fragment of a broad and thick bangle of shell with a
Object: Bangle Material: Shell. rectangular section.
Description: Fragment of a thick bangle of shell with a rectangular
No: 645 Trench: A:TP6 Context: 987 Period: 3
section.
Object: Bangle Material: Shell.
No: 18 Trench: A:TP1 Context: 312 Period: 3 Description: Fragment of a thin and slender bangle of shell with
Object: Bangle Material: Shell. a rectangular section.
Description: Fragment of a slender bangle of shell with a plano-
No: 736 Trench: A:TP6 Context: 1051 Period: 3
convex section.
Object: Bangle Material: Shell.
No: 257 Trench: A:South Area Context: 589 Period: 3 Description: Fragment of a broad and thick bangle of shell with a
Object: Bangle Material: Shell. pentagonal section.
Description: Fragment of a slender bangle of shell with a rectangu-
No: 361 Trench: A:TP5 Context: 638 Period: 3-temple
lar section and plano-convex ends.
Object: Bangle Material: Shell.
No: 270 Trench: A:South Area Context: 591 Period: 3 Description: Fragment of a shell bangle with a roughly biconvex
Object: Bangle Material: Shell. section.
Description: Fragment of a shell bangle with a rectangular section.
No: 368 Trench: A:TP5 Context: 642 Period: 3-temple
No: 271 Trench: A:South Area Context: 591 Period: 3 Object: Bangle Material: Shell.
Object: Bangle Material: Shell. Description: Fragment of a thin and broad shell bangle with a
Description: Fragment of a broad and thin shell bangle with a len- roughly rectangular section.
ticular section.
No: 458 Trench: A:TP5 Context: 667 Period: 3-temple
No: 273 Trench: A:South Area Context: 591 Period: 3 Object: Bangle Material: Shell.
Object: Bangle Material: Shell. Description: Fragment of a thin bangle of shell.
Description: Fragment of a shell bangle with a rectangular section.
No: 514 Trench: A:TP5 Context: 918 Period: 3-temple
No: 275 Trench: A:South Area Context: 591 Period: 3 Object: Bangle Material: Shell.
Object: Bangle Material: Shell. Description: Fragment of a thin bangle of shell with a rectangular
Description: Fragment of a shell bangle with a rectangular section. section.

No: 307 Trench: A:South Area Context: 608 Period: 3 No: 539 Trench: A:TP5 Context: 918 Period: 3-temple
Object: Bangle Material: Shell. Object: Bangle Material: Shell.
Description: Fragment of a broad and thin bangle of shell with a Description: Fragment of a shell bangle with a rectangular section.
rectangular section.
No: 588 Trench: A:TP5 Context: 918 Period: 3-temple
No: 309 Trench: A:South Area Context: 608 Period: 3 Object: Bangle Material: Shell.
Object: Bangle Material: Shell. Description: Fragment of a thin bangle of shell(?) with a rectangu-
Description: Fragment of a slender shell bangle with a rectangular lar section.
section and plano-convex ends.
No: 641 Trench: A:TP5 Context: 978 Period: 3-temple
No: 330 Trench: A:TP5 Context: 622 Period: 2 Object: Bangle Material: Shell.
Object: Bangle Material: Shell. Description: Fragment of a thick bangle of shell with a rectangular
Description: Fragment of a broad and thin bangle of shell with a section.
lenticular section, decorated on top by a circle in low relief followed
No: 691 Trench: A:TP9 Context: 1010 Period: 3-temple
by a series of incised vertical grooves. Fig. 8.5.
Object: Bangle Material: Shell.
No: 500 Trench: A:W Area, TP6 Context: 650 Period: 3 Description: Fragment of a broad and thin bangle of shell with a
Object: Bangle Material: Shell. roughly rectangular section.
Description: Fragment of a bangle of shell with a rectangular
No: 848 Trench: A:TP5 Context: 978 Period: 3-temple
section.
Object: Bangle Material: Shell.
No: 568 Trench: A:South Area Context: 931 Period: 3 Description: Fragment of a thin and slender bangle of shell with a
Object: Bangle Material: Shell. rectangular section.
Description: Fragment of a broad and thin bangle of shell with a
No: 862 Trench: A:TP3 Context: 1112 Period: 3-temple
rectangular section.
Object: Bangle Material: Shell.
No: 628 Trench: A:TP6 Context: 373 Period: 3 Description: Fragment of a bangle of shell with an irregular
Object: Bangle Material: Shell. section, pecked and not ground.
Description: Fragment of a bangle of shell with a rectangular
No: 34 Trench: A:NWNT, TP9 Context: 328 Period: 4
section.
Object: Bangle Material: Shell.
No: 630 Trench: A:TP6 Context: 373 Period: 3 Description: Fragment of a decorated shell bangle.
Object: Bangle Material: Shell.
 Chapter 8 The Small Finds and Glass    179

No: 53 Trench: A:S of S Temple Context: 303 Period: 4 Description: Fragment of a broad and thick bangle of shell with a
Object: Bangle Material: Shell. squarish section.
Description: Fragment of a bangle of shell with a rectangular
No: 749 Trench: E Context: 841 Period: No period
section.
Object: Bangle Material: Shell.
No: 224 Trench: B Context: 437 Period: 4 Description: Fragment of a bangle of shell(?).
Object: Bangle Material: Shell.
No: 781 Trench: E Context: 844 Period: No period
Description: Fragment of a shell bangle with a rectangular section
Object: Bangle Material: Shell.
and plano-convex ends.
Description: Thick and broken bangle of shell with an embossed
No: 227 Trench: B Context: 438 Period: 4 elongated oval decoration on the top.
Object: Bangle Material: Shell.
No: 782 Trench: E Context: 844 Period: No period
Description: Fragment of a slender bangle of shell(?) with a rec-
Object: Bangle Material: Shell.
tangular section and plano-convex ends.
Description: Broken bangle of shell with a rectangular section.
No: 292 Trench: A:South Area Context: 605 Period: 4
No: 783 Trench: E Context: 844 Period: No period
Object: Bangle Material: Shell.
Object: Bangle Material: Shell.
Description: Two fragments of thin slender bangles of shell with
Description: Fragment of a broad bangle of shell with a roughly rec-
a rectangular section, one piece having group of circles painted in
tangular section.
violet colour on top.
No: 786 Trench: E Context: 844 Period: No period
No: 360 Trench: A:W Area Context: 636 Period: 4
Object: Bangle Material: Shell.
Object: Bangle Material: Shell.
Description: Broken and thick bangle of shell with a squarish
Description: Fragment of a slender and thin shell bangle with a
section.
rectangular section and plano-convex ends.
No: 832 Trench: E Context: 1207 Period: No period
No: 406 Trench: A:TP6 Context: 649 Period: 2
Object: Bangle Material: Shell.
Object: Bangle Material: Shell.
Description: Broken bangle of shell with a plano-convex section.
Description: Fragment of a broad and thin bangle of shell with an
irregular section, and decoration of a band of incised lines on top.

No: 483 Trench: A:South Area Context: 645 Period: 4

Object: Bangle Material: Shell. Stone bangle
Description: Fragment of a broad bangle of shell with a rectangular
section. No: 533 Trench: D2 Context: 726 Period: 1
Object: Bangle Material: Stone.
No: 491 Trench: A:TP6 Context: 905 Period: 4
Description: Broken and pulley-shaped bangle(?) of grass-green
Object: Bangle Material: Shell.
stone(?). Fig. 8.5.
Description: Fragment of a bangle of shell with a rectangular
section.

No: 112 Trench: A:NWNT Context: 504 Period: 4?

Object: Bangle Material: Shell. Terracotta bangle
Description: Fragment of a broad bangle of shell(?) decorated on
top by excellent incised design. Fig. 8.5. No: 268 Trench: D1 Context: 752 Period: 2
Object: Bangle Material: Terracotta.
No: 124 Trench: A:NWNT Context: 529 Period: 4? Description: Fragment of a bangle of terracotta with gadrooned
Object: Bangle Material: Shell. design on top. Fig. 8.5.
Description: Fragment of a thick bangle of shell(?) decorated on
top by ovalish relief on the outer surface. Fig. 8.5.
Ref: A similar shell bangle was found from Period V of Nevasa
(Sankalia et al. 1960: 462, fig. 196, 8). Finger-rings
No: 41 Trench: A:TP2 Context: 323 Period: No period
Object: Bangle Material: Shell. Copper finger-rings
Description: Fragment of a slender and thick shell bangle with a
rectangular section. No: 683 Trench: B Context: 484 Period: 1
Object: Finger-ring Material: Copper.
No: 549 Trench: E Context: 812 Period: No period
Description: Fragments of a broad and thin finger-ring of copper
Object: Bangle Material: Shell.
alloy with greenish patina.
Description: Broken, broad and thick bangle with a roughly rectan-
gular section. No: 729 Trench: B Context: 493 Period: 1
Object: Finger-ring Material: Copper.
No: 701 Trench: E Context: 831 Period: No period
Description: Fragment of a finger-ring(?) of copper, corroded.
Object: Bangle Material: Shell.
180   M. Kasturi Bai

No: 482 Trench: B Context: 444 Period: 2 No: 314 Trench: A:South Area Context: 608 Period: 3
Object: Finger-ring Material: Copper. Object: Finger-ring Material: Shell.
Description: Broken finger-ring of copper alloy with wavy edge Description: Broken, thin finger-ring of shell with a plano-convex
and faceted design. Fig. 8.6. section.

No: 598 Trench: A:TP1 Context: 944 Period: 2

Object: Finger-ring Material: Copper.
Description: Fragment of a finger-ring(?) of copper alloy, corroded.
Fig. 8.6.
Finger-ring stones
No: 16 Trench: A:TP1 Context: 312 Period: 3 No: 404 Trench: A:South Area Context: 647 Period: 3
Object: Finger-ring Material: Copper. Object: Ring stone Material: Stone.
Description: Heart-shaped finger-ring of copper with elongated Description: Finger-ring stone of cream colour with plano-convex
diamond decoration on the top. Fig. 8.6. section.

No: 194 Trench: B Context: 435 Period: 4 No: 3 Trench: A:TP1 Context: 300 Period: 4
Object: Finger-ring Material: Copper. Object: Ring stone Material: Stone.
Description: Finger-ring of copper alloy with a slight projection at Description: Finger-ring stone(?) of white colour with plano-con-
the top. Fig. 8.6. vex section.

No: 471 Trench: A:S Wall of N Context: 584 Period: 4

Object: Finger-ring Material: Copper.
Description: Slightly broken finger-ring of copper alloy. Toe-rings
No: 140 Trench: A:NWNT Context: 550 Period: 4? No: 193 Trench: B Context: 435 Period: 4
Object: Finger-ring Material: Copper. Object: Toe-ring Material: Copper.
Description: Fragment of a finger-ring(?) of copper alloy, corroded. Description: Toe-ring of copper alloy, which is a twisted thin wire.
No: 26 Trench: None Context: 0 Period: No period No: 443 Trench: A:South Area, TP6 Context: 635 Period: 4
Object: Finger-ring Material: Copper. Object: Toe-ring Material: Copper.
Description: Finger-ring of copper alloy with a projecting disc as Description: Toe-ring of copper alloy. Fig. 8.6.
decoration, encrusted.

No: 391 Trench: E Context: 801 Period: No period

Object: Finger-ring Material: Copper.
Description: Broken finger-ring of copper alloy. Ear-studs
No: 38 Trench: A:TP9 Context: 320 Period: 4
Object: Ear-stud Material: Copper.
Glass finger-rings Description: Ear-stud(?) of copper alloy with a concentric spool
shape.

No: 487 Trench: A:TP5 Context: 700 Period: 3-temple No: 284 Trench: D1 Context: 751 Period: 3
Object: Finger-ring Material: Glass. Object: Ear-stud Material: Terracotta.
Description: Two fragments of a finger-ring of pale grey opaque, Description: Thick, disc-shaped ear-stud of terracotta with a cir-
soft glass with circular section. cular section, slightly damaged and black in colour. Fig. 8.7.

No: 407 Trench: A:TP6 Context: 649 Period: 4 No: 347 Trench: A:South Area, TP6 Context: 635 Period: 4
Object: Finger-ring Material: Glass. Object: Ear-stud Material: Terracotta.
Description: Broken finger-ring of yellow opaque glass over a cream Description: Cylinder-shaped ear-stud of terracotta with a circu-
base with triangular section. lar section and slightly concave on either ends.

No: 395 Trench: E Context: 805 Period: No period No: 88 Trench: A:S Temple Context: 366 Period: No period
Object: Finger-ring Material: Glass. Object: Ear stud Material: Terracotta.
Description: Broken finger-ring of grass-green opaque glass. Description: Cylinder-shaped ear-stud of terracotta with one lobe
flat and the other lobe slightly concave, perforated.

No: 681 Trench: F Context: 788 Period: 2

Object: Ear-stud Material: Terracotta.
Shell finger-ring
Description: Fragment of a reel-shaped ear-stud of terracotta. Fig.
8.7.
No: 14 Trench: A:TP1 Context: 308 Period: 3
Object: Finger-ring Material: Shell.
Description: Fragment of a thick finger-ring(?) of shell with a rec-
tangular section.
 Chapter 8 The Small Finds and Glass    181

Fig. 8.7: Ear-studs, pendants and other terracotta objects.

182   M. Kasturi Bai

Pendants Description: Kohlstick of ivory(?) with a bud-shaped head at the top

and slightly broken at the bottom, circular section. Fig. 8.8.
No: 151 Trench: B Context: 402 Period: 4 Ref: A similar but complete specimen was found in Period IV at
Object: Pendant Material: Copper. Nevasa with a little variation in the decoration (Sankalia et al.
Description: Vertically broken tubular pendant having two loops 1960: fig. 194, 8) and also from Period II at Nasik (Sankalia and
on top. Made of copper alloy. Fig. 8.6. Deo 1955: fig. 55, 6).
No: 109 Trench: A:NWNT Context: 394 Period: 4? No: 602 Trench: F Context: 778 Period: 3
Object: Pendant Material: Agate. Object: Kohlstick Material: Ivory.
Description: Drop pendant of banded agate. Figs 8.2 and 8.4. Description: Kohlstick of ivory, broader flat end at the top with
No: 754 Trench: F Context: 851 Period: 1 excellent incised decoration and slightly broken, ground and pol-
Object: Pendant Material: Terracotta. ished. Fig. 8.8.
Description: Conical-shaped pendant of terracotta, slightly concave Ref: Similar specimens were recovered from Period II of Nasik
at the base, tip broken. Fig. 8.7. (Sankalia and Deo 1955: fig. 55, 1 and 4, Plate XXIX, 10, 12 and 17),
but with some differences in the decoration.
No: 756 Trench: F Context: 851 Period: 1
Object: Pendant Material: Terracotta. No: 675 Trench: F Context: 788 Period: 2
Description: Conical-shaped pendant of terracotta, slightly broken Object: Kohlstick Material: Ivory.
at the distal end. Description: Kohlstick of ivory(?) with a narrow blunt pointed end
and broad decorated broken head, ground and polished. Fig. 8.8.
No: 336 Trench: D2 Context: 706 Period: 2
Object: Pendant Material: Terracotta.
Description: Conical-shaped pendant of terracotta, slightly
broken at the base, grits visible. Fig. 8.7. Inlay objects
No: 698 Trench: A:TP1 Context: 1052 Period: 2 No: 104 Trench: A:NWNT Context: 376 Period: 4
Object: Pendant Material: Terracotta. Object: Inlay Material: Gold.
Description: Cylinder-shaped pendant of terracotta, slightly Description: Very thin, flat, spheriod-shaped inlay(?) object of
tapering towards upper end. gold.
No: 131 Trench: A:TP2 Context: 538 Period: No period No: 281 Trench: A:TP5 Context: 599 Period: 3
Object: Pendant Material: Terracotta. Object: Inlay Material: Shell.
Description: Conical-shaped pendant(?) of terracotta, broken at Description: Squarish, flat inlay object of shell(?) with incised
the top. cross design on top. Fig. 8.5.
No: 867 Trench: E Context: 1210 Period: No period No: 408 Trench: A:South Area, W Area Context: 651 Period: 3
Object: Pendant Material: Terracotta. Object: Inlay Material: Glass.
Description: Conical-shaped pendant of terracotta with a circular Description: Inlay piece of turquoise glass.
section, flat at the base and slightly broken at the top.

No: 326 Trench: A:South Area Context: 619 Period: 4

Object: Pendant Material: Terracotta.
Description: Perforated triangle-shaped pendant of a grey ware
Gold foil and band
sherd, slightly broken. No: 806 Trench: A:TP1 Context: 1144 Period: 1
No: 478 Trench: E Context: 809 Period: No period Object: Foil Material: Gold.
Object: Pendant Material: Terracotta. Description: Fragment of a very thin foil of gold.
Description: Broken, thick perforated sherd of buff ware used as No: 792 Trench: A:TP1 Context: 1126 Period: 1
a pendant for animals. Object: Foil Material: Gold.
Description: Small and thin foil of gold.

No: 238 Trench: A:NWNT Context: 503 Period: 4?

Hair-pin Object: Foil Material: Gold.
Description: Fragment of a very thin foil of gold.
No: 373 Trench: A:W Area Context: 643 Period: 4
No: 364 Trench: D2 Context: 712 Period: 2
Object: Hair-pin Material: Bone.
Object: Band Material: Gold.
Description: Broken hair-pin(?) of bone.
Description: Fragment of thin band of gold. Glass wire.

No: 262 Trench: D1 Context: 751 Period: 3

Object: Wire Material: Glass.
Kohlsticks Description: Greenish glass wire(?) with a circular section and
white encrustation on surface.
No: 264 Trench: D1 Context: 751 Period: 3
Object: Kohlstick Material: Ivory.
 Chapter 8 The Small Finds and Glass    183

Fig. 8.8: Bone and ivory objects.

184   M. Kasturi Bai

Metal objects No: 615 Trench: A:TP1 Context: 966 Period: 2

Objects: Nail Material: Copper.
Description: Fragments of a thin and slender nail(?) of copper alloy.
A total of 203 metal objects came to light during the excava-
tions (see Table 8.7). Metal objects recovered from the exca- No: 75 Trench: A:NWNT Context: 332 Period: 4?
vations occupy the second place numerically among the Object: Nail Material: Copper.
Description: Small curved nail(?) of copper, corroded. Fig. 8.6.
small finds. Iron is the chief metal used for the preparation
of various objects, followed by copper. Brass, bronze and No: 110 Trench: A:TP9 Context: 398 Period: No period
silver were the least represented metals. The usage of iron Object: Nail Material: Copper.
Description: Fragment of a thin and narrow nail-cleaner(?) of
and copper objects increased from Period 1 to Period 4.
copper.

Brass and bronze objects

Rivet
No: 247 Trench: A:South Area Context: 568 Period: 4
Object: Unidentified Material: Brass. No: 287 Trench: D1 Context: 754 Period: 1
Description: Thin, circular inlay object of brass with an embossed Object: Rivet Material: Copper.
floral motif on top and a negative impression on the bottom, Description: Broken ribbon-shaped rivet of copper alloy, one end
slightly broken at the edge. curved. Fig. 8.6.

No: 137 Trench: A:NWNT Context: 546 Period: 4?

Object: Rod Material: Bronze.
Description: Fragment of a slender rod of bronze(?) with a squarish Fastener
section. Fig. 8.6.
No: 166 Trench: B Context: 414 Period: 4
Object: Fastener Material: Copper.
Description: Ribbon-shaped, thin fastener of copper alloy, slightly
Copper objects disfigured and corroded. Fig. 8.6.

A total of 22 copper metal objects came to light during

the excavations (not including the copper objects that
Needle
have been categorized as jewellery; Table 8.8). For an
illustration of some of these, see Fig. 8.6. No: 225 Trench: B Context: 437 Period: 4
Object: Needle Material: Copper.
Description: Broken needle(?) of copper, corroded. Fig. 8.6.
Nails

No: 575 Trench: B Context: 458 Period: 1

Object: Nail Material: Copper.
Description: Fragment of a nail of copper alloy, corroded and
encrusted.

Table 8.7: Summary of metal objects by type of metal and period.

1 2 3 3-temple 4 4? No period Total

Brass 1 1
Bronze 1 1
Copper 4 1 1 6 7 1 2 22
Iron 14 13 27 22 77 8 12 173
Iron? 1 1
Lead 3 1 4
Silver 1 1
Total 19 14 28 28 89 10 15 203
 Chapter 8 The Small Finds and Glass    185

Rods Unidentified

No: 48 Trench: A:S of S Temple Context: 303 Period: 4 No: 799 Trench: B Context: 736 Period: 1
Object: Rod Material: Copper. Object: Unidentified Material: Copper.
Description: Long cylindrical rod of bronze with plano-convex Description: Fragment of a thin unidentified object of copper alloy,
section, one end broad with incised grooves on either margin and badly damaged.
narrowing to a blunt point on the other end, perhaps used as a
No: 734 Trench: A:TP6 Context: 1051 Period: 3
nail-cleaner. Fig. 8.6.
Object: Unidentified Material: Copper.
No: 163 Trench: B Context: 414 Period: 4 Description: Fragment of a thin and flat unidentified object of
Object: Rod Material: Copper. copper alloy, perhaps a nail-cleaner.
Description: A long, slender and broken rod of copper with one end
No: 590 Trench: A:GG North Context: 938 Period: 3-temple
tapering to sharpness and the other end broadening. Fig. 8.6.
Object: Unidentified Material: Copper.
Description: Fragment of an unidentified object of copper alloy,
corroded.
Bowl
No: 818 Trench: A:TP9 Context: 1156 Period: 3-temple
Object: Unidentified Material: Copper.
No: 696 Trench: A:TP9 Context: 1050 Period: 3-temple Description: Unidentified object of copper alloy.
Object: Bowl Material: Copper.
Description: Fragment of a bowl(?) of copper alloy, encrusted. No: 828 Trench: A:TP9 Context: 1170 Period: 3-temple
Object: Unidentified Material: Copper.
Description: Fragments of an unidentified object of copper, green-
ish patina.
Pellate
No: 231 Trench: B Context: 439 Period: 4
Object: Unidentified Material: Copper.
No: 517 Trench: A:TP5 Context: 918 Period: 3-temple
Description: Unidentified copper object.
Object: Pellate Material: Copper.
Description: Fragments of a pellate of copper alloy, badly damaged No: 437 Trench: E Context: 811 Period: No period
due to corrosion. Object: Unidentified Material: Copper.
Description: Cut piece of an unidentified object of copper.

Plates
Iron objects
No: 380 Trench: D2 Context: 718 Period: 1
Object: Plate Material: Copper. A total of 173 iron objects came to light during the exca-
Description: Fragment of a copper-plate(?) with greenish patina.
vations (see Table 8.9). Nails are predominant, followed
No: 375 Trench: A:W Area Context: 643 Period: 4 by knives and rings. The nails, rivets, fasteners and rings
Object: Plate Material: Copper.
might have been used for construction and mending pur-
Description: Fragment of a thin plate of copper with greenish
patina.
poses. The chisels and borers were perhaps used in car-
pentry. The arrowheads, spears and spikes were used for
defence purposes. The sickles, spatula, buckle, clamp,
latch and wires might have been used for household pur-
Handle of a bell
poses (see Fig. 8.9).
No: 230 Trench: B Context: 439 Period: 4
Object: Handle Material: Copper.
Description: Tubular handle of a bell of copper alloy, with ‘u’- Arrowheads
shaped loop attached to one end, broad and slightly concave disc-
shaped on the other end, corroded. Fig. 8.6. No: 794 Trench: A:TP5 Context: 1130 Period: 3-temple
Object: Arrowhead Material: Iron.
Description: Broken arrowhead(?) of iron, corroded.

Wire No: 177 Trench: B Context: 430 Period: 4

Object: Arrowhead Material: Iron.
Description: Tanged arrowhead of iron, broken at the tip, blade is
No: 609 Trench: A:TP5 Context: 964 Period: 3-temple
four-faced and spike-like, encrusted.
Object: Wire Material: Copper.
Description: Fragment of a very thin wire of copper alloy.
186   M. Kasturi Bai

Fig. 8.9: Iron objects.

 Chapter 8 The Small Finds and Glass    187

No: 308 Trench: A:South Area Context: 608 Period: 3 No: 296 Trench: A:TP5 Context: 610 Period: 2
Object: Arrowhead Material: Iron. Object: Knife Material: Iron.
Description: Leaf-shaped arrowhead(?) of iron, other end broken. Description: Fragment of an iron knife(?), broken on either end,
encrusted.
No: 362 Trench: A:TP5 Context: 638 Period: 3-temple
Object: Arrowhead Material: Iron. No: 565 Trench: A:TP1 Context: 930 Period: 3
Description: Fragment of a leaf-shaped arrowhead(?) of iron, cor- Object: Knife Material: Iron.
roded. Description: Fragment of a knife(?) of iron, corroded and flaky.

No: 204 Trench: B Context: 435 Period: 4 No: 369 Trench: A:TP5 Context: 642 Period: 3-temple
Object: Arrowhead Material: Iron. Object: Knife Material: Iron.
Description: Small, leaf-shaped arrowhead of iron with a tang, Description: Fragment of an iron knife, broken on either end, cor-
pointed end slightly broken, tang curved like a loop. roded.

No: 638 Trench: A:TP3 Context: 981 Period: 3-temple

Object: Knife Material: Iron.
Spear Description: Iron knife(?) broken on either end, corroded.

No: 203 Trench: B Context: 435 Period: 4

No: 532 Trench: D2 Context: 726 Period: 1 Object: Knife Material: Iron.
Object: Spear Material: Iron. Description: Iron knife with a tang, broken on either end, encrusted.
Description: Two fragments of an iron spear with tang, encrusted.
No: 210 Trench: B Context: 435 Period: 4
Object: Knife Material: Iron.
Description: Broken barber’s knife of iron with handle, corroded.

Tanged point No: 232 Trench: B Context: 440 Period: 4

Object: Knife Material: Iron.
No: 570 Trench: A:TP1 Context: 933 Period: 3 Description: Fragment of a blade of iron knife, heavily corroded.
Object: Point Material: Iron.
No: 248 Trench: A:S Wall of N Context: 575 Period: 4
Description: Tanged point(?) of iron with a broken tip, handle
Object: Knife Material: Iron.
curved and corroded.
Description: Iron knife with handle, broken on either end, cor-
roded.

Knives

No: 433 Trench: D2 Context: 721 Period: 1

Handles
Object: Knife Material: Iron.
Description: Fragment of a knife of iron, encrusted. No: 208 Trench: B Context: 435 Period: 4
Object: Handle Material: Iron.
No: 656 Trench: B Context: 469 Period: 1 Description: Tubular handle of iron object, corroded.
Object: Knife Material: Iron.
Description: Iron knife(?), broken on either end, encrusted. No: 243 Trench: A:South Area Context: 568 Period: 4
Object: Handle Material: Iron.
No: 658 Trench: B Context: 469 Period: 1 Description: Vertically broken tubular handle of an iron knife
Object: Knife Material: Iron. with a hole in the centre.
Description: Broken iron knife, encrusted.
No: 293 Trench: A:South Area Context: 605 Period: 4
No: 719 Trench: A:TP1 Context: 1074 Period: 2 Object: Handle Material: Iron.
Object: Knife Material: Iron. Description: Vertically broken tubular handle(?) of an iron, cor-
Description: Broken blade of an iron knife with handle missing, roded.
encrusted.

No: 747 Trench: A:TP1 Context: 1099 Period: 2

Object: Knife Material: Iron.
Description: Broken blade of a knife of iron, encrusted. Chisels

No: 771 Trench: A:TP1 Context: 1110 Period: 2 No: 399 Trench: A:South Area, TP6 Context: 635 Period: 4
Object: Knife Material: Iron. Object: Chisel Material: Iron.
Description: Broken blade of a knife of iron, corroded. Description: Fragment of an iron chisel(?), corroded and encrusted.
No: 266 Trench: D1 Context: 751 Period: 3 No: 844 Trench: A:TP3 Context: 974 Period: 3
Object: Knife Material: Iron. Object: Chisel Material: Iron.
Description: Fragment of an iron knife with encrustation. Description: Fragment of a chisel(?) of iron, corroded.
188   M. Kasturi Bai

No: 863 Trench: A:TP3 Context: 1112 Period: 3-temple Description: Fragment of a small nail of iron(?) with curved pointed
Object: Chisel Material: Iron. end, corroded.
Description: Fragments of a ribbon-shaped iron chisel(?), corroded
No: 692 Trench: A:TP1 Context: 1040 Period: 2
and encrusted.
Object: Nail Material: Iron.
No: 147 Trench: B Context: 402 Period: 4 Description: Fragment of a nail of iron attached to a pot-sherd, cir-
Object: Chisel Material: Iron. cular section, corroded.
Description: Thin fragments of a chisel(?) of iron.
No: 715 Trench: A:TP1 Context: 1058 Period: 2
No: 205 Trench: B Context: 435 Period: 4 Object: Nail Material: Iron.
Object: Chisel Material: Iron. Description: Fragment of a nail of iron with circular section, cor-
Description: Fragment of a chisel of iron, encrusted. roded.

No: 639 Trench: F Context: 784 Period: 2 No: 716 Trench: A:TP1 Context: 1058 Period: 2
Object: Chisel Material: Iron. Object: Nail Material: Iron.
Description: Ribbon-shaped chisel(?) of iron, slightly bent in the Description: Iron nail with pointed end bent, encrusted.
middle, corroded.
No: 717 Trench: A:TP1 Context: 1063 Period: 2
Object: Nail Material: Iron.
Description: Broken nail of iron, highly corroded.
Borers No: 808 Trench: A:TP3 Context: 1148 Period: 2
Object: Nail Material: Iron.
No: 184 Trench: B Context: 432 Period: 4 Description: Iron nail broken towards the head, corroded.
Object: Borer Material: Iron.
No: 860 Trench: A:TP1 Context: 1099 Period: 2
Description: Broken borer(?) of iron, encrusted.
Object: Nail Material: Iron.
No: 222 Trench: B Context: 435 Period: 4 Description: Iron nail with rectangular section and heavily
Object: Borer Material: Iron. encrusted.
Description: Broken borer(?) of iron, corroded.
No: 7 Trench: A:TP1 Context: 307 Period: 3
No: 223 Trench: B Context: 435 Period: 4 Object: Nail Material: Iron.
Object: Borer Material: Iron. Description: Fragment of a nail of iron, encrusted.
Description: Fragment of a borer(?) of iron, corroded.
No: 82 Trench: A:EEST Context: 348 Period: 2
Object: Nail Material: Iron.
Description: Fragment of a nail of iron with ovalish section,
Nails encrusted.

No: 86 Trench: A:EEST Context: 350 Period: 3

No: 417 Trench: D2 Context: 719 Period: 1
Object: Nail Material: Iron.
Object: Nail Material: Iron.
Description: Fragment of an iron nail broken towards the head.
Description: Iron nail(?) curved in the middle, corroded and
encrusted. Fig. 8.9. No: 297 Trench: A:TP5 Context: 610 Period: 3
Object: Nail Material: Iron.
No: 572 Trench: B Context: 456 Period: 1
Description: Fragment of a nail of iron with cylindrical section.
Object: Nail Material: Iron.
Description: Fragment of a nail of iron, flaked and encrusted. No: 363 Trench: A:W Area Context: 640 Period: 3
Object: Nail Material: Iron.
No: 619 Trench: B Context: 458 Period: 1
Description: Iron nail with a round head, bent in the middle,
Object: Nail Material: Iron.
encrusted.
Description: Broken, thick nail of iron, encrusted.
No: 569 Trench: A:South Area Context: 931 Period: 3
No: 622 Trench: B Context: 465 Period: 1
Object: Nail Material: Iron.
Object: Nail Material: Iron.
Description: Fragment of a nail of iron with squarish section, cor-
Description: Fragment of a nail of iron(?), encrusted.
roded.
No: 627 Trench: B Context: 467 Period: 1
No: 599 Trench: A:TP3 Context: 947 Period: 3
Object: Nail Material: Iron.
Object: Nail Material: Iron.
Description: Fragment of a nail of iron, corroded.
Description: Fragment of a nail of iron, corroded.
No: 657 Trench: B Context: 469 Period: 1
No: 604 Trench: F Context: 778 Period: 3
Object: Nail Material: Iron.
Object: Nail Material: Iron.
Description: Iron nail, broken at pointed end, encrusted.
Description: Two broken nails of iron, corroded.
No: 597 Trench: A:TP1 Context: 944 Period: 2
No: 605 Trench: F Context: 778 Period: 3
Object: Nail Material: Iron.
Object: Nail Material: Iron.
 Chapter 8 The Small Finds and Glass    189

Description: Slightly broken long nail of iron, corroded and Description: Fragment of a nail of iron, bent in the middle,
encrusted. Fig. 8.9. encrusted.

No: 629 Trench: A:TP6 Context: 373 Period: 3 No: 74 Trench: A:TP5 Context: 330 Period: 4
Object: Nail Material: Iron. Object: Nail Material: Iron.
Description: Fragment of a nail of iron, encrusted. Description: Iron nail with square section, encrusted.

No: 646 Trench: A:TP6 Context: 987 Period: 3 No: 78 Trench: A:TP5 Context: 334 Period: 4
Object: Nail Material: Iron. Object: Nail Material: Iron.
Description: Two fragments of nails of iron. Description: Fragment of a nail of iron, bent in the middle, heavily
encrusted.
No: 410 Trench: A:GG North Context: 658 Period: 3-temple
Object: Nail Material: Iron. No: 94 Trench: A:NWNT Context: 376 Period: 3
Description: Fragment of a slender nail of iron, corroded. Object: Nail Material: Iron.
Description: Iron nail with a spear-shaped head and pointed end,
No: 440 Trench: A:TP5 Context: 919 Period: 3-temple
encrusted.
Object: Nail Material: Iron.
Description: Iron nail, broken towards the head. No: 96 Trench: A:NWNT Context: 376 Period: 4
Object: Nail Material: Iron.
No: 442 Trench: D2 Context: 718 Period: 1
Description: Fragment of a nail of iron with a circular section,
Object: Nail Material: Iron.
head broken, corroded. Fig. 8.9.
Description: Two fragments of nails of iron, encrusted.
No: 123 Trench: A:TP9 Context: 519 Period: 4
No: 450 Trench: A:TP5 Context: 667 Period: 3-temple
Object: Nail Material: Iron.
Object: Nail Material: Iron.
Description: Fragment of a slender nail of iron, corroded.
Description: Fragment of a nail of iron, corroded.
No: 148 Trench: B Context: 402 Period: 4
No: 511 Trench: A:TP5 Context: 918 Period: 3-temple
Object: Nail Material: Iron.
Object: Nail Material: Iron.
Description: Broken nail of iron with a bent head, encrusted.
Description: Fragment of a nail of iron with the head missing, cor-
roded. No: 149 Trench: B Context: 402 Period: 4
Object: Nail Material: Iron.
No: 513 Trench: A:TP5 Context: 918 Period: 3-temple
Description: Iron nail with a square head and slightly curved.
Object: Nail Material: Iron.
Description: Fragment of a nail of iron, encrusted. No: 150 Trench: B Context: 402 Period: 4
Object: Nail Material: Iron.
No: 587 Trench: A:TP5 Context: 918 Period: 3-temple
Description: Fragment of a nail of iron, encrusted.
Object: Nail Material: Iron.
Description: Slender and small nail of iron with the pointed end No: 164 Trench: B Context: 414 Period: 4
curved and encrusted. Object: Nail Material: Iron.
Description: A small, broken nail of iron with a square head and
No: 591 Trench: A:GG North Context: 938 Period: 3-temple
bent in the middle, encrusted.
Object: Nail Material: Iron.
Description: Fragment of a nail of iron, encrusted. No: 169 Trench: B Context: 418 Period: 4
Object: Nail Material: Iron.
No: 601 Trench: A:GG North Context: 949 Period: 3-temple
Description: Slightly broken nail of iron with bent working end,
Object: Nail Material: Iron.
corroded and flaky.
Description: Two broken nails of iron with encrustation.
No: 178 Trench: B Context: 430 Period: 4
No: 642 Trench: A:TP5 Context: 978 Period: 3-temple
Object: Nail Material: Iron.
Object: Nail Material: Iron.
Description: Small nail of iron with a square head, corroded.
Description: Fragment of a nail of iron, corroded.
No: 179 Trench: B Context: 430 Period: 4
No: 699 Trench: A:TP6 Context: 1054 Period: 3-temple
Object: Nail Material: Iron.
Object: Nail Material: Iron.
Description: Fragment of a nail of iron, encrusted.
Description: Small nail of iron with encrustation.
No: 180 Trench: B Context: 430 Period: 4
No: 831 Trench: A:TP5, TP9, GG North Context: 1186 Period: 3-temple
Object: Nail Material: Iron.
Object: Nail Material: Iron.
Description: Slender nail of iron with a bent head, encrusted.
Description: Fragment of a nail of iron, corroded.
No: 181 Trench: B Context: 430 Period: 4
No: 27 Trench: A:W Area Context: 321 Period: 4
Object: Nail Material: Iron.
Object: Nail Material: Iron.
Description: Fragment of a nail of iron, corroded.
Description: Slightly broken nail of iron.
No: 198 Trench: B Context: 435 Period: 4
No: 54 Trench: A:S of S Temple Context: 303 Period: 4
Object: Nail Material: Iron.
Object: Nail Material: Iron.
Description: Iron nail with a round head, flaky and encrusted.
190   M. Kasturi Bai

No: 199 Trench: B Context: 435 Period: 4 Description: Iron nail with a thick circular head, corroded.
Object: Nail Material: Iron.
No: 246 Trench: A:S Wall of N Context: 573 Period: 4
Description: Long nail of iron with a round head, flaky and
Object: Nail Material: Iron.
encrusted.
Description: Iron nail with a squarish section, corroded.
No: 200 Trench: B Context: 435 Period: 4
No: 349 Trench: A:South Area, TP6 Context: 635 Period: 4
Object: Nail Material: Iron.
Object: Nail Material: Iron.
Description: Long and slender nail of iron with the head missing.
Description: Fragment of a slender nail of iron.
No: 202 Trench: B Context: 435 Period: 4
No: 358 Trench: A:South Area, TP6 Context: 635 Period: 4
Object: Nail Material: Iron.
Object: Nail Material: Iron.
Description: Iron nail with a squarish head, curved and slightly
Description: Iron nail with a rectangular section, slightly bent
broken pointed end.
head and broken pointed end.
No: 206 Trench: B Context: 435 Period: 4
No: 426 Trench: A:TP6 Context: 649 Period: 4
Object: Nail Material: Iron.
Object: Nail Material: Iron.
Description: Fragment of a nail of iron, heavily corroded and flaky.
Description: Broken nail of iron with a round top, corroded and
No: 207 Trench: B Context: 435 Period: 4 encrusted.
Object: Nail Material: Iron.
No: 428 Trench: A:W Area Context: 659 Period: 4
Description: Fragment of a nail of iron, encrusted.
Object: Nail Material: Iron.
No: 211 Trench: B Context: 435 Period: 4 Description: Fragment of a nail of iron.
Object: Nail Material: Iron.
No: 431 Trench: A:TP5, South Area Context: 663 Period: 4
Description: Small nail of iron with encrustation.
Object: Nail Material: Iron.
No: 213 Trench: B Context: 435 Period: 4 Description: Fragment of a thin and small nail of iron, corroded.
Object: Nail Material: Iron.
No: 445 Trench: A:South Area, TP6 Context: 635 Period: 4
Description: Slightly broken, small nail of iron with a rectangular
Object: Nail Material: Iron.
head, encrusted.
Description: Small nail of iron with slightly bent rear head, cor-
No: 214 Trench: B Context: 435 Period: 4 roded.
Object: Nail Material: Iron.
No: 526 Trench: A:TP10 Context: 928 Period: 4
Description: Small nail of iron with a square head, encrusted.
Object: Nail Material: Iron.
No: 215 Trench: B Context: 435 Period: 4 Description: Iron nail broken towards the head.
Object: Nail Material: Iron.
No: 544 Trench: A:TP6 Context: 689 Period: 4
Description: Broken nail of iron, flaky and encrusted.
Object: Nail Material: Iron.
No: 216 Trench: B Context: 435 Period: 4 Description: Iron nail with a slightly broken head, corroded.
Object: Nail Material: Iron.
No: 546 Trench: A:TP6 Context: 689 Period: 4
Description: Small and slender nail of iron with a square head,
Object: Nail Material: Iron.
encrusted.
Description: Fragment of a small nail of iron, corroded and
No: 217 Trench: B Context: 435 Period: 4 encrusted.
Object: Nail Material: Iron.
No: 31 Trench: A:NWNT Context: 327 Period: 4?
Description: Small nail of iron with a projecting square head,
Object: Nail Material: Iron.
encrusted.
Description: Iron nail broken at the head.
No: 218 Trench: B Context: 435 Period: 4
No: 65 Trench: A:NWNT Context: 331 Period: 4?
Object: Nail Material: Iron.
Object: Nail Material: Iron.
Description: Iron nail with a hook-shaped bent head, encrusted.
Description: Iron nail broken at the distal end, encrusted.
No: 219 Trench: B Context: 435 Period: 4
No: 115 Trench: A:NWNT Context: 505 Period: 4?
Object: Nail Material: Iron.
Object: Nail Material: Iron.
Description: Slightly broken and slender nail of iron.
Description: Fragment of a nail of iron, bent in the middle,
No: 220 Trench: B Context: 435 Period: 4 encrusted.
Object: Nail Material: Iron.
No: 117 Trench: A:NWNT Context: 507 Period: 4?
Description: Fragment of a nail of iron, encrusted.
Object: Nail Material: Iron.
No: 221 Trench: B Context: 435 Period: 4 Description: Fragment of a nail of iron, corroded and crackled.
Object: Nail Material: Iron.
No: 125 Trench: A:NWNT Context: 529 Period: 4?
Description: Fragment of a nail of iron, flaky and encrusted.
Object: Nail Material: Iron.
No: 242 Trench: A:South Area Context: 568 Period: 4 Description: Fragment of a small nail of iron, encrusted.
Object: Nail Material: Iron.
 Chapter 8 The Small Finds and Glass    191

No: 128 Trench: A:NWNT Context: 535 Period: 4? Fasteners

Object: Nail Material: Iron.
Description: Two fragments of nails of iron, encrusted. No: 340 Trench: D2 Context: 711 Period: 2
No: 19 Trench: A:TP2 Context: 315 Period: No period Object: Fastener Material: Iron.
Object: Nail Material: Iron. Description: Small ribbon-shaped fastener of iron, encrusted. Fig.
Description: Fragment of a nail of iron, encrusted. 8.9.

No: 22 Trench: A:TP2 Context: 317 Period: No period No: 279 Trench: A:TP6 Context: 596 Period: 4
Object: Nail Material: Iron. Object: Fastener Material: Iron.
Description: Iron nail broken at the distal end, corroded. Description: Fragment of a ribbon-shaped fastener of iron,
encrusted.
No: 40 Trench: A:TP2 Context: 323 Period: No period
Object: Nail Material: Iron. No: 545 Trench: A:TP6 Context: 689 Period: 4
Description: Broken nail of iron, encrusted. Object: Fastener Material: Iron.
Description: Ribbon-shaped fastener(?) of iron, one end slightly
No: 80 Trench: A:TP6 Context: 343 Period: No period narrow, other end broad and slightly curved, corroded. Fig. 8.9.
Object: Nail Material: Iron.
Description: Fragment of a nail of iron, encrusted.

No: 710 Trench: E Context: 816 Period: No period

Object: Nail Material: Iron. Rings
Description: Iron nail with pointed end curved.
No: 286 Trench: D1 Context: 754 Period: 1
No: 770 Trench: E Context: 816 Period: No period
Object: Ring Material: Iron.
Object: Nail Material: Iron.
Description: Broken ring of iron with a rectangular section. Fig. 8.9.
Description: Iron nail curved at the broader end, corroded.
No: 473 Trench: D1 Context: 767 Period: 1
No: 835 Trench: E Context: 1208 Period: No period
Object: Ring Material: Iron.
Object: Nail Material: Iron.
Description: Broken iron ring, encrusted. Fig. 8.9.
Description: Slightly curved nail of iron, broken at the pointed end.
No: 559 Trench: A:TP1 Context: 912 Period: 3
Object: Ring Material: Iron.
Description: Broken, small ring of iron, corroded.
Rivets No: 592 Trench: A:TP3 Context: 942 Period: 3
Object: Ring Material: Iron.
No: 8 Trench: A:TP1 Context: 307 Period: 3 Description: Fragment of a thin ring(?) of iron. Fig. 8.9.
Object: Rivet Material: Iron.
Description: Fragment of a ribbon-shaped rivet of iron, encrusted. No: 512 Trench: A:TP5 Context: 918 Period: 3-temple
Object: Ring Material: Iron.
No: 249 Trench: A:South Area Context: 578 Period: 3 Description: Broken, small ring of iron, encrusted. Fig. 8.9.
Object: Rivet Material: Iron.
Description: Door rivet(?) of iron, flaky and encrusted. Fig. 8.9. No: 36 Trench: A:TP9 Context: 320 Period: 4
Object: Ring Material: Iron.
No: 254 Trench: A:South Area Context: 589 Period: 3 Description: A small broken iron ring, corroded. Fig. 8.9.
Object: Rivet Material: Iron.
Description: Slender rivet of iron, encrusted. Fig. 8.9. No: 174 Trench: B Context: 429 Period: 4
Object: Ring Material: Iron.
No: 269 Trench: D1 Context: 751 Period: 3 Description: Broken ring of iron, encrusted.
Object: Rivet Material: Iron.
Description: Door rivet(?) of iron. No: 209 Trench: B Context: 435 Period: 4
Object: Ring Material: Iron.
No: 272 Trench: A:South Area Context: 591 Period: 3 Description: A small iron ring, corroded and encrusted.
Object: Rivet Material: Iron.
Description: Slightly curved small iron rivet with a bent project- No: 354 Trench: A:South Area, TP6 Context: 635 Period: 4
ing squarish head and pointed end. Object: Ring Material: Iron.
Description: Fragment of a thin iron ring with rectangular section,
No: 95 Trench: A:NWNT Context: 376 Period: 4 corroded.
Object: Rivet Material: Iron.
Description: Iron rivet with squarish section, one end curved, cor- No: 529 Trench: E Context: 816 Period: No period
roded and flaky. Fig. 8.9. Object: Ring Material: Iron.
Description: Broken ring of iron, encrusted. Fig. 8.9.
No: 176 Trench: B Context: 429 Period: 4
Object: Rivet Material: Iron.
Description: Slender iron rivet(?) with a circular section, encrusted.
192   M. Kasturi Bai

Table 8.8: Summary of copper objects by type of object and period.

Type 1 2 3 3-temple 4 4? No period Total

Nail 1 1 1 1 4
Rivet 1 1
Fastener 1 1
Needle 1 1
Rod 2 2
Bowl 1 1
Pellate 1 1
Plate 1 1 2
Handle 1 1
Wire 1 1
Unidentified 1 1 3 1 1 7
Total 4 1 1 6 7 1 2 22

Fish hooks
No: 77 Trench: A:NWNT Context: 332 Period: 4?
Object: Rod Material: Iron.
No: 201 Trench: B Context: 435 Period: 4
Description: Broken and slender cylindrical iron rod.
Object: Fish hook Material: Iron.
Description: Fish hook(?) of iron, encrusted. No: 21 Trench: A:TP2 Context: 317 Period: No period
Object: Rod Material: Iron.
No: 212 Trench: B Context: 435 Period: 4
Description: Fragment of a slender iron rod.
Object: Fish hook Material: Iron.
Description: Fish hook(?) of iron, encrusted.

Spike

Sickles No: 374 Trench: A:W Area Context: 643 Period: 4

Object: Spike Material: Iron.
No: 244 Trench: A:TP5 Context: 570 Period: 4 Description: Small spike(?) of iron with squarish section.
Object: Sickle Material: Iron.
Description: Small iron sickle with slightly broken handle and
working end, corroded and flaky. Fig. 8.9.

No: 87 Trench: A:S Temple Context: 353 Period: No period Clamp

Object: Sickle Material: Iron.
Description: Fragment of a small iron sickle(?), corroded. No: 233 Trench: B Context: 440 Period: 4
Object: Clamp Material: Iron.
Description: Broken iron clamp, fixed in a broad rectangular plate,
corroded.
Spatula

No: 20 Trench: A:TP2 Context: 315 Period: No period

Object: Spatula Material: Iron. Latch
Description: Fragment of an iron spatula(?), corroded.
No: 427 Trench: A:TP6 Context: 649 Period: 4
Object: Latch Material: Iron.
Description: Head portion of a broken latch(?) of iron.
Rods

No: 485 Trench: A:South Area Context: 645 Period: 4

Object: Rod Material: Iron.
Description: Fragment of a thin and slender rod(?) of iron,
encrusted.
 Chapter 8 The Small Finds and Glass    193

Axle Description: Unidentified perforated disc-like object of iron with

slight projection, corroded.
No: 606 Trench: F Context: 778 Period: 3 No: 667 Trench: A:TP9 Context: 1010 Period: 3-temple
Object: Axle Material: Iron. Object: Unidentified Material: Iron.
Description: Iron axle(?), corroded and encrusted. Fig. 8.9. Description: Fragments of unidentified objects of iron, encrusted.

No: 789 Trench: A:TP9 Context: 1121 Period: 3-temple

Object: Unidentified Material: Iron.
Wire Description: Small and elongated oval-shaped unidentified object
of iron, corroded.
No: 697 Trench: A:TP9 Context: 1050 Period: 3-temple No: 859 Trench: A:TP3 Context: 1095 Period: 3-temple
Object: Wire Material: Iron. Object: Unidentified Material: Iron.
Description: Two fragments of iron wire(?). Description: Two slender, ribbon-shaped unidentified objects of
iron.
No: 30 Trench: A:NWNT, TP9 Context: 326 Period: 4
Object: Wire Material: Iron. No: 446 Trench: A:South Area, TP6 Context: 635 Period: 4
Description: Fragment of a thin iron wire. Object: Unidentified Material: Iron.
Description: Thick, unidentified squarish iron object, crackled and
No: 295 Trench: A:South Area Context: 605 Period: 4
encrusted.
Object: Wire Material: Iron.
Description: Two fragments of iron wire, encrusted. No: 91 Trench: A:S Temple Context: 366 Period: No period
Object: Unidentified Material: Iron.
No: 114 Trench: A:NWNT Context: 504 Period: 4?
Description: Thin, ribbon-shaped unidentified object of iron.
Object: Wire Material: Iron.
Description: Fragment of a cylindrical iron wire.

Lead objects
Buckle
A total of four lead objects came to light during the exca-
No: 229 Trench: B Context: 439 Period: 4 vations.
Object: Buckle Material: Iron.
Description: Iron buckle, corroded. No: 158 Trench: B Context: 410 Period: 4
Object: Ball Material: Lead.
Description: Small spherical ball of lead.

No: 522 Trench: A:W Area Context: 920 Period: 4

Unidentified Object: String Material: Lead.
Description: Small curved string of lead(?).
No: 441 Trench: D2 Context: 718 Period: 1
No: 318 Trench: A:TP6, South Area Context: 613 Period: 4
Object: Unidentified Material: Iron.
Object: Band Material: Lead.
Description: Big and irregular-shaped unidentified object of iron
Description: Small band of lead.
with a lot of encrustation.
No: 839 Trench: A:TP7 Context: 1221 Period: No period
No: 682 Trench: B Context: 475 Period: 1
Object: Unidentified Material: Lead.
Object: Unidentified Material: Iron.
Description: Fragment of a small metal piece of lead with some
Description: Fragment of an unidentified object of iron, corroded.
indistinct design.
No: 650 Trench: A:TP1 Context: 988 Period: 2
Object: Unidentified Material: Iron.
Description: Fragment of an unidentified object of iron, encrusted.
Silver objects
No: 252 Trench: A:TP6 Context: 370 Period: 3
Object: Unidentified Material: Iron.
Description: Flat, ribbon-shaped unidentified object of iron, forked
Only a single silver object came to light during the exca-
at one end, broken on other end. vations.
No: 637 Trench: A:TP3 Context: 974 Period: 3 No: 430 Trench: A:TP5, South Area Context: 663 Period: 4
Object: Unidentified Material: Iron. Object: Decorative object Material: Silver.
Description: Big, unidentified iron object of irregular shape with Description: A shallow, eye-shaped object of silver, perhaps used as
concavity on one side, encrusted. decoration to an ‘eye’ of a stone idol with a religious purpose.

No: 488 Trench: A:TP5 Context: 700 Period: 3-temple

Object: Unidentified Material: Iron.
194   M. Kasturi Bai

Other materials Description: Vertically broken tubular object of bone, perhaps a

handle.
A total of 67 household objects came to light during the
excavations. Of these, nine objects were made of bone,
four of ivory, two of glass and one of porcelain. There Ivory objects
were also 26 objects made of stone and 25 of terracotta.
There are four objects made of ivory, which can be cat-
egorised as a perforated disc, a handle of a knife and
Bone objects two lids of a container from Periods 4?, 3-temple and 4
respectively.
There were nine objects made of bone. Six were simple No: 370 Trench: A:TP5 Context: 642 Period: 3-temple
points, two were double-ended points and one was Object: Handle Material: Ivory.
the broken handle of a knife (Fig. 8.8). Of the simple Description: Slightly broken and vertically sleeved tubular
points, four were recovered from Period 1, one from handle(?) of a knife of ivory with excellent incised decoration on
surface, channel slightly broad at one end and slightly broken. Fig.
Period 4 and one from a layer of unattributable period.
8.8.
Among the double-ended points, one was found from
Period 1 and one from Period 3. No: 448 Trench: A:South Area, TP6 Context: 635 Period: 4
Object: Lid Material: Ivory.
No: 365 Trench: D2 Context: 715 Period: 1 Description: Small lid of ivory(?) with a concave base and plano-
Object: Point Material: Bone. convex top with smeared end, decorated by grooves. Fig. 8.10.
Description: Bone point, broken into pieces.
No: 727 Trench: B Context: 435 Period: 4
No: 383 Trench: D1 Context: 766 Period: 1 Object: Lid Material: Ivory.
Object: Point Material: Bone. Description: Lid of ivory with a flat and circular base, and a project-
Description: Bone point, broken towards the head, ground and ing round knob decorated with incised circles on the top portion.
slightly polished. Fig. 8.8. Fig. 8.10.
No: 384 Trench: D1 Context: 766 Period: 1 No: 126 Trench: A:NWNT Context: 532 Period: 4?
Object: Point Material: Bone. Object: Perforated disc Material: Ivory.
Description: Bone point, broken on either end, pecked and ground. Description: Perforated disc of ivory, slightly broken, spindle
No: 418 Trench: D1 Context: 767 Period: 1 whorl(?). Fig. 8.10.
Object: Point Material: Bone.
Description: Bone point, broken towards head, pecked and ground.
Fig. 8.8.
Glass objects
No: 444 Trench: A:South Area, TP6 Context: 635 Period: 4
Object: Point Material: Bone. There are two household objects made of glass, which are
Description: Bone point, broken at the head, ground and polished
rim fragments of bowls, one from Period 2 and one from a
with lime encrustation. Fig. 8.8.
layer of unattributable period.
No: 784 Trench: E Context: 844 Period: No period
Object: Point Material: Bone. No: 677 Trench: F Context: 788 Period: 2
Description: Fragment of a bone point with broader end missing. Object: Bowl Material: Glass.
Description: Rim fragments of bowls of dark blue translucent glass.
No: 382 Trench: D1 Context: 766 Period: 1
Object: Double-ended point Material: Bone. No: 85 Trench: A:TP6 Context: 343 Period: No period
Description: Double-ended point of bone, pecked and ground. Fig. Object: Bowl Material: Glass.
8.8. Description: Fragment of a bowl with tubular rim of pale grey glass.

No: 603 Trench: F Context: 778 Period: 3

Object: Double-ended point Material: Bone.
Description: Double-ended point of bone, slightly broken at one Porcelain objects
end, pecked and ground. Fig. 8.8.
No: 394 Trench: E Context: 801 Period: No period
Ref: Similar specimens were obtained from Satavahana layers at
Object: Base Material: Porcelain.
Nasik (Sankalia and Deo 1955: fig. 56, a, b and I) and also obtained
Description: Base of a goblet(?) of porcelain.
from Period V at Nevasa (Sankalia et al. 1960: fig. 194, 1).

No: 817 Trench: E Context: 1206 Period: No period

Object: Handle Material: Bone.
 Chapter 8 The Small Finds and Glass    195

Fig. 8.10: Bone, kaolin and porcelain objects.

196   M. Kasturi Bai

Table 8.9: Summary of iron objects by type of object and period.

Type 1 2 3 3-temple 4 4? No period Total

Arrow head 3 2 5
Spear 1 1
Point(tanged) 1 1
Knife 3 3 3 2 4 15
Handle 3 3
Chisel 1 1 1 3 6
Borer 3 3
Nail 7 7 11 11 43 6 7 92
Rivet 5 2 7
Fastener 1 2 3
Ring 2 2 1 4 1 10
Fish hook 2 2
Sickle 1 1 2
Spatula 1 1
Rod 1 1 1 3
Spike 1 1
Clamp 1 1
Latch 1 1
Axle 1 1
Wire 1 2 1 4
Buckle 1 1
Unidentified 2 1 2 4 1 1 11
Total 15 13 26 23 77 8 12 174

No: 476 Trench: E Context: 809 Period: No period

Stone objects Object: Ball Material: Stone.
Description: Big, ovalish ball of stone with a roughly plano-convex
There are nine bowls in the above category, one each section.
from Period 1 and Period 3 and seven from Period  4.
One small vase of stone was recovered from Period 1.
Of the three querns, one each was found in Period  1,
Bowls
Period 2 and Period 4. The other household objects
of stone are four grinding stones, three mullers, one No: 411 Trench: D2 Context: 718 Period: 1
pestle, one lamp and a weight. Object: Bowl Material: Stone.
Description: Fragment of a bowl of soap stone.

No: 854 Trench: A:TP3 Context: 1021 Period: 3

Balls Object: Bowl Material: Stone.
Description: Fragment of a convex sided big bowl of sand stone.
No: 474 Trench: A:TP5 Context: 642 Period: 3-temple Fig. 8.11.
Object: Ball Material: Stone.
No: 68 Trench: A:TP1 Context: 301 Period: 4
Description: Big, ovalish ball of stone with a polished surface.
Object: Bowl Material: Stone.
No: 475 Trench: E Context: 809 Period: No period Description: Fragment of a bowl of soap stone.
Object: Ball Material: Stone.
No: 185 Trench: B Context: 435 Period: 4
Description: Half-broken, big spherical ball of basalt(?).
Object: Bowl Material: Stone.
Description: Fragment of a bowl of lime stone.
 Chapter 8 The Small Finds and Glass    197

Fig. 8.11: Stone objects.

198   M. Kasturi Bai

No: 294 Trench: A:South Area Context: 605 Period: 4 No: 853 Trench: A:TP1 Context: 1013 Period: 2
Object: Bowl Material: Stone. Object: Grindstone Material: Stone.
Description: Fragment of a deep bowl of trap(?) stone with feature- Description: Slightly broken spherical grinding stone of sand-
less rim. stone(?).

No: 301 Trench: A:TP6, South Area Context: 613 Period: 4

Object: Bowl Material: Stone.
Description: Fragment of a deep bowl of trap stone (this is part of
object no. 294). Fig. 8.11.
Mullers

No: 469 Trench: A:South Area Context: 568 Period: 4 No: 282 Trench: A:South Area Context: 578 Period: 3
Object: Bowl Material: Stone. Object: Muller Material: Stone.
Description: Fragment of a bowl of soap stone. Description: Broken cylindrical muller of red sandstone.
No: 470 Trench: A:South Area Context: 568 Period: 4 No: 153 Trench: B Context: 406 Period: 4
Object: Bowl Material: Stone. Object: Muller Material: Stone.
Description: Fragment of a thin bowl of soap stone with horizontal Description: Disc-shaped muller of sandstone, slightly broken.
lines on body.
No: 468 Trench: A:South Area Context: 568 Period: 4
No: 25 Trench: A:TP9 Context: 320 Period: 4 Object: Muller Material: Stone.
Object: Bowl Material: Stone. Description: Broken cylindrical muller of basalt(?).
Description: Fragment of a small bowl of soap stone(?) with pro-
jecting flat base.

Pestle

Vase No: 186 Trench: B Context: 435 Period: 4

Object: Pestle Material: Stone.
No: 421 Trench: D1 Context: 767 Period: 1 Description: Rectangular pestle of dolerite(?) with lenticular
Object: Vase Material: Stone. section.
Description: Fragment of a small vase of soap stone. Fig. 8.11.

Lamp
Querns
No: 146 Trench: B Context: 402 Period: 4
No: 385 Trench: D2 Context: 718 Period: 1 Object: Lamp Material: Stone.
Object: Quern Material: Stone. Description: Thick and big lamp made of trap stone with pecked
Description: Leg of a quern of basalt(?) stone with iron rivet. marks.

No: 306 Trench: D2 Context: 711 Period: 2

Object: Quern Material: Stone.
Description: Fragment of circular legged quern of trap(?) stone. Weight
Fig. 8.11.

No: 302 Trench: A:TP6, South Area Context: 613 Period: 4 No: 93 Trench: A:NWNT Context: 376 Period: 2
Object: Rotary quern Material: Stone. Object: Weight(?) Material: Stone.
Description: Fragment of bottom part of a rotary quern of stone. Description: Squarish cube-shaped weight(?) of stone.

Grindstones Terracotta objects

No: 32 Trench: A:NWNT, TP9 Context: 328 Period: 4 Among the terracotta objects, there are two lamps,
Object: Grindstone Material: Stone. three vases, one handle of a vase, one bowl, two lids,
Description: Grinding stone of sandstone in pecked condition.
three stoppers, four dabbers, three skin rubbers, one
No: 162 Trench: B Context: 414 Period: 4 cigar pipe, one possible weight, three balls and one cru-
Object: Grindstone Material: Stone.
cible. A detailed description is given below.
Description: Ovalish grinding stone of trap stone, ground.

No: 850 Trench: A:TP5 Context: 1000 Period: 3-temple

Object: Grindstone Material: Stone.
Description: Fragment of a grinding roller(?) of stone.
 Chapter 8 The Small Finds and Glass    199

Lamps No: 303 Trench: D2 Context: 710 Period: 2

Object: Stopper Material: Terracotta.
No: 142 Trench: B Context: 402 Period: 4 Description: Cylindrical-shaped stopper(?) of terracotta, slightly
Object: Lamp Material: Terracotta. crackled on one end. Fig. 8.7.
Description: Slightly broken lamp of terracotta with a handle No: 714 Trench: E Context: 832 Period: No period
having a flat base and a hollow stand. Fig. 8.12. Object: Stopper Material: Terracotta.
No: 145 Trench: B Context: 402 Period: 4 Description: Small cylindrical stopper(?) of terracotta with a circu-
Object: Lamp Material: Terracotta. lar section.
Description: Terracotta lamp with a handle having a concave base
with hollow stand.

Dabbers

Vases No: 463 Trench: D2 Context: 722 Period: 1

Object: Dabber Material: Terracotta.
Description: Broken dabber(?) of terracotta with a convex base,
No: 742 Trench: A:TP9 Context: 1096 Period: 3-temple
upper end broken. Fig. 8.7.
Object: Vase Material: Pottery.
Description: Decorated pinkish redware sherd of a vase (religious). No: 841 Trench: A:TP5 Context: 918 Period: 3-temple
Object: Dabber Material: Terracotta.
No: 581 Trench: F Context: 778 Period: 3
Description: Thick, plano-convex-shaped dabber(?) of terracotta in
Object: Vase Material: Pottery.
a brick-red colour, smooth at the bottom.
Description: A small vase of red ware with a projecting disc base
and slightly broken at the rim. Fig. 8.12. No: 283 Trench: A:S Wall of N Context: 575 Period: 4
Object: Dabber Material: Terracotta.
No: 583 Trench: F Context: 778 Period: 3
Description: Ovalish dabber(?) of terracotta with a porous surface
Object: Vase Material: Pottery.
and a plano-convex section, black in colour.
Description: A very small, toy vase of red ware with irregular pro-
jecting base. Fig. 8.12. No: 785 Trench: E Context: 844 Period: No period
Object: Dabber Material: Terracotta.
Description: Big and thick dabber(?) of terracotta with sand grains
and vegetable grits.
Handles

No: 235 Trench: B Context: 440 Period: 4

Object: Handle Material: Pottery. Skin rubbers
Description: Handle of a red ware pot in the shape of elephant figu-
rine (religious purpose). Fig. 8.13. No: 746 Trench: A:TP1 Context: 1099 Period: 2
Object: Skin rubber Material: Terracotta.
No: 582 Trench: F Context: 778 Period: 3
Description: Rectangular skin rubber of terracotta with a herring
Object: Bowl Material: Pottery.
bone pattern on either face and on one side a slight depression in
Description: Small, toy bowl of dull red ware. Fig. 8.12.
the centre for grip. Fig. 8.7.
No: 777 Trench: F Context: 851 Period: 1
No: 787 Trench: A:TP1 Context: 1120 Period: 2
Object: Lid Material: Terracotta.
Object: Skin rubber Material: Terracotta.
Description: Very small toy lid(?) of terracotta, slightly broken at
Description: Fragment of a skin rubber of terracotta with incised
the top projecting end. Fig. 8.12.
decoration on all three sides.
No: 798 Trench: F Context: 851 Period: 1
No: 560 Trench: F Context: 778 Period: 3
Object: Lid Material: Pottery.
Object: Skin rubber Material: Terracotta.
Description: Knob of a lid(?) of polished black ware. Fig. 8.12.
Description: Fragment of a skin rubber of terracotta with a series of
crescentic decoration on all sides. Fig. 8.7.

Stoppers
Weight
No: 774 Trench: F Context: 851 Period: 1
Object: Stopper Material: Terracotta. No: 718 Trench: A:TP1 Context: 1063 Period: 2
Description: Cylinder-shaped stopper of terracotta with a circular Object: Weight(?) Material: Terracotta.
section, slightly concave in the middle of the body. Fig. 8.7. Description: Two cylindrical-shaped weights(?) of terracotta with a
circular section and a smooth surface, slightly tapering to one end
and flat on the other end.
200   M. Kasturi Bai

Fig. 8.12: Small jars, crucibles and lamps.

 Chapter 8 The Small Finds and Glass    201

Balls No: 50 Trench: A:S of S Temple Context: 303 Period: 4

Object: Ball Material: Stone.
No: 66 Trench: A:NWNT Context: 331 Period: 4? Description: Small stone ball with pecked marks and not ground.
Object: Ball Material: Terracotta. No: 236 Trench: B Context: 440 Period: 4
Description: Small terracotta ball, perhaps used as a weight. Object: Ball Material: Terracotta.
No: 71 Trench: A:TP1 Context: 306 Period: 3-temple Description: Half-broken spherical ball of terracotta with irregu-
Object: Ball Material: Terracotta. lar lines on broken part.
Description: Very small terracotta ball, perhaps used as weight. No: 107 Trench: A:NWNT Context: 386 Period: 4?
No: 67 Trench: A:NWNT Context: 331 Period: 4? Object: Ball Material: Terracotta.
Object: Ball Material: Terracotta. Description: Terracotta ball of brick-red colour having a slightly
Description: Very small terracotta ball, perhaps used as weight. porous surface.

No: 663 Trench: E Context: 829 Period: No period

Object: Ball Material: Terracotta.
Crucible Description: Small, spherical ball of terracotta.

No: 688 Trench: E Context: 831 Period: No period

No: 830 Trench: B Context: 441 Period: 2 Object: Ball Material: Stone.
Object: Crucible Material: Terracotta. Description: Small, spherical ball of stone.
Description: Broken round based crucible of terracotta. Fig. 8.12.

Discs
Cigar pipe
No: 413 Trench: D2 Context: 718 Period: 1
Object: Disc Material: Pot-sherd.
No: 388 Trench: E Context: 801 Period: No period
Description: Small, edge-ground disc made of dull red ware sherd.
Object: Cigar pipe Material: Terracotta.
Description: Two fragments of cigar pipe of terracotta with occa- No: 573 Trench: B Context: 458 Period: 1
sional shining gold paint. Object: Disc Material: Pot-sherd.
Description: Edge-ground, small disc of Black and Red ware.

No: 757 Trench: F Context: 851 Period: 1

Object: Disc Material: Pot-sherd.
Play objects Description: Small, edge-ground disc of red ware sherd. Fig. 8.7.

No: 758 Trench: F Context: 851 Period: 1

A total of 98 play objects came to light during the Object: Disc Material: Pot-sherd.
excavations (Table 8.10). The number of play objects Description: Small, edge-ground disc of buff ware sherd, slightly
increased from Period 1 to Period 4. They are mainly broken.
related to children’s play; some of the perforated discs No: 760 Trench: F Context: 851 Period: 1
might have been used as toy wheels and also some as Object: Disc Material: Pot-sherd.
spindle whorls. Except for a few, all the discs, hop- Description: Small, edge-ground disc of buff ware sherd.
scotch and toy wheels are made of broken potsherds. No: 761 Trench: F Context: 851 Period: 1
The toy wheels, perforated discs and hopscotch are Object: Disc Material: Pot-sherd.
edge-ground. Description: Slightly damaged, edge-ground disc of red ware sherd.

No: 763 Trench: F Context: 851 Period: 1

Object: Disc Material: Pot-sherd.
Balls Description: Two edge-ground and broken discs of dull red ware
sherds of different sizes.
No: 624 Trench: B Context: 467 Period: 1
No: 305 Trench: D2 Context: 711 Period: 2
Object: Ball Material: Terracotta.
Object: Disc Material: Pot-sherd.
Description: Small, faceted ball of terracotta.
Description: Edge-ground disc of red ware sherd, slightly damaged.
No: 84 Trench: A:EEST Context: 348 Period: 3 Fig. 8.7.
Object: Ball Material: Terracotta.
No: 616 Trench: A:TP1 Context: 966 Period: 2
Description: Small ovalish ball of terracotta.
Object: Disc Material: Pot-sherd.
No: 576 Trench: A:TP6 Context: 342 Period: 3-temple Description: Fragment of an edge-ground thick disc of red ware
Object: Ball Material: Terracotta. sherd.
Description: Spheroid ball of terracotta, faceted.
202   M. Kasturi Bai

Table 8.10: Summary of play objects by type of object and period.

1 2 3 3-temple 4 4? No period Total

Ball 1 1 1 2 1 2 8
Disc 7 10 10 5 5 1 8 46
Perforated disc 3 2 1 4 10
Hopscotch 2 4 1 2 4 1 14
Marble 1 5 1 1 2 10
Toy 1 1 2
Toy Wheel 2 1 1 1 1 6
Dice 1 1
Pan 1 1
Total 16 18 18 10 19 3 14 98

No: 643 Trench: A:TP1 Context: 986 Period: 2 No: 672 Trench: A:TP3 Context: 1012 Period: 3
Object: Disc Material: Pot-sherd. Object: Disc Material: Pot-sherd.
Description: Big and thick edge-ground disc of red ware sherd. Description: Broken, edge-ground disc of black ware sherd.

No: 644 Trench: A:TP1 Context: 986 Period: 2 No: 673 Trench: A:TP3 Context: 1012 Period: 3
Object: Disc Material: Pot-sherd. Object: Disc Material: Pot-sherd.
Description: Small, edge-ground, ovalish disc of dull red ware Description: Fragment of an edge-ground disc of dull buff ware
sherd. sherd.

No: 679 Trench: F Context: 788 Period: 2 No: 790 Trench: A:TP5, TP3 Context: 1124 Period: 3
Object: Disc Material: Pot-sherd. Object: Disc Material: Pot-sherd.
Description: Slightly broken, edge-ground disc of red ware sherd. Description: Slightly broken and edge-ground disc of red ware
sherd.
No: 705 Trench: F Context: 798 Period: 2
Object: Disc Material: Pot-sherd. No: 803 Trench: A:TP3 Context: 1140 Period: 3
Description: Edge-ground, ovalish disc of red ware sherd. Object: Disc Material: Pot-sherd.
Description: Fragment of a big, edge-ground disc of dull red ware
No: 721 Trench: F Context: 846 Period: 2
sherd.
Object: Disc Material: Pot-sherd.
Description: Edge-ground, ovalish disc of dull red ware sherd with No: 804 Trench: A:TP3 Context: 1140 Period: 3
a squarish section. Object: Disc Material: Pot-sherd.
Description: Fragment of an edge-ground disc of dull red ware.
No: 739 Trench: A:TP1 Context: 1063 Period: 2
Object: Disc Material: Pot-sherd. No: 842 Trench: A:TP1 Context: 921 Period: 3
Description: Slightly damaged disc of red ware sherd, partly Object: Disc Material: Pot-sherd.
ground. Description: Thick and ovalish disc of burnished red ware sherd.

No: 851 Trench: A:TP1 Context: 1009 Period: 2 No: 865 Trench: A:TP3 Context: 1129 Period: 3
Object: Disc Material: Pot-sherd. Object: Disc Material: Pot-sherd.
Description: Edge-ground disc of red ware sherd. Description: Big and thick disc of dark brownish red ware sherd,
well burnt showing shell grains and vegetable temper.
No: 263 Trench: D1 Context: 751 Period: 3
Object: Disc Material: Pot-sherd. No: 543 Trench: A:TP9, GG North Context: 927 Period: 3-temple
Description: Edge-ground disc of red ware sherd. Object: Disc Material: Pot-sherd.
Description: Two edge-ground, broken discs of red ware sherds.
No: 467 Trench: A:South Area Context: 668 Period: 3
Object: Disc Material: Pot-sherd. No: 578 Trench: A:TP6 Context: 342 Period: 3-temple
Description: Edge-ground, broken disc of dull red ware sherd. Object: Disc Material: Pot-sherd.
Description: Fragments of an edge-ground disc of red ware sherd.
No: 671 Trench: A:TP3 Context: 1012 Period: 3
Object: Disc Material: Pot-sherd. No: 846 Trench: A:TP3 Context: 981 Period: 3-temple
Description: Broken, edge-ground disc of burnished red ware Object: Disc Material: Pot-sherd.
sherd. Description: Edge-ground, ovalish disc of dull red ware sherd.
 Chapter 8 The Small Finds and Glass    203

No: 847 Trench: A:TP3 Context: 981 Period: 3-temple Description: Thick, edge-ground disc of terracotta.
Object: Disc Material: Pot-sherd.
No: 706 Trench: F Context: 798 Period: 2
Description: Edge-ground and half-broken disc of black ware sherd.
Object: Disc Material: Terracotta.
No: 855 Trench: A:TP6 Context: 1031 Period: 3-temple Description: Fragment of a disc of terracotta.
Object: Disc Material: Stone.
Description: Fragment of an edge-ground disc of trap stone.

No: 39 Trench: A:TP9 Context: 320 Period: 4 Perforated discs

Object: Disc Material: Pot-sherd.
Description: Edge-ground disc of black ware sherd. No: 503 Trench: B Context: 452 Period: 1
Object: Perforated disc Material: Pot-sherd.
No: 99 Trench: A:NWNT Context: 376 Period: 4 Description: Broken perforated disc of dull red ware sherd.
Object: Disc Material: Pot-sherd.
Description: Small, edge-ground thick disc of burnished buff ware No: 547 Trench: D1 Context: 770 Period: 1
sherd. Object: Perforated disc Material: Pot-sherd.
Description: Two fragments of edge-ground perforated discs of bur-
No: 278 Trench: A:TP6 Context: 596 Period: 4 nished red ware sherd. Fig. 8.7.
Object: Disc Material: Pot-sherd.
Description: Edge-ground disc of burnished black ware. No: 654 Trench: B Context: 468 Period: 1
Object: Perforated disc Material: Pot-sherd.
No: 315 Trench: A:TP6, South Area Context: 613 Period: 4 Description: Broken, perforated disc of burnished red ware sherd.
Object: Disc Material: Pot-sherd.
Description: Edge-ground disc of dull red ware sherd. No: 722 Trench: F Context: 846 Period: 2
Object: Perforated disc Material: Pot-sherd.
No: 113 Trench: A:NWNT Context: 504 Period: 4? Description: Broken, edge-ground perforated disc of red ware
Object: Disc Material: Pot-sherd. sherd. Fig. 8.7.
Description: Broken, edge-ground disc of red ware sherd.
No: 852 Trench: A:TP1 Context: 1013 Period: 2
No: 119 Trench: A:South Area Context: 513 Period: No period Object: Perforated disc Material: Pot-sherd.
Object: Disc Material: Potsherd. Description: Partly edge-ground and broken perforated disc of red
Description: Small, edge-ground disc of burnished red ware. ware sherd, perforation made from either side.
No: 129 Trench: A:N Temple Context: 536 Period: No period No: 83 Trench: A:EEST Context: 348 Period: 3
Object: Disc Material: Potsherd. Object: Perforated disc Material: Pot-sherd.
Description: Edge-ground disc of burnished black ware sherd. Description: Edge-ground disc of burnished black ware sherd with
No: 477 Trench: E Context: 809 Period: No period perforation under process.
Object: Disc Material: Pot-sherd. No: 160 Trench: B Context: 413 Period: 4
Description: Edge-ground, triangular-shaped disc of pale grey ware Object: Perforated disc Material: Pot-sherd.
sherd, which is part of a diagonally broken squarish object. Description: Slightly broken perforated disc of red ware sherd,
No: 711 Trench: E Context: 816 Period: No period perforation made from either surface.
Object: Disc Material: Pot-sherd. No: 170 Trench: B Context: 420 Period: 4
Description: Two small, edge-ground discs of buff ware sherds in Object: Perforated disc Material: Pot-sherd.
varied sizes. Description: Broken perforated disc of grey ware, perforation made
No: 750 Trench: E Context: 843 Period: No period from under surface.
Object: Disc Material: Pot-sherd. No: 422 Trench: A:South Area, TP6 Context: 635 Period: 4
Description: Fragment of an edge-ground disc of red ware sherd. Object: Perforated disc Material: Pot-sherd.
No: 751 Trench: E Context: 843 Period: No period Description: Half-broken, perforated disc of red ware sherd, partly
Object: Disc Material: Pot-sherd. edge-ground.
Description: Fragment of an edge-ground disc of burnished red No: 423 Trench: A:South Area, TP6 Context: 635 Period: 4
ware sherd. Object: Perforated disc Material: Pot-sherd.
No: 815 Trench: E Context: 1206 Period: No period Description: Fragment of a perforated disc of red ware sherd, perfo-
Object: Disc Material: Pot-sherd. ration made from either surface.
Description: Thick, squarish disc of red ware sherd with visible
grits and slightly damaged.

No: 816 Trench: E Context: 1206 Period: No period Hopscotches

Object: Disc Material: Pot-sherd.
Description: Fragment of a thin disc of red ware sherd. No: 472 Trench: D1 Context: 766 Period: 1
Object: Hopscotch Material: Pot-sherd.
No: 49 Trench: A:S of S Temple Context: 303 Period: 4
Object: Disc Material: Terracotta.
204   M. Kasturi Bai

Description: Slightly broken disc (hopscotch) of dull red ware Marbles

sherd.
No: 703 Trench: B Context: 485 Period: 1
No: 548 Trench: D1 Context: 773 Period: 1 Object: Marble Material: Terracotta.
Object: Hopscotch Material: Pot-sherd. Description: Broken marble of terracotta.
Description: Small disc (hopscotch) of dull red ware sherd.
No: 12 Trench: A:TP1 Context: 308 Period: 3
No: 618 Trench: A:TP1 Context: 966 Period: 2 Object: Marble Material: Stone.
Object: Hopscotch Material: Pot-sherd. Description: Small, round marble of stone(?).
Description: Fragment of a disc (hopscotch) of burnished black
ware sherd. No: 647 Trench: A:TP6 Context: 987 Period: 3
Object: Marble Material: Stone.
No: 680 Trench: F Context: 788 Period: 2 Description: Half-broken marble of stone.
Object: Hopscotch Material: Pot-sherd.
Description: Thick disc (hopscotch) of red ware sherd. No: 731 Trench: A:TP6 Context: 1051 Period: 3
Object: Marble Material: Stone.
No: 685 Trench: F Context: 789 Period: 2 Description: Slightly broken spherical marble of stone(?).
Object: Hopscotch Material: Pot-sherd.
Description: Broken disc (hopscotch) of dull Black and Red ware No: 13 Trench: A:TP1 Context: 308 Period: 3
sherd. Object: Marble Material: Terracotta.
Description: Spheroid-shaped marble of terracotta in brick-red
No: 738 Trench: A:TP1 Context: 1063 Period: 2 colour.
Object: Hopscotch Material: Pot-sherd.
Description: Big and thick broken disc (hopscotch) of red ware No: 523 Trench: A:TP1 Context: 921 Period: 3
sherd. Object: Marble Material: Terracotta.
Description: Spherical marble of terracotta with incised decora-
No: 566 Trench: A:TP1 Context: 930 Period: 2 tion in the form of ‘+’.
Object: Hopscotch Material: Pot-sherd.
Description: Two small broken discs (hopscotch) of red ware No: 540 Trench: A:TP5 Context: 918 Period: 3-temple
sherds. Object: Marble Material: Stone.
Description: Small spherical marble of stone.
No: 577 Trench: A:TP6 Context: 342 Period: 3-temple
Object: Hopscotch Material: Pot-sherd. No: 456 Trench: A:TP9 Context: 641 Period: 4
Description: Slightly broken disc (hopscotch) of burnished red Object: Marble Material: Terracotta.
ware sherd. Description: Spherical marble of terracotta, red in colour.

No: 579 Trench: A:TP6 Context: 342 Period: 3-temple No: 769 Trench: E Context: 816 Period: No period
Object: Hopscotch Material: Pot-sherd. Object: Marble Material: Stone.
Description: Big disc (hopscotch) of red ware sherd. Description: Small, spheriod marble of stone.

No: 168 Trench: B Context: 418 Period: 4 No: 825 Trench: E Context: 1206 Period: No period
Object: Hopscotch Material: Pot-sherd. Object: Marble Material: Stone.
Description: Edge-ground disc (hopscotch) of burnished buff ware. Description: Small spherical stone marble.

No: 197 Trench: B Context: 435 Period: 4

Object: Hopscotch Material: Pot-sherd.
Description: Partly edge-ground disc (hopscotch) of dull red ware Toys
sherd.
No: 316 Trench: A:TP6, South Area Context: 613 Period: 4
No: 228 Trench: B Context: 439 Period: 4
Object: Toy Material: Terracotta.
Object: Hopscotch Material: Pot-sherd.
Description: Broken toy rotary grinder of terracotta.
Description: Edge-ground disc (hopscotch) of red ware sherd.
No: 678 Trench: F Context: 788 Period: 2
No: 156 Trench: B Context: 410 Period: 4
Object: Toy Material: Porcelain.
Object: Disc Material: Terracotta.
Description: Head of a peacock figurine of porcelain. Fig. 8.10.
Description: Edge-ground terracotta disc (hopscotch?) with porous
Note: This object is clearly intrusive in Period 2.
surface and slightly damaged on one surface.

No: 702 Trench: E Context: 831 Period: No period

Object: Hopscotch Material: Pot-sherd.
Description: Two fragments of discs (hopscotch) of grey ware Toy wheels
sherds(?).
No: 367 Trench: D1 Context: 760 Period: 1
Object: Toy wheel Material: Pot-sherd.
Description: Edge-ground half-broken toy wheel of burnished buff
ware sherd with hourglass section. Fig. 8.7.
 Chapter 8 The Small Finds and Glass    205

No: 412 Trench: D2 Context: 718 Period: 1 Museum at Paithan, which were collected from different
Object: Toy wheel Material: Pot-sherd. mounds of Paithan (see Chapter 11). One of the figurines
Description: Edge-ground toy wheel of burnished red ware sherd, (No. 111) from the excavation resembles very much a
perforation made from either surface. Fig. 8.7.
similar type of figurine from Ter (see below).
No: 593 Trench: A:TP1 Context: 944 Period: 2
No: 748 Trench: A:TP1 Context: 1105 Period: 2
Object: Toy wheel Material: Pot-sherd.
Object: Figurine Material: Kaolin.
Description: Big and thick edge-ground toy wheel of red ware
Description: Left side portion of a face of a female figurine of
sherd, perforation made from either surface.
kaolin having high forehead with circular beaded ornament
No: 861 Trench: A:TP3 Context: 1112 Period: 3-temple (chudamani), broad prominent eyes, plump cheek, gentle smiling
Object: Toy wheel Material: Terracotta. expression, hair hanging down and turning backwards covering
Description: Thick and crude toy wheel of terracotta with grits of the left ear. The right-side portion of the face is missing. Double
stone, shell and vegetable temper. moulded. Figs 8.10 and 8.14.

No: 35 Trench: A:NWNT, TP9 Context: 328 Period: 4 No: 111 Trench: A:NWNT Context: 504 Period: 4?
Object: Toy wheel Material: Pot-sherd. Object: Figurine Material: Kaolin.
Description: Edge-ground, toy wheel of red ware sherd, perfora- Description: Broken torso of a human figurine of kaolin in a sitting
tion made from under surface. posture, stunted left leg spread apart with left hand palm resting
partly on thigh and partly on ornate waist band, wearing three
No: 106 Trench: A:NWNT Context: 381 Period: 4?
rows of ornaments in anklet. Double moulded. Figs 8.10 and 8.14.
Object: Toy wheel Material: Pot-sherd.
Ref: A similar figurine was recovered from Period II at Ter (Chapekar
Description: Toy wheel of red ware sherd, perforation made on
1969: 88, fig. 28, 42). The upper portion of the torso is missing in the
either side.
Paithan example; otherwise, both are similar.

No: 600 Trench: A:GG North Context: 949 Period: 3-temple

Object: Figurine Material: Kaolin.
Dice Description: Bust of a male(?) figurine of kaolin depicting the
body bending slightly towards the left with a prominent navel and
No: 182 Trench: B Context: 430 Period: 4
a beaded pendent of a chain below the neck. Other things on the
Object: Dice Material: Ivory.
body are not identifiable. Double moulded. Figs 8.10 and 8.14.
Description: Long cubical dice of ivory with slightly worn set of ring
incisions on each face, one(?), three and six circles respectively. Fig. No: 694 Trench: A:TP1 Context: 1040 Period: 2
8.8. Object: Figurine Material: Kaolin.
Description: Fragment of an animal figurine of kaolin showing the
feet of the fore legs of an elephant. Double moulded. Fig. 8.10.

Pan No: 809 Trench: A:TP3 Context: 1148 Period: 2

Object: Figurine Material: Kaolin.
No: 664 Trench: E Context: 829 Period: No period Description: Fragment of an unidentified portion of a figurine of
Object: Pan Material: Ivory. kaolin. Double moulded.
Description: Two fragments of a cylindrical pan of ivory(?) with a
circular section, ground and polished.

Terracottas

Figurines Of the 26 terracotta figurines, five are human and the rest
are broken animal or bird figurines and their body parts.
Some are hand modelled and some are single moulded.
A total of 32 figurines came to light during the excava-
tions. Of these, 26 were of terracotta, one of unbaked
clay and five of kaolin.
Human figurines

No: 144 Trench: B Context: 402 Period: 4

Kaolin figurines Object: Figurine Material: Terracotta.
Description: Boy in sitting posture with excellent facial features,
Among the five figurines of kaolin, three come from hair tied on the top of the head in a circular fashion, legs and one
Period 2, one from Period 3-temple and one from ‘Period hand slightly broken. Hand modelled in terracotta. Fig. 8.15.

4?’. All are double moulded. Kaolin figurines recov- No: 161 Trench: B Context: 413 Period: 4
ered from the excavations compare well with the kaolin Object: Figurine Material: Terracotta.
figurines displayed in the Balasaheb Patil Government
206   M. Kasturi Bai

Description: Crude female figurine (anthropomorphic) with a cylin- Description: Broken leg of an animal(?) figurine of terracotta with
drical body and with prominent breasts, legs stretched sidewise. a coating of white paint.
The head, hands and left leg are broken. Hand modelled from ter-
No: 567 Trench: A:TP1 Context: 930 Period: 3
racotta. Figs 8.15 and 8.16.
Object: Figurine Material: Terracotta.
No: 289 Trench: A:South Area Context: 605 Period: 4 Description: Leg(?) portion of an animal figurine of terracotta
Object: Figurine Material: Terracotta. with a coat of white paint and a patch of pink paint.
Description: Seated male figurine with head and hands broken,
No: 317 Trench: A:TP6, South Area Context: 613 Period: 4
resting on barrel sided back rest. Crudely hand-made in terracotta.
Object: Figurine Material: Terracotta.
Fig. 8.16.
Description: Slightly broken leg of an animal figurine of terracotta
No: 386 Trench: E Context: 801 Period: No period with partly visible white paint.
Object: Figurine Material: Terracotta.
No: 843 Trench: A:TP10 Context: 928 Period: 4
Description: Torso of a male figurine of terracotta.
Object: Figurine Material: Terracotta.
No: 24 Trench: None Context: 0 Period: No period Description: Leg fragment of an animal(?) figurine of terracotta.
Object: Figurine Material: Terracotta.
No: 396 Trench: E Context: 801 Period: No period
Description: Head of a male figurine with prominent nose, bulged
Object: Figurine Material: Terracotta.
eyes, and a croff on the forehead. Single moulded in terracotta. Fig.
Description: Leg portion of an animal figurine of terracotta.
8.16.
No: 157 Trench: B Context: 410 Period: 4
Object: Figurine Material: Terracotta.
Description: Leg(?) of an animal figurine of terracotta with traces
Animals and birds of white paint.

No: 152 Trench: B Context: 402 Period: 4

The animals represented in terracotta are a bull, goat,
Object: Figurine Material: Terracotta.
horse and tiger. The body parts of the animals found at Description: Horn of a bull figurine of terracotta.
Paithan were represented by horns and legs. One was
No: 253 Trench: A:South Area Context: 589 Period: 3
found in Period 1, two in Period 2, two in Period 3, one Object: Figurine Material: Terracotta.
from Period 3-temple, ten from Period 4 and five from Description: Horn(?) portion of an animal figurine(?) of terracotta
layers that have not been allocated to a period. with a lime coat.
No: 237 Trench: None Context: 0 Period: No period No: 752 Trench: E Context: 843 Period: No period
Object: Figurine Material: Terracotta. Object: Figurine Material: Terracotta.
Description: Head of a tiger figurine of terracotta with one ear Description: Conical-shaped horn(?) of an animal figurine of ter-
partly damaged and missing body portion. Fig. 8.15. racotta. Slightly broken.
No: 90 Trench: A:S Temple Context: 366 Period: No period No: 241 Trench: A:South Area Context: 568 Period: 4
Object: Figurine Material: Terracotta. Object: Figurine Material: Terracotta.
Description: Small humped bull figurine of terracotta (votive Description: Horn of an animal figurine of terracotta, coated with
Nandi?). Slightly broken, handmade. Fig. 8.16. pale pinkish paint and violet painted lines at the end.
No: 143 Trench: B Context: 402 Period: 4 No: 778 Trench: F Context: 851 Period: 1
Object: Figurine Material: Terracotta. Object: Figurine Material: Terracotta.
Description: Goat figurine of terracotta with slightly broken legs, Description: Tail portion of an animal(?) figurine of terracotta.
mouth and tail. Hand modelled. Fig. 8.13.
No: 845 Trench: A:TP3 Context: 981 Period: 3-temple
No: 240 Trench: A:South Area Context: 568 Period: 4 Object: Figurine Material: Terracotta.
Object: Figurine Material: Terracotta. Description: Fragment of a bird(?) figurine of terracotta with a lime
Description: Broken horse figurine of terracotta with occasional wash. Fig. 8.16.
white paint. Double moulded. Fig. 8.16.
No: 159 Trench: B Context: 410 Period: 4
No: 290 Trench: A:South Area Context: 605 Period: 4 Object: Figurine Material: Terracotta.
Object: Figurine Material: Terracotta. Description: Broken, unidentified animal figurine of terracotta
Description: Conical-shaped leg of an animal figurine of terracotta with a seat-arrangement on its back.
with tail representation. Coated in white paint.
No: 346 Trench: A:South Area, TP6 Context: 635 Period: 4
No: 617 Trench: A:TP1 Context: 966 Period: 2 Object: Figurine Material: Terracotta.
Object: Figurine Material: Terracotta. Description: Trunk(?) of an elephant figurine of terracotta, slightly
Description: Leg fragment of an animal figurine of terracotta. broken at the narrower end and faintly visible white paint.
No: 693 Trench: A:TP1 Context: 1040 Period: 2 No: 531 Trench: E Context: 816 Period: No period
Object: Figurine Material: Terracotta. Object: Figurine Material: Clay.
 Chapter 8 The Small Finds and Glass    207

Fig. 8.13: Animal terracotta figurines.

208   M. Kasturi Bai

Fig. 8.14: Kaolin figurines.

 Chapter 8 The Small Finds and Glass    209

Fig. 8.15: Terracotta figurines.

210   M. Kasturi Bai

Fig. 8.16: Human and animal terracotta figurines.

 Chapter 8 The Small Finds and Glass    211

Fig. 8.17: Stone sculpture 234.

212   M. Kasturi Bai

Fig. 8.18: Stone sculpture 234.

 Chapter 8 The Small Finds and Glass    213

Description: Three fragments of a broken serpent figurine with a No: 779 Trench: B Context: 735 Period: 1
hood, made of unbaked clay. Object: Unidentified Material: Tortoise shell.
Description: Unidentified object(?) of tortoise shell.

No: 810 Trench: A:TP3 Context: 1148 Period: 2

Object: Unidentified Material: Glass.
Sculpture fragment Description: Fragment of an unidentified object of dark honey-
coloured soft glass.
No: 234 Trench: B Context: 440 Period: 4
No: 695 Trench: A:TP9 Context: 1050 Period: 3-temple
Object: Sculpture Material: Stone.
Object: Unidentified Material: Glass.
Description: Bust of a small female sculpture of greenish lime-
Description: Fragment of an unidentified object of glass.
stone with the head missing, arms suspended sidewise with a
series of ornaments executed in low relief. Robust breasts with No: 402 Trench: A:South Area, TP6 Context: 635 Period: 4
pointed nipples, traces of an ear ornament hanging down on the Object: Unidentified Material: Porcelain.
right-side of the body above the breast. The back of the sculpture Description: Fragment of an unidentified object of porcelain.
shows hair hanging down freely and knotted in the middle with
No: 304 Trench: D2 Context: 710 Period: 2
ends of the hair coming out at the sides. Figs 8.17 and 8.18.
Object: Tablet Material: Stone.
Description: Unidentified tablet of agate(?).

No: 674 Trench: A:TP3 Context: 1023 Period: 3

Iron slag Object: Unidentified Material: Stone.

Description: Fragment of an unidentified squarish object of stone
with a leg.
No: 571 Trench: B Context: 453 Period: 1
Object: Slag Material: Iron. No: 827 Trench: A:TP9 Context: 1151 Period: 3-temple
Description: Iron slag. Object: Unidentified Material: Stone.
Description: Unidentified circular object of stone(?) with indistinct
No: 621 Trench: B Context: 458 Period: 1 symbols on either surface.
Object: Slag Material: Iron.
Description: Iron slag. No: 127 Trench: A:NWNT Context: 532 Period: 4?
Object: Unidentified Material: Stone.
No: 640 Trench: F Context: 784 Period: 2 Description: Fragment of an unidentified polished circular object
Object: Slag Material: Iron. of granite(?).
Description: Iron slag.
No: 42 Trench: A:TP2 Context: 323 Period: No period
No: 813 Trench: A:TP1 Context: 1150 Period: No period Object: Unidentified Material: Stone.
Object: Slag Material: Iron. Description: Broken unidentified polished dolerite stone.
Description: Iron slag with encrustation. Date: This object appears to have a Mauryan polish.

No: 132 Trench: A:TP2 Context: 538 Period: No period

Object: Unidentified Material: Stone.
Unidentified objects Description: Plano-convex-shaped object of sandstone with a
shallow depression at the bottom.

No: 81 Trench: A:EEST Context: 348 Period: 3 No: 69 Trench: A:TP1 Context: 300 Period: 4
Object: Sheet Material: Lead. Object: Tablet Material: Terracotta.
Description: Small and thin squarish sheet of lead with traces of Description: Edge-ground rectangular tablet of terracotta.
Ujjain(?) symbol on one side.
No: 762 Trench: F Context: 851 Period: 1
No: 429 Trench: A:TP5 Context: 661 Period: 3-temple Object: Unidentified Material: Terracotta.
Object: Unidentified Material: Ivory. Description: Broken unidentified object of terracotta.
Description: Unidentified object of ivory, perhaps handle of a comb
No: 452 Trench: A:TP5 Context: 667 Period: 3-temple
with excellent cut design of human representation followed by
Object: Unidentified Material: Terracotta.
incised design. Fig. 8.10.
Description: Small, unidentified disc-shaped object of smoky
No: 501 Trench: B Context: 452 Period: 1 quartz(?) with a polished top.
Object: Unidentified Material: Shell.
No: 814 Trench: A:TP9 Context: 1151 Period: 3-temple
Description: Fragment of a ribbon-shaped unidentified shell object
Object: Unidentified Material: Terracotta.
with incised decoration. Fig. 8.10.
Description: Fragment of unidentified object of terracotta with
No: 737 Trench: A:TP6 Context: 1051 Period: 3 incised decoration of dots and lines.
Object: Unidentified Material: Shell.
No: 857 Trench: A:TP3, TP5 Context: 1062 Period: 3-temple
Description: Broken ribbon-shaped unidentified shell object.
Object: Unidentified Material: Terracotta.
214   M. Kasturi Bai

Description: Fragment of an unidentified portion of a terracotta No: 389 Trench: E Context: 801 Period: No period
object with decoration. Double moulded. Object: Unidentified Material: Terracotta.
Description: Broken unidentified object of terracotta with traces
No: 277 Trench: A:South Area Context: 592 Period: 4 of white coating.
Object: Unidentified Material: Terracotta.
No: 868 Trench: A:TP6 Context: 1214 Period: No period
Description: Broken unidentified object of terracotta.
Object: Unidentified Material: Terracotta.
No: 291 Trench: A:South Area Context: 605 Period: 4 Description: Fragment of an unidentified object of terracotta.
Object: Unidentified Material: Terracotta.
Description: Slightly broken unidentified object of terracotta.

No: 300 Trench: A:W Area Context: 616 Period: 4

Object: Unidentified Material: Terracotta. Glass vessels
Description: Fragment of an unidentified terracotta object deco-
rated with appliqué petal motif.
Apart from the glass bangles and beads that are
No: 352 Trench: A:South Area, TP6 Context: 635 Period: 4 described above with the small finds, only six frag-
Object: Unidentified Material: Terracotta.
ments of glass vessels were recorded from the exca-
Description: Broken unidentified object of terracotta with a lime
coating.
vations (Table 8.11). All but one of these comes from
Period 4. Some of them appear to be very recent,
No: 353 Trench: A:South Area, TP6 Context: 635 Period: 4
although P2184, P2181 and P1618 look as if they could
Object: Unidentified Material: Terracotta.
Description: Unidentified and slightly curved cylindrical terracotta be considerably older. P2184 comes from a stratified
object, broken on either end. context allocated to Period 2.
No: 447 Trench: A:South Area, TP6 Context: 635 Period: 4
Object: Unidentified Material: Terracotta.
Description: Edge-ground, unidentified triangular object of dull
red ware sherd, slightly broken at one end.
Fig. 8.19: Glass fragments.

Table 8.11: Glass fragments.

No. Context Period Description Figure

P1616 435 4 Body fragment. Thick translucent glass with a light green-blue tint. Not
weathered. 4 mm thick.
P1617 435 4 Body fragment. Opaque clear glass with slight yellow tint. Some weathering.
Less than 1 mm thick. Almost flat. Traces of painted or enamelled decoration
on the exterior surface.
P1618 435 4 Rim fragment. Semi-opaque purple-blue glass. Some weathering. 3 mm Fig. 8.19
thick.
P2181 414 4 Small body fragment. Opaque white colour, weathered. 1.5 mm thick.
P2184 846 2 Rim fragment. Black opaque glass, weathered. 3 mm thick. Fig. 8.19
P2554 429 4 Tiny body fragment. Translucent light-yellow glass. Not weathered. 1 mm
thick.
M. Kasturi Bai and Shailendra Bhandare (with an appendix by D. Kennet)

Chapter 9 
The Coins

Introduction ‘Hollow cross’-type coins

The excavations at Paithan yielded a total of 56 coins, This type consists of a hollow cross on the obverse.
including two that were found on the surface. Many These are square in shape. They are also corroded,
of these are in a very poor condition, making identi- cracked, worn out and sometimes flaked. All of these
fication difficult and tentative or, in 17 cases, impos- coins can be attributed to the post-Mauryan and pre-
sible. Those coins that can be identified are broadly Satavahana period.
classified in the catalogue below as pre-Satavahana
Coin 1. Find No. 574 (Context: 458), Tr. B.
(c. 370–150/100 bc), Satavahana (c. 140 bc–ad 220),
Period 1. Fig. 9.1.
post-Satavahana (2nd–7th centuries ad), Islamic (c.
Copper alloy, squarish, broken, corroded and worn out.
ad 1300–19th century) and British (ad 1818 and later),
Size: 1.75 × (-) × 0.50 cm, weight: 3.49 g.
some of which are subdivided into types. Table 9.1
Obv: Traces of a hollow cross are present.
shows the occurrence of each of these categories and
Rev: Indistinct due to corrosion.
types by period.
Coin 2. Find No. 504 (Context: 453), Tr. B.
Period 1. Fig. 9.1.

Pre-Satavahana coins Copper alloy, squarish, corroded, greenish patina and

cracked.
Size: 1.94 × 1.85 × 0.57 cm, weight: 5.85 g.
The pre-Satavahana coins are subdivided into ‘hollow
Obv: Hollow cross is faintly visible.
cross’ and ‘tree’ types. All are made of copper alloy.
Rev: Corroded, no details visible, perhaps blank.

Table 9.1: The occurrence of coins by period.

Period 1 Period 2 Period 3 Period Period 4 No period Total

3-temple

Pre-Satavahana (total) 8 2 1 3 14
‘Hollow cross’ 8 2 1 11
‘Tree’ 1 2 3
Satavahana (total) 1 3 6 5 3 18
‘Elephant’ 1 2 2 4 3 12
‘Bull’ 1 4 5
‘Lion’ 1 1
Post-Satavahana 2 1 3
Unidentified 2 1 6 5 3 17
Late Med/Islamic 2 1 3
British 1 1
Total 8 5 5 17 14 7 56

https://doi.org/10.1515/9783110653540-009
216   M. Kasturi Bai and Shailendra Bhandare (with an appendix by D. Kennet)

Fig. 9.1: Coins 1 (574), 3 (505), 5 (537), 2 (504), 4 (506) and 6 (298).
 Chapter 9 The Coins    217

Coin 3. Find No. 505 (Context: 453), Tr. B. Coin 10. Find No. 338 (Context: 711), Tr. D2.
Period 1. Fig. 9.1. Period 2.
Copper alloy, squarish, corroded, slight greenish patina, Copper alloy, broken into two pieces, corroded, cracked.
cracked. Size: 1.80 × (-) × 0.50 cm, weight: 2.04 g.
Size: 1.78 × 1.76 × 1.00 cm, weight: 8.23 g. Obv: No details visible.
Obv: Hollow cross faintly visible. Rev: No details visible.
Rev: Core visible with uneven surface.
Coin 11. Find No. 536 (Context: 923), Tr. A.
Coin 4. Find No. 506 (Context: 453), Tr. B. Period 3-temple.
Period 1. Fig. 9.1. Copper alloy, squarish, broken, corroded and cracked.
Copper alloy, squarish, slightly broken, highly corroded, Size: 1.52 × (-) × 0.43 cm, weight: 2.10 g.
flaked, cracked. Obv: No details visible due to corrosion.
Size: 2.06 × 1.69 × 0.75 cm, weight: 4.42 g. Rev: No details visible due to corrosion.
Obv: Hollow cross is visible. Note: Fabric is the same as that of ‘hollow cross’ type coins.
Rev: Corroded, no details visible.

Coin 5. Find No. 537 (Context: 453), Tr. B. ‘Tree’-type coins

Period 1. Fig. 9.1.
There are three coins of this type with a tree associated
Copper alloy, squarish, highly corroded, flaked.
with other symbols. The type occurs in the pre-Sata-
Size: 1.88 × (-) × 0.55 cm, weight: 2.87 g.
vahana and Satavahana periods, but all three coins of
Obv: No details visible.
this type found at Paithan are pre-Satavahana (Nos.
Rev: No details visible, core exposed.
12, 13 and 14). Of these, No. 12 was found in Period 3,
Note: This coin can be categorized as a ‘hollow cross’
the other two, Nos. 13 and 14, were recovered from the
type based on the fabric.
deposits of Period 3-temple. Coin 14 is similar to coin 12
Coin 6. Find No. 298 (Context: 754), Tr. D1. but differs in ancillary symbols.
Period 1. Fig. 9.1.
Coin 12. Find No. 535 (Context: 921), Tr. A.
Copper alloy, squarish with projection, corroded, flaked.
Period 3. Fig. 9.2.
Size: 1.82 × 1.65 × 0.77 cm, weight: 4.38 g.
Bronze, squarish, worn out.
Obv: Hollow cross faintly visible.
Size: 0.77 × (-) × 0.15 cm, weight: 0.46 g.
Rev: Not discernable.
Obv: Branched tree with taurine symbol on its right-side
Coin 7. Find No. 299 (Context: 754), Tr. D1. top corner, some indistinct symbols at the bottom that are
Period 1. illegible due to the worn out condition.
Copper alloy, rectangular(?), broken irregularly, cor- Rev: Some worn, indistinct symbols are visible.
roded to core, worn out.
Coin 13. Find No. 648 (Context: 956), Tr. A.
Size: 1.55 × (-) × 0.20 cm, weight: 1.08 g.
Period 3-temple. Fig. 9.2.
Obv: Worn out.
Copper, rectangular, badly worn out.
Rev: No details visible.
Size: 0.84 × (-) × 0.14 cm, weight: 0.35 g.
Coin 8. Find No. 812 (Context: 1150), Tr. A. Obv: Indistinct symbols inside a square frame.
Period 1. Rev: Worn out, part of a square frame is visible.
Copper alloy, fragment, corroded and encrusted.
Coin 14. Find No. 538 (Context: 919), Tr. A.
Size: (-) × (-) × 0.45 cm, weight: not taken.
Period 3-temple. Fig. 9.2.
Obv: No details visible.
Bronze, rectangular, obliquely broken.
Rev: No details visible due to corrosion.
Size: 0.91 × (-) × 0.16 cm, weight: 0.46 g.
Coin 9. Find No. 337 (Context: 711), Tr. D2. Obv: Branched tree with partly visible railing, taurine
Period 2. Fig. 9.2. symbol on right-side top corner of the tree, indistinct
Copper alloy, squarish, broken into two pieces, cor- symbol on the left side of the tree at the bottom, the left
roded, cracked. side upper portion is cut off.
Size: 1.90 × 1.78 × 0.78 cm, weight: 6.28 g. Rev: A tree with oval leaves to right.
Obv: Hollow cross is faintly visible. Note: This coin is similar to No. 12 in fabric and in the
Rev: Unidentifiable due to flaking of the surface. main symbol.
218   M. Kasturi Bai and Shailendra Bhandare (with an appendix by D. Kennet)

Fig. 9.2: Coins 9 (337), 12 (535), 14 (538), 52 (493), 13 (648) and 25 (379).
 Chapter 9 The Coins    219

Satavahana coins can be safely attributed to Kosikiputa Siri Satakani,

by comparison with published specimens (Chumble
1991: 41–50) and also on the basis of a distinct ancil-
‘Elephant’-type coins in copper
lary symbol, a Shrivatsa placed below the mouth of the
elephant.
Coin 15. Find No. 724 (Context: 1072), Tr. A.
Period 3-temple. Fig. 9.3. Coin 20. Find No. 57 (Context: 320), Tr. A.
Copper, rectangular, worn out. Period 4. Fig. 9.4.
Size: 0.87 × 0.73 × 0.20 cm, weight: 0.71 g. Copper alloy, round, corroded and worn out, cracked.
Obv: Traces of an elephant facing to right, traces of some Size: 1.90 × 0.51 cm, weight: 4.76 g.
indistinct letters above the animal. Obv: Traces of an elephant facing to right, trunk hanging,
Rev: Traces of Ujjain symbol. swastika (or Shrivatsa) above the head of the animal,
indistinct letters.
Coin 16. Find No. 495 (Context: 700), Tr. A.
Rev: A series of dots in groups, a crescent like symbol is
Period 3-temple.
visible but unidentifiable.
Copper alloy, squarish, fragment, corroded, worn out.
Size: (-) × (-) × 0.37 cm, weight: 0.43 g. Coin 21. Find No. 55 (Context: 303), Tr. A.
Obv: Traces of some indistinct symbol. Period 4. Fig. 9.4.
Rev: Traces of part of an Ujjain symbol. Copper alloy, squarish, slightly broken with greenish
patina, corroded.
Coin 17. Find No. 745 (Context: 1099), Tr. A.
Size: 1.25 × 1.15 × 0.35 cm, weight: 2.02 g.
Period 2. Fig. 9.3.
Obv: Elephant facing to right, trunk hanging is visible.
Copper alloy, broken, squarish with greenish patina.
Rev: Worn out and traces of indistinct symbols.
Size: 1.70 × (-) × 0.45 cm, weight: 5.29 g.
Obv: Indistinct symbols and not discernable due to Coin 22. Find No. 46 (Context: 303), Tr. A (TP–3).
worn out condition and patina. Period 4. Fig. 9.4.
Rev: Corroded and worn out, however traces of two orbs Copper alloy, round with greenish patina, worn out,
of an Ujjain symbol are visible. slightly cracked.
Size: 1.45 × 1.40 × 0.30 cm, weight: 1.47 g.
Coin 18. Find No. 02 (Context: surface), Tr. A.
Obv: Faintly visible elephant facing to right with trunk
Surface. Fig. 9.3.
hanging, enclosed in a dotted circle that is partly visible,
Copper alloy, round, corroded, cracked, worn out.
legend along the flan is illegible.
Size: 1.49 × 1.50 × 0.36 cm, weight: 2.02 g.
Rev: Worn out, nothing can be made out.
Obv: Faint traces of an elephant.
Rev: Faint traces of an Ujjain symbol. Coin 23. Find No. 118 (Context: 513), Tr. A.
Unattributable period.
Coin 19. Find No. 188 (Context: 435), Tr. A.
Copper alloy, round, slightly broken, cracked to the
Period 4. Fig. 9.3.
core, worn out.
Copper alloy, ovalish, slightly worn out.
Size: 1.10 × (-) × 0.21 cm, weight: 0.48 g.
Size: 1.54 × 1.34 × 0.23 cm, weight: 1.98 g.
Obv: Indistinct symbols, unidentifiable.
Obv: Elephant walking to right, trunk and back partly
Rev: Indistinct symbols, unidentifiable.
off the flan, Shrivatsa symbol in front above the fore-
legs and below the mouth, legend along the flan is Coin 24. Find No. 390 (Context: 801), Tr. E.
partly cut off, the extant legend starting at 2 o’clock Unattributable period. Fig. 9.4.
reads ‘(PU) TASA SIRISADA (KA)…’ Copper alloy, round, slight greenish patina, corroded,
Rev: Tree with five oval leaves was found towards the worn out.
left side of the reverse of the coin. Size: 1.52 × 1.35 × 0.30 cm, weight: 1.88 g.
Note: This coin belongs to a type markedly oriented Obv: Traces of elephant standing to right, trunk hanging.
to the Nevasa-Paithan region in terms of its distribu- Rev: Traces of some indistinct symbols.
tion (see the discussion at the end of this chapter).
The inscription on it can be read as ‘(PU) TASA
SIRISADA (KA)…’, leaving out the complete metro-
nymic. Although the full legend is not clear, the coin
220   M. Kasturi Bai and Shailendra Bhandare (with an appendix by D. Kennet)

Fig. 9.3: Coins 30 (740), 18 (02), 15 (724), 33 (820), 17 (745) and 19 (188).
 Chapter 9 The Coins    221

Copper, rectangular.
‘Elephant’-type coins in lead
Size: 1.07 × 0.84 × 0.22 cm, weight: 1.08 g.
Coin 25. Find No. 379 (Context: 642), Tr. A. Obv: A humped bull with prominent horns walking to
Period 3. Fig. 9.2. left. A symbol comprising of four interconnected circles
Lead, roundish, small. seen partially above its hump.
Size: 1.15 × 1.07 × 0.17 cm, weight: 1.21 g. Rev: Traces of some indistinct symbols.
Obv: Elephant facing to the right, trunk partly cut
off, legend along the flan partly cut off and worn out.
However, the letter ‘sa’ is visible below the fore legs. ‘Lion’-type coins
Rev: Four unconnected circles, one of the circles is sur-
mounted by a crescent and part of a swastika is visible Coin 32. Find No. 120 (Context: 519), Tr. A.
in between the two circles. Period 4. Fig. 9.5.
Lead, squarish, thin, worn out.
Coin 26. Find No. 494 (Context: 650), Tr. A. Size: 1.50 × 1.30 × 0.14 cm, weight: 1.70 g.
Period 3. Obv: Lion facing to right, fore leg lifted up as if jumping,
Lead, round, broken, worn out. tail raised up, triratna motif above the neck.
Size: 1.15 × (-) × 0.20 cm, weight: 0.62 g. Rev: Worn out.
Obv: Traces of an elephant facing to right, legend worn out.
Rev: Ujjain symbol faintly visible.

Post-Satavahana coins
‘Bull’-type coins
Three coins can be dated to the post-Satavahana period,
Coin 27. Find No. 651 (Context: 993), Tr. A.
two are from Period 3-temple and one from Period 4.
Period 3.
Copper alloy, broken, corroded up to core, cracked. Coin 33. Find No. 820 (Context: 1161), Tr. A.
Size: 1.18 (diam.), 0.31 cm, weight: 0.62 g. Period 3-temple. Fig. 9.3.
Obv: Traces of bull(?). Copper alloy, ovalish, corroded almost up to core, worn
Rev: Worn out. out.
Size: 1.30 × 1.00 × 0.23 cm, weight: 0.67 g.
Coin 28. Find No. 829 (Context: 1172), Tr. A.
Obv: Legend along the flan is indistinct and illegible.
Period 3-temple. Fig. 9.4.
Rev: Worn out and not discernable.
Copper alloy with lead coating, squarish, tiny, worn out.
Note: Possibly a Western Kshatrapa coin.
Size: 0.60 × 0.55 × 0.17 cm, weight: 0.28 g.
Obv: Traces of humped bull facing right. Coin 34. Find No. 496 (Context: 700), Tr. A.
Rev: Traces of three-arched hill. Period 3-temple. Fig. 9.5.
Note: Before conservation this coin was identified as lead Copper alloy, round, cracked, one flake peeled off on
rather than copper. the obverse, worn out.
Size: 1.60 × 1.55 × 0.20 cm, weight: 1.72 g.
Coin 29. Find No. 492 (Context: 667), Tr. A.
Obv: Some indistinct symbols are found but are uni-
Period 3-temple, Fig. 9.4.
dentifiable due to the flaking of metal and the worn
Lead, round, corroded.
out condition.
Size: 1.45 × 1.37 × 0.27 cm, weight: 2.50 g.
Rev: Trident with traces of crude Brahmi characters,
Obv: Humped bull facing to right.
enclosed within traces of a dotted border.
Rev: Corroded and indistinct symbol.
Note: For dating and possible attribution see discussion
Coin 30. Find No. 740 (Context: 1083), Tr. A. below.
Period 3-temple. Fig. 9.3.
Coin 35. Find No. 29 (Context: 320), Tr. A.
Lead, round, thick coin.
Period 4. Fig. 9.5.
Size: 1.30 × 1.20 × 0.22 cm, weight: 2.27 g.
Silver, round, heavily worn out.
Obv: Bull walking to right, partly visible.
Size: 1.64 × 1.60 × 0.20 cm, weight: 3.00 g.
Rev: Faint traces of Ujjain symbol.
Obv: Bust of a king (?), legend illegible.
Coin 31. Find No. 462 (Context: 666), Tr. A. Rev: Worn out.
Period 3-temple. Fig. 9.5. Note: For dating and attribution see discussion below.
222   M. Kasturi Bai and Shailendra Bhandare (with an appendix by D. Kennet)

Fig. 9.4: Coins 22 (46), 21 (55), 29 (492), 20 (57), 24 (390) and 28 (829).
 Chapter 9 The Coins    223

Unidentifiable coins Coin 44. Find No. 424 (Context: 649), Tr. A.
Period 4.
Among the unidentifiable coins, some are complete and
Copper alloy, squarish, half-broken, worn.
others are fragments. These coins cannot be attributed to
Obv & Rev: No details visible.
particular type due to the poor condition, although, to judge
by the fabric, many of these coins may be Satavahana. Coin 45. Find No. 01 (Context: 303), Tr. A.
Period 4.
Coin 36. Find No. 259 (Context: 703), Tr. D2.
Copper alloy, squarish, slightly broken, corroded to
Period 2.
the core and worn out.
Thin fragment of a coin, greenish patina and metal
Size: 1.15 × 1.00 × 0.13 cm, weight: 0.70 g.
eaten away to core.
Obv & Rev: No details visible due to corrosion.
Obv & Rev: No details visible.
Coin 46. Find No. 43 (Context: 303), Tr. A (TP-3).
Coin 37. Find No. 524 (Context: 921), Tr. A.
Period 4.
Period 3.
Copper alloy, round, corroded and worn out, cracked.
Copper alloy, fragment, corroded.
Size: 1.63 × 1.56 × 0.36 cm, weight: 2.61 g.
Obv & Rev: No details visible.
Obv: Some indistinct animal and letters are visible.
Coin 38. Find No. 819 (Context: 1157), Tr. A. Rev: Indistinct symbols.
Period 3-temple.
Coin 47. Find No. 47 (Context: 303), Tr. A (TP-3).
Copper alloy, ovalish, broken, corroded, worn out,
Period 4.
cracked.
Copper alloy, ovalish with projection, greenish patina,
Obv & Rev: No details visible.
worn out, slightly cracked.
Note: This was identified as lead before conservation.
Size: 1.55 × 1.25 × 0.47 cm, weight: 1.65 g.
There was a lead coating, which survives as patches on
Obv & Rev: Some indistinct symbols are visible.
the copper.
Coin 48. Find No. 44 (Context: 303), Tr. A (TP-3).
Coin 39. Find No. 796 (Context: 1130), Tr. A.
Period 4.
Period 3-temple.
Copper alloy, round, broken, corroded, cracked.
Copper alloy, thin fragment of a coin.
Size: 1.37 × 1.20 × 0.30 cm, weight: 1.59 g.
Obv: Traces of a symbol but indistinct.
Obv & Rev: No details visible due to corrosion.
Rev: No details visible.
Coin 49. Find No. 836 (Context: 1191), Tr. A.
Coin 40. Find No. 649 (Context: 964), Tr. A.
Unattributable period.
Period 3-temple.
Copper alloy with lead coating, squarish, corroded and
Copper, squarish, broken and worn out on either side.
worn out.
Size: 0.83 × 0.81 × 0.17 cm, weight: 0.40 g.
Size: 0.90 × 0.90 × 0.20 cm, weight: 0.55 g.
Obv & Rev: Trace of indistinct symbols.
Obv & Rev: Indistinct symbols.
Note: Based on the fabric this coin may be ‘tree’ type.
Note: Before conservation this coin was identified as lead.
Coin 41. Find No. 516 (Context: 918), Tr. A.
Coin 50. Find No. 837 (Context: 1191), Tr. A.
Period 3-temple.
Unattributable period.
Copper alloy, fragment, highly corroded, worn out.
Copper alloy, fragment, corroded.
Obv & Rev: No details visible.
Obv & Rev: No details visible.
Coin 42. Find No. 497 (Context: 700), Tr. A.
Coin 51. Find No. 660 (Context: 827), Tr. E.
Period 3-temple.
Unattributable period.
Copper alloy, rectangular, corroded up to the core.
Copper alloy, rectangular, corroded to the core, badly
Size: 1.24 × 0.95 × 0.35 cm, weight: 1.54 g.
worn out.
Obv & Rev: No details visible due to corrosion.
Size: 1.55 × 1.33 × 0.20 cm, weight: 1.71 g.
Coin 43. Find No. 580 (Context: 342), Tr. A. Obv & Rev: No details visible.
Period 3-temple.
Coin 52. Find No. 493 (Context: 451), Tr. B.
Copper alloy, fragment, highly corroded.
Period 2. Fig. 9.2.
Obv & Rev: No details visible.
Copper alloy, octagonal.
224   M. Kasturi Bai and Shailendra Bhandare (with an appendix by D. Kennet)

Fig. 9.5: Coins 31 (462), 34 (496), 55 (187), 32 (120), 35 (29) and 56 (56).
 Chapter 9 The Coins    225

Size: 1.77 × 1.69 × 0.23 cm, weight: 4.23 g. British coins

Obv: Hollow cross symbol is executed in incised dots.
Rev: Taurine symbol is executed in incised dots. Coin 56. Find No. 56 (Context: 303), Tr. A (TP–3).
Note: This is almost certainly not a coin, but a roughly Period 4. Fig. 9.5.
octagonal metallic object made of copper bearing the Copper alloy, round, thin, worn out.
hollow cross on one side and a taurine symbol on the Size: 1.75 × 1.73 × 0.10 cm, weight: 1.93 g.
other. Both these symbols have been incised using a Obv: Bust of Queen Victoria facing to left with legend in
pointed tool. On the taurine-side, two strikes have front.
resulted in scratches, thus giving an indication for Rev: 1/12 ana, India, 1897 within a circle encircled by a
the slipping of the tool while the punches were being floral pattern.
struck. The exact purpose of this object is difficult to
ascertain, but it is more likely to be intended for a non-
monetary use, given the curious method of its manu-
facture. But the motifs on it are clearly derived from
Discussion
coins.
The occurrence of the coin categories and types by period
are shown in Table 9.1. The pattern of occurrence sug-
gests that Period 1 should be dated to the pre-Satava-
Islamic coins hana period and Period 2 to the Satavahana period,
but the dating of Period 3 and Period 3-temple is more
Of the three Islamic coins, two were recovered from complicated. According to the pottery, these two
Period 4 deposits and one from the surface. periods are contemporary, but only Period 3-temple
levels contain post-Satavahana coins. This may simply
Coin 53. Find No. 45 (Context: 303), Tr. A (TP–3).
be due to statistical chance because post-Satavahana
Period 4.
coins are very much rarer and it may be that the greater
Copper alloy, round, thick, heavily worn out.
total number of coins from Period 3-temple deposits
Size: 1.72 × 1.58 × 0.30 cm, weight: 4.81 g.
has made it more likely that rare coins would be found.
Obv: Traces of Arabic inscription.
It is not therefore clear, on the basis of the coins alone,
Rev: Arabic inscription partly visible.
whether Period 3 should be dated to the Satavahana or
Note: For attribution see discussion below.
post-Satavahana period. Finally, Period 4 can be dated
Coin 54. Find No. 141 (Context: surface), Tr. B. to the Late Medieval and British periods. Table 9.2 shows
Surface. the occurrence of the main categories of coins as percent-
Copper alloy, ovalish, thick, heavy, depression at the ages of the total number of coins found in each period.
centre on one of the surfaces, greenish patina, worn The chronology of the four periods has already been dis-
out. cussed in Chapter 6.
Size: 2.06 × 1.95 × 0.45 cm, weight: 10.74 g. The tables also show high levels of residuality in all
Obv: No details visible. periods bar Period 1. In the Period 3-temple deposits, at
Rev: Arabic inscription partly visible. least 53% of the coins are residual (i.e. all of the pre-
Note: Attributed to Mughals: Aurangzeb (see discussion Satavahana and Satavahana coins), but this number
below). could be as high as 88% if all of the six unidentified
coins were Satavahana or earlier. Were it not for the two
Coin 55. Find No. 187 (Context: 435), Tr. B.
post-Satavahana coins that have been found in these
Period 4. Fig. 9.5.
deposits, this period might easily have been errone-
Copper alloy, squarish, edges slightly cut, worn out.
ously dated to the Satavahana period on the basis of the
Size: 1.84 × 1.88 × 0.40 cm, weight: 9.87 g.
other coins that it has yielded. The deposits of Period 4
Obv: Arabic letters faintly visible.
also contain very high levels of residual coins. Such high
Rev: Arabic letters faintly visible.
levels of residuality make it clear that the simple pres-
Note: Attributed to the Qutb Shahi Sultanate of Golkonda
ence of large numbers of coins in an excavated layer or
(see discussion below).
period must always be treated with caution when estab-
lishing a date. This is especially true in the post-Satava-
hana period during which time post-Satavahana coins
226   M. Kasturi Bai and Shailendra Bhandare (with an appendix by D. Kennet)

appear to have been relatively rare, whilst Satavahana are lumped together are of the same type as the soli-
coins continued to be deposited, either as part of sec- tary specimen, but, to judge by the shape and fabric, it
ondary deposits or because they continued in use. seems very likely that they are. If this is the case, then
the reverse of the ‘hollow cross’ coins would carry a
tree-in-railing motif, seen clearly on one of the coins
Pre-Satavahana coins from the lump illustrated by Yusuf. This tree-in-railing
motif, which Yusuf identified as a Bodhi tree (Yusuf
Silver punch-marked coins of the Magadha-Maurya ‘Kar- 1939: 42, pl. XXd), is flanked by a taurine symbol on
shapana’ series are noticeably absent, although they do its left.
appear to have been found elsewhere at Paithan or in the Taking a broader perspective, ‘hollow cross’-type
vicinity (Chapter 11, coins 1 to 7). They are classified as coins have been found in the excavations at Adam
GH VIb-570 (two coins), GH IVd-416 (2 coins), GH VIb- (Nath 1993: 20–21). They have also been reported along
566, GH IVd-456 and GH IVd-450 (GH numbers refer to a Vidarbha-Telingana regional axis, lying to the east
Gupta and Hardaker 1985). According to the dating of the Marathwada region, wherein Paithan is located
scheme proposed by Gupta and Hardaker (1985: 10, (Maheshwari 1977: 4–7, coins 7–10; Todywalla/Hurmuz
table 1), all these coins are dateable to the ‘Nanda’ Kaus 2004: lots 11a–c). These are generally dated to
and ‘Maurya-Sunga’ periods, or c. 370–150 bc. Three between 200 and 150 bc (Maheshwari 1977). But there
punch-marked coins are illustrated in Yusuf ’s excava- are significant differences between the Paithan and
tion reports (Yusuf 1938: pl IXa top; 1939: pl. XXc) but the Vidarbha-Telingana coins. Firstly, the Paithan
as the picture is of poor quality, nothing more can be specimens are square or irregular, whereas most of the
said about them, except that they are all rectangular others are round. Secondly, most other specimens of
in shape. this type are cast copper coins, whereas even though
The ‘hollow cross’ coins therefore seem to be the the condition of the Paithan specimens is too poor to
earliest of the coins from the excavations, being the allow any firm conclusions as to how they were manu-
only coins that occur in Period 1 deposits and are also factured, from their appearance it seems very probable
the most common type in Period 2 deposits (Table 9.1). that they are die-struck. The prevalence of die-struck
Yusuf reported ‘numerous copper coins’ from his square coins of this type at Paithan may suggest the
excavations at Paithan and his report contains two existence of a contemporary regional sub-series of
illustrations; the first, a lump of ten or so coins and uninscribed post-Mauryan copper coinage in Marath-
the second, a single specimen, which clearly has the wada.
hollow cross on it (Yusuf 1938: pl. IXa right, IXb; 1939: Apart from the ‘hollow cross’-type coins, the
pls XXd, e, f). Yusuf, however, identified the hollow other coins of a possibly pre-Satavahana date to be
cross seen on the solitary coin as a swastika and attrib- found at Paithan belong to the ‘tree’ type. In fabric
uted the coins to the Andhras. He also drew attention and weight, they conform to a genre of post-Mauryan
to coins with a similar motif found at Taxila by Alex- coins, widely recognized as ‘Ujjain’ coins and encoun-
ander Cunningham (Yusuf 1939: 42, pl. XXf). He does tered at Maheshwar and Prakash (Sankalia et al. 1958:
not make it sufficiently clear whether the coins that 73–74; Thapar 1967: 130, pl. XXVII-C 1, 3). They are gener-

Table 9.2: Occurrence of coins by period as an approximate percentage of the total number of coins found in each period.

Period 1 Period 2 Period 3 Period 3-temple Period 4

Pre-Satavahana 100% 40% 20% 18%

Satavahana 20% 60% 35% 36%
Post-Satavahana 12% 7%
Unidentified 40% 20% 35% 36%
Late Med/Islamic 14%
British 7%
Total coins 8 5 5 17 14
 Chapter 9 The Coins    227

ally dated to between 200 and 100 bc (Bopearachchi and by the Wategaon hoard, which included coins of Vasith-
Pieper 1998: 21–22). Similar coins, but of different types, iputa Siri Pulumavi and his successors (Mirashi 1972).
have also been found in the excavations at Bhokardan Coin 26 is a small denomination of the universal ‘ele-
(Deo and Gupte 1974: 63–65). It cannot be said with cer- phant × Ujjain symbol’ type in potin, which was struck
tainty whether the Paithan coins belong to a distinct by Satavahana kings post-Gautamiputa Siri Satakani
regional series of uninscribed die-struck copper coins (Sarma 1980: 97–104).
of the post-Mauryan period. The ‘bull’ type Satavahana coins found at Paithan
belong to two distinct varieties. The first, which is rep-
resented by coin 31, has the bull walking to the left with
Satavahana coins a symbol, similar to that seen above the back of the ele-
phant in the regio-specific type just described, placed
Amongst the coins classified as belonging to the ‘elephant’ above its hump. Similar coins, with the bull walking to
type, the most common at Paithan is one that occurs the right, have been published bearing legends attrib-
specifically in the Newasa-Paithan-Bhokardan region uting them to the Satavahana ruler Kochhiputa Siri
and which occurs from Period 2 onwards in the Paithan Satakani (Maheshwari 1979). Coins with legends refer-
sequence (see Table 9.1). The type may be described as ring to the generic ‘Satavahana’ name are also known
follows: (Jha and Chumble 2003/4: type 6). It is likely that
–– Obverse: The obverse has an elephant walking coin 31 is an uninscribed smaller denomination of the
to the right with an upraised trunk. It also has a inscribed coins already published and, as such, is to be
legend inscribed in exergue, usually beginning at dated to the pre-Gotamiputa period. The second, rep-
7 o’clock, with the word Rajno followed in order resented by coins 29 and 30, may date somewhat later
by the metronymic, the ruler’s first name and the despite retaining a regio-specific orientation to Nevasa-
dynastic appellation, usually ending below the feet Paithan. These coins are of lead and are anepigraphic.
of the elephant. A symbol composed of four circles Exactly similar coins have been found at Bhokardan (Deo
touching each other tangentially, with one modi- and Gupte 1974: 26–27, pl. XVI–11,12). Two similar coins
fied with a crescent, rests above the elephant’s have also been reported as having a ‘Western India’
back. Variations include the presence of a Shrivatsa provenance in the British Museum Catalogue of Satava-
symbol below the elephant’s mouth. hana coins (Rapson 1908: 54, no. 219 and 56, no. 232).
–– Reverse: The reverse has a tree with (usually) five Both are from the collection of Pt. Bhagwanlal Indraji
broad leaves, flanked by symbols such as Shrivatsa (cf. Sarma 1980: 170–171).
and the Nandipada. Prominent variations include Two more Satavahana coins, of an anepigraphic
the way the leaves emanate out from the stem and variety, were found amongst those excavated. One of
the arrangement of ancillary symbols. them, coin 32, is a lead coin of the ‘leaping lion × tree’
type. Similar coins were found in the Nevasa excava-
Coins 18, 20, 19 and 33 belong to this type. Addition- tions (Sarma 1980: 187). Coin 15 belongs to the ‘tree ×
ally, coins 22 and 24 may be included, but, as they are in Ujjain symbol’ type. Some coins of this type are known
a bad condition, this attribution is tentative. Coin 19 is from the Bhokardan excavations (Deo and Gupte 1974:
the only coin of this type that bears any legible inscrip- 50–52). Some others appear to have been counterstruck
tion. As described in the catalogue, it is most probably by the Kshaharata Kshatrapa ruler Nahapana with his
an issue of the Satavahana king Kosikiputa Siri Sata- dynastic emblem (Jha and Rajgor 1994: 105–106, nos.
kani. Coins of this particular regio-specific type have 159–161), and the undertype can therefore be safely
only been encountered in significant numbers in the dated to the mid-1st century ad.
Nevasa and Bhokardan excavations (Deo and Gupte From a chronological perspective, all of the identi-
1974: 40–49; Sarma 1980: 200–202, 204). There are also fiable Satavahana coins at Paithan can be dated to the
two coins of this type amongst the specimens preserved immediate pre- and post-Gotamiputa Satakani period,
in the Balasaheb Patil Collection (Chapter 11, coins 10 that is to say from the late 1st century bc to the late 1st
and 11). century ad based on the dates proposed for this ruler
Coins 25 and 26 can also be categorized as ‘elephant’- (Cribb 1992: 132–136). Specifically, the earliest coin is
type coins, but judging by type characteristics, they probably coin 31 and the latest is coin 26, both of which
belong to a later chronological horizon. Coin 25 appears have been discussed above. It is noteworthy that no
to be a smaller denomination of lead coins exemplified coins that can be attributed conclusively to Gotamiputa
228   M. Kasturi Bai and Shailendra Bhandare (with an appendix by D. Kennet)

Siri Satakani were found although the Balasaheb Patil Drammas. It is uncertain who struck them, but Deyell
Collection is known to include one silver coin of Nahap- believes they were struck prior to the mid-9th century
ana (c. ad 33–78) counterstruck by Gotamiputa, of a ad and gives a scheme of the distribution of their finds
variety that is found in the Jogalthembi Hoard (Scott in the Gangetic Plains (Deyell 1990: 262–268). The Patil
1908; Chapter 11, coin 14). The Patil collection also collection includes one more ‘Indo-Sasanian’ coin,
includes two silver portrait coins of Satavahana rulers belonging to the ‘Gadhaiya Paisa’ series. This coin,
Vasithiputa Siri Pulumavi and Gotamiputa Siri Yana as is evident from its wide flan and visibly high silver
Satakani (Chapter 11, coins 12 and 13). A fragment of a content, dates to an early period (c. 9th–10th century
silver coin of Vasithiputa Siva Siri Pulumavi has also ad) in the scheme of the evolution of the series (Deyell
been reportedly found at Paithan (Dhopate 2004: 51, 1990: 115–123; Chapter 11, coin 18).
coin 1).

Late Medieval and Islamic coins

Post-Satavahana coins
Late Medieval and Islamic coins are represented in the
The post-Satavahana to Early Medieval period at excavated finds by only three specimens: one attrib-
Paithan is represented by very few coins. Coins of the uted to the Bahmani Sultans of Gulbarga (coin 53),
Western Kshatrapas are not represented amongst those one to the Mughal Empire (coin 54) and the third to
excavated, except for a fragment, coin 33, which can the Qutb Shahi Sultanate of Golkonda (coin 55). The
only tentatively be attributed to one of these rulers and Bahmani coin was struck during the rule of Ahmed
should therefore be datable to the 2nd to 4th century Shah II, who is also known as ‘Wali’ (ad 1435–1457), a
ad. A few Western Kshatrapa specimens exist in the noted patron of Sufism during whose reign Khuldabad
Patil collection (Chapter 11). Three coins, struck in the and Daulatabad, both located in the vicinity of Paithan,
name of Kumara Gupta (c. ad 415–455), but belong- flourished as centres of Sufi worship. The coin (Goron
ing to the posthumous debased series, are also known and Goenka 2001: type BH88) has Arabic inscriptions
from the Patil collection (Chapter 11, coins 15, 16, 17). on both sides alluding to the Sultan’s political and reli-
Amongst the excavated finds presented here, coin 34 is gious prowess. The Mughal coin can be attributed to
particularly noteworthy. This coin is in a bad state of the Emperor Aurangzeb (ad 1658–1707) and is a copper
preservation, but distinct traces of a trident symbol sur- Fulus struck at the mint of Surat. This is evident from
rounded by crude Brahmi letters and a circular border remnants of the reverse inscription (the Persian letter
of interrupted dots are apparent. These aspects help us ‘Te’ clearly seen below parts of the word Sanah, refer-
tentatively to identify it as a degraded copy of the silver ring to the emperor’s regnal year). The Qutb Shahi coin
coins of the Sarvva-Bhattaraka and Maitraka dynasties bears the inscription Khatama b’il-Khair wa al-Sa’adat
of peninsular Gujarat. The dating of these coins with and is anonymous. It also gives the mint name on the
reference to the Gupta conquest of Gujarat and from the reverse as Darb Dar al-Saltanat Haidarabad. Coins
analytical perspective of ‘type vis-à-vis metallic deg- of this type (Goron and Goenka 2001: type Q73) were
radation’ has recently been discussed in some detail struck during the reigns of the last two Sultans, namely
(Maheshwari and Rath 1996: 190–194; Maheshwari Abdullah and Abu’l Hasan Qutb Shah (ad 1626–1672
and Rath 1998: 310–311; Bhandare 2006: 98–100). If we and ad 1672–1684, respectively).
accept the chronological schema outlined in these con- The latest coin amongst the excavated finds is a 1/12th
tributions, the extent of deterioration in type character- Anna of British India of 1897, bearing the effigy of Queen
istics might suggest that the Paithan specimen dates to Victoria.
about the 7th century ad. Degraded Maitraka coins of The low numbers of Late Medieval and Islamic
a similar type have also been illustrated by Mitchiner coins that were found in the excavations are probably
(1978: 609, nos. 4897–98). due to the fact that very few layers of this period were
Another Early Medieval coin, this time found in excavated during the course of the four seasons of field-
Period 4 layers, is coin 35, which belongs to the ‘Sri work.
Vigra’ variety of the ‘Indo-Sasanian’ genre of debased
silver ‘Dramma’ coins. These coins bear the abbrevi-
ated form of the name ‘Vigrahapala’ and are referred to
in many north Indian inscriptions as Vigrahapaliya
Derek Kennet

Appendix 9.1 
Coin Deposition in the Temple Foundation
Deposits and Coin Residuality
As has been shown above, 39 coins were recovered This data and the calculation of coins per cubic
from Trench A, 18 of which came from temple founda- metre, sherds per cubic metre and the ratio of coins to
tion deposits, 14 from the South Temple and 4 from sherds are presented in Tables 9.4 and 9.5.
the North Temple (Table 9.3). The fact that so many On the basis of this analysis, it is possible to make
of the coins found in Trench A come from the foun- the following observations: (1) the two Period 3-temple
dation deposits raises the possibility that these coins assemblages are at the higher end of the range of coins
may have been deliberately added to the foundations per cubic metre; (2) the two Period 3-temple assem-
during construction of the temples. blages are at the middle to lower end of the range of
However, it is not simple to demonstrate this sherds per cubic metre; (3) the two Period 3-temple
because a far greater amount of earth was excavated assemblages are at the lower end of the range of ratios
from the temple foundation deposits than from any of coins to sherds (Fig. 9.6) and (4) the two Period
other type of deposit in Trench A, and this may explain 3-temple assemblages are close or adjacent to each
the high number of coins from the foundations. Obvi- other in each of the three rankings in Table 9.5.
ously, the number of coins needs to be considered in
relation to the quantity of earth excavated if an accu- Table 9.4: The number of coins, sherds and the amount of earth
excavated from a sample of period-sequences from across the site
rate impression is to be gained of the relative pattern of
(Trench A Test Pit 1 (TP1): Periods 1–4; Trench B (TrB): Periods 1, 2
coin deposition. and 4; Trench D1 (TrD1): Period 1; Trench D2 (TrD2): Periods 1 and
In order to investigate this, the amount of exca- 2; Trench F (TrF): Periods 1, 2 and 3; Trench A Garbha Griha of
vated earth was calculated for a sample of 15 period- the North Temple (GGN): Period 3-temple; and Trench A Test Pit 5
sequences from trenches and test pits across the site, (TP5): Period 3-temple).
including two from foundation deposits. This data
allows the density of coins per cubic metre to be com- Period Coins Sherds Cubic
pared. The number of sherds was also recorded so metres

that comparisons of the ratio of coins to pottery sherds TrB-1 1 5 10,602 5.15
could also be made from the same period-sequences. A TrB-2 2 1 2,646 4.11
‘period-sequence’ is the sequence of layers from a par- TrB-4 4 2 15,832 16.64
ticular period in a particular trench.
TP5-3T 3T 8 10,184 8.60
TrF-1 1 0 5,787 1.49
Table 9.3: The coins from Trench A summarized by period.
TrF-2 2 0 6,423 4.46
TrF-3 3 0 1,286 3.53
Period Total coins % of Trench A
coins TrD1-1 1 2 6,772 3.53

Unphased 3 7.7% TrD2-1 1 0 7,658 2.54

1 1 2.6% TrD2-2 2 3 4,382 4.92

2 1 2.6% TP1-1 1 1 3,500 0.32

3 4 10.3% TP1-2 2 1 17,516 4.19

3-temple 18 46.2% TP1-3 3 2 6,263 2.42

4 12 30.8% TP1-4 4 0 837 2.02

Total 39 GGN-3T 3T 2 3,088 2.45

230   Derek Kennet

These observations suggest that the Period 3-temple Some of the coins in the temple foundation depos-
deposits have closely related depositional histories. its almost certainly found their way into these layers
Whilst they both contain below-average amounts of accidentally, along with redeposited earth containing
anthropogenic material more generally (as indicated by older, anthropogenic material from elsewhere at the
the low amount of sherds per cubic metre), they both site. However, the analysis presented here strongly
contain a higher-than-average ratio of coins per cubic suggests that a significant proportion of the coins in
metre. This is clearly reflected in the fact that they have the temple foundation deposits were deliberately
amongst the highest ratios of coins to sherds at the site added to those deposits at the time the temples were
and, perhaps most importantly, a much higher ratio of being constructed, possibly as part of a formal or infor-
coins to cubic metre than Period 3 deposits generally (Fig. mal ritual.
9.6).

Table 9.5: Coins per cubic metre, sherds per cubic metre and the ratio of coins to sherds from a sample of period-sequences from
across the site (based on the data in Table 9.4). The period-sequences from the temple deposits are indicated in bold.

Coins per m³ Sherds per m³ Sherds per coin

TP1–1 3.13 TP1–1 10,938 TP5–3T 1,273

TrB–1 0.97 TP1–2 4,180 TrD2–2 1,461
TP5–3T 0.93 TrF–1 3,884 GGN–3T 1,544
TP1–3 0.83 TrD2–1 3,015 TrB–1 2,120
GGN–3T 0.82 TP1–3 2,588 TrB–2 2,646
TrD2–2 0.61 TrB–1 2,059 TP1–3 3,132
TrD1–1 0.57 TrD1–1 1,918 TrD1–1 3,386
TrB–2 0.24 TrF–2 1,440 TP1–1 3,500
TP1–2 0.24 GGN–3T 1,260 TrB–4 7,916
TrB–4 0.12 TP5–3T 1,184 TP1–2 17,516
TrF–1 0.00 TrB–4 951 TrF–1 -
TrF–2 0.00 TrD2–2 891 TrF–2 -
TrF–3 0.00 TrB–2 644 TrF–3 -
TrD2–1 0.00 TP1–4 414 TrD2–1 -
TP1–4 0.00 TrF–3 364 TP1–4 -

Fig. 9.6: Rank analysis of the ratio of coins to sherds from a sample of period-sequences from across the site. The two Period 3-temple
period-sequences (TP5-3T and GGN-3T) are at the lower end of the range.
 Appendix 9.1 Coin Deposition in the Temple Foundation Deposits and Coin Residuality    231

This provides useful insights into two things: firstly size, shape and fabric – this indicates that a significant
the practice of deliberately depositing coins in temple proportion of the coins that were put into the founda-
foundation deposits during temple construction and tion deposits when the temples were constructed were
secondly the type of coins that were in circulation at Satavahana coins that were already several hundred
the time the temples were constructed. Amongst the 18 years old. This adds weight to a suggestion that has
coins from the temple foundations, three are pre-Sata- already been made by other scholars based on differ-
vahana, seven are Satavahana and six are unidentifi- ent evidence that Satavahana coins continued to circu-
able due to their poor a state of preservation. In addi- late widely into much later periods (e.g. Shastri 1992:
tion, there are two post-Satavahana coins, both from 142 n. 23, 144).
the South Temple deposits. One may be a fragment Apart from being of considerable interest
of a Western Kshatrapa coin datable to 2nd to the 4th for the political and economic history of the Late
century ad (No. 33), and the second a degraded copy of Historic/Early Medieval period, this conclusion has
the silver coins of the Sarvva-Bhattaraka and Maitraka profound implications for the dating of archaeological
dynasties, which can be dated to the 7th century (No. layers that may have been attributed to the Satavahana
34). The latter coin provides an important terminus period by excavators on the basis of the coins con-
post quem for the construction of the South Temple. If tained within them, but which may in fact have been
most of the unidentified coins mentioned above are of deposited much later.
Satavahana date – as is supposed on the basis of their
M. Kasturi Bai

Chapter 10 
Summary of the Finds from the 1996 and 1997
Excavations

The excavations carried out during the 1996 and 1997 illegible symbol on the obverse. Due to the bad con-
seasons in Trench C yielded a number of small finds dition of the obverse, it cannot be attributed to a par-
that were stored in the office of the Aurangabad Circle of ticular king, although it clearly belongs to the Satava-
the ASI. Because these excavations were carried out hana period. There are also round copper alloy coins
according to a different methodology and because (Field Nos. 225, 361 and 375), which have an elephant
detailed trench reports and a site matrix are not avail- on the obverse and a Ujjain symbol on the reverse.
able, a study of these finds is not included in this pub- Another round coin (Field No. 455) from the surface
lication. However, based on a list of small finds made contains a horse on the obverse and a crude represen-
during the 1997 season, the finds can be categorised tation of a bow and arrow. Mention may also be made
into coins, beads, ear-studs, pendants of various mate- of a terracotta seal obtained from the excavation, which
rials, hair pins, points of bone and ivory, household has the reverse motif of a thunderbolt and arrow and
objects of stone, terracotta, iron etc., toy objects such which can be attributed to the Kshatrapas. Two copper
as marbles, discs, hopscotch, toy wheels, toys, bangles coins that were obtained from the surface belong to
and rings etc. the Islamic period. In summary, the coin collection of
A total of 91 coins were recorded during the 1996– the 1996–1997 seasons contains coins belonging to the
1997 seasons, including 18 coins from the surface; two post-Mauryan, Satavahana and Kshatrapa periods in
of them are lead, the rest being copper alloy. Most of addition to two coins of the Muslim period.
the copper coins are badly corroded and encrusted, Other small finds were also found during the exca-
and sometimes, they are cracked and flaked. The vations, beads being the most frequent. They occur in
copper coins exhibit different shapes and are found various shapes and sizes and are made of materials such
in different sizes, whilst some are thick and squarish as carnelian, agate, chalcedony, jasper, crystal, glass,
or rectangular in shape. A few round or ovalish coins paste, shell and terracotta. The carnelian beads exhibit
are also found. Although corroded and cracked, some spherical, barrel-faceted, bicone barrel, pentagon cyl-
of the squarish coins (Field Nos. 480, 500, 774 and inder and disc shapes. One of the spherical beads has
849) are decorated with a hollow cross symbol on the etched decoration. Crystal beads are also spherical,
obverse and a blank reverse. The fabric, the shape, the barrel, faceted, barrel-eared and disc shaped. Both
condition and the symbols on the obverse compare banded and plain agate beads were found in small
well with the coins obtained from Period 1 of the 1998 quantities with long cylinder, barrel and disc shapes.
and 1999 seasons, which belong to the post-Mauryan Barrel-faceted beads of garnet and almond-shaped
but pre-Satavahana period. A small, squarish copper chalcedony were also found, as well as rectangular
coin (Field No. 429) has an Ujjain symbol, a sun and an and spherical beads of jasper. Among the shell beads,
arched hill on one side and is blank on the other side. spherical, cylindrical, barrel-eared and disc shapes
Some of the coins bear symbols like elephants, were recorded. A few paste beads of the barrel-eared
horses and possibly lions. One of the ovalish copper variety were also found. Spherical beads of chert and
alloy coins from the surface (Field No. 425) has an rectangular and oblong-shaped beads of lapis lazuli
elephant facing right, trunk uplifted with the extant were also present. Glass and terracotta beads are also
legend ‘…MAVISA’ on the obverse and a tree with very common. The shapes noted in glass are spherical,
big leaves in a railing on the reverse. This certainly short cylinder truncated, barrel-eared, bicone and a
belongs to one of the Pulumavi’s of the Satavahana bird shape. Arecanut is a predominant shape amongst
period. A similar reverse symbol is found on one large the terracotta beads. The other shapes in terracotta
lead coin (Field No. 282), which has an indistinct and are globular, squat globular, hemispherical, conical

https://doi.org/10.1515/9783110653540-010
 Chapter 10 Summary of the Finds from the 1996 and 1997 Excavations    233

with flat shaped, bicone, long cylindrical, rectangu- Broken elephant and bird-like terracotta objects, a
lar eared and grooved, flat based with conical profile, plaque fragment of a god and a kaolin fragment were
barrel and bell shaped. Pulley-shaped ear-studs and also found. A few bangles of terracotta and glass, along
flat-based pendants with a conical profile were found with a few finger-rings, were noticed.
in small numbers. Among the iron objects, nails predominate, a total
Small cylinders used as stoppers, skin rubbers of 96 being found, and occur in different sizes. These
with a herringbone pattern and cylindrical ceramic are mostly broken, corroded and encrusted. Other iron
weights were also recorded. A number of pottery discs, objects include tanged, thick arrowheads, needles,
hopscotch discs, toy wheels, spindle whorls and toys chisels and blades of knives.
are the play objects found at the site. Among the bone
objects, mention may be made of pins, double-ended
points, discs, carved bangles and a stand with a knob-
like design.
M. Kasturi Bai

Chapter 11 
Objects Displayed in the Balasaheb Patil
Government Museum

The Balasaheb Patil Government Museum at Paithan Of course, it should be noted that the precise
contains the private antiquity collection of the late find spots of these objects is unknown. Although
Shri Balasaheb Patil, after whom it is named. He most objects are believed to come from Paithan and
was a local dignitary and coin and art collector who its immediate surroundings, it is impossible to be
acquired objects randomly from different mounds of certain of their provenance.
Paithan and the surrounding area over a period of
many years and donated them to the State Depart-
ment of Archaeology, Maharashtra. The collection
holds a variety of objects of different periods, which
Coins
are of interest to the general public as well as to his-
Coins displayed in the museum can be categorized
torians and archaeologists. The collection has never
as punch-marked (8), uninscribed (1), inscribed (2)
been catalogued, comprehensively studied or pub-
and portrait coins of the Satavahana kings (2) and
lished, although objects have been mentioned in
Kshatrapa kings (8).
print. Some brief notes on some of the objects from
the collection were also published by R. S. Morwan-
chikar in his book on the history and archaeology
Punch-marked coins
of Paithan (Morwanchikar 1985: 137–154, plates II–
XVIII), which are further discussed below.
Seven of the punch-marked coins are made of silver and
The present report on the collection is based solely
one of copper. These are squarish, rectangular, round,
on a viewing of the museum display cases where only
ovalish and irregular in shape. The symbol numbers
a part of the collection was on display, whilst the rest
used in the descriptions of the punch-marked coins
was kept in store and was not seen. For this reason, the
below are the numbers given by Gupta (1960: 39–61,
weights of coins and other details could not be measured
pl. I–II), based on symbols or combinations of symbols
due to lack of access. Some of the coins and two of the ter-
found on silver punch-marked coins of the Amaravati
racotta moulds have already been studied and published
Hoard.
by other scholars (see below).
The objects of archaeological interest are coins; 1. Silver, squarish with clipped edges, worn out. Fig.
moulds, medals and bullae of terracotta; figurines 11.1.
of kaolin and terracotta; stone sculptures; different Size: 1.435 × 1.246 × 0.241 cm.
types of pottery; household objects of various mate- Obv: Five symbols; solar symbol (Symbol 1), sadachakra
rials such as pottery, kaolin, stone and metals; and (Symbol 7; i.e. three arrows and three ovals alternately
some miscellaneous objects. There is also Proto His- placed around a circle with a conspicuous dot in the
toric and Early Historic pottery. The complete spouted centre), elephant facing right (Symbol 71) and two other
vase of Jorwe ware (Fig. 11.6, 1) is an interesting find indistinct symbols.
belonging to the Chalcolithic period and recalls Jorwe Rev: Worn out, two indistinct marks are found.
pottery found at Nevasa (Sankalia et al. 1960: 213, fig.
2. Silver, rectangular with clipped edges. Fig. 11.1.
85b), although the painted decorations are different in
Size: 1.986 × 1.181 × 0.183 cm.
this case. Some of the Red Polished spouted vases and
Obv: Five symbols; solar symbol (Symbol 1), sad-
a broken kaolin pot with astamangala symbols are also
achakra (Symbol 7), elephant facing right (Symbol 71)
worth mentioning.
and two other indistinct symbols.

https://doi.org/10.1515/9783110653540-011
 Chapter 11 Objects Displayed in the Balasaheb Patil Government Museum    235

Rev: Four symbols; solar symbol (Symbol 1), sadachakra Obv: Four symbols type; solar symbol (Symbol 1), sad-
and two other worn out symbols. achakra (Symbol 7), animal (?), and one unidentified
symbol.
3. Silver, roughly rectangular with clipped edges. Fig.
Rev: Fish in a tank (?), other indistinct symbols.
11.1.
Size: 2.182 × 1.139 × 0.201 cm.
Obv: Five symbol type: solar symbol (Symbol 1), sad-
Uninscribed coins
achakra (Symbol 23, i.e. three arrows and three taurines
alternately placed around a circle with a conspicu-
A single copper uninscribed coin was found in the col-
ous dot in the centre), crescent on a three-arched hill
lection.
(Symbol 48), bull(?) facing right (Symbol 62), cadu-
ceus (Symbol 84). 9. Copper alloy, squarish, good condition. Fig. 11.2.
Rev: Caduceus with some worn out and indistinct marks. Size: 1.540 × 1.395 × 0.341 cm.
Obv: Elephant facing to left as if walking towards a
4. Silver, round, worn out. Fig. 11.1.
triangle-headed standard, trunk hanging with a tusk,
Size: 2.021 × 1.765 × 0.126 cm.
a swastika and inversed taurine above the animal,
Obv: Five symbol type: solar symbol (Symbol 1), hare
ladder below.
facing right standing on five hills (Symbol 43) or two
Rev: Hollow cross at the bottom left, a crescent on a
hills (Symbol 99), bull facing right (Symbol 62), ele-
three-arched hill above it and a tree in railing to its right.
phant facing right (Symbol 71), heavily worn out and
Note: This coin can be compared with Mauryan coins
indistinct symbol perhaps a fish symbol.
of similar type recovered from middle Gangetic valley
Rev: Fish in tank (?) with some indistinct marks.
as stated by Mitchiner (1972: nos. 18–26). The symbols
5. Silver, rectangular with clipped edges. Fig. 11.1. and the size ranges are the same in both areas.
Size: 1.662 × 1.300 × 0.161 cm.
Obv: Five symbol type; solar symbol (Symbol 1), a wavy
line (Symbol 80), three other indistinct symbols are Inscribed coins
visible.
Rev: Traces of two symbols in square and circular There are two inscribed circular coins of elephant/tree
incuses, a few worn out marks. type, one each in copper and lead.

6. Silver, square with clipped edges, slightly worn out. 10. Copper alloy, round, slightly worn out. Fig. 11.2.
Fig. 11.1. Size: 1.880 × 0.330 cm.
Size: 1.841 × 1.784 × 0.133 cm. Obv: Elephant walking to the right, trunk uplifted, four
Obv: Four symbol type; solar symbol (Symbol 1), sad- circles placed together above the animal, two Brahmi
achakra (Symbol 6. i.e. three arrows and three taurines letters in front of mouth reads as ‘SI RI------’ and above
in an oval, alternately placed around a circle with a on the back of the animal ‘RA JNO ------’ faintly visible.
conspicuous dot in the centre), six-arched hill (Symbol Rev: Branched tree in railing with oval leaves, srivatsa
46), some other symbol, perhaps a triangular headed symbol to its left and an indistinct symbol (arched
standard with four taurines (Symbol 61). hill?) to its right.
Rev: Worn out, crescent on a three-arched hill (Symbol
11. Lead, round, worn out. Fig. 11.2.
136) and two other indistinct symbols.
Size: 2.289 × 0.320 cm.
7. Silver, irregularly clipped square, worn out. Fig. 11.1. Obv: Elephant walking right, trunk uplifted with four
Size: 1.501 × 1.279 × 0.346 cm. unconnected circles with a crescent on one of the circles
Obv: Five symbol type; Solar symbol (Symbol 1), sad- above the animal, the legend along the flan is faintly
achakra (Symbol 7), crescent on a three-arched hill visible and reads ‘SA RA JNO SI RI SA TA---------’.
(symbol.48) (?) and two other indistinct symbols, cadu- Rev: A tree with five oval leaves is faintly visible.
ceus(?). Note: This can be compared with a circular copper coin at
Rev: Crescent on arched hill (Symbol 48), some indistinct Nevasa having the legend ‘RA NO SI RI SA (TA KA) NI SA’
marks. and five oval leaves in a railing of six compartments on the
reverse (Sankalia et al. 1960: 180–181, fig. 78, 2, pl. II, 4).
8. Copper alloy, squarish, slightly worn out. Fig. 11.2.
The same type of coin was also issued in lead at Nevasa.
Size: 1.441 × 1.260 × 0.188 cm.
236   M. Kasturi Bai

Fig. 11.1: Coins from the Balasaheb Patil Collection (1–7).

 Chapter 11 Objects Displayed in the Balasaheb Patil Government Museum    237

definitely assigned to Satakarni II or Gautami Putra

Satakarni etc.

Portrait coins of the Satavahana kings

There are two silver portrait coins in the collection.

12. Silver, round, good condition. Fig. 11.3.

Size: 1.563 × 1.535 × 0.175 cm.
Obv: Portrait of a king in the centre, inscription along
the flan reads ‘S I R I Y A N A S A T A K A N I S A R A J N
O’ then chudamani begins, again the legend reads ‘G O
T A M I P U T A S A’.
Rev: Ujjain symbol to the left and six-arched hill to
the right; both are surmounted by crescent, a rayed
sun symbol in between the two on top and a wavy line
below the whole device within a dotted circle, which
is partly off the flan. A vertical line passing through the
Ujjain symbol is found on the flan. The legend starts
from about 5 o’clock and reads ‘(KU) H A R I YA NA HA
TA KA NI KU ARA….’.
Note: This coin has already been published by Ranade
(1964: 98, pl. I, 8) but his reading on the reverse is slightly
different. The initial letters preceding ‘YA N A’ he reads
as ‘SI RI’, but they are ‘(KU) HI RI’ and also ‘SA’ instead
of ‘HA’.

13. Silver, round, good condition. Fig. 11.3.

Size: 1.592 × 1.504 × 0.160 cm.
Obv: Portrait of a king in the centre, legend along the
flan reading ‘SI RI PU DU MA VI SA (RA JNO)’, then
chudamani begins and afterwards the legend reads
‘(VA) S I T H I PU TA SA’. The upper portion of the letters
Fig. 11.2: Coins from the Balasaheb Patil Collection (8–11). ‘RAJNO’ and ‘VA’ are cut off.
Rev: The devices on the reverse are the same as that of the
The above two coins of elephant/tree type may be attrib- Yajnasiri Satakarni coin mentioned above, but here, an
uted to King Satakarni but were issued in different arched hill is effaced except for the crescent above within
metals. Both the Paithan and Nevasa coins of this type a partly visible dotted circle. The legend along the flan
may belong to the same King Satakarni but cannot be reads ‘…….T(I) RU PU DU SHA VI KU A RA HA NA…….’.

Fig. 11.3: Coins from the Balasaheb Patil Collection (12–13).

238   M. Kasturi Bai

Note: Here the letter that seems to be ‘SHA’ is ‘MA’. The terstruck by Gotamiputa Siri Satakani, Satavahana
top portion of ‘MA’ is given in straight lines instead king (c. ad 70–85). Satavahana devices of an ‘Ujjain’
of cursive so it looks like ‘SHA’. It may also reflect the symbol and a ‘three-arched hill surmounted by cres-
influence on the scribe of scribbling the legends on the cent’ are seen on reverse and obverse, respectively.
western Kshatrapa coins. Fig. 11.4.

15, 16, 17. Debased ‘peacock’-type drachm struck post-

The legend on the reverse of the above two coins reads
humously in the name of Kumara Gupta, the Gupta
‘HA’ for ‘SA’, the sixth case ending suffix ‘KU’ in each
emperor (c. ad 415–455).
word and ‘TI RU’ as elsewhere on portrait coins of
Obverse: Profile bust of the king facing to the right;
Satavahanas as has been pointed out by Sarma (2000:
reverse: stylised peacock in the centre, surrounded
79–80).
by Brahmi legend ‘/Parama Bhagavata Rajadhiraja
Sri Kumara Gupta Mahendradityah/’, which is seen in
parts on these coins. Fig. 11.4.
Kshatrapa coins
18. Silver Drachm (Dramma) of the ‘Indo-Sasanian’
There are eight portrait coins of Kshatrapa kings (six series, early ‘Gadhaiya’ variety, c. 10–12th century ad. Fig.
silver and two copper) in the collection. But since they 11.4.
are worn out and illegible, and due to lack of access
and good photographs, they are not included in this
study.
Moulds
There are three interesting terracotta moulds of circu-
Further coins
lar shape. One of them contains three separate incuses
interconnected by a channel and measures 48.84 mm
The following five coins were later identified and
in diameter and 9.67 mm in thickness. It consisted of
described from photographs by S. Bhandare.
devices in negative on the obverse of the King Gau-
14. Silver coin of Nahapana, Western Ksatrapa (Ksha- tamiputra Yajnasri Satakarni depicting the three
harata) ruler of Gujarat-Saurashtra (c. 33–78 AD), coun- stages of life of the king (Fig. 11.5, 1). The other side of

Fig. 11.4: Coins from the Balasaheb Patil Collection (14–18).

 Chapter 11 Objects Displayed in the Balasaheb Patil Government Museum    239

the mould is flat and blank. It was earlier reported by played in the museum are master works of the crafts-
Gokhale (1995: 138–139) as a mould belonging to the men from the Satavahana/Kshtrapa times and of later
three stages of king Gautamiputra Yajnasri Satakarni. Early Historic times. They are all of double-moulded
The second terracotta mould referred to here is type. The human figurines depicted in Fig. 11.6, 8 are
circular in shape with a channel on both the surfaces the heads of the male and female figurines. A number
having symbols (Fig. 11.5, 2). The third terracotta mould of other miniature figurines were also found, including
has a deep incuse without a channel having the reverse a couple, a Chinese man with a beard and the heads of
motif of a Satavahana king (Fig. 11.5, 3). It was not pos- monks (saints), all showing different hairstyles and
sible to study either of these moulds in more detail due adorned with a variety of ornaments. A complete kaolin
to lack of access. One of the clay moulds of the collec- figurine in a sitting posture (Fig. 11.6, 7) delineates
tion that was not on display has been studied and iden- excellent artistic features that can be compared with a
tified as a proof piece of a silver coin of Vasisthiputra fragment of a kaolin figurine from the Paithan excava-
Satakarni (Shastri 1991). tion (Chapter 8, No. 111). The animals include heads
of goats and horses and the torso of a bull with wheel
decoration on its mane (seat).

Art Objects Terracotta objects of human (Fig. 11.6, 9) and

animal figures are also displayed in the museum. The
dark red colour terracotta faces of a male figurine in a
The collection contains a variety of kaolin representa-
laughing posture are interesting. The other terracotta
tions of humans and animals. The kaolin objects dis-

Fig. 11.5: Coin moulds from the Balasaheb Patil Collection (1–3).
240   M. Kasturi Bai

Fig. 11.6: Pottery and terracotta objects from the Balasaheb Patil Collection (1–9).
 Chapter 11 Objects Displayed in the Balasaheb Patil Government Museum    241

Fig. 11.7: Figurines from Paithan (from Morwanchikar 1985: pls X–XIII).
242   M. Kasturi Bai

objects include the bust of a human being, a full figure The ear ornaments of terracotta are crescent-
of a man, a monk’s head, two terracotta plaques of shaped (Fig. 11.6, 2), which are broad in the middle and
human beings and animals such as monkeys, cocks, taper at the ends. Ear ornaments of similar shape were
elephants, a royal horse, Nandi and a cart. The head recovered from Bhokardan (Deo and Gupte 1974: pl.
of a lady depicted in a side pose is a very interesting LIV, 11).
piece, having parallels with Roman art. The terracotta There are full and broken bangles of shell and
bulls with portraits of kings are also very interesting ivory displayed in the museum and manufacturing
objects. There are a few terracotta medals of circular waste from conch shells. Among these, there are broad
shape depicting a sitting stag and a Nandi within a and thin varieties and plain and decorated bangles. A
floral motif, which are interesting finds (Fig. 11.6, 5–6). broad bangle having a deep groove with projections on
One of the ivory objects of an umbrella shape is a either side is interesting. Ivory bangles are decorated
unique piece of art with decoration executed on the with circles, diamonds and some other decorations.
body. Perforation was made in the middle of the body There are a number of terracotta bangles of various
after making the decoration. There are other interest- sizes and decorations (Fig. 11.6, 3–4). A few broad
ing ivory objects like a stupa-like motif and one with bangles contain diamond-shaped cuttings and other
nut-powder-box shape. designs. A few finger-rings were also found. There are
kohl sticks and hairpins of ivory and also bone points.
Among the ivory objects, different sizes of dices with

Ornaments different depictions are interesting finds.

The ornaments displayed in the museum include

beads, ear-studs, pendants, bangles and rings of dif- Objects published by Mowanchikar
ferent materials in a variety of shapes, along with
other ornaments. The beads are made up of semi-pre- Morwanchikar published a number of objects that are
cious stones like agate, carnelian, chalcedony, garnet, also apparently from the Balasaheb Patil Collection
jasper, rock crystal, steatite (paste), shell, glass and (1985: chapter 7). These include querns, skin rubbers,
terracotta. Among the carnelian beads, the etched bullae (at least one of which appears to be based on
example of cylindrical shape with a circular section a Roman coin; Morwanchikar 1985: pl VII a), medal-
is a remarkable find, which can be dated to between lions, sealings, coin moulds, and a number of kaolin,
approximately 100 bc and ad 100. Other carnelian terracotta and stone figurines that are not amongst
beads are spherical, barrel shaped, cylindrical and those shown in Fig. 11.6 (Morwanchikar 1985: 143–153,
rectangular. There is also a top-shaped pendant in this pls X–XIII). Twelve of these objects are illustrated in
material. The crystal beads are spherical, rectangular- Fig. 11.7, one of which (2) is very similar to a kaolin
cylindrical, barrel-faceted and disc shaped (tablet). head from Ter in the British Museum, dated there
Beads of shell and ivory are found here in different to the 4th century ad on stylistic grounds (British
shapes and sizes. Beads that are squarish, flat, barrel Museum registration number 1958,1017.1) suggesting
shaped, truncated eared, cylindrical-circular, spheri- that the Satavahana date proposed by Morwanchikar
cal, star shaped, oval shaped with a flat bottom and for this object may be too early (1985: 144). As with
pointed top are the most important shapes. The terracotta all objects in the Balasaheb Patil Collection, it is not
beads are predominantly of arecanut shape. The others absolutely certain that they actually originated in
are gadrooned, eared and tablet shaped. Paithan as it is possible they were purchased from
dealers who brought them from other localities.
Part IV
Environmental Evidence
Derek Kennet

Introductory Remarks on the Environmental

Evidence
Part IV of this volume presents two specialist reports Near East and Europe, to be relatively low by global
on the environmental evidence from the excavations standards. This may be a regional phenomenon related
(archaeobotanical and faunal). to diet or culture (although not bone survival as the
As with the collection and study of the artefacts, a bone that was collected was all in good condition), but
key aim of the present project was to use a systematic because data on density of occurrence are rarely pub-
and quantified approach to both the collection and lished from sites anywhere in the world, it is impos-
study of environmental evidence. This included, on sible to make reliable comparisons.
the advice of Dr Mark Nesbitt of Kew Gardens, London, The results of these collection and study strategies
the construction of a ‘Siraf-type’ flotation machine are two chapters that make an outstanding and ground-
that was set up on the banks of the Godavari and breaking contribution to our understanding of the
allowed the processing of very large quantities of flota- development of Paithan and its agricultural economy
tion samples from key trenches and soundings. The and also make the most substantial contribution to
advantage of this approach is manifest in the assem- our understanding of Early Historic and Early Medieval
blage of more than 8,300 carbonized seeds that form agriculture that has yet been published from anywhere
the basis of the archaeobotanical study in Chapter 12. in India. These studies take the first steps in plugging
This is the largest archaeobotanical assemblage from a problematic and gaping hole in our knowledge of the
any Early Historic or Medieval site that has so far been economy of these periods that has been lamented by
published from anywhere in India. At the same time, a almost every scholar who has written on the subject.
rigorous programme of sieving of excavated earth was The results of the two studies presented in Part IV
implemented. This resulted in the collection of a good are further discussed and contextualised in the con-
faunal assemblage that provides insights into some clusions (Chapter 14).
very useful historical trends in animal husbandry at It should be noted that no systematic environmen-
the site. Having said this, the general density of animal tal sampling was carried out in 1996 or 1997, but a small
bone by cubic metre of excavated earth was noted by pilot project of bucket flotation recovered some plant
the excavators, many of whom have excavated in the remains in 1997.

https://doi.org/10.1515/9783110653540-012
Dorian Q. Fuller

Chapter 12 
Archaeobotany

Introduction tices and in some cases dietary preferences (e.g. Jones

1991; Colledge 1998, 2001; Pearsall 2000; Stevens 2003;
Wilkinson and Stevens 2003). One of the aims of this
Archaeobotany is the study of plant remains from
report is to show how quantification can contribute to
archaeological contexts, with aims to contribute to
our understanding of the Early Historic plant economy
archaeological interpretations of past economies,
on the peninsula.
foodways and in some cases environmental change.
Archaeobotanical remains also provide evidence for
the evolution of domesticated plants that have changed
in the past as a result of human activities, in particu- Samples and their distribution
lar cultivation. As is most often the case, the present
report will focus on seed and fruit remains (carpologi- Most samples were collected systematically in the
cal analysis), which provides evidence mainly about 1998 and 1999 seasons and processed with a flotation
food plants and agricultural systems. Other archaeo- machine that was set-up at the site. A few samples had
botanical datasets, such as wood charcoal and some been processed in 1997 by simple bucket flotation using
charred parenchymatous remains (probably from the wash-over method (for details on various flota-
tubers), are also present at Paithan and would repay tion methods, see Pearsall 2000). For many samples,
future study. Archaeobotanical studies have a long initial sorting was carried out in the field, supervised
history in India and have become increasingly system- by Peter Rowley-Conwy, and the separated seeds and
atic in terms of methods of recovery and sophistication non-wood material were then exported to the UK for
of analytical and interpretative methodologies over further analysis. In addition, some unsorted flots were
the past few decades (Weber 2001; Fuller 2002). Nev- sorted in full in the UK. As such, the latter could be
ertheless, the archaeobotanical record of India (and separated in a microscopy laboratory with rather
Pakistan) remains inconsistent. There are major vari- more care. Seed density and diversity may have been
ations between periods and regions that have been affected by sampling method in these samples. As a
studied, with generally more evidence being available result, sorted samples vary greatly in richness, and it
from the Neolithic and Chalcolithic periods, and often is not possible to consider how sample richness might
from regions that were foci of civilizations, such as relate to past activity patterns; although it might repay
the Indus and Ganges valleys, although the Southern further analysis to consider the impact of sampling
Deccan is also well sampled (see Kajale 1991; Saraswat and sorting procedures on data, it is beyond the scope
1992; Fuller 2002, 2006a). Historical periods are gener- of this interim report to do. Despite these concerns, the
ally more poorly represented, while Medieval and more overall sample size from Paithan is considerable and
recent times are almost entirely unsampled. Paithan, allows us to consider patterns through time at the site
as a well-sampled site of Early Historic through Early and provides a firm basis for comparison to the rather
Medieval times, is therefore of some significance. more limited published Early Historic datasets from
There are differences in the degree to which remains this region.
are counted and analysed quantitatively. In many pub- The raw data are presented in Table 12.1. The
lished reports, only the presence of species is reported maximum seed count in one sample is 1,124 items, the
from individual samples or even site phases. A large minimum is 2 and the average is around 100 items per
body of research worldwide indicates that statistically sample. Forty samples contain more than 30 items and
significant quantitative patterns exist in archaeobot- 20 samples contain more than 70. In terms of samples
anical assemblages, structured by a combination of that are reliably phased on the basis of stratigraphic and
archaeological formation processes and past human artefactual evidence, the three main phases are fairly
activities, such as crop processing, agricultural prac- evenly represented by sample numbers, although the
246   Dorian Q. Fuller

total seed counts are quite variable. We can therefore age of samples in which a species is present. Thus,
be more confident about the patterns in Period 1 and ubiquity is less prone to biases due to high versus low
Period 3. A number of samples, especially from bucket seed input into the record and preservation biases. It
flotation in 1997, come from contexts which could not can be regarded as an estimate of recurrence of discard
be phased. They are included in the total site patterns of species and for economic species recurrence of use.
and in a few cases contain well-preserved, identifiable Relative frequency may be more useful for assessing
remains. For most discussion, however, samples will how much of a species was used relative to another,
be considered only where they are assigned to phases: but this may be heavily biased by numerous factors,
especially differential preservation, which probably
Period 1 (Pre-Satavahana, perhaps the 3rd century bc
leads to higher rates of destruction for oily seeds as
through to about the 1st century bc): 17 samples and
opposed to dry starchy seeds (like cereal grains and
2,529 seeds studied.
dried pulses), and differential rates of accidental loss
Period 2 (Satavahana period and later, probably 1st during processing (e.g. more hulled cereals and pod-
century bc/ad–4th/5th century ad or perhaps earlier): 21 threshing pulses versus free-threshing types). For
studied samples, but only 762 seeds identified. weed seeds which are likely to be incidental inclusions
in crop-processing waste, seed productivity may also
Period 3 (construction of the earliest preserved Hindu
be important, and thus, prolific seed producers such
temples on the site, probably 4th/5th century ad to 8th
as Chenopodium may be over-represented in relation to
century ad or later): 17 samples and 1,128 seeds.
smaller seed producers such as Convovulaceae or Mal-
Period 4 (post-Medieval, 14th/15th to 17th century and vaceae. Despite such biases, we expect assemblages
later): only one archaeobotanical sample of this period prior to charring to have been structured by human
has been studied. Those remaining in the Paithan archive activities, including the quantities of species used and
would repay future analysis. processing practices, as the latter structure crop-chaff-
weed assemblages systematically (see e.g. Hillman
1984; Fuller 2002: 262–264; Stevens 2003; Harvey
General patterns in the evidence and Fuller 2005; Fuller and Weber 2005). Relative fre-
quencies might therefore include more of a signal of
Broad patterns in the data can be illustrated by ‘rela- activities, especially when considered on a sample-by-
tive frequencies’ and ‘ubiquities’, which are calcu- sample basis, whereas ubiquity will be less informative
lated for Periods 1–3 (based only on samples which except to indicate that taxa have been used.
were phased) and on site totals (including unphased In general terms, there is a higher occurrence of
samples). Relative frequency refers to the percentage of grain crops (cereals and pulses) during Paithan Period
identified seeds (including indeterminate seeds but not 1 than in later periods, with relative increases in the
indeterminate fragments) which are summed across all occurrence of rice, wheat and barley and winter pulses
the samples for a phase. Ubiquity refers to the percent- in Period 3. This is indicated by relative frequency (Fig.

Fig. 12.1: Relative frequency of major crop groups.

 Chapter 12 Archaeobotany    247

12.1), where it can be seen that through all phases, 1 towards other millets and a further decline in kodo
millets dominate the archaeobotanical assemblage. millet in Period 3 as other non-millet crops increased.
This can be considered further through the ubiquity of There is also a general tendency towards increas-
some of the same crop groups together with some sub- ing cultivation of winter crops in the later phases of
divisions of the millets (Fig. 12.2). In this, it can be seen the site, seen especially in winter pulses. This is of
that kodo millet (Paspalum scrobiculatum) is the most interest because the predominant rainfall pattern in
ubiquitous crop overall, followed by rice and the com- central Maharastra is the summer monsoon, and cul-
posite group of summer pulses. Kodo millet is also the tivation of winter crops poses some challenges and
dominant crop in terms of overall relative frequency, normally implies some form of artificial irrigation, as
accounting for about 45% of all identified seeds. It can noted by Kajale (1988). While winter crops had been
be noted, however, that the ubiquity of kodo millet major staples in the region in the Chalcolithic (Kajale
declines through the three periods, whereas the small 1991; Fuller 2003c), there are indications that agricul-
‘S/E/B’ millets, which are mainly Brachiaria ramosa ture went into some decline after the Jorwe period. The
and Setaria italica, increase, as do African millets. particular emphasis on an increase in winter pulses
Summer pulses show a similar trend and a Period 3 may imply increasing double-cropping of the same
increase in winter pulses is also clear. This suggests an lands. Pulses are naturally nitrogen-fixing plants and
overall trend away from kodo millet cultivation towards thus do not remove nitrates from the soil (Smartt 1990).
a diverse range of millets, including the introduced Nitrates are one of the limiting nutrients for plant
African millets and Setaria italica. It is also worth growth, and growing of legumes can help to replenish
noting the presence and persistence of the Southern ‘tired’ soils. Thus, soils which were heavily and annu-
Neolithic Ashmound millet, Brachiaria ramosa, as this ally cropped for staple cereals in the summer might be
is very much a relict and is, today, nearly extinct as a planted with pulses in the winter to help replenish
crop (Kimata et al. 2000) but was clearly a major staple for use again the next year. This has two likely impli-
in the earliest agriculture on the peninsula (Fuller et cations, which are not mutually exclusive: first, that
al. 2001, 2004). The decline of Brachiaria ramosa can cropping became increasingly continuous and fallow
therefore be hypothesized to post-date these Paithan times were reduced between reuse of the same plot of
samples. The trend towards more small millets in this land; and second, that cropping was expanded onto
long-embryo group is highlighted by a consideration more marginal soils. In general the black regur soils
of relative frequency (Fig. 12.3). Once again, increas- of much of Maharastra are quite rich and thus move
ing frequency is also seen in African millets and winter towards a regular rotation of cereals and pulses might
pulses, while the trend for summer pulses is less clear. imply use of locally poorer soils. Taken together, this
With rice, wheat and barley, no single directional trend evidence suggests moves to increased agricultural
can be seen, but there is a decline in Period 2 and an output, both through expansion (onto new lands and
increase that is more than a recovery in Period 3. On a wider range of soils) and intensification (by decreas-
the whole, we might summarize the trends as indicat- ing fallow periods and increasing repeated cropping
ing a move away from kodo millet dominance in Period on the same lands).

Fig. 12.2: Ubiquity of major crop groups and selected subdivisions of millets.
248   Dorian Q. Fuller

Table 12.1a: Archaeobotanical data from Paithan. P: present; T: total.

Period 1 1 2 2 3 3
Site Freq. P Ubiquity T P T P T P
total

Pulse %
Pulses items
Vigna cf. aconitifolia 5 0.06% 3 3.90% 7 0.12 0 0% 0 0% 2 0% 1 5% 3 0% 2 13%
(whole)
Vigna cf. aconitifolia (coty- 1 0.01% 1 1.30% 1 0% 1 6% 0 0% 0 0% 0 0% 0 0%
ledon)
Vigna cf. aconitifolia (frag- 1 0.01% 1 1.30% 0 0% 0 0% 1 0% 1 5% 0 0% 0 0%
ment)
Vigna cf. mungo (whole) 4 0.05% 3 3.90% 12 0.20 0 0% 0 0% 0 0% 0 0% 3 0% 2 13%
Vigna cf. mungo (cotyle- 8 0.10% 6 7.79% 0 0% 0 0% 2 0% 1 5% 4 0% 4 25%
don)
Vigna cf. mungo (frags) 0 0.00% 0 0.00% 0 0% 0 0% 0 0% 0 0% 0 0% 0 0%
Vigna cf. radiata (whole) 10 0.12% 7 9.09% 29 0.50 0 0% 0 0% 2 0% 2 10% 8 1% 5 31%
Vigna cf. radiata (cotyle- 19 0.23% 13 16.88% 4 0% 3 18% 5 1% 3 15% 8 1% 6 38%
don)
Vigna cf. radiata (frags) 0 0.00% 0 0.00% 0 0% 0 0% 0 0% 0 0% 0 0% 0 0%
Vigna sp. (whole) 180 2.16% 24 31.17% 299 5.15 138 5% 10 59% 7 1% 3 15% 13 1% 6 38%
Vigna sp. (cotyledon) 84 1.01% 16 20.78% 40 2% 7 41% 1 0% 1 5% 3 0% 2 13%
Vigna sp. (frag) 35 0.42% 12 15.58% 10 0% 5 29% 0 0% 0 0% 11 1% 3 19%
Vigna unguiculata (coty- 1 0.01% 1 1.30% 0 0% 0 0% 0 0% 0 0% 1 0% 1 6%
ledon)
Macrotyloma (whole) 3 0.04% 3 3.90% 23 0.39 1 0% 1 6% 1 0% 1 5% 1 0% 1 6%
Macrotyloma (cotyledon) 13 0.16% 3 3.90% 11 0% 2 12% 0 0% 0 0% 2 0% 1 6%
Macrotyloma (frag) 7 0.08% 3 3.90% 5 0% 1 6% 1 0% 1 5% 1 0% 1 6%
Lablab (whole) 6 0.07% 1 1.30% 23 0.39 0 0% 0 0% 0 0% 0 0% 6 1% 1 6%
Lablab (cotyledon) 8 0.10% 4 5.19% 5 0% 2 12% 2 0% 1 5% 0 0% 0 0%
Lablab (frag) 9 0.11% 5 6.49% 1 0% 1 6% 2 0% 1 5% 2 0% 1 6%
Cajanus (whole) 8 0.10% 6 7.79% 18 0.31 0 0% 0 0% 4 1% 2 10% 3 0% 3 19%
Cajanus (cotyledons) 1 0.01% 1 1.30% 0 0% 0 0% 0 0% 0 0% 1 0% 1 6%
Cajanus (frags) 9 0.11% 2 2.60% 0 0% 0 0% 9 1% 2 10% 0 0% 0 0%
Cicer (whole) 12 0.14% 6 7.79% 36 0.62 0 0% 0 0% 3 0% 1 5% 9 1% 5 31%
Cicer (cotyledons) 21 0.25% 9 11.69% 0 0% 0 0% 3 0% 1 5% 17 2% 7 44%
Cicer (frags) 3 0.04% 2 2.60% 0 0% 0 0% 1 0% 1 5% 2 0% 1 6%
cf. Cicer (reticulate surface) 3 0.04% 1 1.30% 0 0% 0 0% 3 0% 1 5% 0% 0 0%
Lathyrus sativus (whole) 4 0.05% 3 3.90% 7 0.12 0 0% 0 0% 0 0% 0 0% 3 0% 2 13%
Lathyrus sativus (cotyle- 3 0.04% 1 1.30% 0 0% 0 0% 0 0% 0 0% 3 0% 1 6%
don)
Lens (whole) 48 0.58% 14 18.18% 71 1.22 0 0% 0 0% 0 0% 0 0% 38 3% 9 56%
Lens (cotyledon) 12 0.14% 5 6.49% 0 0% 0 0% 0 0% 0 0% 7 1% 4 25%
Lens (frags) 11 0.13% 4 5.19% 1 0% 1 6% 0 0% 0 0% 10 1% 3 19%
Pisum sativum (whole) 2 0.02% 2 2.60% 0 0% 0 0% 0 0% 0 0% 0 0% 0 0%
Indet. large legume 63 0.76% 8 10.39% 16 1% 1 6% 15 2% 1 5% 20 2% 1 6%
Millets
Brachiria/Setaria caryopsis 74 0.89% 5 6.49% 0 0% 0 0% 27 4% 1 5% 0 0% 0 0%
Brachiaria ramosa/Setaria 269 3.23% 16 20.78% 5 0% 3 18% 24 3% 4 20% 111 10% 3 19%
italica
Brachiaria ramosa type 67 0.80% 11 14.29% 0 0% 0 0% 13 2% 2 10% 39 3% 6 38%
(w/o husk)
B. ramosa (w/ husk) 4 0.05% 4 5.19% 0 0% 0 0% 1 0% 1 5% 1 0% 1 6%
B. ramosa spikelets (silici- 0 0.00% 0 0.00% 0 0% 0 0% 0 0% 0 0% 0 0% 0 0%
fied)
 Chapter 12 Archaeobotany    249

Echinochloa sp. caryopsis 10 0.12% 6 7.79% 0 0% 0 0% 1 0% 1 5% 6 1% 3 19%

Eleusine coracana 2 0.02% 1 1.30% 2 0% 1 6% 0 0% 0 0% 0 0% 0 0%
cf. Eleusine coracana 4 0.05% 3 3.90% 2 0% 1 6% 0 0% 0 0% 0 0% 0 0%
Panicum sp. (small) 5 0.06% 2 2.60% 0 0% 0 0% 0 0% 0 0% 0 0% 0 0%
Panicum sp. (small imma- 2 0.02% 2 2.60% 0 0% 0 0% 0 0% 0 0% 1 0% 1 6%
ture)
Paincum miliaceum 4 0.05% 2 2.60% 0 0% 0 0% 0 0% 0 0% 4 0% 2 13%
Panicum cf. sumatrense 27 0.32% 10 12.99% 2 0% 2 12% 4 1% 2 10% 14 1% 3 19%
Panicum cf. sumatrense 1 0.01% 1 1.30% 0 0% 0 0% 0 0% 0 0% 0 0% 0 0%
(spikelet)
Panicum sp. (immature) 5 0.06% 1 1.30% 0 0% 0 0% 0 0% 0 0% 5 0% 1 6%
Paspalum scrobiculatum 3,724 44.72% 60 77.92% 1,842 73% 16 94% 192 25% 17 85% 69 6% 12 75%
Paspalum scrobiculatum 4 0.05% 3 3.90% 3 0% 2 12% 0 0% 0 0% 0 0% 0 0%
(w/ husk)
cf. Paspalum husk frags 121 1.45% 2 2.60% 31 1% 1 6% 0 0% 0 0% 0 0% 0 0%
(charred)
Paspalum spikelet bases 8 0.10% 1 1.30% 0 0% 0 0% 0 0% 0 0% 0 0% 0 0%
cf. Paspalum rachis 1 0.01% 1 1.30% 0 0% 0 0% 0 0% 0 0% 0 0% 0 0%
segment
Pennisetum glaucum 75 0.90% 9 11.69% 2 0% 1 6% 8 1% 1 5% 58 5% 3 19%
Setaria italica type 41 0.49% 5 6.49% 0 0% 0 0% 2 0% 2 10% 39 3% 3 19%
S. italica type w/ husk 4 0.05% 2 2.60% 0 0% 0 0% 2 0% 1 5% 2 0% 1 6%
S. italica spikelet 1 0.01% 1 1.30% 0 0% 0 0% 0 0% 0 0% 0 0% 0 0%
Setaria verticillata type 35 0.42% 9 11.69% 0 0% 0 0% 6 1% 2 10% 26 2% 6 38%
S. verticillata type w/ husk 4 0.05% 4 5.19% 0 0% 0 0% 2 0% 2 10% 1 0% 1 6%
Setaria cf. pumila 19 0.23% 3 3.90% 0 0% 0 0% 0 0% 0 0% 0 0% 0 0%
Setaria cf. sphaceleata 2 0.02% 2 2.60% 0 0% 0 0% 0 0% 0 0% 1 0% 1 6%
Setaria/Brachiaria imma- 5 0.06% 4 5.19% 0 0% 0 0% 1 0% 1 5% 2 0% 2 13%
ture
Sorghum bicolor 44 0.53% 16 20.78% 0 0% 0 0% 13 2% 4 20% 22 2% 8 50%
S. bicolor cf. race bicolor 3 0.04% 3 3.90% 0 0% 0 0% 1 0% 1 5% 1 0% 1 6%
S. bicolor cf. race durra 4 0.05% 3 3.90% 0 0% 0 0% 0 0% 0 0% 4 0% 3 19%
Indet. small millet 40 0.48% 5 6.49% 0 0% 0 0% 0 0% 0 0% 0 0% 0 0%
Indet. millet 6 0.07% 1 1.30% 0 0% 0 0% 0 0% 0 0% 6 1% 1 6%
Big cereals
Oryza cf. sativa (grains) 163 1.96% 39 50.65% 55 2% 8 47% 27 4% 9 45% 60 5% 12 75%
Oryza grain frags 137 1.65% 23 29.87% 75 3% 5 29% 5 1% 4 20% 13 1% 4 25%
Oryza husk fragments 3 0.04% 3 3.90% 0 0% 0 0% 0 0% 0 0% 0 0% 0 0%
(charred)
Oryza husk fragments 0 0.00% 0 0.00% 0 0% 0 0% 0 0% 0 0% 0 0% 0 0%
(silicified)
Oryza spikelet bases 12 0.14% 1 1.30% 12 0% 1 6% 0 0% 0 0% 0 0% 0 0%
Oryza sp. (small wild) 6 0.07% 2 2.60% 3 0% 1 6% 0 0% 0 0% 3 0% 1 6%
Hordeum vulgare (Grains/ 56 0.67% 14 18.18% 33 1% 6 35% 1 0% 1 5% 1 0% 1 6%
frags)
Hordeum vulgare rachis 1 0.01% 1 1.30% 0 0% 0 0% 0 0% 0 0% 0 0% 0 0%
Triticum sp. (grain) 13 0.16% 6 7.79% 0 0% 0 0% 0 0% 0 0% 6 1% 3 19%
Triticum cf. dicoccum 15 0.18% 7 9.09% 0 0% 0 0% 2 0% 1 5% 11 1% 4 25%
(grain)
Triticum FT (grain) 19 0.23% 5 6.49% 0 0% 0 0% 0 0% 0 0% 18 2% 4 25%
Indet Hordeun/Triticum 27 0.32% 5 6.49% 1 0% 1 6% 0 0% 0 0% 0 0% 0 0%
(grain)
Culm node (large grass/ 6 0.07% 4 5.19% 0 0% 0 0% 2 0% 1 5% 4 0% 3 19%
cereal)
250   Dorian Q. Fuller

Table 12.1b: Archaeobotanical data from Paithan. P: present; T: total.

Period 1 1 2 2 3 3
Site Freq. P Ubiq- T P T P T P
total uity

Other crops
Sesamum indicum 225 2.70% 15 19.48% 6 0% 1 6% 165 22% 5 25% 7 1% 4 25%
cf. Sesamum radiatum type 1 0.01% 1 1.30% 0 0% 0 0% 0 0% 0 0% 1 0% 1 6%
Linum usitatissimum 1 0.01% 1 1.30% 0 0% 0 0% 0 0% 0 0% 1 0% 1 6%
cf. Linum frags. 3 0.04% 1 1.30% 0 0% 0 0% 0 0% 0 0% 3 0% 1 6%
Carthamus tinctorius 2 0.02% 1 1.30% 0 0% 0 0% 0 0% 0 0% 0 0% 0 0%
cf. Carthamus tinctorius 6 0.07% 2 2.60% 3 0% 1 6% 0 0% 0 0% 0 0% 0 0%
(frags)
Cannabis sativa cf. subsp. 1 0.01% 1 1.30% 0 0% 0 0% 0 0% 0 0% 1 0% 1 6%
sativa
Brassica cf. juncea 8 0.10% 1 1.30% 0 0% 0 0% 8 1% 1 5% 0 0% 0 0%
Gossypium sp. 25 0.30% 12 15.58% 2 0% 2 12% 3 0% 3 15% 16 1% 6 38%
Gossypium testa frags 13 0.16% 5 6.49% 2 0% 1 6% 5 1% 2 10% 5 0% 1 6%
Abelmoschus sp. 8 0.10% 5 6.49% 0 0% 0 0% 1 0% 1 5% 6 1% 3 19%
Cucumis cf. melo agrestis 1 0.01% 1 1.30% 0 0% 0 0% 0 0% 0 0% 1 0% 1 6%
Cucumis melo/sativus 2 0.02% 2 2.60% 0 0% 0 0% 1 0% 1 5% 1 0% 1 6%
Cucurbitaceae: small 7 0.08% 4 5.19% 0 0% 0 0% 0 0% 0 0% 6 1% 3 19%
Melothria heterophylla
Cucurbitaceae 2 0.02% 2 2.60% 1 0% 1 6% 0 0% 0 0% 1 0% 1 6%
cf. Cuminum 3 0.04% 1 1.30% 0 0% 0 0% 0 0% 0 0% 3 0% 1 6%
Tree fruits
Fruit/nut indet. frags 12 0.14% 7 9.09% 1 0% 1 6% 6 1% 2 10% 3 0% 3 19%
Ziziphus sp. (wh+frags) 35 0.42% 10 12.99% 4 0% 2 12% 3 0% 2 10% 22 2% 4 25%
Ficus sp. 118 1.42% 14 18.18% 5 0% 1 6% 2 0% 2 10% 56 5% 5 31%
Emblica sp. 2 0.02% 2 2.60% 0 0% 0 0% 1 0% 1 5% 1 0% 1 6%
cf. Vitaceae small 1 0.01% 1 1.30% 0 0% 0 0% 1 0% 1 5% 0 0% 0 0%
cf. Acacia sp. 2 0.02% 2 2.60% 0 0% 0 0% 1 0% 1 5% 1 0% 1 6%
cf. Prosopis sp. 4 0.05% 2 2.60% 2 0% 1 6% 2 0% 1 5% 0 0% 0 0%
Indet fruit seed/nutlet 1 0.01% 0 0.00% 0 0% 0 0% 0 0% 0 0% 0 0% 0 0%
Weedy grasses (excluding
millet genera)
Alopecurus sp. 1 0.01% 1 1.30% 0 0% 0 0% 0 0% 0 0% 1 0% 1 6%
Coix caryopsis 12 0.14% 1 1.30% 0 0% 0 0% 12 2% 1 5% 0 0% 0 0%
cf. Coix caryopsis(?) 1 0.01% 1 1.30% 0 0% 0 0% 1 0% 1 5% 0 0% 0 0%
Dactyloctenium cf. aegyp- 1 0.01% 1 1.30% 0 0% 0 0% 0 0% 0 0% 0 0% 0 0%
tium
Digitaria sp. 31 0.37% 8 10.39% 0 0% 0 0% 5 1% 2 10% 23 2% 5 31%
Eleusine indica 3 0.04% 2 2.60% 2 0% 1 6% 0 0% 0 0% 1 0% 1 6%
Ischaemum cf. rugosum 21 0.25% 7 9.09% 0 0% 0 0% 3 0% 2 10% 18 2% 5 31%
cf. Lolium temulentum 2 0.02% 1 1.30% 0 0% 0 0% 0 0% 0 0% 2 0% 1 6%
(large Hordeae)
cf. Pennisetum sp. small 2 0.02% 1 1.30% 0 0% 0 0% 0 0% 0 0% 0 0% 0 0%
wild
cf. Stipa(?) 1 0.01% 1 1.30% 0 0% 0 0% 1 0% 1 5% 0 0% 0 0%
Urochloa/Braciahia eruci- 20 0.24% 4 5.19% 0 0% 0 0% 0 0% 0 0% 14 1% 1 6%
forme
Indet small grass 46 0.55% 6 7.79% 0 0% 0 0% 2 0% 2 10% 19 2% 1 6%
Aizooaceae 1 0.01% 1 1.30% 0 0% 0 0% 1 0% 1 5% 0 0% 0 0%
Aizooaceae: Trianthema 35 0.42% 5 6.49% 1 0% 1 6% 5 1% 2 10% 3 0% 2 13%
Aizooaceae: Zaleya decan- 84 1.01% 23 29.87% 0 0% 0 0% 25 3% 6 30% 28 2% 6 38%
dra
 Chapter 12 Archaeobotany    251

Amaranthaceae 2 0.02% 2 2.60% 0 0% 0 0% 0 0% 0 0% 0 0% 0 0%

Apiaceae 1 0.01% 1 1.30% 0 0% 0 0% 0 0% 0 0% 1 0% 1 6%
Asteraceae 19 0.23% 3 3.90% 17 1% 1 6% 0 0% 0 0% 1 0% 1 6%
Boraginaceae type 3 0.04% 1 1.30% 3 0% 1 6% 0 0% 0 0% 0 0% 0 0%
Brassicaceae 26 0.31% 7 9.09% 1 0% 1 6% 0 0% 0 0% 13 1% 3 19%
Capparaceae type 1 0.01% 1 1.30% 0 0% 0 0% 0 0% 0 0% 0 0% 0 0%
Caryophyllaceae 10 0.12% 1 1.30% 0 0% 0 0% 0 0% 0 0% 2 0% 1 6%
Chenopodiaceae: Cheno- 5 0.06% 3 3.90% 0 0% 0 0% 0 0% 0 0% 0 0% 0 0%
podium sp.
Commelinaceae: Com- 9 0.11% 7 9.09% 1 0% 1 6% 5 1% 3 15% 3 0% 3 19%
melina sp.
Cyperaceae: cf. Carex 2 0.02% 2 2.60% 0 0% 0 0% 0 0% 0 0% 0 0% 0 0%
Cyperaceae: Cyperus 1 0.01% 1 1.30% 0 0% 0 0% 1 0% 1 5% 0 0% 0 0%
Cyperaceae: Fimbristylis 10 0.12% 1 1.30% 0 0% 0 0% 0 0% 0 0% 0 0% 0 0%
Cyperaceae 7 0.08% 4 5.19% 0 0% 0 0% 0 0% 0 0% 0 0% 0 0%
Euphorbiaceae type 6 0.07% 4 5.19% 0 0% 0 0% 0 0% 0 0% 0 0% 0 0%
Leguminosae small 109 1.31% 19 24.68% 12 0% 1 6% 1 0% 1 5% 24 2% 5 31%
Lamiaceae 2 0.02% 2 2.60% 0 0% 0 0% 0 0% 0 0% 1 0% 1 6%
Malvaceae (large) CROP?/ 44 0.53% 15 19.48% 0 0% 0 0% 8 1% 5 25% 35 3% 9 56%
below
Malvaceae (large) cf. 23 0.28% 3 3.90% 0 0% 0 0% 11 1% 1 5% 12 1% 2 13%
Hibiscus
Malavaceae (small) cf. 21 0.25% 10 12.99% 3 0% 1 6% 5 1% 2 10% 10 1% 4 25%
Malva
Papavaraceae type 2 0.02% 1 1.30% 0 0% 0 0% 0 0% 0 0% 0 0% 0 0%
Poaceae indet. 4 0.05% 2 2.60% 0 0% 0 0% 0 0% 0 0% 0 0% 0 0%
Polygonacaeae: cf. Polygo- 1 0.01% 1 1.30% 0 0% 0 0% 0 0% 0 0% 0 0% 0 0%
num sp.
Portulaceae type 4 0.05% 1 1.30% 0 0% 0 0% 0 0% 0 0% 4 0% 1 6%
Rubiaceae mericarp, 4 0.05% 2 2.60% 0 0% 0 0% 0 0% 0 0% 3 0% 2 13%
Spermacoce cf. hispida
Rubiaceae cf. Rubia sp. 2 0.02% 2 2.60% 0 0% 0 0% 0 0% 0 0% 1 0% 1 6%
Solanaceae: Solanum sp. 4 0.05% 3 3.90% 2 0% 1 6% 0 0% 0 0% 1 0% 1 6%
Schrophulariaceae: cf. 4 0.05% 3 3.90% 0 0% 0 0% 0 0% 0 0% 4 0% 3 19%
Verbascum
Indet large seeds 14 0.17% 6 7.79% 0 0% 0 0% 2 0% 1 5% 2 0% 2 13%
Indet small seeds/sclerotia 73 0.88% 8 10.39% 0 0% 0 0% 3 0% 1 5% 32 3% 3 19%
Indet 39 0.47% 3 3.90% 37 1 6% 1 1 5% 1 1 6%
Indet testa frags 39 0.47% 5 6.49% 0 0 0% 0 0 0% 0 0 0%
Indet. fragments 1,330 15.97% 15 19.48% 122 2 12% 10 1 5% 5 1 6%
Large frags (cf. paren- 94 1.13% 4 5.19% 0 0 0% 63 1 5% 22 2 13%
chyma)
Dung fragments 1 0.01% 1 1.30% 1 1 6% 0 0 0% 0 0 0%
Total items 8,327 77 2,529 762 1,128
No. samples 78 17 21 17
252   Dorian Q. Fuller

Fig. 12.3: Relative frequency of major crop groups.

Fig. 12.4: Comparative diversity indices.

Overall, in terms of staple crop production, this might words, an assemblage with ten species but in which
be interpreted as some tendency towards agricul- only one or two of these account for most individuals is
tural diversification. Measured merely in terms of the less diverse than an assemblage with ten species that
number of cereal/pulse species, there is directional are all found in similarly low frequencies. To measure
change: from 13 taxa in Period 1 to 18 taxa in Period 2 this, three simple indices have been drawn from the
and 22 in Period 3 (this count includes Setaria verticil- ecologists’ toolkit (Krebs 1985: 519ff). These include
lata, assumes the presence of both Brachiaria ramosa the Simpson index and the Shannon index, which are
and Setaria italica throughout and groups together in common use amongst paleontologists and archaeo-
sorghum subdivisions but separates free-threshing zoologists. The Shannon index represents entropy on
and diococcum-type wheat grains). This increase in a scale from 0 (when a single species represents all indi-
diversity is not reflected in the weed flora, of which viduals) to higher numbers as more species are encoun-
each period has 38 identifiable species present tered and are more evenly distributed. The Simpson
(although the total for all periods is 43, indicating index represents evenness with an index between 0 and
that there were some minor changes in the presence 1, by calculating a ratio of dominance (for each species
of weed species). In biological terms, diversity is often the number of individuals of the species is divided by
considered to include not just the number of species the total of all individuals squared, and then summed
but also how evenly distributed they are. In other across all species). The Menhinick richness index takes
 Chapter 12 Archaeobotany    253

into account the number of species and sample size 1991): the winter cereals (wheat and barley), or rabi
as an attempt to correct for influence of sample size cereals, and summer (monsoon) millets (a wide range
(i.e. that larger samples normally contain more taxa). of species, including sorghum), or kharif crops, as well
These indices were calculated using the program PAST as rice, which is probably most often a kharif crop,
(Hammer et al. 2001). By these measures, the increase although winter varieties exist and may have already
in diversity is clear (Fig. 12.4). been available in the past. For the most part, these sea-
In the subsequent sections of this report, dis- sonalities are constrained by physiology and genetics
cussion is focussed around particular quantitative of the crops themselves, as flowering (and subsequent
aspects of occurrence, issues of identification and seed production) is normally triggered by changes
significance of the species in terms of the economy at in day-length (Willcox 1992). Thus, winter crops are
Paithan. The evidence from Paithan is situated in rela- long-day plants that flower as a result of days getting
tion to the archaeobotanical record for South Asia as a longer after the shortest day (21 December), leading to
whole. Table 12.1 presents the raw data. harvests in February or March, while summer crops are
short-day plants that flower as days shorten towards
the end of summer. In modern agriculture, exceptions
Limitations of the interim report are found, and these are usually ‘photoperiod neutral’
plants, i.e. plants in which sensitivity to day-length has
There is still further work to be done on the archaeobo- been deactivated through plant breeding. Thus, there
tanical assemblage at Paithan. This includes prepar- are winter sorghums and pearl millets and summer
ing a comprehensive set of representative illustrations wheats and barleys (in Europe and in the Himala-
for all taxa, systematic measurements on many taxa, yas). How long ago these alternative season varieties
rechecking and refining some identifications especially evolved is unknown, although in many cases, these
of small millets and putting further efforts into identi- alternative states may exist in different wild popula-
fying some unknown seed types. Of particular impor- tions, for example, in sorghum and rice. Wild rices in
tance is refining the identification of weed seeds to southern regions, such as Indonesia or Sri Lanka, are
genus and/or species level. At present, available refer- often photoperiod neutral and not tied to the summer,
ence collection materials are inadequate, and unlike whereas those further north in south China or the
many world regions, good seed atlases are not avail- Ganges and Orissa are (Jennings 1966). The same is
able for southern India. In the present report, a com- true of Southern African sorghums, both domesticated
prehensive discussion of the weed taxa recovered is and wild. Thus, it is possible that winter varieties were
not provided. already available to the inhabitants of Paithan, and
at present, we just do not know. However, since the
establishment of millet and sorghum agriculture in

Cereals Southern Africa is rather later – probably closer to the

time that Paithan was founded (Marshall and Hilde-
brandt 2002) – it seems likely that the sorghum vari-
Cereals are the mainstays of most world agricultural
eties available at Paithan would have been predomi-
economies today as they were in the past. These are
nantly, if not entirely, tied to summer cultivation. This
cultivated grasses in which production is focussed on
was probably also the case with early indica rices, prior
the seed, which is a dry, hard grain that is readily stor-
to the diffusion of tropical japonica varieties from the
able and high in carbohydrates. Cereal grains, once
Malaysia region. In the case of summer wheats and
mature, are not directly edible but require preparation
barleys in the Himalayas, in all likelihood, the evolu-
through pulverisation (grinding), which makes flours
tion of these varieties was necessary for agricultural
for breads or gruels, or through boiling or steam-
settlement and may be hypothesized to have occurred
ing whole, as is familiar for rice, or boiling and fer-
by the start of the Kashmir Neolithic (Fuller 2006a:
menting to make beers. Paithan has a rich record for
36), but such forms are unlikely to have ever spread to
cereals and includes a wide range of species, includ-
southern India.
ing most of those known to have been cultivated in
India in ancient times. Cropping in general and cereals
in particular are often divided into two main seasonal
groups in South Asia (see e.g. Kajale 1988; Weber
254   Dorian Q. Fuller

Table 12.2: The presence/absence of cereal crops on peninsular Indian sites of Iron Age or Early Historic date. Sources: Cooke et al. (2005),
Kajale (1974, 1977b, 1982, 1984, 1989, 1994), Fuller et al. (2004), Fuller (unpublished data) (Piklihal), Vishnu-Mittre (1968), Vishnu-Mittre
and Gupta (1968a,b) and Vishnu-Mittre et al. (1971, 1984). (IA = Iron Age; IV = Hallur period IV; EH = Early Historic).

Barley

No. cereal crops

Wheat(s)

Rice
Sorghum bicolor

Pennisetum glaucum

Eleusine coracana

Brachiaria ramosa

Setaria italica

Setaria verticillata

Panicum sumatrense

Paspalum scrobiculatum

Echinochloa colona
Period Site

800 bc IA Adam Cave - - + - - -? +? + +? - - - 5

IA Bhagimohari - + + - - - - - - - - - 2
IA Bhokardan + - + + - -? +? - ? - - - 4
IA Naikund + - - - - - - - - - - - 1
IA Veerapuram - - + - - - - - - - + - 2
IV/IA Hallur - - +? - - + + - + - - - 4
300 bc–? EH Piklihal - - - + - - + - - + + - 4
EH Veerapuram - - + - - - - - - - + - 2
EH Adam Cave + + + - - - - - - - - - 3
EH Bhatkuli - - - + - - - - - - - - 1
EH Bhokardan + - - + - -? ? ? ? - - - 2
EH Dangwada + - + - - - - - - - - - 2
EH Kaundinyapura - - + - - - - - - - - - 1
EH Nevasa + + + + + + ? - ? - + - 7
EH Paunar - - + + - - - - - - - - 2
EH Ter + + + - - - - - - - + - 4
EH Kodumanal - - + - - - + - - + - - 3
EH Mangudi - - + - - - + - - - - + 3
EH Perur - - + - - + + - + + + + 7
300–0 bc Paithan I - + + - + + + +? - - + - 7
ad 0–400 Paithan II + + + + + - + + + + + + 11
ad 400–700 Paithan III + + + + + - + + + + + + 11
Total sites 9 7 18 8 4 4 10 4 5 5 9 4

Fig. 12.5: Scanning electron micrograph of charred Paspalum scrobiculatum (kodo millet) from Paithan; dorsal view (left) and ventral
view (right).
 Chapter 12 Archaeobotany    255

In comparison to other peninsular sites of Iron Age as a weed of cultivated rice and is still today a perni-
or Early Historic date, Paithan has the most diverse cious weed of rice (Moody 1989; Galinato et al. 1999). It
range of cereal species (Table 12.2). This may, however, may then have been recognized as a useful fall back or
merely reflect the impact of a large-scale flotation supplement food and began to be cultivated on poorer
program rather than any real difference in agricultural non-rice soils. This would make kodo millet a true sec-
diversity. ondary domesticate in the Vavilovian sense, as postu-
lated for several Indian millets by Kimata et al. (2000).
Kodo millet, Paspalum scrobiculatum (Fig. 12.5), is Subsequently, it may have spread to the Deccan at the
by far the dominant plant in the Paithan archaeobo- same period as rice, but in the drier savannah zone, it
tanical assemblage; it occurs in 80% of the samples was found to be particularly well suited.
and accounts for 44% of all the seeds identified. As This species is drought-tolerant, does well on
discussed above, its unrivalled dominance in Period 1 poor soils and has a short growing season of 3 to 4
declines through subsequent phases and sees replace- months, thereby being a low-labour demanding crop
ment by other millets, including African millets and (Watt 1889–1893; De Wet et al. 1983). It also provides a
especially Brachiaria ramosa and/or Setaria italica. relatively nutritious grain (FAO 1995; Weber and Fuller
This trend is of interest since kodo millet is not a major 2008). One of its drawbacks however, is that it is prone
crop in the region today, although it is known (Kimata to fungal infection especially if it gets wet during grain
1989; Kobayashi and Kimata 1989), and does not maturation, and such infected grains are toxic for
appear to have been a crop in the region during Chal- people and in animal fodder (FAO 1995). Frequency of
colithic times (Kajale 1991; Fuller 2002). It is present as this would have varied with climatic conditions, and
only a few seeds in the later Southern Neolithic in Kar- realization of this could have contributed to decreased
nataka, which may not represent use as a crop (Fuller et popularity over the course of the Early Historic period,
al. 2004). All of this implies that during the late Iron Age which may have been a time of increasing monsoons
and through the Satavahana period, kodo millet was a between 300 bc and ad 100 (see e.g. increased Oxygen-
major crop in this region. Indeed, kodo millet is known 18 Isotope levels in the Arabian sediment profile of
from several other sites of the Iron Age or Early His- Staubwasser et al. 2003).
toric period on the peninsula, and this period may rep-
resent the apogee of cultivation of this species (Table Rice, Oryza sativa (Fig. 12.6) is the world’s most pro-
12.3). Based on the current archaeobotanical evidence, ductive crop today and a widespread staple food in
we might point towards the Ganges Chalcolithic (e.g. Asia and South Asia. Its origins have been reviewed
Narhan Culture) as a possible locus of domestication elsewhere (Fuller 2006a: 39–41; 2006b; 2007: 911–915),
of this crop. Presumably, this species was first known

Fig. 12.6: Rice spikelet base from context 721; lateral view (left) and ventral view (right). Notice the indented scar indicative of
domesticated/non-shattering morphotype.
256   Dorian Q. Fuller

Table 12.3: The occurrence of kodo millet in South Asian archaeology.

Site Period Paspalum Dominant cereal(s) Others Reference

ubiquity/comments

Kuntasi, Saurashtra Mature Harappan Probable Paspalum Panicum sumatrense, Wheat, barley, Kajale 1996a
Setaria spp.(?) many Coix for (preliminary report
beads only)
Kurugodu Neolithic IIIA One grain, one Brachiaria ramosa, Wheat, barley Fuller et al. 2001,
(1750–1550 bc) sample Setaria verticillata 2004; revised dating:
(of four) Fuller et al. 2007
Rojdi, Saurashtra Period C, Late 1% (of 254 samples) Panicum sumatrense, Sorghum, Weber 1991
Harappan (2000– Setaria spp.(?) Echinochloa
1700 bc) (weedy?)
Rupar, Punjab Late Harappan/Baran ? Wheat Vishnu-Mittre and
(2000–1600 bc?) Savithri 1979
Senuwar Period IB, Late 2% (of 42), absent Rice, barley Wheat, sorghum, Saraswat 2004
Neolithic (2000– from earlier period Setaria pumila
1300 bc)
Daimabad, Maha- Jorwe (1500–1100 Absent from earlier Wheat, barley, Sorghum, rice Kajale 1977a; Vishnu-
rashtra bc) levels, as are rice and Brachiaria ramosa(?)/ Mittre et al. 1986
sorghum Setaria italica
Imlidh-Khurd, Uttar Period II, Narhan ? Rice, wheat, barley Setaria pumila Saraswat 1993
Pradesh Culture (1300–800
bc)
Malhar, Uttar Pradesh Period II, Early Iron 5.5% (of 18) Rice, wheat, barley Sorghum, Eleusine Tewari et al. 2000
Age (1600(?)–800
bc)
Narhan, Uttar Pradesh Chalcolithic-Iron Age 23% (of 26 samples), Rice, wheat, barley Pearl millet Saraswat et al. 1994
(c. 1300–200 bc) absent from later
period
Hallur, Karnataja Neolithic IIIB 12.5% (1 of 8 Brachiaria ramosa, Wheat, pearl Fuller et al. 2001,
(1550–1350 bc) samples; 2% fre- Setaria verticillata millet, Panicum 2004; revised dating:
quency) sumatrense, wild Fuller et al. 2007
Oryza sp.
Magha, Uttar Pradesh Black and Red ware ? Rice Echinochloa sp. Vishnu-Mittre et al.
1984
Hulaskhera, Uttar Iron Age (800–300 17% (of 6 samples), Barley, rice Rare wheat Chanchala 1992
Pradesh bc) mis-reported as
‘Eleusine’
Perur, Tamil Nadu Early Historic/Late 20% (of 10 samples, Rice, Eleusine coracana Panicum Cooke et al. 2005
Megalithic (300 frequency 9% and sumatrense,
bc–300 ad) 26%) Echinochoa cf.
colona
Veerapuram, 500 bc–400 ad 14% (of 21 samples, Rice, kodo millet(?) Barley Kajale 1984
Andhra Pradesh but high frequency)
Hallur, Karnataka 800–200 bc One of two samples Brachiaria ramosa, Panicum Fuller 1999; Fuller et
studied of this Setaria verticillata sumatrense, al. 2004; Vishnu-Mit-
period, low Eleusine tre 1971; for dating:
Frequency (9.6%); coracana, rice Fuller et al. 2007
also preserved by
mineralization with
metal tools
Charda, Uttar Pradesh Period IIA, 600–200 9% (of 11 samples) Rice, barley Wheat, a small Chanchala 2002
bc millet
 Chapter 12 Archaeobotany    257

Saunphari, Uttar 1000 bc–ad 300 ? Rice, barley, wheat Chanchala 2004
Pradesh
Hulaskhera, Uttar Early Historic 83% (of 23 Kodo millet, barley, rice Rare wheat, pearl Chanchala 1992
Pradesh (300 bc–ad 300) samples), mis- millet
reported as ‘Eleu-
sine’)
Manji, Uttar Pradesh Early Historic 50% (of 4 samples), Barley, wheat, rice Rare Setaria cf. Chanchala 2000–2001
(300 bc–ad 300) mis-reported as italica
‘Eleusine’)
Sanghol Early Historic ?, mis-reported as Rice, barley, wheat Sorghum, Setaria Pokharia and Saraswat
(300 bc–ad 300) ‘Eleusine’ sp. 1999
Balathal, Rajasthan Early Historic ? Barley, wheat, rice, Panicum, Setaria, Kajale 1996a
(300 bc–ad 300) kodo millet(?) Coix
Nevasa, Early Historic ? Barley, wheat, rice Pearl millet, Anonymous, in
Maharashtra (150 bc–ad 200) sorghum, Sankalia
Eleusine(?) et al. 1960: 529–30;
Kajale 1977b
Ter (Thair), Early Historic Large quantities Rice, kodo millet(?) Wheat, barley Vishnu-Mittre et al.
Maharashtra (200 bc–ad 400) 1971
Piklihal, Karnataka Early Historic One grain in one Brachiaria ramosa, Sorghum Fuller, unpublished
sample (of two) Panicum sumatrense data; dating: Fuller et
al. 2007
Paithan, 400 bc–ad 700 78% (in 77 samples, Kodo millet, Wheat, barley, This report
Maharashtra declining over dura- Brachiaria ramosa/ sorghum, pearl
tion of site from 94% Setaria italica, rice millet, Panicum
to 75%) sumatrense,
Eleusine coracana

but there is now a strong case for one of several domes- numbers are suggestive. Rice was widely available, but
tication events for Asian rice in northern South Asia, somehow special, and does not appear to have been
probably the Gangetic region, while separate processes the staple crop, but perhaps a high-status or special-
of domestication are inferred for Yangtze river basin use cereal alongside the staple millets and pulses.
of China (Crawford 2006; Fuller 2007). Available evi- This is how we would expect the traditional diet of the
dence suggests that rice became a crop in parts of penin- Deccan to have been until quite recent times, until the
sular India during the Iron Age (Kajale 1989, 1991; Fuller spread of artificial irrigation. Rice is a thirsty crop, and
2006a: 53), with the earliest few finds from the latest without major capital investments (landesque intensi-
level at Inamgaon in Maharashtra, dated to 1200–900 fication, sensu Brookfield 1984; Blaikie and Brookfield
bc (Kajale 1988). The Iron Age adoption of rice was 1987: 9; also Kirch 1994: 19), it is unlikely to have been
accompanied by some changes in ceramic repertoire produced on as large a scale as monsoon-adapted dry
that suggest the influence of northern Indian food crops like millets and summer pulses.
traditions (Fuller 2005: 769). It is therefore no surprise Rice is highly diverse, with a wide range of landraces
that rice was a consistent find at Paithan. It has a high known in India as well as several varieties of wild weedy
ubiquity of c. 65% of samples, which is essentially con- rice that may occur as weeds of cultivation. Varieties in
stant through all phases. In terms of absolute numbers India include both the conventional subspecies indica,
and relative frequencies, however, rice is quite low, with with likely origins in northern India, and japonica of
its total relative frequency of c. 3%. While this is about East Asian origin. Grain morphometric evidence sug-
twice the total frequency of wheat and barley, it is gests that japonica rices were introduced to the north-
less than the frequency of summer pulses or winter western subcontinent, as at Pirak, in the first half of
pulses (when taxa are grouped) and much less than the 2nd millennium bc, together with other evidence
the total frequency of millets, which is over 60%. Given for diffusion from China (Fuller 2006a: 36; Fuller and
that recovery and sorting biases might be expected to Qin 2009). It is therefore of interest to know which
favour large seeds (like rice, wheat and pulses), these subspecies are most likely represented by the Paithan
258   Dorian Q. Fuller

Fig. 12.7: Comparison of the frequency of wheat and barley with other taxa.

finds, and detailed metrical analysis may allow this in could be more carefully tended and watered (Fig. 12.7).
the future. In addition, the presence of apparent wild At Paithan, these species are fairly widespread, with
rices with noticeably smaller and thinner grains must a combined ubiquity of over 40% but with very low
also be noted, some of which may have had long use frequency. However, the patterns in wheat and barley
as occasional foodstuffs in South India, or perhaps differ. In Period 1, only barley is recorded, and this
occurred as weeds of pulse or millet cultivation, as shows a significant subsequent decline into Periods 2
indicated by small quantities of probable Oryza gran- and 3. Wheat first occurs in Period 2 and shows a major
ulata grains from Late Neolithic Hallur in Karnataka rise into Period 3. The latter increase is mirrored in
(2nd millennium bc) (Fuller 2003c: 378, n.2). In addi- the winter pulses, which suggests a general increase
tion, other thin and flattened grains are likely to rep- in winter-cropping; this perhaps implies an increase
resent immature grains, implying that crops did not in the control over water supply to provide irrigation
ripen evenly and that early harvests included some through the winter. This may be related to an overall
ill-formed grains. A full morphometric analysis of the pattern of agricultural intensification, diversification
Paithan rices in the context of the wider context of and expansion aimed at increasing overall productiv-
grain shape variation in archaeobotanical materials in ity (as noted in the general discussion, above).
South Asia is warranted. The evidence from Paithan implies a significant
difference from the regional pattern in the Chalcolithic
period. In addition, in more recent times, it is free-
Wheat and barley threshing bread wheats that are most prominent, with
barley significantly less so. Glumed wheats, namely
The major winter cereals of South Asia have their emmer (Triticum dicoccum), were reported from the
origins in South-West Asia and have been transmitted to region in the 19th century (Watt 1889–1893; Fuller
India via the Indus region, where they were the staple 2006a: 22–24), and it is therefore of interest to know
crops of the Harappan civilization (Fuller and Madella the relative proportions of different winter cereal types
2001; Fuller 2006a: 20–35). In Maharashtra, they were through time. Distinguishing the glume wheat emmer
widespread and important cereals of the Deccan Chal- from free-threshing wheat can only definitively be done
colithic (Kajale 1991; Fuller 2003c: 372–374), with the on the basis of chaff remains, although grains tend to
more drought-tolerant barley in particular dominat- be clearly distinct, and on this basis, some plumper
ing the archaeobotanical record at Inamgaon (Kajale grains have been assigned to the free-threshing cate-
1988). Further south, in the Southern Neolithic, these gory and others to the emmer-type category (Figs 12.8–
cereals were occasional and of low frequency, sug- 12.10). This suggests that there is a fairly even split
gesting that they may have had some special status between emmer and free-threshing wheat at Paithan,
(Fuller et al. 2004; Fuller 2005), and they might have at least in Period 3, where there are sufficient numbers.
been cultivated in small, garden-like plots, where they
 Chapter 12 Archaeobotany    259

Fig. 12.9: Grains of free-threshing wheat (Triticum aestivun/

durum) from context 1036; dorsal view.

within both bread wheat and durum wheat, it is pre-

mature and misleading to refer archaeological mate-
rial to ‘Triticum sphaerococcum’ until more systematic
metrical studies of populations have been carried out
on both modern and archaeological Indian wheats. A
Fig. 12.8: Grains of barley (Hordeum vulgare), hulled variety, from definitive identification would also be aided by the
context 721; ventral view (above) and dorsal view (below). recovery of charred rachis remains, which have proved
elusive in the Paithan samples.
It should be noted that no attempt here is made to
identify Triticum sphaerococcum, ‘Indian shot wheat’.
Although T. sphaerococcum has been widely reported in African millets (sorghum, pearl millet and finger
South Asian archaeobotany, its identification as a dis- millet) (Figs 12.11–12.13)
tinct genotype group of hexaploid breadwheats on the
basis of grain remains is dubious (Fuller 2002: 273– These three cereals (Hindi jowar, bajra and ragi), which
277; 2006a: 23). No clear criteria exist for recognizing were brought into cultivation in three different parts
this species from grain remains. The reported criteria of Africa (Fuller 2003b; cf. Marshall and Hildebrand
of some authors contradict each other and if found in 2002), are of widespread importance in South Asia
Neolithic Europe would fit with known free-threshing today, and especially on the Indian peninsula (FAO
tetraploids (cf. Maier 1996). Given the impact of car- 1995; Possehl 1986). These species are on the whole
bonization, which tends to make grains more spheri- fairly minor at Paithan, although both sorghum and
cal through differential shrinkage of length, and the pearl  millet show major increases in frequency in
potential wide range of variability of grain shapes Period 3. This may suggest that the beginning of the
process by which these species rose to prominence in

Fig. 12.10: Grain of emmer wheat type (Triticum diococcum) from context 919; dorsal view (left) and lateral view (right).
260   Dorian Q. Fuller

Fig. 12.11: Pearl millet grain (Pennisetum glaucum), narrow grain type, from context 930; dorsal view (left) and dorsal view (right).

Fig. 12.12: Sorghum (Sorghum bicolor) of race bicolor type from context 930; dorsal view (right). Sorghum of race durra type from
context 1092; dorsal view (right).

South India began only in the 3rd or 4th century. Some ality to the indigenous millets and have an advantage
years ago, Possehl (1986) postulated that the introduc- in being free-threshing (true of pearl millet and finger
tion of these species from Africa may have been funda- millet, and advanced durra type sorghums; see Reddy
mental to the development of double-cropping in the 2003; Fuller and Weber 2005), and this implies fewer
greater Indus region and especially Gujarat, and paved crop-processing stages and labour needed to take
the way for the development of agriculture in peninsu- them from the field to the plate. Nevertheless, cultural
lar India. Subsequently, systematic archaeobotanical preference may have operated against them in favour
work in Gujarat suggested that the African millets of the traditional small millets, or else the persistent
were added to cultivation based on indigenous millets, husks of the small millets may have been favoured
especially Panicum sumatrense (Weber 1991, 1998), for the advantages that the husks provide in storage
and research in South India established the primacy as extra protection against fungus and to some extent
of a suite of native crops with Brachiaria ramosa and against insects. In this regard, it is of interest to note
Setaria verticillata millets (Fuller et al. 2004). What that earlier finds of sorghum in Periods 1 and 2 – when
is intriguing about the evidence from Paithan is that referable to variety – appear to be hulled bicolor type
it implies that even though these species were intro- sorghums, whilst the free-threshing larger-grained
duced already in the Chalcolithic period (see Fuller durra is present only in Period 3. It should also be noted
2003b), they may have remained fairly minor crops for that recent research suggests that a local trajectory of
the first two millennia or so that they were in India. evolution from bicolor to durra sorghum took place in
The reasons for this are not immediately apparent. In the Middle Nile Valley (Nubia) in the early centuries
general terms, they have a similar ecology and season- ad, and thus, it may be that this free-threshing form
 Chapter 12 Archaeobotany    261

became available by diffusion from Africa only in the itself more readily to destruction by fire. At present,
3rd century or so (see Clapham and Rowley-Conwy there is so little good archaeobotanical evidence for
2007; Shinnie and Anderson 2004: 366). this species, from either East Africa or India, that its
early cultivation remains truly obscure.

The Chinese millets (broomcorn millet and common

foxtail millet) (Figs 12.14 and 12.15)

These two species, Panicum miliaceum and Setaria

italica, are both widespread in India today, and the
latter, in particular, is a major crop in South India.
At Paithan, there are only a few occurrences of broom-
corn millet in Periods 2 and 3, while Setaria italica is
confirmed for both of these periods and probably in
significant proportions, having a ubiquity of 10–20%.
A complicating factor is that charred grains of Setaria
italica are very difficult to separate from grains of the
indigenous Brachiaria ramosa, even their rugose husk
patterns are similar, although the rugae are slightly
coarser in B. ramosa, which has grains that tend to
be somewhat flatter and with a relatively larger hilum.
These distinctions, however, are made more problem-
atic by the distortions introduced by charring.

Fig. 12.13: Finger millet (Eleusine coracana), fragmented, from

context 721; basal view of embryo. Approximately half of the grain
is preserved.

There have been recent concerns over identification

criteria for distinguishing these cereals (see Fuller
2003b for a review). While some reports of sorghum on
the basis of ceramic and mudbrick impressions appear
to have been in error, most reports on charred macro-
remains appear valid, and specimens from Paithan
are well preserved and readily recognizable. This is
also true of pearl millet. Finger millet appears to have
been the most widely mis-reported. Clear grains, with
characteristic globular-angular shapes and minute
Fig. 12.14: Grain of common broomcorn millet (Panicum mili-
surface warts, were recovered only from one sample in aceum) from context 1128; dorsal view.
Period 1 at Paithan (Fig. 12.13). Previously, the author
has found one distinct specimen from Iron Age Hallur
Further re-examination of this material is warranted;
(Fuller et al. 2004) and ten from Early Historic Perur
only for a few samples has the author attempted
in Tamil Nadu (Cooke et al. 2005). These extremely
these divisions, and thus for many samples, only a B.
small numbers raise the question as to whether there
ramosa/S. italica category is counted. As will be seen,
is some inherent preservation bias acting against this
this accounts for about 10% of all seed finds in Period
species. Compared to other cereals, it has a distinc-
3, indicating major significance. Future work needs to
tively thicker pericarp, which is somewhat sticky and
focus on determining the relative contribution of these
oily in appearance. It is conceivable that this lends
262   Dorian Q. Fuller

Fig. 12.15: Grain of common foxtail millet (Setaria italica) from context 1135; dorsal view (left) and ventral view (right), both with
adhering lemma. Notice on ventral view the indentation of charred away hilum which is relatively short compared to overall grain
length (in contrast to Brachiaria ramosa).

two species, as at some stage, there must have been a somewhere on the subcontinent (see Fuller 2006a: 18).
transition from prehistoric dominance of Brachiaria At least five of these were present at Paithan, including
ramosa, the native crop of the Southern Neolithic kodo millet (see above) and the other genera discussed
and probably the Deccan Chalcolithic (Fuller 2003b, here (Panicum, Brachiaria, Echinochloa and Setaria).
2006a; Fuller et al. 2004; Cooke et al. 2005), to recent One difficulty with all of these species is that they
historical dominance by Setaria italica, with Brachiaria are wild components of the flora in India and occur
persisting in cultivation only in isolated peripheral as weeds of other crops (including other millets and
pockets, where it is maintained apparently for ritual rice), but they are also known in cultivated or domes-
uses (Kimata et al. 2000). This transition amongst the ticated forms. For the archaeobotanist, we do not yet
small millets represents a major, still undocumented, have clear and unambiguous means for determining
change in the history of Indian agriculture. whether finds of these represent domesticated crops
Foxtail and Broomcorn millets are associated with or weeds. For early periods, such as the Southern Neo-
the earliest agriculture in North China in the Yellow lithic, this has been inferred on the basis of high rela-
River basin (Crawford 2006), and both species appear tive frequency and ubiquity of occurrence (see Fuller
to have been domesticated there. Nevertheless, recent 2003c; Fuller et al. 2004), but at an agriculturally
genetic studies on modern foxtail millet varieties diverse site such as Paithan, this is somewhat more
suggest more than one domestication centre, and a problematic. Nevertheless, these species are likely to
distinctive wild race found today in Afghanistan and have been cultivars, and most of them (with the excep-
northern Pakistan may have given rise independently tion of Brachiaria ramosa) can be found in cultivation
to some domesticated landraces (Fukunaga et al. 2002, in Maharashtra today (Kimata 1989; Kobayashi and
2006). This being the case, it need not surprise us that Kimata 1989; Kimata and Sakamoto 1992). B. ramosa
Setaria italica reports, for example at Rojdi A (Weber was the staple cereal of the Southern Neolithic (Fuller
1991), predate those of Panicum miliaceum or other et al. 2004), and it may also have been cultivated in the
potential Chinese imports, which appear mainly after Neolithic Ganges, Gujarat and Maharashtra, although
1900 bc (Fuller 2006a: 36). Nevertheless, the identifi- difficulties with precise identification and determining
cation challenge of separating S. italica and B. ramosa whether it might be present in the wild form, as a weed,
needs to be tackled more widely. raise interpretative problems. Its presence at Paithan,
as well as Early Historic Tamil Nadu (Cooke et al. 2005),
indicates that it was more widespread in the past than
The Indian small millets (little millet, browntop today (cf. Kimata et al. 2000; Kingwell-Banham and
millet, sawa millet and other Setaria spp.) (Fig. 12.16) Fuller 2014). As noted in the discussion of S. italica
(above), determining the timing and regions in which
There is a wide range of small millets cultivated in this crop declined remains an important issue. In the
South Asia today, and many of these were domesticated Southern Neolithic, the recurrent presence of Setaria
 Chapter 12 Archaeobotany    263

Fig. 12.16: Grain of browntop millet (Brachiaria ramosa) from context 1127; ventral view (left) showing adhering palea (centre) and dorsal
view (right). Notice (left) faint traces of hilum, which is relatively long compared to overall grain length (1/4–1/3 length).

verticillata has been interpreted as having been a (Smartt 1990; Zohary and Hopf 2000: 92). They can
co-staple with Brachiaria (Fuller et al. 2004). It is not generally be grown on most soils except those which
clear whether this species was ever domesticated, but are waterlogged. In general, those pulses native to the
it may have been a tolerated and consumed weed. Its tropics (India or Africa) have a monsoon-adapted sea-
co-occurrence at Paithan in some quantity in Period 3, sonality, while those that spread in prehistory from
together with B. ramosa and S. italica, suggests that this the Near East (including lentil, pea, chickpea and
continued to be the case into Medieval times. Panicum grasspea) are winter crops. They are important also in
sumatrense was the staple cereal of Protohistoric Sau- human nutrition as sources of protein and thus comple-
rashtra (Weber 1991: 84–85) and was also cultivated at ment the carbohydrate-rich cereals. Indeed, in much
Harappa (Weber 1999). The process and timing by which of India today, pulses are probably the main protein
it spread to the peninsula remain unclear, but it persists source for vegetarians, and in the form of dhal, they
as a widespread minor cereal today. The early history are a regular part of Indian meals. Pulses can also be
of sawa millet (Echinochoa colona) remains obscure. It made into flour, a key ingredient in many distinctive
may be, like kodo millet, a secondary domestication southern Indian food products, such as dosa, idli and
of an early rice or millet weed. Early finds of probable vadai. It is noticeable that archaeologically, pulses are
weedy type include those from the Southern Neolithic a prominent part of the archaeobotanical evidence in
(Fuller et al. 2004) and Rojdi Phase C (Weber 1991). peninsular India already in prehistory, in many cases
occurring in equal or greater quantity than cereals
(Fuller et al. 2004; Fuller and Harvey 2006). This is a

Pulses notable contrast from the quantitative frequency of

pulses in other world regions, such as the Near East,
Europe or the Indus Valley. It can be hypothesized that
As already noted above, pulses are important in crop
this resulted from the fact that the distinctive impor-
rotations because they are nitrogen-fixing plants

Fig. 12.17: Relative frequency of pulses.

264   Dorian Q. Fuller

Table 12.4: The presence/absence of pulse crops on peninsular Indian sites of Iron Age or Early Historic date. Sources as for Table
12.2. (IA = Iron Age; IV = Hallur period IV; EH = Early Historic).

Lentil

No. pulse crops

Pea

Chickpea

Grasspea

Mung bean

Urd bean

Moth bean

Horsegram

Pigeonpea

Hyacinth bean

Cowpea
Period Site

800 bc ΙΑ Adam Cave - - - + - + - + - + - 4

  ΙΑ Bhagimohari + + - + +? + - + + + - 8
  ΙΑ Bhokardan + + + + - + - - + + - 7
  ΙΑ Naikund + + - - - + - - - - - 3
  ΙΑ Veerapuram - - + - - + - + - + - 4
  IV/ΙΑ Hallur - - - - + + - + - + - 4
300 bc–? EH Piklihal - - + + + - - + - + - 5
  EH Veerapuram -  - + - - - - + - + - 3
  EH Adam Cave + + - + + + - - - + - 5
  EH Bhatkuli - - + - - - - - - - - 1
  EH Bhokardan + + + + + + - - + + - 8
  EH Dangwada + - - + - - - - - - - 2
  EH Kaundinyapura - + - +  - - - - - - - 2
  EH Nevasa + + + + + + - + + + - 8
  EH Paunar - - - - - - - - - - - 0
  EH Ter + + - + + + - + - - - 6
  EH Kodumanal - - - - + + - + - + + 5
  EH Mangudi - - - - - - - - - - - 0
  EH Perur - - - - + + - + - + + 5
300–0 bc   Paithan I + ? - - + - + + - + - 5
ad 0–400   Paithan II - ? + - + + + + + + - 7
ad 400–700   Paithan III + ? + + + + + + + + + 10
    Total sites 10 8 9 11 12 14 3 13 6 15 3  

tance of pulses, and especially pulse flours, associated species have, unfortunately, caused a degree of confu-
with southern Indian foods had already developed on sion (see Verdcourt 1971; Fuller 2002: 334, n.6; Fuller
the Indian peninsula and that this distinctive pulse and Harvey 2006). These two species had separate
preparation style, which often involves dry-roasting, led origins from distinct wild progenitors, and current evi-
to more frequent carbonization (Fuller and Harvey 2006). dence suggests that they had non-overlapping origins,
The relative frequency of pulses at Paithan is illustrated with a suggested epicentre of early urd cultivation in
in Fig. 12.17. A comparative table on the presence of pulses Gujarat/north-west Maharashtra and two early centres
on the Indian peninsula during Iron Age and Early His- of mung cultivation in the Eastern Harappan zone and
toric times is shown in Table 12.4. As with cereals, Paithan the South Indian Neolithic (Fuller and Harvey 2006;
provides evidence for a wider range of pulses than other Fuller 2007). Distinguishing these species may prove
sites, a likely reflection of more extensive sampling. difficult, especially if charred seeds are whole (with
the cotyledons still together), but the seed coat and
hilum have been destroyed by charring. If preserved,
Mung and urd beans (Figs 12.18 and 12.19) the hilum clearly distinguishes these taxa, with a much
larger, protruding hilum in V. mungo and a smaller
An important group of related pulses are those of the hilum in V. radiata that is flush with the seed surface,
genus Vigna, subgenus Ceratotropis, which is focussed but very few such examples are available from Paithan,
in Asia (Tomooka et al. 2003). This includes two closely as was also true in the Southern Neolithic (Fuller et
related species of particular importance in India, the al. 2004). The criteria used in this study are those of
mung bean (or green gram), V. radiata, and urd bean Fuller et al. (2004) and Fuller and Harvey (2006), with
(or black gram), V. mungo. The scientific names of these a focus on plumule length (visible on split cotyledons)
 Chapter 12 Archaeobotany    265

to overall seed length. These species have been widely 2004), while urd is also prominent on the northern
recovered from archaeological sites throughout India. peninsula. Despite the Early Historic date of Paithan,
Vigna specimens from the site can contribute to our
understanding of the domestication process in pulse
species. As with cereals, the domestication syndrome
in pulses is expected to include grain size increase
as well as a shift to non-dehiscent pods and loss of
seed dormancy. While the latter two traits must have
evolved early, it has been suggested that seed size
increase did not and that this is a recurrent feature of
pulses (Fuller and Harvey 2006; Fuller 2007). It may
have required more intensive cultivation regimes,
such as ard-ploughing, resulting in deeper burial
of planted seeds to select for seed size increase. This
assumes an unconscious mechanism, but it is also
possible that farmers eventually took the conscious
decision to select those crops that produced bigger
seeds. Evidence for this lag time in size increase
comes from the fact that early pulses, such as Vigna
measured from Southern Neolithic sites, are small
and within the size range of modern wild populations
Fig. 12.18: Interior view of the cotyledon of mung bean (Vigna
radiata) from context 1150. Notice the relatively long plumule.
(which have been corrected for shrinkage due to char-
ring), but that at later sites, particularly of Iron Age
or Early Historic date, larger seeds, which would fit
within the range of modern domesticated crops, are
present. Some measurements from Paithan Vigna (on
an initial representative assemblage rather than all
examples) illustrates this in comparison to Southern
Neolithic measurements (Fig. 12.20).

Moth bean, Vigna aconitifolia (Fig. 12.21), is another

closely related pulse crop domesticated in South Asia.
It can be distinguished from V. radiata and V. mungo
by its minute hilum, which lies in a slight depression;
it also has a distinctively short plumule length visible
on split cotyledons. Unfortunately, it has not been ade-
quately mapped in terms of its wild and weedy distri-
bution, although it is not considered native to southern
India (Cooke 1903–1908; Gamble 1935; Saldanha 1984)
and thus may be a northern or central Indian domes-
ticate (cf. Tomooka et al. 2003). In addition, early
archaeological finds span from Harayana, includ-
ing Harappan Burthana Tigrana and Late Harappan
Fig. 12.19: Interior view of the cotyledon of urd bean (Vigna mungo)
from context 1150. Notice the shorter, half-seed-length plumule.
Mitathal, through the middle Ganges, at several sites
in the 1300–600 bc horizon (Narhan, Imlidih-Khurd
Period II, Raja-Nala-Ka-Tila Period II, Senuwar Period
At Paithan, the Indian Vigna species dominate the
II, Kharadih Period II and Saunphari; see Chanchala
pulse assemblage (Fig. 12.17). This is not unexpected
2004; Saraswat 2005). Together, these archaeological
as mung bean is one of the most frequent and ubiq-
finds suggest a north Indian origin. Finds on the pen-
uitous pulses in Neolithic South India and Chalco-
insula are lacking (see Table 12.4), and it may be that
lithic Maharashtra (Kajale 1988, 1991; Fuller et al.
266   Dorian Q. Fuller

Fig. 12.20: Comparison of measurements of Paithan Vigna with samples from Neolithic contexts in southern India.

the appearance of this species at Paithan indicates its Chalcolithic reports of Vigna unguiculata at Hulas and
initial spread to the south. Daimabad (Fuller 2003b), although in neither case is
the evidence unambiguous. More recently, additional
finds have been reported from the Late Chalcolithic/
Iron Age Ganges at Kharadih and Raja-Nala-Ka-Tila in
Uttar Pradesh (Saraswat 2005), while those of Early
Historic Sanghol are unambiguous (Pokharia and Sar-
aswat 1999).

Horsegram (Macrotyloma uniflorum (Lam.) Verdcourt)

(Fig. 12.22), known in Hindi as kulthi, is one of the most
widespread pulses (grain legumes) of modern and ancient
Fig. 12.21: Moth beans (Vigna aconitifolia) from context 1136.
India. It is grown as a small shrub throughout the semi-arid
or savannah regions of South Asia. It is highly drought-
resistant and yields a crop in 3–4 months. It is thought
to be native to a region with such a climate, although
very little botanical research on wild populations has
been carried out. A limited herbarium survey by the
author indicates wild populations in Rajasthan, such
as around Mount Abu, a few north peninsular finds
and several in the Southern Deccan, all in the drier
semi-arid savannah zone (Fuller and Harvey 2006).
Archaeologically, horsegram is one of the most widely
Fig. 12.22: Horsegram (Macrotyloma uniflorum) from context 467;
complete seed (left) and interior of split cotyledon (right).
reported pulses in prehistoric/protohistoric India (cf.
Fuller and Harvey 2006), and it is the most widespread
pulse find in peninsular India in Iron Age or Early His-
Cowpea (Vigna unguiculata), Hindi lobiya, is repre-
toric times (Table 12.4). Its seeds are distinctly rectan-
sented by a single unambiguous specimen from Period
gular-trapezoidal with a sharp edge. It is a pod-thresh-
3. This species is native to Africa and was most likely
ing pulse, and as such, seeds may be more prone
brought into cultivation in the West African savannah
to loss during processing. Its flat seed shape may also
woodland transition, perhaps within modern Ghana
provide for preservation bias under charring, as seems
(Vaillancourt and Weeden 1992; Ng 1995; Fuller 2003b:
to be the case with lentils. Archaeobotanical finds
241; D’Andrea et al. 2007). This author’s previous review
indicate that it was cultivated from Southern India
of crops of African origin in South Asia accepted two
 Chapter 12 Archaeobotany    267

Fig. 12.23: Pigeonpea (Cajanus cajan) from context 1036; lateral view (left) and interior view of split cotyledon (right).

to Haryana and the middle Gangetic basin by 2200 occurs further east into South-East Asia and China. The
bc. While this species has been suggested to be part form at Paithan was presumably a pulse type.
of the indigenous Southern Neolithic package (Fuller
2002: 296; Fuller et al. 2004), current archaeological
evidence might suggest independent domestication(s) Pigeonpea Cajanus cajan (Hindi toor dhal and arhar)
in northern/western India (Fuller & Murphy 2018). (Fig. 12.23)

This is a major pulse throughout India and often a

Hyacinth bean staple dhal variety. Nevertheless, it has only been
reported from relatively few sites in India to date,
Finds of Lablab purpureus (Hindi sem) in India are including Tuljapur Garhi, Peddamudiyam and San-
earlier than any available from Africa, although it ganakallu, all from the Deccan in roughly the mid-2nd
is generally regarded as originating in Africa (Fuller millennium bc, with similar age finds from Gopalpur
2003b; Fuller and Harvey 2006). Early finds in Africa in Orissa (Fuller et al. 2004; Fuller and Harvey 2006;
include early-centuries-ad finds in Nubia (Clapham Harvey et al. 2006). It is a rare find at Paithan, as on
and Rowley-Conwy 2007) and in Southern Africa (Smith other sites where it occurs, and it may be that there are
and Jacobson 1995). By contrast, Lablab appears to persistent preservation biases working against this
have arrived in India relatively early, probably in the species.
first half of the 2nd millennium. Out of 90 sites with
evidence for pulses, Lablab has been reported from 19,
of which 15 are on the peninsula (Fuller and Harvey Winter pulses (chickpea, grasspea, pea, lentil) (Figs
2006). It appears that this crop became widespread 12.24 and 12.25)
in peninsular India by c. 1500 bc, with an early direct
AMS date of c. 1600 bc (Fuller and Harvey 2006; Fuller Cicer arietinum (Hindi chana), Lathyrus sativus (Hindi
et al. 2007). A few earlier finds may come from the khesari), Pisum sativum (Hindi matar), Lens culinaris
Eastern Harappan zone, with reports from Late Harap- (Hindi masur) – these four pulse species have their
pan Sanghol and Mahorana, but dating evidence is origins in the Near East in the same general regions
inadequate to place these securely much earlier than as wheat and barley (Smartt 1990; Zohary and Hopf
1700–1500 bc. In Africa, this species is grown mainly 2000). While they spread east to South Asia in prehis-
as a pulse, for its dried seeds, whilst in India, it has tory, the earliest evidence for these species comes from
both pulse and green vegetable varieties, the latter the Early and particularly the Mature Harappan period
being grown for its green pods, which are harvested (Fuller and Madella 2001). All species were present
before seed maturation as a vegetable (Smartt 1990). in Maharashtra in the Chalcolithic period, but they
The latter variety presumably evolved in India and pre- appear to have been less favoured and largely absent
dominates more towards the north and east and also further south (Kajale 1991; Fuller 2005). They remain
268   Dorian Q. Fuller

frequent finds on the northern peninsula through the are also a few tree fruits, including the ubiquitous
Iron Age and Early Historic periods. As already dis- Ziziphus, emblic myrobalan (Phyllanthus emblica) and
cussed, one aspect of their importance lies in nitro- fig seeds (not necessarily of a cultivated species) (Fig.
gen-fixation, which allows them to be cropped in the 12.27). Possible okra seeds (Abelmoschus sp.) deserve
winter on soils that have already been used for cereals note and further investigation in order to determine
or other crops in the summer; their use in this way can whether domesticated versus wild progenitor species
help to maintain soil fertility. might be involved. Also worthy of further investigation
are Solanum seeds, listed in this report under weeds
(as there are numerous weed taxa in this genus), in
relation to the possible presence of eggplants (Solanum
melongena), for which clear archaeobotanical identi-
fication criteria are not yet available. It should also
be noted that there is evidence of probable ‘mouse
cucumber’, Melothria heterophylla (Fig. 12.28), a wild
species throughout India and much of South-East Asia
that is eaten (from wild or weedy plants) but not cul-
tivated. The Melothria-type seeds had been previously
referred to as ‘cf. Cucumis prophetarum’ in reports on
the Southern Neolithic (Fuller et al. 2004: Fig. 6H),
which testifies to the fact that the gamut of cucurbi-
taceae species in South Asia, both wild and domesti-
cated, still requires further comparative study.
Fig. 12.24: An example of chickpea (Cicer arietinum); ventral view.

Fig. 12.26: Seed of mustard (Brassica cf. juncea) from context 985,
Fig. 12.25: An example of lentil (Lens culinaris); lateral view. with part of seed coat cell pattern visible.

Sesame (Sesamum indicum L. syn. S. orientale L.)

Some other field crops (Fig. 12.29) is an oilseed crop of global importance.
This crop is widely cultivated in South Asia through
Burma, and somewhat more patchily eastwards. It is
There is a range of other crops at Paithan, and the
now firmly established that sesame has a South Asian
section below will include five oilseed/fibre crops of
origin (Bedigian and Harlan 1986; Bedigian 1998, 2003;
particular note. In addition to these, there is evidence
Fuller 2003a), although details about when, and more
for probable brown mustard (Brassica juncea) culti-
precisely where, and the ecological circumstances
vation (Fig. 12.26) and possible cucurbit cultivation,
of domestication require research. Origins in the
including possible melons or cucumbers (which can
greater Indus region seem likely. At present, the ear-
be difficult to distinguish on charred seeds). There
liest archaeological evidence comes from the mature
 Chapter 12 Archaeobotany    269

urban phase of the Harappan civilization (2500–2000

bc), at which time sesame would appear to have been
an important oilseed crop (Fuller and Madella 2001;
Fuller 2003a). Finds are nevertheless few, probably
owing to the low preservation potential of oily seeds,
and include finds from Miri Qalat (Tengberg 1999) and
Harappa (Vats 1940: 467; Weber 1999, 2003) and Late
Harappan Sanghol. In the case of Miri Qalat, sesame
is absent from earlier levels, back to 4000 bc, indicat-
ing that it is likely to have been introduced to this site
during the Harappan period. South Asian finds from
the 2nd millennium bc and later are fairly numerous,
especially from the Middle Ganges region, including
Imlidh-Khurd Period II, Narhan, Kharaidih Period II,
Senuwar Period II and Raj-Nala-Ka-Tila Period II (Sar-
aswat 2004, 2005). The cultivation of this species can Fig. 12.28: Seed of mouse cucumber (Melothria cf. heterophylla,
be considered widespread in the subcontinent by c. family Cucurbitaceae) from context 1116.
1000 bc.
As an oilseed, there are strong biases against preser-
vation by charring; thus, it is not surprising that find-
counts are low. At Paithan, some 67 seeds (<1% by
frequency) have been recovered from 12 samples (16%
ubiquity), suggesting fairly widespread use. Unfor-
tunately, a large number of these come from unphased
samples. In phased samples, Sesamum increased in ubiq-
uity and frequency through Periods 2 (Satavahana) and 3
(Early Medieval) (Fig. 12.30).
Sesame grows best in well-drained soils with long-
warm growing seasons and is highly drought-resistant.
It generally has similar climatic requirements to cotton
and is well suited to savannah environments. It is a
Fig. 12.27: Seeds of fig type (Ficus sp.) from context 619. short-day plant growing to maturity more quickly
through the summer (Griffee 2001; Burden 2005).

Fig. 12.29: Sesame seed (Sesamum indicum) from context 930, whole seed (left) and detail of seed coat (right).
270   Dorian Q. Fuller

Fig. 12.30: Ubiquity of selected crop types and weeds.

Cotton (Gossypium sp.) is represented by charred by finds at Mehrgarh in Pakistan in the 6th millennium
whole seeds, seed fragments and sometimes a minute bc (Costantini 1983; Moulherat et al. 2002). Archaeo-
funicular ‘cap’ structure (Fig. 12.31). Although identi- logical evidence for cotton in South Asia is fairly exten-
fication at species level seems impossible on the basis sive (see Table 12.5; Fig. 12.32). Cotton is presumed to
of preserved anatomical features, it is fairly safe to have been an important ‘cash crop’ of the Indus region
presume that these early cottons represent the indig- during the Harappan civilization, and several sites
enous species tree cotton (G. arboreum). have yielded finds of cotton seeds or possible textile
Cotton is grown primarily for its unique seed hairs remains, including Harappa and Mohenjo-daro (see
that are readily spun and turned into threads and tex- Weber 1999; Fuller and Madella 2001). During this
tiles. However, cotton seeds can also be pressed for era, cotton was unknown in Mesopotamia or Egypt,
oil once the fibre is removed, and with refining, this where textile production was focussed on flax (Zohary
is edible. The pressed seed cakes also make a tradi- and Hopf 2000). Beyond the Harappan zone, cotton
tional fodder (Rehm and Epsig 1991: 343). There are appears to have spread only in the post-urban period,
four cotton species that are cultivated for their seed including Late Harappan era finds on the eastern and
fibres today, including two of New World origin (G. south-eastern zones of the Harappan civilization, and
hirsutum and G. barbadense), which have become of from the mid to late 2nd millennium bc, finds are
global importance in the past 300 years. Indigenous reported from a number of sites in the middle Ganges
to the Old World are tree cotton (G. arboreum) and region. The earliest hard evidence from South India is
short-staple cotton (G. herbaceum) (Wendel 1995). Tree that from the later levels at Hallur, Karnataka, which
cotton, G. arboreum L., is now considered most likely produced a direct AMS date of c. 950 bc (Fuller et al.
to be of South Asian origin (Zohary and Hopf 2000; 2007).
Fuller 2002), and its early use is probably represented

Fig. 12.31: Cotton (Gossypium sp.) remains from context 930, including fragmented seed (left), cross-section of seed coat (centre)
and funicular cap from inferior seed (right).
 Chapter 12 Archaeobotany    271

Table 12.5: Archaeobotanical and textile remains of cotton from South Asia.

Site Evidence Period Reference(s)

Mehrgarh, Baluchistan Seeds (uncharred); Ceramic Neolithic, 6000–4500 bc Costantini 1983;

mineralized thread in Costantini and Biasini 1985: 24;
copper bead Moulherat et al. 2002
Mohenjo-daro Cloth Mature Harappan, 2600–2000 bc Gulati and Turner 1929
Balakot, Sindh Malavaceae pollen type, Mature Harappan, 2500–2000 bc McKean 1983;
comparable to Gos- also mentioned in Dales 1986
sypium
Harappa Seed(s)/imprints Mature Harappan(?), 2600–1900 Weber 1999: 818;
bc Vats 1940: 466.
Kunal Seed(s) Mature Harappan, ?2500–2000 bc, Saraswat and Pokharia 2003
perhaps equivalent to Harappa 3C(?),
2200–1900 bc
Banawali Seed(s) Mature Harappan (?=Harappa 3C), Saraswat et al. 2000;
2200–1900 bc Saraswat 2002
Sanghol Seed(s) Late Harappan, 1900–1400 bc Saraswat 1997
Hulas Seed(s) Late Harappan, 1800–1300 bc Saraswat 1993
Kanmer, Kacchh Seed(s) Late Harappan, 2000–1700 bc Pokharia in Kharakwal et al. 2007
Imlidh-Khurd, Gorakhpur, Seed(s) Period II, 1300–800 bc Saraswat 2005
Uttar Pradesh
Waina, Ballia, Uttar Pradesh Seed(s) Period I, 1600–800 bc Saraswat 2005
Sringaverapura, Dist Alla- Fibres Late Ochre-Coloured Pottery, 1200–700 bc Saraswat 1986
habad, Uttar Pradesh
Hallur, Upper Tungabhadra, Seeds and fragments Early Iron Age, AMS direct date: 950–900 Fuller et al. 2004;
Karnataka bc for dating: Fuller et al. 2007
Sanghol, Ludhiana Dist., Seed(s) Early Historic, Kushana, 200 bc–ad 300 Pokharia and Saraswat 1999
Indian Punjab
Hund, Peshwar Dist., Pakistan Seeds and fragments Kushana through Mughal (all periods), 200 Fuller, unpublished data;
bc–ad 1600 Cooke 2002
Kausambi Seed(s) NBP horizon, 550–250 bc Chanchala 1995
Nevasa (1954–1956 season) Seed(s) Early Historic, 250 bc–ad 250 Anonymous, in Sankalia et al. 1960:
529–530
Kodumanal, Coimbatore Dist., Seeds and fragments Early Historic/Late Megalithic, 300 bc–ad Cooke et al. 2005
Tamil Nadu 300
Perur, Coimbatore Dist., Tamil Seed and fragments Early Historic/Late Megalithic, 300 bc–ad Cooke et al. 2005
Nadu 300
Mangudi, Madurai Dist., Tamil Seed and fragments Early Historic/Late Megalithic, 300 bc–ad Cooke et al. 2005
Nadu 300
Ufalda, Garhwal, Uttaranchal Seeds and fragments Early Historic(?), ad 0–600 Fuller, unpublished data
Singh-Bhagwantpur, Rupna- Seed(s) Medieval, ad 800–1100 Vishnu-Mittre et al. 1984
gar Dist., Indian Punjab
Mangali Luduwala, Haryana Seed(s) Sub-recent, ad 1500–1900 Willcox 1992
272   Dorian Q. Fuller

Fig. 12.32: The distribution of archaeological finds of cotton and flax seed in South Asia, indicated by broad time horizons. Sites
numbered with data references: 1. Mehrgarh (Moulherat et al. 2002); 2. Nausharo (Costantini 1990); 3. Pirak (Costantini 1979); 4. Miri
Qalat (Tengberg 1999); 5. Mohenjo-daro (Gulati and Turner 1929); 6. Balakot (McKean 1983); 7. Hund (Fuller, unpublished data); 8. Loebanhr
3 (Constantini 1987); 9. Harappa (Weber 1999, 2003); 10. Kunal (Saraswat and Pokharia 2003); 11. Banawali (Saraswat 2002); 12. Sanghol
(indicating Late Harappan and Early Historic evidence) (Saraswat 1997; Pokharia and Saraswat 1999); 13. Hulas (Saraswat 1993); 14. Bala-
thal (Kajale 1996b); 15. Ojiyana (Pokharia 2008); 16. Kanmer (Pokharia 2007); 17. Babor Kot (Reddy 1994, 2003); 18. Rojdi (Weber 1991);
19. Hulaskhera (see Chanchala 1992, plate 5, 24, interpreted by the present author as cotton seed fragment); 20. Charda (Saraswat
1993; Chanchala 2002); 21. Imlidh-Khurd (Saraswat 2005); 22. Narhan (Saraswat et al. 1994); 23. Waina (Saraswat 2005); 24. Sringavera-
pura (Saraswat 1986); 25. Kausambi (Chanchala 1995); 26. Senuwar (Saraswat 2004); 27. Raja-Nala-Ka-Tila (Saraswat 2005); 28. Navdatoli
(Vishnu-Mittre 1961); 29. Paithan (this report); 30. Daimabad (Kajale 1977a); 31. Nevasa (Sankalia et al. 1960: 529–530); 32. Hallur (Fuller
et al. 2004); 33. Perur (Cooke et al. 2005); 34. Kodumanal (Cooke et al. 2005); 35. Mangudi (Cooke et al. 2005); 36. Mangali/Luduwala
(Willcox 1992); 37. Singh-Bhagwantpur (Vishnu-Mittre et al. 1984); 38. Ufalda (Fuller, unpublished data).
 Chapter 12 Archaeobotany    273

Fig. 12.33: The spread of cotton beyond South Asia: finds of seeds and textiles. Sites in India and Pakistan are unlabelled but repre-
sent those in Fig. 12.32 and Table 12.5. Other sites, with citations, are labelled as follows, with approximate dates indicated and nature
of evidence if not seed remains. Question marks indicate uncertain evidence in terms of identifications, dating or indications of textile use.
1. Volubilis, ad 780–1000 (Fuller, unpublished data; cf. Pelling 2005); 2. Jarma, c. ad 200 (Pelling 2005); 3. Essouk (Murray and Nixon 2006);
4. Ogo, pollen, ad 1100–1200 (Chavane 1986); 5. Tellem burial caves, Bandiagara escarpment, textiles, ad 1100–1500 (Marr 1978; Boland et
al. 1991); 6. Dia, ad 1000–1600 (Murray 2004, 2007); 7. Gao, ad 1200–1500 (Fuller 2000); 8. Dakleh Oasis, ad 200–400 (Thanheiser 2002);
9. Khargeh Oasis, ad 300–400 (Clapham and Rowley-Conwy 2009); 10. Tinnis, pre-ad 1200 (Clapham and Rowley-Conwy 2009); 11.
Karanis, thread, ad 300–500 (Wilson 1933); 12. Abu Sha’ar, textile, ad 300–700 (Bender-Jorgensen and Vogelsang-Eastwood 1991);
13. Quesir al-Qadim, textile (Eastwood 1982); 14. Berenike, textiles, ad 0–500 (Wild and Wild 1998, 2001, 2005); 15. Wadi Qitna, ad
300–400 (Strouhal 1984); 16. Afyeh, c. 3000 bc, seeds in dung (Chowdhury and Buth 1971: 2005); 17. Karanog, textiles, ad 100–300 (Grif-
fith and Crowfoot 1934); 18. Arminna West, textiles, ad 200–500 (Fuller, unpublished data); 19. Ballanna, ad 0–500 (Mayer-Thurman
and Williams 1979); 20. Qasr Ibrim, seeds, capsules, fibres and textiles, 25 bc–ad 1800 (Clapham and Rowley-Conwy 2009); 21.
Qustul, ad 0–500 (Mayer-Thurman and Williams 1979); 22. Second cataract cemeteries, ad 100–500 (Bergman 1975); 23. Nauri, seeds,
ad 1000–1400 (Fuller and Edwards 2001); 24. Meroe West cemetery, textiles (Griffith and Crowfoot 1934); 25. Axum, ad 400–800 (Board-
man 1999); 26. Dhuweila, fibres in plaster, 4400–3000 bc (Betts et al. 1994); 27. Sheheil, ad 700–800, Guftan, ad 1000–1100, Medād,
ad 1100–1200, Hrim, ad 1200–1300, charred seeds (Samuel 2001); 28. Maykop kurgan, textiles, ‘cotton-like’, 3700–3200 bc (Shishlina et
al. 2003); 29. Merv, seeds, ad 300–500 (Nesbitt 1993, 1994); 30. Sampul (Mallory and Mair 2000: 155); 31. Ordek (Mallory and Mair 2000:
152); 32. Niya (Mallory and Mair 2000: 165); 33. Arrow pointing in direction of Pemba island sites, Tumbe, ad 600–1000, and Chwaka, ad
1100–1600 (Walshaw 2005). For additional recent finds, see Viot 2019.
274   Dorian Q. Fuller

The quantity of cotton found at Paithan shows a marked scale for household needs, such production would be
increase over the lifespan of the site (Fig. 12.30), sug- unlikely to produce recurrent finds in archaeobotani-
gesting an increasing investment to cash-cropping and cal samples from an urban excavation. Thus, we can
a probably local textile production. conclude that there is likely to have been production
Tree cotton is normally harvested in the summer. It for consumption above the household level. Craft
is a crop that likes high temperatures but also requires crops such as cotton therefore constitute important
sufficient water. It is also important that in its final evidence of early ‘cash-cropping’ (sensu Sherratt
stages, as the cotton bolls form, that it is kept dry and 1999).
free of rain, or else fibres will become discoloured and
may become mouldy. In much of India, cultivation can
rely on the monsoon, but in the Indus region, it is likely
that systems of irrigation were developed. It is there-
fore conceivable that the spread of this crop to other
regions also involved irrigation.
The production of textiles is an important part of
craft-production economies in two ways: firstly to do
with craft and secondly with agriculture. In terms
of craft, textiles are labour-intensive and time-con-
suming to produce. They require labour for spin-
ning and weaving after the initial fibre processing,
which itself involves several stages: dehusking, seed
removal (ginning), cleaning and smoothing, bowing
or carding to separate the fine fibres, which are then Fig. 12.34: Safflower (Carthamus tinctorius) from context 841.
rolled and ready to be spun. Only then can weaving of
the threads or yarn be done. Accounts of traditional That cotton was an important trade commodity of
cotton processing from different regions are similar, peninsular India in Early Historic times is implied
including those from ancient India (Schlingloff 1974). by evidence from Roman importation of cotton cloth,
In recent centuries, much of this process has been including specifically from Paithan. Importation is
aided by mechanical developments, but traditional indicated archaeologically by cotton textile finds at
ginning was a labour-intensive method involving a the Roman-era port of Berenike on the Red Sea coast of
rolling pin (often metal) and a wooden board which Egypt, which are inferred to have come from India on
was used to force out seeds. Bowing is done with a the basis of Z-spun threads; these finds dominate the
bow-like instrument in which a vibrating cord helps archaeological textile record at the port (Wild and Wild
to loosen the separation of the fibres. The implica- 1998, 2001, 2005, 2008). Cotton is also indicated as an
tion of all this is that cotton production requires both import from India in Papyrus Vinod from Egypt (Casson
the expenditure of labour in cultivation, beyond 1990; Sidebotham 1991). Importation is also discussed
subsistence requirements, and in the household, in ancient Roman texts, including the Periplus Maris
in processing, but with the result of a commodity by Erythraei, a 1st century ad Roman mariner’s travel
which wealth can be accumulated and fairly easily guide, written in Greek, probably in Egypt, which is
transported, or sequestered by elites. The produc- associated with several sites in the Deccan. The site
tion of textile crops implies additional surplus pro- of Ter (reached overland from Paithan) is specifically
duction beyond what is required to feed families and named, as are Minnagara and Ujjain to the north of
communities. It is therefore necessarily production Paithan (Wild and Wild 2008; see Casson 1989). While
for trade, as fields of textile crops produce fibre far Paithan was not specifically named as a source of
beyond what individual households are likely to use cotton cloth, the archaeobotanical evidence suggests
or need. In addition, it seems unlikely that all farming that it was part of a broader Indian geography of cotton
households that produced cotton crops would have production. The distribution of archaeological finds,
also had the time and skill necessary to process raw whether in north-eastern Africa or Central Asia, is pre-
cotton through to finished textiles, and this implies dominantly from the Roman period (the Chinese Han
the trade of cotton to craftspeople. While it is conceiv- period) (Fig. 12.33), testifying to the widespread trade
able that cotton might be grown on a small, garden in cotton during the Indian Early Historic period.
 Chapter 12 Archaeobotany    275

Flax/linseed (Linum usitatissimum) is present in one late as the end of the 5th or 4th millennium bc (Zohary
sample. This species originated in the Near East and and Hopf 2000). By the end of the 4th millennium bc,
is widely cultivated for its edible oily seeds, as well as it had spread in India, as it has been found in quantity
for its fibres. Divergent varieties have been developed in a jar from an early level at Balathal in Rajasthan
for each of these specialized functions (Zohary and (Misra and Mohanty 2001). During the 2nd millennium
Hopf 2000). The seed oil is important for human con- bc, it had become widespread in South Asia, with finds
sumption but is not easy to store once extracted from reported from Daimabad and Nevasa during the Malwa
the seeds. Extracted from the seeds, the oil goes rancid period (Kajale 1991), and several late 2nd millennium
but has other uses such as a lighting fuel or lubricant. or early 1st millennium bc sites in the middle Ganges,
Identifying Linum archaeologically is fairly straight- including Imlidh-Khurd (Period II), Narhan, Waina
forward, although congeneric wild species, which are (Period II), Senuwar (Period II) and Raja-Nala-Ka-Tila
widespread in Eurasia, may be difficult to distinguish (Period II) (Saraswat et al. 1994; Saraswat 2004, 2005).
definitively. In South India, for example, there is an Its archaeological distribution therefore parallels that
indigenous L. mysurense (Saldanha 1996), but seeds of of flax and cotton.
this species are expected to be much smaller. At Paithan, safflower is represented by a few frag-
Within India, the spread of Linum correlates with ments in Period 1 and two well-preserved fruits from an
that of cotton, moving east and south of the Harappan unphased sample.
zone in post-Harappan/Chalcolithic times, especially
towards the latter part of that period (Fig. 12.32). This
correlation suggests that Linum moved as a fibre crop,
although oilseed varieties are likely to have also been
present. The meagre evidence of this species at Paithan
should not be taken as evidence that it was not present
in the region, given the strong biases against seeds of
this species surviving charring, although it is probably
safe to conclude that this was in much less frequent use
as an edible oilseed than sesame.
As discussed in relation to cotton, the spread of
flax as a fibre crop implies a level of ‘cash-crop’ pro-
duction and agricultural production beyond the basic
surplus level. It is also a labour-intensive species to
process. Processing plants for fibres involves rippling
and retting (soaking with partial fermentation of stems),
followed by pounding and combing to separate the
internal (bast) fibres (for a synopsis, see McCorriston
1997: 522–524), after which spinning and weaving are
possible. Unlike cotton, which had an international
export market, linen may have been for more local use.

Fig. 12.35: Lone specimen, poorly preserved, of possible hemp (cf.

Safflower (Carthamus tinctorius) (Fig. 12.34) is used Cannabis sativa) from context 1128.
for its edible, oily seeds as well as its colourful flowers
and pollen, which can be used in dying as well as in
Hemp, Cannabis (Fig. 12.35), is a highly polymorphic
cooking. Safflower is a thistle-like winter annual, pro-
group of plants, which have been widely spread through
ducing yellow, orange or red flowers (capitulum). Due
cultivation and naturalization. Traditional taxonomy
to its large taproot, it tolerates dry conditions and is
groups all of these as a single variable species as all
thus suitable for cultivation during the dry winters of
forms are interfertile (Small and Cronquist 1976; Zhou
South Asia. It produces a cypsela with a fibrous shell,
and Bathelomew 2003). Plants are monoeious, produc-
within which is an edible oily kernel, like that in sun-
ing either male or female flowers. Hemp is the distinc-
flower seeds. Like the main winter cereals and pulses,
tive thick-stemmed and large-seeded form that have
safflower has its origins in the Near East, but early
been selected as fibre-producing and also have large,
finds suggest domestication rather late, perhaps as
276   Dorian Q. Fuller

edible oily seeds. By contrast, a much more herbaceous the site with rice, such as sedges (Cyperaceae) and
form, with much smaller seeds, is typically cultivated Ischaemum (cf. Galinato et al. 1999) (Fig. 12.36), while
as a drug plant with higher concentrations of the psy-
chotomimetic compound, especially in female flowers,
conventionally called Cannabis sativa ssp. indica, or C.
indica Lamarck. Hillig (2005) recognizes three species:
a west Eurasian C. sativa sensu stricto, which included
fibre of oilseed landraces from Central Asia, West
Asia, Europe and northeast Asia; C. indica, including
all of the various drug cultivars, feral populations in
South Asia, which are common in the Himalayas, and
Chinese fibre cultivars; and the wild species Canna-
bis ruderalis. Although criteria for separating these
species have not yet been developed for charred seeds,
we expect the C. indica type to have been present, with
its main uses for the drug, some use for its edible oily
seeds and perhaps as a minor fibre source, much as Fig. 12.36: Weedy grass, wrinklegrass (Ischaemum rugosum) from
is true in recent times (Watt 1889–1893; Pandey and context 1116, dorsal view.
Chanda 1996). It is a winter crop.
The few finds in South Asia are from the 2nd mil-
lennium bc, when it was probably introduced from
Central Asia (cf. Fuller 2006a: 38). Phytoliths of Can-
nabis have been reported from Harappa (Fuller and
Madella 2001: 338; Madella 2003), from the Late
Harappan period (after 2000 bc). Charcoal evidence
and charred seeds were found at Chalcolithic levels
at Senuwar, 1300–600 bc (Saraswat 2004). The single
specimen from Paithan, which is probable Cannabis, is
therefore worthy of note. As oily seeds, there are major
preservation biases against this species.

The potential of archaeobotanical Fig. 12.37: Weedy grass, Indian goosegrass (Eleusine indica) from
context 721, dorsal view.
weeds
Most of these species probably entered the archaeo-
logical record as part of crop-processing waste, as
weedy contaminants are a normal part of any harvest
(Hillman 1984; Reddy 1997, 2003; Stevens 2003; Harvey
and Fuller 2005). Some species might also have entered
the site with fuel, such as remains of fodder or grazed
plant in dung fuel, although this is regarded generally
as a less likely scenario in this case (see Fuller et al.
2014; Fuller 2003c). Therefore, weeds provide a record
of the ecology of arable fields, which in turn may
provide insights into the nature of agriculture in the
past. Unfortunately, this is confounded by two issues:
lack of species-level resolution in identification and
uncertainty about which crops particular weeds are to Fig. 12.38: Weedy grass, fingergrass (Digitaria sp.) from context
be associated with. Some weeds most likely came onto 1116, dorsal view.
 Chapter 12 Archaeobotany    277

others are more plausibly millet crop weeds, includ- Carpetweeds

ing the carpetweeds (discussed below) and Rubiaceae,
as well as grasses such as Eleusine indica (Fig. 12.37), Three species of carpet weeds (family Aizooaceae) recur
Digitaria (Fig. 12.38) and Urochloa. Other weed species in the Paithan assemblage: Zaleya decandra, Trianthema
may cut across several crops. portulacastrum and Trianthema triquetra (Fig. 12.39).
Another potential use of weed data is for inferring Zaleya cf. decandra (L.) Burm. f. (syn. Trianthema decan-
aspects of crop-processing, as weed seeds of particu- dra) is a widely encountered seed type. It is a curled dis-
lar size and weight characteristics are systematically coidal seed with an ornamented surface, which is typical
removed from crop-processing at different stages. of several major Caryophyllid families. As with many
Once again, confounding factors include the large Aizooaceae, the seed coat ornamentation tends to be rep-
diversity of crops at Paithan, which may not have all resented by wavy ridges that are concentric with the curl
been treated the same way. Nevertheless, one trend is of the seed. In this case, they are represented by short,
clearly evident which is towards increasing quantities round rugouse ridges. Archaeobotanical examples iden-
of weeds, from less than 2% relative frequency (18% tified to this species were illustrated by Vishnu-Mittre
ubiquity) in Period 1, to 12% (and 62%) in Period 2, to (1990) at Surkotada (also Chanchala 1994: pl. 1.20) and
23% (and 82%) in Period 3. The total number of weed were also reported throughout the sequence at Rojdi, but
species, however, does not change (43 taxa). These in large quantities only from the Late Harappan Phase C
data strongly suggest that more routine crop-process- (Weber 1991: 96). These seeds are commonly encountered
ing stages are being carried out on site in Periods 2 and in Southern Neolithic material (Fuller 1999). Comparable
3 compared to Period 1. It is also notable that in the two reference material is illustrated from a congeneric species
later periods there is a greater range of small-seeded by Kirkbride et al. (2006). This appears to be the same
weeds, such as Digitaria, Eragrostis, Urochloa and the species identified as ‘Portulaca cf oleracea’ at Inamgaon
Aizooaceae species. We expect many weed seeds, espe- (Kajale 1988: 818, fig. 27), and possibly also at other sites.
cially small seeds, to be removed early in the crop-pro- The large hilum of Portulaca is, however, missing, as are
cessing sequence, i.e. through winnowing. The absence the distinctive beads of the surface decoration, suggest-
of this from the site during Period 1 could imply a ing earlier misattribution of Zaleya specimens. This is a
more centralized organization of agricultural labour common weed in the open, dry areas of the inner Deccan
during this period, with larger workgroups available that flowers after the monsoons, with reported seed sets
for processing in the fields immediately after harvest in October–November or during the winter (Tadulingam
and fewer stages of processing carried out on site in and Venatanarayana 1985; Singh et al. 2001). Its distribu-
domestic contexts on a day-to-day basis (see discus- tion and archaeobotanical finds imply an association with
sion of reasoning in Stevens 2003; Harvey and Fuller millets. Common names include Belekomme or Gaija-
2005; Fuller et al. 2014). It may also be the case that soppe (Kannada), Tella-Galijeru (Telugu), Vellai Sharanai
this change relates to the move away from an empha- (Tamil) and Gadabani (Hindi) (Saldanha 1984; Tadul-
sis on kodo millet towards a more diversified cropping ingam and Venatanarayana 1985; Ambasta et al. 1986).
system. It may be that agricultural diversification and Some medicinal uses are reported for its roots, while the
intensification were linked with an increasing empha- leaves have served as a famine food (Ambasta et al. 1986).
sis on smaller, household units of productions. Trianthema portulacastrum L. (syn. T. monogyna L.)
Three recurrent and recognizable weeds in the has plump curled seeds, with narrow beaks near the hilum
family Aizooaceae are singled out for discussion and wider, somewhat square, ends. From the square end
together with a Rubaiceae type as these have been radiate faint wavy ridges, which are periodically discon-
widely found in South Asian archaeobotany and it tinuous (cf. Noda et al. 1985). This species is also known
should now be possible to identify them accurately to from Chalcolithic and Iron Age sites in the Ganges basin,
species and to consider their significance. In the dis- e.g. Narhan (Saraswat et al. 1994), Hulaskhera (Chanchala
cussion some ethnographic uses are also mentioned, 1992) and later Early Historic sites, e.g. Manji (Chanchala
although it should be stressed that such uses are 2000–2001). This species has not been noted in the South-
unlikely to have resulted in recurrent exposure to fire ern Neolithic assemblage and may be later addition to the
and archaeological recovery. Nevertheless, the pres- peninsular weed flora. It is reported from Rojdi but only
ence of the species raises the possibility of multivari- from the Late Harappan Phase C (Weber 1991: 95). Today
ous uses in the past. this is a widespread and ‘aggressive’ weed of cultivation
in India and elsewhere (Tadulingam and Venatanaray-
278   Dorian Q. Fuller

ana 1985; Singh et al. 2001). The seeds are readily spread (Saldanha 1984; Tadulingam and Venatanarayana
through animal dung. Nevertheless, it has some medici- 1985; Ambasta et al. 1986; Singh et al. 2001). It is not
nal uses, can be fodder and is useful as a green manure regarded as a significant weed under regular cultiva-
(Ambasta et al. 1986). Its common names include Pund- tion (i.e. regular tillage), being more common on field
haroighentuli or Ghol (Marathi), Lal-sabuni (Hindi), Santhi edges (Tadulingam and Venatanarayana 1985), and it
(Gujarati), Ambatimadu or Galjeru (Telugu), Muchchugoni may be poisonous to livestock (Ambasta et al. 1986).
or Pasale-soppu (Kannada) and Shaaranaj (Tamil). It normally seeds from September to November. Local
names include Naye-soppu (Kannada), Kukkapala-
koora (Telugu), Siru-Sharanai (Tamil) and Pathar-phor
(Rajasthani).
Borreria/Spermacoce (Rubiaceae) (Fig. 12.40) con-
sists of small mericarps in which seeds are assymet-
rical, ovoid and one smooth convex back opposite a
linear indentation in which a tubular attachment is
present. These might be compared with miniature
elongate coffee beans. This type is well represented
in Southern Neolithic samples (Fuller 1999: C39) and
occurs throughout the sequence at Harappan Rojdi
(Weber 1991: 66). The widespread weed Spermacoce
articularis L. f. (syn. Borreria hispida (L.) K. Schum.) is
the most likely species, although this has not yet been
confirmed against reference material, and two other
species from the genus are reported from Maharashtra
(Singh et al. 2001). These ‘buttonweeds’ are all low
(<30 cm) to prostrate annual herbs that fruit at the
end of the monsoon and are reported from light soils
(Tadulingam and Venatanarayana 1985; Singh et al.
2001). This type is a likely millet weed; it is also tra-
ditionally used as fodder and the seeds and the leaves
have medicinal uses (Ambasta et al. 1986). Common
names include Ghanti-chi-bhaji (Maharati), Guthari
Fig. 12.39: Two species of carpetweeds (Aizooaceae) from context (Hindi), Madana-budama (Telugu), Natha-churi (Tamil)
933: Trianthema portulacastrum (above) and Trianthema triquetra and Madanabudu-maegiddah (Kannada).
(below), lateral views. Caryophyllaceae/Molluginaceae types are small
curled discoidal seeds with distinct warted surfaces
and small beaks above the hilum. They are rounder
Trianthema triquetra Willd. Ex Rottl than the typical temperate Silene or Stellaria but a
precise match is not yet available from reference mate-
These seeds are smaller and more flattened than either rial or seed atlases. This same type is known from the
of the previous two species and have distinctive con- Southern Neolithic (Fuller 1999: C37). Another Caryo-
centric ridges that run as uninterrupted rings around phyllaceae or perhaps Molluginaceae type has distinct
the seeds (cf. Kirkbride et al. 2006). These occurred radiating ridges (Fig. 12.41) and should be identifiable
occasionally in Southern Neolithic samples, as the when reference material is available. The situation is
author has examined, and are also reported from similar with the seeds of a mallow (cf. Malva sp.) (Fig.
Late Harappan Rojdi Phase C (Weber 1991: 96) and 12.42), and further comparative work is needed on
the Early Historic period at Shirkapur (cf. Chanchala seeds of this group from India, although several taxa
1994: pl. 2.16). The species is also reported from the are known as weeds. These latter weeds highlight the
Early Historic Punjab at Sanghol (Pokharia and Saras- pressing need to develop a more comprehensive refer-
wat 1999), but not in early Chalcolithic levels (Saras- ence collection of wild herbaceous plants from penin-
wat 1996–1997). It is reported as occasional in scrub- sular India.
land and moist soils, but favouring saline sandy soils
 Chapter 12 Archaeobotany    279

Fig. 12.40: Buttonweed mericarp (Borreira/Spermacoce) from context 1116: dorsal (left) and ventral view (right).

Fig. 12.41: Indeterminate weed type of Caryophyllaceae/Mollugi- Fig. 12.43: Weedy dayflower seed (Commelina cf. benghalensis)
naceae type from context 721, lateral view. from context 486, dorsal view.
An additional weed type of note is the day-
flower (Commelina cf. benghalensis) (Fig. 12.43). This
species is common today as a weed of both irrigated
and rainfed rice (Galinato et al. 1999). This species is
reported as an occasional weed in the Neolithic and
Chalcolithic Ganges, presumably of rainfed rice, but
became increasingly common in Iron Age and Early His-
toric times probably with the increase in irrigated rice
(see Fuller and Qin 2009: table 2). A very similar type,
however, has also been found in Southern Neolithic
samples, suggesting that this may also be a weed of
millets (Fuller 1999: fig. C4,B).

Fig. 12.42: Weedy mallow type seed (Malvaceae, cf. Malva), lateral Concluding remarks
view.
The Paithan excavations have provided a rich archive of
plant use from the Early Historic to Medieval periods.
This represents the botanically richest archaeological
280   Dorian Q. Fuller

assemblage to date from the peninsular region. This region. Thus, we have archaeobotanical evidence that
is a result of routine sampling and flotation as part of Paithan was one of many Early Historic centres of
the excavations but is also a product of the diversified cotton textile production for trade. The general impor-
plant economy practiced by the ancient inhabitants of tance of cotton cloth export from the Deccan is implied
the Godavari. Paithan has contributed important evi- in Roman historical sources, as well as archaeological
dence on the history of several major crops. As should finds at Berenike on the Red Sea coast of Egypt (Wild
be evident from the discussion above, there are many and Wild 2008).
issues which require further investigation and further The patterns of change testified to through the
evidence, and therefore, more programs of archaeo- site’s occupation sequence suggest various strategies
botanical sampling and laboratory investigation are for increasing agricultural production, which in turn
needed on more sites. The staple cereals appear to may imply a growing population. There is an unam-
have been millets, well adapted to the semi-arid savan- biguous increase in the biodiversity of plant species,
nah zone of peninsular India, and in particular native including a wider range of crops. There is also more
millets, of which kodo millet was pre-eminent in the emphasis on winter pulses through time. This implies
Satavahana period. By the Early Medieval period, the both diversification of cropping strategies and proba-
African crops, sorghum and pearl millet had risen bly cultivation of a wider range of soil types, including
in importance together with foxtail millet (Setaria more marginal areas. In addition, it suggests increased
italica), but the South Indian native Brachiaria ramosa double-cropping and reduction of the fallow periods
remained in cultivation. Rice was consistently present, between crops, i.e. true intensification (see Morrison
but in relatively low amounts, suggesting that it may 1994). It may be that agricultural diversification and
have been a high-status cereal. Another important intensification were linked with an increasing empha-
crop throughout the sequence at Paithan was cotton. sis on smaller, household units of producers, as is sug-
The presence of charred cotton seeds implies that pro- gested by the evidence for increased on-site processing
cessing of cotton bolls for fibre took place on site and of crops.
that cotton was cultivated in the surrounding
Peter Rowley-Conwy, with a contribution by P. K. Thomas

Chapter 13 
The Faunal Remains

Introduction include Nasik (George 1955), Nevasa (Eapen 1960) and

Bhorkardan (Rao 1974), all the reports providing con-
siderable detail about the bones present. Because of
The study of faunal remains from archaeological sites
the nature of the site of Paithan, the present report will
has a long history. Methodological advances over the
set itself two additional goals. First, the fact that four
years have meant that zooarchaeological studies can
major periods are represented means that it is possible
yield a great deal of information. It has increasingly
to examine change through time. Second, to achieve the
been understood that such information has been
first goal, precise quantification must be undertaken of
more useful to archaeologists than to zoologists. This
the remains in each phase. As will be seen below, the
is for two reasons. Firstly, the zoological history of
outcome provides information of the greatest interest.
more recent times is now quite well known, so zoo-
One factor which has a major effect on any sample
archaeological studies do not add much to the sum of
of archaeological animal bones is the method by which
current knowledge. Secondly, animal remains from
they were recovered. Contrary to popular opinion,
archaeological sites are usually present because of
archaeologists and excavation labourers do not detect
human action. They were acquired, killed, processed,
and recover all items in the deposits they are excavat-
consumed and their remains discarded entirely within
ing: many objects are overlooked. Not surprisingly, it
the cultural context of the humans who inhabited the
is the smallest objects that are most often overlooked.
site. This cultural context is the very thing archaeolo-
This has a predictable effect on zooarchaeological
gists seek to reconstruct, and the animal remains form
samples: bones of smaller animals are overlooked more
a valuable and powerful means to assist in this task.
frequently, and smaller animal species are therefore
The following report is written entirely from this per-
under-represented in the identified totals. This is true
spective.
whether the excavators are fully trained archaeologists
This report discusses the faunal remains from
(Payne 1972, 1975) or labourers specializing in archaeo-
Paithan. A grand total of 4,777 fragments of bone,
logical work (Rowley-Conwy 1994). The proportions of
tooth, scute and shell were examined, although as is
smaller bones overlooked vary greatly, depending not
usual, only a small proportion of these could be iden-
just on the degree of skill and diligence of the excava-
tified to species. A total of 490 items were identified as
tors, but also on factors such as soil type and mois-
coming from the four most common domestic species:
ture, amount of daylight and weather conditions and
sheep (Ovis aries), goat (Capra hircus), pig (Sus scrofa)
time of day. It is therefore generally recognized that it
and cow (Bos indicus). Smaller numbers of items were
is necessary to sieve archaeological deposits in order
referred to other taxa. They will be discussed grouped
to ensure unbiased representation of zooarchaeologi-
into the four major periods represented at Paithan
cal remains (and indeed of all kinds of archaeological
(Periods 1 to 4, see Chapter 6).
items). The Paithan excavation was conducted to fully
An added complexity is that some layers and con-
modern standards. All deposits were sieved through
texts in Period 3 are directly associated with the con-
5-mm mesh, so the faunal samples may be considered
struction of the temples, while others are not. The
unbiased by recovery.
Period 3 remains are therefore divided into Period
Another factor that can have a major biasing effect
3-temple contexts and Period 3 contexts. If no subdivi-
on assemblages of animal bones is gnawing by dogs.
sion of Period 3 is specified, then the overall period
If dogs have access to the bones after their discard
total comprising both 3-temple and 3 is being referred
by people, they may damage or destroy many frag-
to.
ments. Unfortunately, such destruction is not random:
Some previous bone reports are available from
some elements within the animal skeleton are harder
sites in the region which fall into this time range,
than others, and dogs tend to destroy the softer ele-
and these are useful for comparative purposes. They
https://doi.org/10.1515/9783110653540-013
282   Peter Rowley-Conwy, with a contribution by P. K. Thomas

ments more frequently. The bones of younger animals

are also softer than those of adults, so juveniles also
suffer greater losses (e.g. Brain 1981; Binford 1981;
Legge 1992). Gnawing by dogs may be recognized in
two ways. First, tooth marks may be present. Second,
if a bone has passed through a dog’s digestive system,
the powerful digestive acids often leave characteris-
tic etching on the surface of the bone (if they do not
destroy it completely) (Payne and Munson 1985). At
Paithan, very few bones exhibited such damage (Fig.
13.1). In Period 1, one fragment was gnawed and one
digested; in Period 2, one was gnawed; in Period 3,
three were digested; and in Period 4, a single fragment
showed traces of both gnawing and digestion. These
very small totals suggest that dogs did not have much
access to the animal bones at Paithan, so the effects of
this biasing factor are likely to be insignificant.
Two of the common biasing factors may thus
largely be discounted. We may therefore approach the
Paithan assemblage with a degree of confidence.

Species represented
The animal bones from Paithan were transported to
Fig. 13.1: Phalanges of blackbuck (Antilope cervicapra) showing
Goa along with the rest of the excavated material for the characteristic etching caused by the digestive acids of dogs
post-excavation study and were identified there by the (top) and holes caused by gnawing (both specimens). The speci-
author in 2000. It is desirable wherever possible to mens were identified to species by Dr P. K. Thomas.
identify animal bones with the help of a comparative
collection. However, no such collection was available
cow and pig, and basic classification of the fragments
in Goa, so the specimens were identified using the
as sheep/goat, cow and pig was usually unproblematic.
author’s drawings and identification notes. Most of the
The 490 specimens identified as coming from these taxa
identified specimens were from species with which the
are listed in Table 13.1; the totals are unmodified in any
author has long been familiar, namely sheep, goat,
way, thus conforming to the ‘Number of Identified Spec-

Table 13.1: Number of Identified Specimens (NISP) of the main domestic taxa at Paithan, divided by layer and expressed as percent-
ages of the layer total.

Period 1 2 3 4

temple ordinary total

N % N % N % N % N % N %
Sheep/goat 52 35 73 51 38 56 20 56 58 56 64 67
Cow 40 27 37 26 25 37 10 28 35 34 28 29
Pig 55 37 33 23 5 7 6 17 11 11 4 4
Total 147 143 68 36 104 96
 Chapter 13 The Faunal Remains    283

imens’ (NISP) method of quantification (other methods Table 13.3: Other specimens identified by the author, not included
of quantification will be discussed below). in Table 13.1.
The separation of sheep from goat is a consider-
able problem, and the NISPs of the two species are Period 1 2 3 4
amalgamated in Table 13.1. Only a minority of skeletal
temple ordinary total
elements can be reliably and consistently separated,
although methodological improvements continue to Dog or jackal 2 3 5 1 6 6

be made (see Boessneck 1969 for a general review). Equid 0 0 0 1 0 0

Among the most useful elements are distal metacarpal Deer 0 1 0 0 0 0
and metatarsal (Boessneck 1969; Payne 1969; Rowley- Bird 5 12 3 0 3 1
Conwy 1998), the deciduous dentition (Payne 1985) Fish 5 1 4 2 6 21
and the distal tibia (Kratochvil 1969), and since the Frog or toad 0 0 0 1 1 0
Paithan material was examined, a method has been
Gastropod 15 14 3 9 12 4
published using the permanent dentition (Halstead et
Bivalve 40 13 0 1 1 3
al. 2002). The small number of bones that could be def-
Cowrie 0 0 0 4 4 24
initely identified as sheep or goat using these methods
is listed in Table 13.2. The sample is too small to permit
precise conclusions to be drawn, but it may tenta- Once the identifications described above had been
tively be suggested that (a) both species were present made, there remained a number of specimens which
throughout; (b) sheep consistently outnumbered goats were clearly potentially identifiable, but which came
by between 2.75 and 8 to 1; and (c) there is no indication from species with which the author was not familiar.
that the proportions changed through time. The small It was most fortunate that the author was able to visit
size of the identified sample must however be born in Deccan College in Pune and consult Dr P. K. Thomas
mind. about these specimens. Dr Thomas kindly identified
Apart from the major food taxa, various other items the various specimens, which are listed in Table 13.4.
were identified (Table 13.3). Canids were the most common A fragment of crocodile mandible containing two
mammal. Dog (Canis familiaris) and jackal (C. aureus) are teeth (Fig. 13.2) was identified as belonging to Croco-
notoriously hard to distinguish, especially without com- dilus palustris. Twenty-one turtle scute fragments were
parative skeletons, and no attempt to do so was made in found, all but two in Period 4; some of these could def-
this instance. Under the circumstances, however, it is initely be identified as Trionyx gangeticus, and the rest
likely that most or all of the bones come from domes- were tentatively referred to the same species. Three
tic dogs: they would most probably have been present of the scute fragments from Period 4 were cut with a
and would have acted to keep jackals away from the sharp knife (Fig. 13.3), suggesting that craft or manu-
human settlement. Single fragments of teeth from facturing activities may be one reason for the pres-
members of the horse family (Equidae) and deer family ence of the turtle scute fragments. A few specimens
(Cervidae) were also identified, but species could not of other species were present (Table 13.4), including
be determined. Small numbers of bones were identi- the blackbuck phalanges illustrated in Fig. 13.1. The
fied as coming from birds and fish, but again, species two monkey bones could not be identified to species.
could not be determined. Finally, 130 shells of various A pelvis fragment from Period 3 came from a juvenile
kinds were counted. They were referred to three main animal, while an ulna from Period 2 came from a species
groups (see Table 13.3). visibly larger than the langur (Presbytis entellus).
A few of the more complete fragments listed as
Table 13.2: Number of sheep/goat specimens definitely identified as ‘cattle’ in Table 13.1 were also taken to Deccan College,
either sheep or goat.
with a view to attempting to establish whether they
were all zebu (Bos indicus) or whether any water
Period 1 2 3 4
buffalo (Bubalus bubalis) might be present. The biolog-
temple ordinary total ical tribe Bovini constitutes the largest members of the
family Bovidae, comprising principally the genera Bos
Sheep 10 11 2 6 8 10
(cattle including zebu), Bison (European and Ameri-
Goat 2 4 0 1 1 2
can), Syncerus (the African buffalo) and Bubalus (the
water buffalo). Distinction between all these species
284   Peter Rowley-Conwy, with a contribution by P. K. Thomas

is difficult. Published criteria exist for distinguishing

between Bos and Bison (Olsen 1960) and between Bos
and Syncerus (Peters 1986). No criteria for distinguish-
ing between Bos and Bubalus are, however, known to
the author. The two genera are quite distinct in evolu-
tionary terms and have long been distinct species (Ritz
et al. 2000; Hernández Fernández and Vrba 2005), so it
is to be expected that reliable criteria might well exist.
A complicating factor is that many breeds of water
buffalo exist (Kikkawa et al. 1997; Kumar et al. 2006),
and how much osteological variation there might be
within Bubalus is not known to the author. As regards
Paithan, eight specimens were compared to the Bos
and Bubalus comparative skeletons at Deccan College.
Six of these appeared to align with Bos. These were
Fig. 13.2: Fragment of mandible of Crocodilus palustris, identified
fragments of a distal metacarpal and a distal metacar- by Dr P. K. Thomas.
pal from Period 1, a distal tibia and a distal metatar-
sal from Period 2 and an astragalus and proximal radius
Once all the various items discussed above had been
from Period 3. The other two appeared to align with
identified, there remained the majority of bone frag-
Bubalus. These were a distal humerus from Period 1
ments which could not be identified. In many cases,
and an astragalus from Period 3. These very tentative
it was possible to determine that the animals from
identifications can do no more than suggest the pos-
which the fragments had come were relatively small
sible presence of Bubalus, not unexpected in view of
(sheep/goat and pig size), while others came from
its presence at both Bhokardan (Rao 1974) and Nasik
larger animals (cow size). Given that the overwhelm-
(George 1955) – although it was apparently not present
ing majority of identified bones come from sheep/goat,
at Nevasa (Eapen 1960). A comparative osteological
pig and cow, it is highly probable that the overwhelm-
study of Bos and Bubalus, and a survey of archaeo-
ing majority of unidentified fragments also do so, even
logical material containing these species, would be of
though this cannot be demonstrated. Table 13.5 pre-
great benefit to South Asian zooarchaeology.
sents the totals for the two size categories, divided into

Table 13.4: Dr P. K. Thomas’ identifications of bones and scute fragments from Paithan.

Period 1 2 3 4

temple ordinary total

crocodile 0 0 1 0 1 0
Crocodilus palustris
turtle 2 0 0 0 0 19
Trionyx gangeticus
hare 0 2 0 1 1 3
Lepus nigricollis
bandicoot 0 0 0 1 1 0
Bandicota indica
blackbuck 0 1 1 2 3 1
Antilope cervicapra
monkey 0 1 0 1 1 0
Cercopithecoidea
swamp cat 0 0 0 1 1 0
Felis chaus
 Chapter 13 The Faunal Remains    285

the anatomical region of the skeleton, for each of the Change through time
phases. Finally, there were numerous fragments that
could not be classified at all, and these are also listed The examination of changing species frequency
at the bottom of Table 13.5. through time is a main goal of this contribution. To
be able to examine this, precise quantification of the
bone assemblage is required (see above). One step
towards precise quantification has been taken, with
the summing of fragments presented in Table 13.1.
There has, however, been considerable discussion
about whether the NISP total presented there is the
best method of quantification to use in such studies.
For the following discussion, see Fig. 13.5 for the loca-
tion of the various bones mentioned; ‘distal’ refers to
the end of the bone away from the vertebral column,
‘proximal’ to the end nearer the vertebral column.
Various factors may lead to the basic NISP total
being an imperfect method of quantification. One is
the simple fact that different species have different
numbers of bones in their bodies. Pigs, for example,
have four metacarpals in each forefoot and four meta-
tarsals in each hindfoot, while cattle, sheep and goats
have only one in each case. This problem is increased
when considering animals such as fish, which have a
completely different bodily conformation; and clearly,
no meaningful comparison can be made between (say)
the number of sheep bones and the number of cowrie
shells found on a site. Another factor may be differen-
tial fragmentation. A distal humerus of a cow might
Fig. 13.3: Fragments of scute of turtle (Trionyx gangeticus) cut with
a knife.
be broken into (say) five fragments, all of which might
be identifiable, while a sheep distal humerus might be

Table 13.5: The unidentified bone fragments from Paithan, divided into size class and anatomical region where possible.

Period 1 2 3 4

temple ordinary total

Sheep/goat or pig sized animal

limb splinter 129 182 72 128 200 144
vertebral fragment 43 39 16 15 31 10
skull fragment 21 12 7 20 27 11
rib fragment 122 170 82 82 164 86
total 315 403 177 245 422 251

Cow-sized animal
limb splinter 61 52 31 41 72 47
vertebral fragment 19 15 1 5 6 0
skull fragment 21 6 0 19 19 2
rib fragment 65 69 11 17 28 23
total 166 142 43 82 125 72
Unclassified fragments 421 421 493 443 936 372
286   Peter Rowley-Conwy, with a contribution by P. K. Thomas

found in a complete state. If a simple NISP count was the site phasing at the time the recording is carried
applied in such an instance, cattle would incorrectly out. Shaft fragments that do not have articular ends
appear to be five times more common than sheep (for are discounted. For mandibles and maxilla, the same
general discussions, see Casteel 1977; Watson 1979; principle is used: complete mandibles are counted,
Gautier 1984; Grayson 1984; Davis 1987). and loose teeth or partial mandibles then grouped
One method of quantification that seeks to into a theoretical minimum number that is added to
get round these problems is the calculation of the the complete ones. (2) The resulting MNE totals are
Minimum Number of Individuals (MNI) total. This brought into line with the artiodactyl skeleton. Each
seeks to establish the minimum number of individual such skeleton has two distal humeri. Each such skel-
animals that it took to generate the bone assemblage. eton has two distal humeri but only one atlas vertebra
At its simplest, a total of (say) five left and three right (see Fig. 13.5), so the MNE for atlas is doubled. Each
distal humeri of cattle must derive from a minimum artiodactyl skeleton has two distal metacarpals, so the
of five cattle – because the right-side elements might MNE of these bones is left unchanged; each pig skel-
come from the same animals that produced the left eton, however, has four distal metacarpals, so the two
side elements. There are, however, problems with easily recognizable lateral metacarpals are discounted,
the MNI method. The MNI for each species is usually and the MNE for the central pair is halved. Phalanges
a rather small number, a particular problem in small (toes) cannot be divided between fore and hind foot.
assemblages like that from Paithan, and the frequency There are eight first phalanges in each artiodactyl skel-
of species represented by a single bone is exaggerated. eton (two in each foot), so the total is divided by 4; and
It is also based on an unrealistic assumption. Using the same procedure is used for second and third pha-
the hypothetical example of five left and three right langes.
distal humeri in the context of an archaeological site
occupied for two centuries, it is clear that the MNI is
in any literal sense meaningless: many more than five
cattle would have been killed and consumed in two
centuries. We cannot know how many; but even if only
one animal was killed each year, making 200 cattle in
all, it is evidently highly unlikely that any left and right
elements in the archaeological assemblage actually come
from the same animal.
No method of quantification is perfect. Perhaps the
safest option is one that lies between NISP and MNI.
This is one that sums the so-called Minimum Animal
Units (MAU) for each taxon. The MAU is intended to
bring all elements into line with the basic artiodactyl
(e.g. cow, sheep/goat) skeleton, in the following two
steps (Binford 1984). (1) Differential fragmentation is,
to some extent, overcome by recording which part of
(say) the distal humerus is present. Thus, the Paithan
records show whether any distal humerus fragment
is complete (i.e. the entire articulation is present) or
whether just the medial or just the lateral portion of
the articulation is present. This allows determination
of the Minimum Number of Elements (MNE) to be cal-
culated: it is the sum of complete articulations, plus
either the medial or the lateral portions, whichever is
the most common. In an ideal world, the analyst would
attempt to refit medial and lateral fragments or deter- Fig. 13.4: Minimum Animal Units (MAU) of the main domestic taxa
mine whether they could come from the same bone. In at Paithan, divided by layer and expressed as percentages of the
practice, the time constraint usually means that this is layer total (see text for the method of deriving the MAU).

impossible, especially when the analyst does not know

 Chapter 13 The Faunal Remains    287

Fig. 13.5: Diagram of the skeleton of an artiodactyl, showing the various bones.

The resulting MAUs for each element are listed in is more prominent in the MAU chart. Finally, and
Appendix 13.1. The sum of MAUs for each species in perhaps most interestingly, pig decreases through time
each period should, in theory, give the most secure basis in both charts. The outline is rather different depend-
for the relative quantification of species frequency, ing on method of quantification, however: in the
and thus for considering change through time. These NISP chart, the decrease is fairly even, while in the
MAU totals are listed and calculated as percentages in MAU chart, it appears as a more abrupt step between
Table 13.6. Periods 2 and 3.
Using the MAU totals, we are in a position to examine There is no independent way to determine which
change through time at Paithan. Fig. 13.4 plots the per- method of quantification gives the result which is
centages of the three main taxa, based on MAU (top), closest to reality, although for the reasons discussed
with the NISP percentages from Table 13.1 also pre- above, this author prefers the MAU method. Quite
sented for comparative purposes (bottom). It is imme- apart from anything else, the Paithan sample is rela-
diately apparent that there is substantial change tively small, and stochastic variations may therefore
through time. In both charts, sheep/goat increase sub- appear more marked. But the most important conclu-
stantially through the four periods represented at the sion is that change through time is clearly visible, and
site; the pattern is broadly similar using both methods it is consistent across the four periods: sheep and goat
of quantification, except that in Period 1, the NISP increase in frequency, while pigs decrease, throughout
percentage is lower, which accentuates the visible the time of occupation of Paithan. This will be further
increase. Cattle do not vary directionally through time, discussed in the conclusions.
although they appear to peak in Period 3; this peak

Table 13.6: Minimum Animal Units (MAU) of the main domestic taxa at Paithan, divided by layer and expressed as percentages of the
layer total (see text for the method of deriving the MAU).

Period 1 2 3 4

temple ordinary total

N % N % N % N % N % N %
Sheep/goat 42 41 53.25 52 22.5 58 12.25 57 34.75 57 42.75 68
Cow 29.75 29 22.25 22 13.5 35 10.25 44 23.75 39 16 25
Pig 30 29 27.25 27 2.5 6 1 4 3.5 5 4 6
Total 101.75 102.75 38.5 23.5 62 62.75
288   Peter Rowley-Conwy, with a contribution by P. K. Thomas

The changing frequencies of sheep/goat and sals are remarkably rare (and since they are among the
pig are clearly the most important change occurring most diagnostic elements for separating sheep from
through time at Paithan, but a couple of more minor goat, their rarity contributes to the small number of
points may also be made in this section. Neither birds definitely identifiable bones of these species: see Table
nor fish are ever very common, but bird bones reach 13.2). In the forelimb, proximal metacarpal is more
a minor peak in Period 2, while fish do so in Period 4. common than distal metacarpal; it may be that this
Among the shells, bivalves decrease through time; if bone was chopped through during butchery so that the
their presence reflects their use as a food source, their distal end and the phalanges could be discarded. Fig.
dietary importance was apparently reducing. Gastro- 13.6 shows a metacarpal exhibiting marks resulting
pods may also show a decrease, but the trend is less from two chops that have not cut through the bone. In
clear. Cowries, in contrast, appear to increase through the hindlimb, metatarsals are rare, while astragalus
time. Since cowries are marine species, all such shells and calcaneum are more common; possibly the leg was
must have been imported from coastal regions, and it cut through between the astragalus and the naviculo-
may be that the increase in Period 4 signals an increase cuboid, and the foot subsequently discarded. The
in trade or other connections at this time. None of the samples of cattle and pig bones are even smaller, so
other taxa listed in the tables appear to show any trend conclusions are even more tentative. Among the cattle
through time. bones, feet are generally rare, as are heads except
for the five maxillas (upper jaws) in Period 3-temple.
Among the pigs, feet are also rare, but heads are sub-

The nature of the samples, and the stantially more common in Periods 1 and 2.

temple deposits
One interesting aspect of any animal bone assemblage
is the question of whether the assemblage is dominated
by butchery waste, food waste or a mixture of the two.
The starting point for any analysis is the MAU total,
described above and listed by period for each species
in Appendix 13.1. Each table in Appendix 13.1 starts
with the head and neck (mandible through to axis –
see Fig. 13.5), then runs down the forelimb (scapula to
metacarpal), then down the hindlimb (pelvis to meta-
tarsal) and ends with the phalanges. The outline of the
soft tissues in Fig. 13.5 shows the decreasing amount
of meat towards the feet of the animal. The lower limbs
are therefore generally classified as butchery waste, Fig. 13.6: Metacarpal shaft fragment of sheep or goat, showing
discarded at an early stage in the butchery process. two heavy chop marks.
The upper limbs and trunk carry most of the edible
meat. The Paithan samples thus appear to represent rather
The samples from Paithan are mostly too small for ordinary waste material, with the exception that
significant conclusions to be drawn, but a few sugges- the feet and in some cases the heads may have been
tions may be offered. Considering sheep/goat first, detached and dumped elsewhere. From this perspec-
the feet are rather rare, while meat-bearing bones such tive, we can approach the nature of the temple depos-
as the pelvis are relatively common, particularly in its. The tables of bones in Appendix 13.1 do not suggest
Periods 1 and 2. This is unusual, since the pelvis is not that the contents of the Period 3-temple and Period 3
a very hard bone and often does not survive in large deposits differ materially from each other, or from the
numbers. This might suggest that the Paithan sheep/ other periods, with the possible exception of the pres-
goat bones derive predominantly from food waste ence of the cattle maxillas in Period 3-temple already
rather than butchery waste, although the pattern is less noted. The small size of the samples must again be
clear in Periods 3 and 4, and the small size of the sample stressed, however, and the presence of the cattle max-
must be born in mind. Distal metacarpals and metatar- illas is plausibly due simply to chance. The unidenti-
 Chapter 13 The Faunal Remains    289

fied bones recovered from Period 3-temple and listed rare at Paithan, and recovery was excellent, so these
in Table 13.5 similarly resemble those from the other factors may be discounted. The small sample therefore
deposits. Little or nothing in the animal bones in fact appears to suggest that dairy products were not a major
marks Period 3-temple out in any way. The only pos- goal at Paithan, although this does not exclude the
sible exception to this is the fragment of crocodile possibility of some less specialized use.
mandible containing two teeth, identified by Dr P. K. In pigs, the focus is usually different: pigs produce
Thomas as Crocodilus palustris (Table 13.4). This frag- half a dozen or more offspring (as opposed to the one
ment was recovered from Period 3-temple deposits. or two in the other major species), and meat is the only
Whether its presence there has any significance that significant product. The high breeding turnover of pigs
does not emerge from the straightforward study of the means that most animals are typically killed in their first
bones carried out here must be a subject for future dis- 2 years of life, and the jaws listed in Appendix 13.2 suggest
cussion. that Paithan was not an exception to this.
Metrical analysis is also an important zooarchaeo-
logical tool. Sometimes, measurements may comple-

Age at death and metrical analysis ment analysis of age at death. For example, in cattle
and goats (but not in sheep), elements such as distal
metacarpal allow males and females to be distin-
Age at death is an informative tool in zooarchaeology.
guished, males forming a separate scatter larger than
Best results are obtained from large statistically reli-
the females. Distal metacarpals fuse only when the
able assemblages. The Paithan assemblage is too small
animals approach adult size, however. Under a herding
to produce such results, but the available mandibular
policy focussed on dairy products, it is likely to be the
data are tabulated in Appendix 13.2. Ageing is based
males that are killed when very young, because of
on the replacement and wear of the mandibular teeth,
the herdsmen’s desire to keep the females alive and in
the most useful being the rear milk premolar (dp4), or
due course join the milking herd. In the archaeologi-
the rear adult molar (P4) if it has replaced dp4, and
cal record, this policy will lead to a predominance of
the three permanent molars (M1, M2 and M3). During
females among the measurable adults. If, on the other
identification, the state of eruption and wear of the
hand, the males are kept alive for longer and are killed
teeth in the mandibles were recorded, and for the more
for meat, the sex ratio visible in the distal metacarpals
complete fragments, this permits an approximate age at
will be more equal (see e.g. Legge 1981; Halstead 1998;
death to be established (see Appendix 13.2 for details).
Rowley-Conwy 2000).
In large assemblages, age at death may enable us
As mentioned above, butchery practices at Paithan
to understand something of the herding practices
unfortunately mean that distal metacarpals and meta-
employed in the past. For example, in sheep, goats
tarsals are virtually absent, and none could be meas-
and cattle, maximum meat weight is usually reached
ured. The Paithan measurements do not therefore play
somewhere around 2–4 years of age, and this is a good
a primary part in the analysis presented here, but are
time to slaughter animals for meat. Animals of this age
listed in Appendix 13.3 in the hope that they may be of
will not put on much more meat, and their removal
use to other researchers in the future.
allows more lambs, kids or calves to be reared. However,
if dairy products are what is desired, it is necessary to
kill some animals at a much younger age, usually in
the first month or two of life. A female will only lactate Conclusions
after producing offspring, and if these offspring are
kept alive, they will consume at least some of the milk The relatively small zooarchaeological assemblage
(see e.g. Payne 1973; Legge 1981, 1992; Halstead 1987, from Paithan has proved to be of the greatest interest,
1998). and due to the excellent method of recovery has gener-
The small number of sheep/goat mandibles from ated a considerable amount of information.
Paithan do not include any animals in the first 2 months The major conclusion is that there was a consist-
of life. Such jaws are smaller and softer than those ent increase in the importance of sheep and goat, and
of adults and are thus more prone to loss due to dog a concomitant decrease in the importance of pigs,
gnawing in ancient times or poor recovery during exca- through the period of occupation of the site. This is a
vation. As mentioned above, however, dog gnawing is conclusion of considerable interest, and it would be of
290   Peter Rowley-Conwy, with a contribution by P. K. Thomas

the greatest importance to know whether this is mir- Appendix 13.1 

rored at other sites in India and indeed in South Asia
in general. The rise of Islam probably meant that pig
Minimum Animal Units (MAU)
keeping declined in much of central and western Asia. Tables
The reasons for a parallel decline, although not dis-
appearance, in the pig in the non-Muslim context of Note: P indicates the Proximal, D the Distal, articulation of each
Paithan raises wider issues concerning the changing longbone

importance of this species: was the Islamic prohibi-

tion superimposed on a more general decline taking
place for other reasons? A. Sheep/goat: Minimum Animal Units
This is a question for future research. In the mean-
time, it is hoped that the present report has achieved the Period 1 2 3 4
aims set out in the introduction and also that it shows
temple ordinary total
something of the kinds of information that can be
extracted from animal bones, and which makes them Mandible 1 2 2 4 6 4
so worthwhile an object of study. Maxilla 1 2 2 2 4 3
Atlas 2 2 0 0 0 0
Axis 2 4 0 2 2 4
Acknowledgements Scapula 6 2 0 0 0 0
Humerus P 1 0 0 0 0 0
I would like to thank Dr Derek Kennet for asking me
Humerus D 1 4 1 2 3 2
to study this most interesting assemblage and Dr P. K.
Thomas for all his help and for the identifications Radius P 0 1 0 3 3 0

listed in Table 13.4. Dr Ruby Ceron-Carrasco discussed Ulna 3 5 1 1 2 1

cowrie shells with me. Radius D 0 3 0 1 1 2
Metacarpal P 2 3 1 1 2 4
Metacarpal D 0 1 0 0 0 1
Pelvis 4 6 1 2 3 2
Femur P 1 1 0 0 0 2
Femur D 2 2 0 2 2 2
Tibia P 3 2 1 1 2 1
Tibia D 3 6 0 0 0 2
Astragalus 5 3 2 0 2 1
Calcaneum 2 2 1 1 2 6
Naviculo- 1 0 0 0 0 0
cuboid
Metatarsal P 0 1 0 0 0 2
Metatarsal D 1 0 0 0 0 2
Phalanx 1 0.75 1 0 0.5 0.5 1.75
Phalanx 2 0 0.25 0 0 0 0
Phalanx 3 0.25 0 0.25 0 0.25 0
 Chapter 13 The Faunal Remains    291

B. Pig: Minimum Animal Units C. Cattle: Minimum Animal Units (MAU)

Period 1 2 3 4 Period 1 2 3 4

temple ordinary total temple ordinary total

Mandible 3 3 0 0 0 0 Mandible 1 1 0 0 0 2
Maxilla 7 5 0 0 0 0 Maxilla 1 1 5 0 5 3
Atlas 0 0 0 0 0 0 Atlas 0 0 0 2 2 0
Axis 0 0 0 0 0 0 Axis 0 0 0 0 0 0
Scapula 5 2 1 0 1 1 Scapula 2 0 0 1 1 1
Humerus P 0 1 0 0 0 0 Humerus P 3 1 0 0 0 0
Humerus D 2 2 0 0 0 0 Humerus D 2 0 0 2 2 1
Radius P 1 0 0 0 0 0 Radius P 1 3 0 1 1 1
Ulna 0 2 0 0 0 0 Ulna 0 0 0 0 0 0
Radius D 0 0 0 0 0 0 Radius D 0 1 1 0 1 1
Metacarpal P 2 1 0 0 0 0 Metacarpal P 2 0 0 1 1 1
Metacarpal D 0 1 0 0 0 0 Metacarpal D 2 0 1 1 2 2
Pelvis 1 0 0 1 1 1 Pelvis 3 0 1 1 2 0
Femur P 0 0 0 0 0 0 Femur P 5 0 0 0 0 0
Femur D 2 1 0 0 0 0 Femur D 2 2 0 0 0 1
Tibia P 0 0 0 1 1 0 Tibia P 0 0 1 0 1 0
Tibia D 2 4 0 0 0 0 Tibia D 1 2 0 0 0 0
Astragalus 1 1 0 0 0 0 Astragalus 0 1 1 2 3 0
Calcaneum 1 1 0 0 0 0 Calcaneum 2 1 0 1 1 1
Naviculo- 0 0 0 0 0 0 Naviculo- 0 3 0 1 1 1
cuboid cuboid
Metatarsal P 1 1 0 0 0 1 Metatarsal P 0 5 0 0 0 0
Metatarsal D 0 2 0 0 0 1 Metatarsal D 1 1 0 0 0 0
Phalanx 1 1.25 0 0 0.25 0.25 0 Phalanx 1 1 0 0.25 0.5 0.75 0.25
Phalanx 2 0.5 0 0 0 0 0 Phalanx 2 0.25 0.25 0 0 0 0.75
Phalanx 3 0.25 0.25 0 0.25 0.25 0 Phalanx 3 0.5 0 0 0 0 0
292   Peter Rowley-Conwy, with a contribution by P. K. Thomas

Appendix 13.2 
Ageable jaw fragments of the
three main species at Paithan
Ages are approximate only for sheep/goat, and they are
taken from Payne (1973); for pigs, they are taken from
Higham (1967).
For all taxa, (P) means tooth present but wear
stage not ascertainable; E = top of tooth erupted to
level of the jawbone; H = tooth half erupted; U = erupted
to full height but unworn; J = tooth just in wear; nye = not
yet erupted.
For sheep/goat, the numbered wear stages are
those distinguished by Payne (1973), the first stage
with no wear being numbered 1. For pig, the lettered
wear stages are those distinguished by Grant (1982).

Period dp4 P4 M1 M2 M3 Age

A. Sheep/goat
2 (P) (P) 6 H 18–24 months
2 U 8 5 U 18–24 months
3(o) 5 (P) 8 6 3–4 years
3(o) 6 (P) 6 (P) ?18–24 months
3(o)* 8 ?
3(t) 10 8 3 2–3 years
4* 7 H nye 2–6 months
4 9 14 14 10 6–8 years
4 (P) E ?2–6 months
* = identified as sheep

B. Cattle
3 (P) U
4 J

C. Pigs
1 J H nye 10–11 months
1 U (P) a 15–16 months
1 a g b 16–18 months
2 d a U 11–12 months
2 e d E 9–10 months
2 a H 19–21 months
 Chapter 13 The Faunal Remains    293

Appendix 13.3 
Measurements of animal bones from Paithan
All measurements are in millimetres. Only fully All measurements are as defined and illustrated
fused adult bones are included. All the cattle bones by von den Driesch (1976), except as follows: distal
measured are tentatively identified as Bos, not humerus HT and HTC, and distal metatarsal Dd,
Bubalus (see text). Bracketed measurements are from Legge and Rowley-Conwy (1988); pig tooth
uncertain. measurements from Payne and Bull (1988).

A. Sheep/goat
Sheep/goat distal humerus

Period Sheep or goat Trochlea breadth (BT) Thickness of trochlea (HT) Trochlea at constriction (HTC)
2 Goat? 29.0 19.0 13.7
3(o) Sheep - 20.8 15.3
3(t) ? - - 15.9

Sheep/goat radius

Period Sheep or goat Proximal breadth (Bp) Distal breadth (Bd)

2 Goat 28.3
2 Sheep 28.8
3(o) Sheep? 33.9
3(o) Sheep 30.1
3(o) Sheep 38.2

Sheep/goat tibia

Period Sheep or goat Distal breadth (Bd)

1 Sheep 25.3
1 Goat 28.8
2 Goat 27.5
2 Sheep 28.1
2 Sheep 27.9
4 Sheep 23.8

Sheep/goat astragalus

Period Sheep or goat Greatest length (GLl) Distal breadth (Bd) depth (Dl)
1 Sheep 30.5 17.9 16.9
1 Sheep 30.4 - 16.9
1 Sheep 34.3 19.7 18.3
1 Sheep? 31.4 20.2 16.4
2 Sheep 33.0 19.9 18.0
2 Goat? - 22.8 -
2 ? - 19.0 -
3(t) Sheep 33.8 21.0 18.7
3(t) Sheep 31.5 19.4 18.0
294   Peter Rowley-Conwy, with a contribution by P. K. Thomas

B. Cattle
Cattle radius

Period Proximal breadth (Bp)

3 86.8
Cattle tibia

Period Distal breadth (Bd)

2 75.2
Cattle astragalus

Period Greatest length (GLl) Distal breadth (Bd) Depth (Dl)

3(o) (69.0) (45.1) 36.9
Cattle metatarsal

Period Distal breadth (Bd) Distal thickness (Dd)

2 (61.2) 34.8

C. Pig
Pig scapula

Period Length of glenoid Breadth of glenoid (BG) Smallest length of collum (SLC)
(GLP)
1 (27.2) (20.0) 17.5
1 33.4 21.6 -
1 - - 20.2
1 (28.5) 20.2 18.2
Pig astragalus

Period Greatest length (GLl)

2 44.9
 Chapter 13 The Faunal Remains    295

In the following table, each jaw fragment is given one line;

thus, if measurements for more than one tooth appear in
the same line, they were found in the same jaw fragment.

Pig mandibular teeth

Period dp4 M1 M2 M3
L WP L WA WP L WA WP L WA
1 15.5 9.9 10.5
1 20.7 13.5 14.1
1 15.2 9.4 9.8 19.1 12.4 13.4
2 17.5 7.8 16.1 – 10.0
2 17.4 7.9 14.3 8.4 9.7

Pig maxillary teeth

Period dp4 M1 M2 M3
L WP L WA WP L WA WP L WA
1 14.8 11.3 16.7 12.7 13.7
1 33.2 19.6
1 16.3 13.5 21.4 16.6 –
1 29.8 19.5
1 21.7 17.8 17.6
2 15.7 – 13.8 23.0 17.2 17.8 29.1 17.6
2 17.7 12.3 12.1 – 15.7 –
Part V
Conclusions
Derek Kennet

Chapter 14 
Paithan in Its Regional, Historical and Cultural
Context

The excavations at Paithan have provided some impor- in Part III, especially those that were retrieved from
tant insights into the cultural and economic transitions layers stratigraphically associated with the temples,
that took place during the Early Historic and Early provides information on the production and distribu-
Medieval periods in central India. In this chapter, we tion of artefacts which contradicts some widely held
will return to some of the key points that have emerged assumptions about the artisanal economy during the
from the excavations, particularly those with reso- Early Historic and Early Medieval periods (see below).
nance beyond Paithan and its immediate locality to The question of the link between the establishment of
the wider arena of Early Historic and Early Medieval the temples and the developing economy is discussed
India more generally, for example the chronology of in some detail further on in the present chapter in rela-
the occupation at the site, the spatial development tion to recent historical work on the subject. On the
of the site, the construction and development of the final question of the nature and extent of occupation
temples and the changing nature of the economy of the at the site during the Early Historic and Early Medieval
Early Historic and Early Medieval periods. periods, the following section sets out what has been
In Chapter 1, the rationale behind the present learned.
project was set out, along with five research questions.
In relation to the first question about the construc-
tion and development of the temples, in Chapter 4 a
detailed outline of the development of the temples has
Part 1: The development of the site
been presented along with a chronological framework and the temples
established on the basis of parallels with dated temples
elsewhere. This question is also further discussed
The development of the site
below in the present chapter. In relation to questions
2 and 3, it has been possible to retrieve a stratigraphi-
The fact that most of the ancient archaeological site of
cally excavated and systematically collected set of arte-
Paithan lies buried beneath the modern town makes a
facts and environmental samples associated with the
comprehensive overview of its development very dif-
temples. This has resulted in 78 studied archaeobotan-
ficult. Although investigating this development was
ical samples that have yielded 8,327 seeds – the largest
not a key aim of the excavations, trenches were exca-
archaeobotanical assemblage from any Early His-
vated in as many different parts of the site as possible
toric site in India by some distance. This has allowed
in an attempt to gain some insight. Such efforts were,
Fuller to present a detailed and compelling analysis in
however, severely hampered by practical constraints
Chapter 12 of the development of agricultural practice
such as finding suitable space and gaining the permis-
throughout the occupation of Paithan. In addition,
sion of landowners. Trench F, located on the northern
the 4,777 fragments of faunal material have allowed
side of the centre of the town, gives an insight into
Rowley-Conwy to present an equally useful picture of
the stratigraphic sequence in that location, but it was
the changing nature of animal husbandry at the site
very restricted in size, as was the trench excavated
in Chapter 13. Together, these two studies present us
not far away by Dikshit many years before. From these
with the most detailed account to date of Early Historic
trenches there is a distance of over 500 m to Trenches
and Early Medieval agricultural practice, supporting a
A and D in the ASI-protected area to the south, and it
detailed discussion of developing crop-livestock inter-
is in this 500 m – currently an archaeological blank
action strategies further on in the present chapter. The
– where the core of the Early Historic settlement is
analysis of the pottery, coins and small finds presented
likely to have been located. Much then still remains to
https://doi.org/10.1515/9783110653540-014
300   Derek Kennet

be learned about the spatial development of Paithan, although they may of course be much earlier (see
beyond the very basic outline that it has been possible Chapter 7). Contemporary occupation appears to have
to surmise below. consisted predominantly of wooden huts with beaten-
earth floors associated with occasional hearths and
rubbish pits.
Earliest levels In the area excavated by Yusuf, a fired-brick drain
and three cylindrical brick structures were located in the
The earliest evidence of occupation that came to lower levels of his sixth stratum (Yusuf 1938: 4, pl. VI;
light during the present excavations was the eroded 1939: 41, pl. XVIIIa, b). The very presence of fired brick
sherds of possible ‘Jorwe ware’ that were excavated in these levels suggests that they should be allocated to
in the lowest levels of Trench D (Chapter 7). Due to Period 2 because fired brick was completely absent
their eroded state, these sherds cannot be attributed from Period 1 deposits in TP1 in Trench A. Yusuf,
a precise date and are taken as being indicative of a however, mentioned that the drain was associated
human presence at Paithan at some time between c. with copper coins, including – it seems – hollow cross
1400 to 700 bc. It should be noted, however, that these coins (Chapter 9), which fall within our Period 1. But
sherds were in stratigraphic association with later Yusuf ’s stratigraphy was not precisely defined, and in
material such as Black and Red ware, indicating that any case, the coins provide only a terminus post quem
they are residual and may have been redeposited for the drain, which could also be later in date. It is
from other areas of the site. It is likely that the seven therefore sensible to allocate Yusuf ’s brick drain to our
microliths that were recorded by the present excava- Period 2, whilst noting the presence of the hollow cross
tions also date from this period, but these are also all coins as evidence of Period 1 activity in the vicinity.
residual and redeposited (Chapter 8). Aside from these Meanwhile, Dikshit’s Period I has similar characteris-
facts, nothing is known at present about the extent tics to our Period 1 and is probably roughly contempo-
and nature of occupation at Paithan during the Proto rary (Dikshit 1973: 28).
Historic period. The presence of Period 1 layers in all of the present
Interestingly, there is no reliable evidence of occu- trenches as well as those excavated by Yusuf and
pation at Paithan during a period of 400 years or more Dikshit suggests that this was perhaps the first period
between 1400 and 700 bc and the Mauryan period. during which occupation became widespread across
Occupation in the Mauryan period, or at least the the Paithan mound. This might suggest the arrival
‘Nanda’ and ‘Maurya-Sunga’ periods (370 to 150 bc), is of new settlers from elsewhere to a location that was
suggested by the eight punch-marked coins on display previously unoccupied, or it might be indicative of the
in the Balasaheb Patil Museum, as well as others that expansion of a smaller, already existing settlement,
were found and published by Yusuf (Chapter 11, nos. the location of which has not yet been found.
1–8; Yusuf 1938: pl. IXa top; 1939: pl. XXc). Some of the The post-Mauryan/pre-Satavahana period is still
NBP sherds found by the present excavations as well as poorly understood in this part of the western Deccan.
by Dikshit (1973: 28) might also conceivably date to this Ray has proposed that political authority in the Deccan
period (Chapter 7). became very localised at this time (Ray 1989: 106). She
has also pointed out that a common denominator of
many of the known settlements of this period is that
Period 1 they are located on overland trade routes (Ray 1989:
(200–100 bc) 105). Period 1 at Paithan – limited as the evidence is
– would therefore appear to be a fairly typical settle-
When we come to Period 1, it is possible to say a little ment of this period, albeit one that seems to have been
more about occupation at the site as layers dating to slowly increasing in size.
this period were excavated by the present project. The
hollow cross-type coins found in these layers suggest
a pre-Satavahana date roughly between 200 and 100 Period 2
bc. Evidence of occupation at this time was found in (100 bc–ad 100 and possibly later)
all of the excavated trenches to varying degrees and
it is also to this period that most of the NBP sherds Fired brick can be said with some confidence to have
recovered by the excavations are likely to be dated, appeared by Period 2. Broken fired bricks were used in
 Chapter 14 Paithan in Its Regional, Historical and Cultural Context    301

Fig. 14.1: A schematic outline of the building materials used in the Paithan sequence and at eight other Early Historic sites in the region.

Trench F to construct a floor, and they also make up a As mentioned above, in the area excavated by
significant part of the dense rubble layers that occur Yusuf, a fired-brick drain and three cylindrical brick
in the Period 2 layers in TP1 in Trench A. Although no structures were located in the lower levels (Yusuf 1938:
buildings or walls were uncovered from this period, 4, pl VI; 1939: 41, pl. XVIIIa, b). To judge by the pub-
it can be surmised, on the basis of excavations under- lished photograph, the drain was roughly constructed
taken at contemporary sites elsewhere in the region whilst the ‘cylindrical brick structures’ appear to be
(see Fig. 14.1), that with the introduction of fired brick, ring-wells that were dug down through these layers
domestic architecture changed its form from circular- at a later date. Nonetheless, on the basis that the fired
shaped wooden huts to rectangular houses with brick brick first appears in Period 2, the drain is indicative of
walls and terracotta tiled roofs. activity in this part of the site at that time.
In addition, it will have been noted that in Trench Dikshit recorded the presence of a 1.22-m-thick
B objects apparently associated with some form of layer of mud overlying his Period I deposits (1973: 29).
manufacturing activity came to light in this period. This resulted, he assumed, from a heavy flood that
Not enough evidence was uncovered to demonstrate covered the whole site at some time. This is an unlikely
more than the presence of activities associated with a explanation and it seems more likely to be the result
furnace. of natural soil accumulation during a time when the
302   Derek Kennet

area was abandoned and overgrown by vegetation. at the site contracted after Period 2, although the avail-
Indeed, it is noteworthy that Yusuf recorded a thick able ‘sample’ area is very small.
earth deposit below his Fifth Stratum, which is strati- Of course, the temples themselves indicate that
graphically much later than Dikshit’s, confirming that activity certainly continued between the 4th/5th
such accumulations result from localised abandon- century ad and about the 8th century in one location,
ment rather than widespread flooding (Yusuf 1938: 4). although nothing is known about the nature of occupa-
Above the mud layer, Dikshit reported finding a ruined tion elsewhere on the site at this time. With the excep-
structure built of ‘extraordinarily large-sized bricks’ that tion of a possible issue of the 4th/5th century and one
he interpreted as the plinth of a ruined temple and which of the 7th century from the foundation deposits of the
he tentatively dated to the Rashtrakuta period, although temples (Chapter 9, coins 33 and 34), no coins of this
he provided no evidence to support either interpreta- period have come to light – a point which is further dis-
tion or dating (Dikshit 1973: 29). Dikshit did not give cussed below in the section on coins and monetization.
the measurements of the bricks, so it is impossible to It is not possible to be certain when the temples were
compare them with those from Trench A. Nonetheless, finally abandoned. It seems most likely that this took
it is notable that the brick drain uncovered by Yusuf is place around the 8th century or perhaps slightly later
also described as having been made of bricks of ‘excep- as there are indications that the temples remained in
tionally large size’ (9 × 20 × 50 cm in this case) and that use for a reasonable period of time, although they did
they are the only really large bricks from the sequence not undergo further embellishments.
(Yusuf 1938: 4; 1939: 41). It therefore seems likely that The nature of occupation in large, urban centres
Dikshit is mistaken in his dating – and also probably during the post-Satavahana period has been the focus
his interpretation – of this brick structure and that it of much debate. It has been argued by some scholars
should be placed in Period 2. If this is correct, Period that this was a period of widespread urban decline but
2 remains were located in all excavated trenches and these arguments, although they need to be carefully
it is therefore likely to have been a period of relatively considered, are not without problems (e.g. Sharma
widespread activity across the site. 1987; Kennet 2013). It is difficult to form a clear picture
The Satavahana period has been characterised as of the nature of occupation at Paithan from the limited
a period of growth and economic prosperity in the area that has been exposed, but the construction
western Deccan (e.g. Ray 1986; 1989: 106–107). As a site of two early Hindu temples in what is likely to have
of some historical significance, it is no surprise to see been an area that was marginal to the earlier site is an
evidence of widespread activity, large brick structures interesting development as it is suggestive of a shift in
and abundant use of coinage at this time. the focus of settlement. This is a point that is signifi-
cant to our understanding of the formal elaboration of
temples at this time and their relationship to existing
Period 3 settlement and power structures.
(4th/5th to 8th/9th centuries ad)

With the exception of a single coin (Chapter 9: coin 33) Period 4

which is datable to between the 2nd and 4th century
ad, there is no datable evidence for activity at Paithan After Period 3, there is no clear evidence for occupation
between the 1st and the 7th centuries ad. This is likely at the site until the 14th/15th centuries, to which period
to be due to an absence of coin deposition at this time, some of the glazed pottery from the present excavations
but it might reflect a reduction of settlement or an can be dated as can a number of coins reported by Yusuf
abandonment. Until a more precise understanding of (1938: 3; 1939: 41).
the pottery chronology is available, it is impossible to It should be noted that Period 4 is a rather loosely
be certain. defined phase and has essentially been taken to rep-
With the possible exception of a single deposit resent all post-Medieval periods at the site. No system-
from Trench D1 (751), Period 3 remains are entirely atic attempt was made to investigate the archaeology of
restricted to Trench A, where they are associated with these periods by the present project. Period 4 levels were
the construction, enlargement and use of the two brick present in the Trench A area, but they had been almost
temples. This might suggest that the extent of activity entirely removed by Yusuf ’s earlier excavations. Period
4 levels were excavated in Trench B, where they con-
 Chapter 14 Paithan in Its Regional, Historical and Cultural Context    303

Table 14.1: Tentative correlation of the excavated sequences from Yusuf (1938; 1939), Dikshit (1973) and the present excavations.

Yusuf 1938; 1939 Dikshit 1973 Present excavations

1st stratum Stone-and-mortar

and brick-and-mud ?
buildings, British
period?
Period 4 Houses,
2nd ?18th–19th cen- Period IV House, 17th century +
stratum turies 18th–19th cen-
turies
3rd stratum ?
Mughal period
4th stratum Jumbled, tilted
walls, ?
?14th–15th cen-
turies ?
5th stratum Thick deposit
with coins, no
structures
6th stratum Brick temples Period 3 Brick temples
(upper)
Period III Brick ‘temple’ Period 2 Fired bricks
6th stratum Brick drain, coins Period II Thick sterile mud
(lower)
Period 1 Post-holes, coins,
Period I Coins, NBP NBP

sisted of the corner of a stone building overlying a covered by the ASI protected area was abandoned, the
building with six pillar bases. Below this, a sequence buildings within it fell into disrepair and their bricks
of uneven floors, ash and rubbish pits came to light. In and stones were removed for use as building materials
Trenches D1 and D2, Period 4 levels had been removed elsewhere.
by brick quarrying. Neither were they present in It would be wrong to attempt to characterise the
Trench F, although it is not clear why not. Dikshit notes nature and extent of occupation in Period 4 any more
the presence of later buildings dating to his Period IV precisely based on the evidence that has been col-
in the trench he excavated. These remains included lected here. Nonetheless, the extensive remains of
the cellar of a house that had cut into underlying levels Mughal and post-Mughal period architecture that
(Dikshit 1973: 29). Yusuf describes an upper level of are still scattered about the site suggest that a more
buildings constructed of stone, lime and brick dated by detailed investigation of this period would yield inter-
coins to the British, Asaf Jahi and Mughal periods (1938 esting and informative results related to the develop-
3, pl. III; 1939: 41, pl. XVI). ment of minor urban centres at this time.
The general impression is that the site was quite
densely built up towards the end of the Mughal period,
reaching a peak perhaps in the Asaf Jahi or early Summary
British period, and it may have been during this time
that the urban plan of the older part of modern Paithan The present excavations did not have as their main aim
came into existence (Fig. 1.5). The whole town was an investigation of the spatial development of the site.
also walled at some time, possibly during the Mughal Indeed, such an aim would require a much larger-scale
period (Govt. of Maharashtra 1977: 1035, 1604). At a project with greater resources than were available to
later date, it is likely that the area of the site that is now the present one. Nonetheless, it has been possible to
304   Derek Kennet

set out a tentative overview based on the information Satavahana towns actually looked like, how they were
from the various trenches that have so far been exca- constructed, whether they included large, monumen-
vated (see Table 14.1). tal architecture and how they were laid out. Excavated
In order to provide a regional context for these sites have revealed evidence for earthen ramparts,
developments, Fig. 14.1 shows a schematic outline of sometimes reinforced with fired brick, and generally
the building materials used at Paithan compared to small and isolated units of small-scale architecture
eight other excavated Early Historic and Medieval such as houses and structures of a similar scale. From
sites in the region. It should be noted that the absolute an archaeological perspective, therefore, the Early
chronology of almost all of the phases and periods Historic town still remains something of a mystery and
shown is tentative and open to question, and this may the question of the nature of Early Historic urbanism
explain some of the apparent anomalies, for example perhaps represents one of the most important areas of
the apparently late introduction of brick at Nasik, Early Historic archaeological research.
Nagara and Nevasa. The phases of abandonment in the
Early Medieval period are also open to question and
reinterpretation being based, as they are, on limited The development of the temples
excavation, biased sampling and problematic chronol-
ogies (e.g. Kennet 2004b; 2013). Nonetheless, Fig. 14.1 One of the most significant aspects of the Paithan exca-
serves to demonstrate that the pattern of development vations is the detailed architectural and stratigraphic
recorded at Paithan is part of a regional trend rather investigation of the two Vakataka to Rashtrakuta
than a local or site-specific development, even if the period brick temples, the details of which have been
precise chronology has not yet been finalised. set out in Chapter 4. Aside from providing an insight
The same can also be said of the general trends of into the nature of brick temples of this date in Maha-
expansion and contraction that have been tentatively rashtra, about which very little is presently known,
identified at Paithan. A phase of growth up until the these are almost the first temples of this period that
end of Period 2, followed by a phase of contraction of have been stratigraphically excavated, recorded and
the settlement, would conform to the general trends analysed to the same level of detail anywhere in India
that have been identified at other sites in the region (e.g. Jayaswal 2001). So, although the Paithan temples
(e.g. Sharma 1987: 60–83). Nonetheless, the evidence were not major monuments benefiting from royal
from Paithan is not strong, and it would be wrong to patronage – as were the temples at Bhitari – they are
make too much of it here. well-documented examples of small to medium-sized
An obvious question that is raised by the analy- Early Hindu temples that were built during the forma-
sis shown in Fig. 14.1 is what do these regional trends tive period of temple architecture in South Asia (e.g.
represent? For example, why was there a transition Meister et al. 1988: 22). For this reason, they provide
from mud-and-wood to brick-and-tile architecture an important insight into the architectural traditions
in around the 1st century bc? Is it indicative of cul- of the time.
tural, technological or economic changes? Does the Both temples underwent a phased architectural
manufacture of tile and brick reflect a more organized, development before they reached their final forms,
commercialised economy or does it simply reflect the by which time they both consisted of what might be
spread of a new technology? Similar questions are referred to as the ‘standard temple plan’, by which is
raised by the transition from brick to stone in the Early meant a square or rectangular garbhagriha (cella) for
Medieval period. Does this reflect an increase in the the deity and an attached mandapa (hall) for shelter-
cost of fuel for firing bricks, a change in cultural atti- ing the devotee (Meister et al. 1988: 22). In the case of
tudes, or a change in the systems that produced and the North Temple, the fact that it has been possible to
distributed these materials? We are not at present in a trace the development from a relatively simple plat-
position to answer these questions, but this analysis form shrine in Phase I to a standard temple plan in
demonstrates that important insights into economic Phase III might provide something of a blueprint for
and social developments may be revealed by further the way in which the standard temple plan is likely to
consideration of these issues. have evolved from the 4th century more generally. This
The lack of evidence for monumental architec- insight is certainly one of the most important contri-
ture, paved streets and dense urban architecture at butions to arise from this study. By contrast, the South
Paithan reminds us how little is known about what Temple, which was probably built in the 7th century –
 Chapter 14 Paithan in Its Regional, Historical and Cultural Context    305

but certainly after the North Temple had already been structure replaced a wooden shrine at the same location,
in existence for 200 years or more – did not undergo but any such evidence would have been destroyed during
the same development but was planned and con- the construction of the Phase I structure itself.
structed according to the standard temple plan, which In considering what the Phase I structure actually
suggests that the standard temple plan had become looked like when it was in use, it is possible that it was the
the accepted norm in the intervening years. basal platform of an open-sided shrine that was covered
In view of the significance of these questions, a by a roof supported on pillars: a one-storied, simplified
few comments are offered below on the key stages in and obviously later version of the concept depicted on
the Paithan temples’ development. stone reliefs from monuments such as Amaravati (Fig.
14.2) and Jaggayyapeta, which Shah has compared to
the 6th-century temple at Gop in Gujarat (Coomaras-
The foundations wamy 1930: figs 30, 33; Shah 1975: fig. 9).

The foundations of both temples are remarkable. It

came as a surprise to the excavators to encounter foun-
dations of this scale and elaborate complexity sup-
porting structures of such a relatively small size. The
foundations are clearly in excess of what is required for
the size of the building, suggesting that the rationale
behind them is sacred rather than structural.
Parallels for such foundations have already been
discussed in Chapter 4, where the suggestion was
made that this aspect of temple construction might be
a formalisation of regional, non-sacred construction
techniques, thus providing a possible insight into the
origins of temple-construction rituals and methods.
It should also be noted that the close correspondence
between the Paithan foundations and the Brahminical
literature describing temple foundations suggests that
the sacred formulae related to temple construction
were widely disseminated and closely adhered to, even
at this relatively early date.

North Temple Phase I: Mulaprasada

Phase I of the North Temple appears to have been a small,

square shrine located on a brick platform. It might be
seen as a more elaborate form of the square brick shrines
that are known from sites such as Yelleshvarum, Nagar-
junakonda and Veerapuram, and which appear to have
been in existence from at least the 3rd/4th centuries
Fig. 14.2: Drawing of a shrine in a relief from Amaravati (from
ad (Khan 1963: 14–15; Sarkar and Misra 1972: 24–31; Coomaraswamy 1930: fig. 33).
Sastri et al. 1984: 38–44). These simple shrines them-
selves might be seen, in the broader scheme of things,
However it looked, Phase I is the earliest evidence of
as representing a development towards a more per-
any cultic activity that has so far come to light in this
manent, brick-built manifestation of earlier wooden
location. No exact parallels can be found for it, but it
shrines of a type that is known, at present, exclusively
is possible that other such brick shrines do still exist
from representations in relief sculpture and on coins
from this period and they are encased within the later
(e.g. Coomaraswamy 1927: 83; 1930: fig. 23; 1992 [1956]:
structures of brick temples, thus making them invis-
105–109; see also Ray 2009). It is possible that the Phase I
306   Derek Kennet

ible, except where there has been damage to the later Chapter 4, the first mandapas that closely resemble
phases of construction. Phase III of the North Temple at Paithan in terms of
proportions and layout first occur at the end of the 6th
or the early 7th century in North India, or a little later
North Temple Phase II: Shikhara in the Karnata-Dravida tradition, whilst by the 8th
century, they had become a feature of most temples
As has been noted in Chapter 4, the thickened walls of and part of the standard temple plan (Meister et al.
Phase II of the North Temple were probably intended 1988: 230–236, 256–260, 265–267; Hardy 1995: 67, 71,
to support a brick shikhara (tower), the introduction fig. 46).
of which – it has been suggested – can be dated to At Paithan, the addition of a mandapa appears to
about the middle of the 5th century on the basis of par- represent a further stage in the monumentalisation
allels with temples such as Bhitargaon, Deogarh and of the shrine and its final transformation into what
the brick temple at Bodhgaya (Meister et al. 1988: 23). can be recognized as a fully formed temple. Once again,
The triratha form (square with slight projections on the architectural development of the structure followed
each side) of the structure and the fact that it was prob- closely developments that were occurring elsewhere in
ably surmounted by a shikhara similar to – but smaller India, including the addition of adhisthana mouldings
than – Bhitargaon, suggest that the shrine underwent (basal plinth mouldings) to the mandapa wall. This is a
increased monumentalisation at this time. As with the further statement of the obvious point that, even in a rela-
foundations mentioned above, this development was tively modest-sized temple such as this, there was close
carried out in close conformity with the prevalent correspondence with the pan-Indian norms and rules of
pan-Indian norms of temple architecture. Hindu temple construction.
It can be imagined that the resulting shikhara,
even if only of relatively modest proportions, would
have formed a prominent landmark in the surround- South Temple Phase I: construction
ing landscape, probably reflecting an increased local
importance of the shrine. The construction of the South Temple took place some
One key question about the Phase II shine is how time shortly after the addition of a mandapa to the
access would have been gained to the deity, given the North Temple. This can be demonstrated by the size
absence of an obvious entrance into the garbhagriha. of the bricks, which had been consistently 7 × 25 × 40
One possibility is that there may have been window- cm through the first three phases of the North Temple
like openings above ground level in one or more of the but which changed to 6.5 × 26 × 42/43 cm before the
walls. Alternatively, it is possible that there was access construction of the South Temple and Phase IV of the
from the west side, the evidence for which has been North Temple and did not change again. Although
obscured by later destruction of the wall on that side. using brick sizes as an absolute dating method at a
regional scale is highly problematic and error prone, it
is likely to be reliable when used for the relative dating
North Temple Phase III: Mandapa of individual structure phases, as is the case here.
At the same time, the similarity in the style of the
As Kramrisch has pointed out in relation to the Hindu adhisthana mouldings in the mandapas of the two
temple more generally, the addition of a mandapa was temples suggests that there was not a long time gap
something of an afterthought, and so it seems was the between the two construction phases. One possible
case in the North Temple at Paithan (Kramrisch 1946: interpretation of this sequence is that between the con-
254). Here, the mandapa can be shown to have been a struction of Phase II of the North Temple and Phase I
later addition. It is built of bricks which are the same of the South Temple, a mandapa had become a recog-
size and quality as the earlier two phases, but they nized and expected part of a temple and was included
clearly abut or are inter-keyed with the Phase II wall. from the planning stage rather than being added later.
In addition, the mandapa foundation cut is different If this is correct, then the development of the North
from, and shallower than, the Phase II foundation cut. and South Temples presents a microcosm of the evolu-
The development of the mandapa as an architec- tion of the standard temple plan through this period.
tural concept is key to the formation of the Hindu Once again, it is impossible to know whether
temple as it is now known. As has been argued in an earlier shrine of some sort stood where the South
 Chapter 14 Paithan in Its Regional, Historical and Cultural Context    307

Temple now stands; any evidence for such a shrine Discussion

would have been completely destroyed by the con-
struction of the temple. In either case, it seems that the Unfortunately, with the exception of the fragmentary
construction of a second temple right next to the North stone sculpture of a female found in Trench B some
Temple represents a further increase in the monumen- 70 m to the south of Trench A (Chapter 8 No. 234, Figs
talisation of this sacred locality and highlights, once 8.17 and 8.18), no images or inscriptions were found
again, its growing importance. directly in association with the temples that might give
some insight into the nature of the cult that was prac-
ticed here. The Balasaheb Patil Government Museum
North Temple Phases IV–V and South Temple Phases at Paithan contains, however, a number of terracotta,
II–V: later developments kaolin and stone figurines that apparently come from
Paithan, although it is not known exactly wherea-
Following the construction of the South Temple, bouts they were found. Some of these are illustrated
further architectural developments consist only of in Chapter 11 (Figs 11.6 and 11.7), and further exam-
repairs, rebuildings, internal subdivisions and the ples are published by Morwanchikar (1985: chapter 7).
construction of a crude perimeter wall some time later. Some of these are datable to about the 4th century ad
The North Temple saw the construction of an inter- based on stylistic comparisons, and these are presuma-
nal subdivision (N-IV), probably a low, raised plat- bly the types of images that might originally have been
form in the middle of the mandapa, that was probably associated with the Paithan temples.
related to a subordinate deity that was placed there. It Whatever the nature of the cult that was practiced,
appears to be very similar in concept to Phase V of the it is clear that, despite the elaborate formulae that
South Temple and it is tempting to suggest that both were involved in the planning and construction of these
alterations were made at the same time. two temples (as exemplified by the careful brickwork, the
The layout of the South Temple was adapted, elaborate foundations and the precise conformity with
extended and the superstructure almost completely the rules as set out in Brahminical texts), the temples
rebuilt in Phase IV. Nonetheless, it is key to note that, were dynamic structures, both of which were constantly
apart from a slight increase in size towards the east, adapted and re-formed throughout their lives.
there was no further significant enlargement or monu- In terms of the development of Hindu temple
mentalisation of either of the structures. This suggests architecture, it is the North Temple that is the most
that the site had reached its zenith, at least in terms interesting and significant. It developed from a simple
of size, investment and architectural development. platform shrine in the 4th/5th century and was then
Indeed, the building techniques used in some of the elaborated through the construction of a shikhara
later extensions and adaptations suggest a decline in around the middle of the 5th century and was finally
the quality of construction and perhaps some degree turned into a larger and more complex standard
of neglect or impoverishment. The latest changes in temple plan incorporating a mandapa with adhisthana
the North Temple (the Phase V platform at the western mouldings in the late 6th or early 7th century. In this
end of the mandapa interior against the wall of the way, it passed through the stages of temple develop-
garbhagriha) are extremely crude and might even be ment that have been mapped out for the Hindu temple
taken to indicate informal re-use of the temple for more generally on the basis of standing buildings of
cultic purposes once the temple itself had been for- different dates (e.g. Meister et al. 1988: 22–24; Pichard
mally abandoned and had fallen into disrepair. 1994: chapter 2, fig. 4; Hardy 1995: chapter 4). The
The stone perimeter wall, which appears to have North Temple at Paithan is, however, the only temple
been constructed some time later, seems to represent at which these developments have been clearly shown
a crude attempt to enclose the temples and to define the to have taken place in a single structure.
plot on which they stand. This might reflect a change Why the North Temple should have developed in
in land use immediately surrounding the temples that this way is an important question. It seems quite pos-
created the requirement to isolate them in this way. So sible that the small Phase I shrine housed a local cult
little of the stratigraphic layers related to these later image whose significance in the area attracted the
phases were left undisturbed by Yusuf, that it is impos- patronage of a local personage, thereby allowing the
sible to say much about these developments. cult and the structure that housed it to grow in stature
throughout the 5th and 6th centuries. In suggesting
308   Derek Kennet

this, it is assumed that the architectural developments Economic change and agricultural
that have been mapped out in the North Temple were intensification
paralleled by an increased formality, permanence and
institutionalisation of the cult that it housed. These One of the most significant findings of the present
trends were continued into the construction of the excavations is the evidence of the way in which agri-
South Temple, which appears to take up the architec- cultural practice and strategy changed through the
tural story that had been started in the North Temple at site’s history. This evidence has been outlined by
a later stage in its development. The South Temple was Fuller in Chapter 12 and can be linked to developments
conceived and constructed as a fully formed standard described by Rowley-Conwy in Chapter 13.
temple plan with mandapa from the start. Its presence, Fuller’s analysis of the archaeobotanical assem-
size and form are also likely to reflect the increasing blage provides evidence for increasing diversification,
permanence, institutionalisation and monumentality intensification, investment in and commercialisation
of the cult that was practiced here. Indeed, it might of cultivation strategies. It also suggests a change in
therefore be argued that, in some ways, the develop- the organization of agricultural labour away from cen-
ment of the two temples can be seen as an architec- tralized or communally organized processing towards
tural palimpsest of the transformation of a ‘village household-based processing. The argument for the
cult’ into a ‘temple cult’, with its attendant formalisa- diversification of cropping strategies is based on the
tion and Hinduization, following the pattern outlined increase in the diversity of crops between Periods 1
by Eschmann and others on the basis of fieldwork and 2 and also between Periods 2 and 3, most notably,
in Orissa (Eschmann et al. 1978: 86–89; see also Ray the trend away from kodo millet, which dominated
2009). the Period 1 assemblage, towards a wider range of
Arguing that such developments took place at millets through both Periods 2 and 3, including
Paithan is therefore to argue that the Paithan cult African millets and the so-called ‘S/E/B’ millets. These
and temples were not a unique case but rather that changes are likely to be explained by the exploitation
the developments mapped out here were linked to of a wider range of soil types, which are likely to have
a broader process that was taking place at this time included lower-grade, marginal areas around the site
as cults, deities and the temples in which they were with poorer soils that would not have been the first
housed became an increasingly important aspect of choice of cultivators but that were exploited in later
the political and cultural institutions of Early Medieval periods due to increased pressure on land and food
states across India (e.g. Bakker 1992; Kulke 1978; 1995b; production. The argument for intensification is based
Willis 2009: 113–122). This theme is further explored in on the evidence for an increase in winter crops, espe-
relation to the associated evidence for Paithan’s eco- cially winter pulses, such as chickpea, grasspea, pea
nomic development that is discussed below. and lentil, that would have allowed double-cropping
of the same land that was used for other crops in the
summer but would also have allowed a reduction in

Part 2: The changing economy the length of the fallow periods between crops due
to their nitrogen-fixing characteristics. As for invest-
of the Early Historic and Early ment, as Fuller points out, the increased exploitation

Medieval periods of winter crops in an area such as Maharashtra, which

experiences relatively low levels of winter rainfall,
would almost certainly have required the construc-
An excavation such as this can provide insights into
tion of some form of irrigation such as tanks, canals
various aspects of the ancient economy. This section
and/or wells, which would have required techni-
presents an integrated overview of the agricultural
cal knowledge as well as the investment of money
economy (as elucidated by the archaeobotanical
and labour. Fuller’s argument for the growing com-
and paleofaunal assemblages), patterns of produc-
mercialisation of the agricultural economy is based
tion and consumption (as elucidated principally by
on the continuing trend of increase in the amount of
pottery deposition) and patterns of coin deposition.
cotton at the site from Period 1 to 3 and the assumption
An attempt is made to place this discussion within the
– because textiles are labour-intensive and therefore
broader historical context of the Early Historic and
imply surplus production beyond subsistence – that
Early Medieval periods in central India.
cotton is indicative of ‘cash-cropping’, or production
 Chapter 14 Paithan in Its Regional, Historical and Cultural Context    309

for commerce rather than for local consumption. This comparable shifts in meat consumption – the likelihood
certainly seems to have been the case in the Mughal is therefore that the rationale behind this change is eco-
period (Habib 1999: 43) and it should also be noted nomic.
that the Periplus specifically mentions cotton from Indeed, if we consider how pigs and sheep/goats
Tagara (Ter) being transported by cart overland to tend to be kept and fed, there are some obvious poten-
Barygaza on the coast (Casson 1989: 82–83), although tial links between a decline in pig husbandry and the
the commerce might also have been directed towards changes in cultivation strategy identified by Fuller.
more localised markets. Of course, it is possible that The increasingly extensive agricultural exploitation
some cotton production was also for household con- of the area around Paithan suggested by Fuller’s evi-
sumption, but, as Fuller points out, the continued dence would very likely have resulted in a reduction
occurrence of charred cotton seeds in archaeobotani- in the amount of scattered micro-environments that
cal samples from urban contexts is probably indica- are suitable for foraging by pigs, for example areas
tive of larger-scale production. In addition, Fuller of woodland and patches of disused, overgrown and
notes that African millets, which had been cultivated poor-grade land. This would have made the feeding of
in low quantities in the region since the Chalcolithic, pigs more difficult and potentially more expensive and
but which show a trend of marked increase through may have led to free-foraging pigs posing an increasing
Periods 2 and 3, would have been cheaper to process threat to crop fields close to settlements. Such tensions
and might therefore indicate an increase in commercial can perhaps be seen in the Laws of Manu (on the dating
considerations amongst cultivators or the landown- of which see below), a text that defines an area around
ers that controlled agricultural production. The argu- villages and towns within which it is the responsibil-
ment for a shift towards household-based production ity of the cultivator to maintain good fencing, without
is based on the presence of weed seeds. Fuller points any gaps ‘that a dog or a pig could put his muzzle or
out that the increased numbers of weed seeds in the snout through’ (Laws of Manu 1991: 8.237–244). At the
archaeobotanical assemblage probably results from an same time, there would have been an accompanying
increased degree of crop-processing being carried out increase in crop residues such as straw and haulm
on site in the latter two periods. This he interprets as from the expanding cultivated areas, and it is exactly
being indicative of a change from a more centralized or these types of residue that are ideal for feeding goats
community-level organization of agricultural labour in and sheep, although they are not at all suitable for
Period 1 towards an increasing emphasis on smaller, pigs. Indeed, crop residues are still important today
household-level organization through Periods 2 and in many parts of India for feeding goats and sheep as a
3, with the most significant shift apparently occurring traditional part of crop-livestock interaction strategies
between Periods 1 and 2. (e.g. Geerlings 2001: 36–38; Singh et al. 2007: 41–36).
As for the palaeofaunal evidence, in his analysis, This makes clear one potential link between the
Rowley-Conwy has set out a number of points, the trends defined in the archaeobotanical and palaeo-
most significant of which is a continuing decline in the faunal assemblages by illustrating how much more
amount of pig and the commensurate increase in the easily sheep and goat husbandry could have been inte-
consumption of sheep/goat between Periods 1 and grated into the more extensified cropping strategies of
4. This very clear, long-term trend raises two obvious Periods 2 and 3 than pig husbandry might have been.
questions: how might this shift be interpreted in terms But this is not the whole story. It may also be that inten-
of broader agricultural and economic strategy? And sified demand for meat made sheep and goats a better
can it in any way be linked to the trends identified by option than pigs. In order to explain this, it is nec-
Fuller in the archaeobotanical assemblage? essary to look at evidence from the Levant, a region
Firstly, it needs to be stated that there are no where a considerable amount of analysis of palaeofau-
grounds for using the decline in pig bones as an ‘ethnic nal remains from archaeological sites has been carried
index fossil’ with which to identify the presence of out, which has led to the development of a number
Muslims at Paithan as the shift away from pig began of theories about the significance of pig husbandry
between Periods 1 and 2, which is obviously far too and the possible reasons behind its decline in certain
early for Islamic practice to have been a contribu- periods, some of which are directly applicable to the
tory factor. Neither is it possible to identify any other Paithan evidence. In the Levant, it has been argued
obvious cultural or social factors that might have acted that pigs played an important a role in certain quite
at this period and over such a duration of time to cause specific types of agricultural strategy and that changes
310   Derek Kennet

in the levels of pig consumption can be indicative of production than would have been possible on the basis
changes in broader economic structures. Generally, of locally available resources alone.
and for reasons that will be further outlined below, The potential value of goat and sheep mobility is
evidence for high levels of pig consumption is thought further increased if the availability of fallow-grazing
to be indicative of localised, relatively low-intensity, around the site was being reduced by double-crop-
subsistence-based, small-holder economies that are ping, as Fuller’s evidence again suggests was the case
poorly integrated into broader market structures, at Paithan. The Raika, a group of sheep pastoralists in
whilst lower levels of pig have been taken to indicate contemporary Rajasthan, provide a modern illustra-
higher-intensity, more commercialised economies that tion of mobile pastoralist strategies and their interac-
are better integrated into broader market structures tion with double-cropping cultivators (Agrawal 1999;
(e.g. Redding 1991; Zeder 1996: 298–299, 306–309; Geerlings 2001). In studies of the Raika, it has been
Hesse and Wapnish 1997: 238–253). noted that a shift to irrigated double-cropping has
To understand why this is, and why it might also led to a reduction in the availability of fallow-grazing
be applicable to Paithan, we need to consider an imag- and a commensurate rise in the distances moved by
inary Paithan small-holder, who kept a few pigs that pastoralists each day, and the numbers of pastoral-
were fed on household refuse and let out to forage ists practicing longer-distance transhumant migra-
during the day. This would have been a cheap and tion (Agrawal 1999: 14; Geerlings 2001: 37–38; Robbins
effective way of producing small amounts of meat for 1994: 10). Of course, strategies such as those used by
the family. Pigs breed and grow quickly, their meat is the Raika involving regular, long-distance movement
high in calories and fat and they could have been kept would be completely impractical with pigs.
and fed at little or no cost if they were kept in small In the light of these points, the decline in pig and
numbers (Zeder 1996: 300–303; Singh et al. 2007: the increase in sheep and goat consumption identified
37). A problem would have arisen, however, if it had by Rowley-Conwy make perfect sense in relation to the
become necessary or desirable to increase meat pro- changes in cropping practices that have been outlined
duction significantly, for example due to the wish or by Fuller: both are likely to have resulted from pres-
need to sell commercially. Keeping a number of pigs sure to expand and increase production and both may
larger than could have been fed from household refuse possibly reflect a shift towards a more commercially
and free foraging would have immediately meant that oriented, cash-cropping economy.
it would have been necessary to provide water, shelter, It also needs to be pointed out that the changes in
containment and especially feed, all of which would agricultural strategy that have been outlined above are
have required expenditure and investment. This would likely to have been accompanied by changes in social
have been especially true if patches of woodland and relationships. For example, according to the argument
other suitable pig-foraging areas around the settle- made above, the increase in mobile goat and sheep
ment were being cleared, fenced and brought under production would have led to the emergence of spe-
cultivation, as it seems from Fuller’s evidence they cialist, mobile or semi-mobile sheep/goat pastoralist
were. Goats and sheep would have had the advantage groups similar to the modern-day Raika mentioned
of being able to fallow graze, feeding on crop residues above. It is unclear whether such groups would have
whilst a large herd of pigs would have consumed foods been those pushed to the margins by the expansion
that would have needed to be specially produced or of cultivation or those being brought in to the margins
purchased. Goats and sheep also have the advantage of cultivation from the outside, as outlined by Ludden
of being more comfortably mobile than pigs. Mobility (1999: 72). In either case, the emergence of specialist
was potentially a key issue as it would have allowed groups in one area of the economy is likely to have
goat and sheep pastoralists to move herds and flocks been accompanied by a degree of specialization more
several kilometres (or even tens or hundreds of kilo- widely across society in order to allow for the produc-
metres), either on a daily basis ‘scout-browsing’ or tion of surplus that could be traded for sheep/goat
seasonally throughout the year as part of a longer-dis- meat. In short, the archaeobotanical and paleofaunal
tance transhumant migration strategy. In both cases, evidence may reflect a long and slow shift from a sub-
herders would have been able to take advantage of sistence-oriented economy to one that was increas-
more widely scattered grazing resources with sheep ingly specialized and possibly commercialised.
and goats, thereby permitting higher levels of livestock In this context, it is interesting to consider the point
made by Fuller in relation to the evidence for a shift
 Chapter 14 Paithan in Its Regional, Historical and Cultural Context    311

towards greater numbers of crops being processed in status, low-value food and upper-caste Hindus have
individual households rather than communally or cen- tended to shun it.
trally. Might this development (the evidence for which,
it must be noted, is still very limited) reflect the sort of
transformation that Chakravarti has in mind when she
describes ‘the emergence of the family as the basic pro-
ducing unit’ in agriculture’? (Chakravarti 1996: 93). Of
course, Chakravarti’s comments were made specifi-
cally in relation to the gana-sanghas of north-eastern
India and she saw the developments she describes
as being linked to a number of other very profound
social and political changes, including the emergence
of private control over land for the first time. Less is
Fig. 14.3: Numbers of Indian livestock (1,000s) from the 17th Live-
known about the social and political configuration of stock Census of 2003 (source: FAO 2001–2003).
the Paithan region than is known about the area of
the gana-sanghas in the earliest periods represented in
Interestingly, pigs are also associated with low com-
this sequence, but it is quite possible that some of the
mercial value and low social status in some ancient
same developments took place in both areas, even if at
texts, for example the Arthashastra and the Laws of
slightly different periods.
Manu. Because of its nature, the Arthashastra has
little to say on the polluting nature of pigs, but it does
set the fine for stealing or killing a pig at 54 panas,
The status of pigs
which is the lowest level for any livestock (the same
as a dog, cat or cock) and is 1/11th of the 600 pana fine
Before moving on, it is worth making a few comments
for the theft of a cow (Arthashastra 1987: 4.10.2–11). In
on the social and cultural status of pigs in India. By
almost all instances where pigs are mentioned in the
world standards, India’s pig population has tradi-
Laws of Manu, they are regarded as low-value or unclean
tionally always been low, although it has been growing
and polluting animals, and this is reflected in the fact
significantly since the 1960s mainly due to modern,
that one of the most common terms for them is vid-
large-scale production systems (Tisdell and Gali 1999:
varaha, literally ‘shit-boar’ or ‘dung-heap pig’ (Laws of
7, table 1). Today, pigs are a relatively insignificant
Manu 1991: 3.190, 3.239, 3.241, 5.14, 5.19, 11.155, 11.157,
part of the Indian livestock industry, making up less
11.200, 12.55). Along with dogs, they are protected by a
than 3% of the total cattle, buffalo, goat, sheep and
fine of only one mashaka, which, based on a mashaka
pig population in 2003 (Fig. 14.3), a figure that is, in
being worth 1/16th of a pana (van Schrötter 1930: 548;
fact, remarkably close to Rowley-Conwy’s figures for
Laws of Manu 1991: 8.132–7; Arthashastra 1987: 109, 327
Period 4 at Paithan. How long this has been the case
(2.12.24), 767 (2.19.2–7)), is 1/3,200th of the 200 pana
is unknown, but low levels of pig husbandry were also
fine for small cow and 1/8,000th of the 500 pana fine for
noted by Fa-Hsien, the Chinese Buddhist traveller who
a large cow (Laws of Manu 1991: 8.290–8).
visited India in the 5th century ad and who was led to
Bearing in mind the often complex symbolism
comment that people in India did not keep pigs or fowls
linking animals and social status in Indian litera-
at all at that time (Fa-Hsien 1923: 21). In most parts of
ture (e.g. Doniger 2009), the Arthashastra and Laws of
modern-day India, largely due to their low-mainte-
Manu suggest that pigs were already the lowest-valued
nance costs, pigs are kept in small numbers by under-
livestock by the 2nd to 4th centuries ad (to which broad
privileged and landless groups for whom they provide
period these texts are to be dated following, most
income and food (Singh et al. 2007: 37, 49). They are
recently, Willis 2009: 62, 204–206 and Bronkhorst
primarily left to free-forage, exploiting all available
2011: 2, 65–74). Of course, as the Paithan sequence has
food sources in exactly the same way as the imaginary
clearly demonstrated, by this time, pig consumption
Paithan small-holder discussed above might have oper-
had already decreased markedly from a much higher
ated, except that, now, these food sources also include
point during the pre-Satavahana period. The ques-
sewage drains and rubbish dumps. Perhaps because of
tion that must therefore be asked is how were pigs per-
these habits, pork has long been regarded as a low-
ceived socially in the pre-Satavahana and Satavahana
312   Derek Kennet

periods when they appear to have been a much more archaeological sequence that has been studied in this
important component of the agricultural economy? It way, making it impossible to know whether Paithan is
should certainly not be assumed that the low social typical of broader regional developments or whether it
status and economic value that have been noted above is, for some reason, a unique case.
can be applied to this earlier period. Indeed, it is quite This is also the reason why the history of agricul-
possible to imagine that, as the economic importance ture in Early Historic and Medieval India is, to date,
of pigs declined in favour of sheep and goat during based almost entirely on historical, epigraphic and
the Satavahana period and later, they became increas- art-historical sources (e.g. Gopal 1980; Randhawa
ingly associated with poor, low-status social groups 1980; Ludden 1999; Gopal and Srivastava 2008). Whilst
and it was this association that was ultimately the these sources provide many important insights into
cause of the low status and low value that was attrib- some agricultural practices, they are – by their nature
uted to them by the time the two texts mentioned above – incapable of providing a sustained and detailed nar-
were composed. rative of the development of agricultural strategies
across long periods of time. The reasons for this are
obvious and can be summarized under the following
Paithan’s agricultural economy in its broader context headings:
–– they generally lack the level of detail required to
From the discussion above, it can be seen that the gain a full insight into the agricultural strategy of
developments documented by the Paithan archaeobo- any one time or place, or indeed to allow compari-
tanical and paleofaunal sequences provide evidence sons between different times and places;
of a dynamic rural economy that evolved in all its –– they are, by their nature, unsystematic and in some
aspects, from cropping and crop-livestock interaction ways biased in the type of information that they
strategies to the processing of harvests. This dyna- record; for example they tend to concentrate on
mism gives the lie to the notion of a stagnant, unchang- the activities of literate groups rather than smaller
ing rural economy that has sometimes been projected rural farming communities;
onto India’s ancient countryside. Instead, we are given –– they are arbitrarily scattered across the histori-
a picture of an economy that seems to have continu- cal time-scale so that some periods are quite well
ally been under pressure to increase production. The covered, whilst others are hardly covered at all;
changes that were made in order to achieve this were –– there are uncertainties about the historiography,
not based on the introduction of new crops – for which date and reliability of many of the key texts.
there is no evidence at all – but on complex re-config-
urations of cropping strategies, crop-livestock interac- Nonetheless, it is worth considering how the develop-
tion strategies, group specializations and infrastruc- ments outlined at Paithan relate to the broader histori-
ture development. cal understanding of the development of Indian agri-
The insights that have been provided into Early His- culture during the Early Historic and Early Medieval
toric and Early Medieval agricultural strategies are pos- periods that have been gleaned from historical sources.
sible because the archaeobotanical and paleofaunal Whilst there are a number of studies of specific periods
assemblages were rigorously and systematically collected and places in the scholarly literature, there are rela-
through a carefully implemented sampling programme tively few attempts at a comprehensive history of Indian
– which included flotation for the archaeobotanical agriculture that spans the Early Historic and Medieval
material – and were quantitatively analysed in relation periods. In this respect, M. S. Randhawa’s A History of
to each other and to the site’s stratigraphic sequence. Agriculture in India (1980) and D. Ludden’s An Agrarian
Such an approach has hardly ever been applied to History of South Asia (1999) are well-known and widely
Early Historic and Medieval levels in India. The few cited studies which draw together the most significant
published reports from excavations of these periods research up to the time when they were each written
are based on small samples and consist mostly of basic (e.g. Ludden 1999: 231–248). Using these two works, let
presence-or-absence reporting of species. These allow us start by summarizing the consensus on the historical
the identification of newly introduced species, but they development of Indian agriculture through the period
do not allow any insight into changing agricultural with which we are concerned. The development is gen-
strategy (e.g. Kajale 1994: table 2.2; Fuller 2002: 249– erally portrayed as being based around a number of dis-
269). At the present time, there is therefore no other tinct historical phases. Initially, the so-called ‘Vedic’
 Chapter 14 Paithan in Its Regional, Historical and Cultural Context    313

and ‘later Vedic’ periods are seen as encompassing a Mauryan (e.g. Chakravarti 1996: 16–20), Mauryan
general shift away from pastoralism towards settled (e.g. Thapur 1984: 123–125), Satavahana (e.g. Ray 1986:
agriculture and a peasant economy, specifically in the chapter 4), Gupta/Vakataka (e.g. Sharma 1987: chapter
Ganges valley and northern India (Randhawa 1980: 10) and Medieval (e.g. Kulke 1995b: 252, 262). Is this
chapters 19–20; Ludden 1999: 61–3; see also Thapur a reliable picture? It might accurately reflect a long,
1984). The Magadhan and Mauryan periods are seen steady underlying process of population growth and
as a time during which agriculture was expanded land clearance that spread across the subcontinent,
and intensified under state authority and collection of beginning with the first farming communities and con-
surplus around core urban areas, whilst the transition tinuing, with possible scattered interruptions, until
from pastoralism to settled farming went on gradually relatively recent times – a process that is, of course,
elsewhere – ‘islands of farming in a sea of pastoralism’ only unevenly recorded in historical texts and inscrip-
as Ludden puts it (1999: 65; Randhawa 1980: chapters tions. However, it might not be accurate; it might
21–26; see also Thapur 1984). Little is said on the Sata- simply reflect the fact that new land clearances, set-
vahana period by either work (Randhawa 1980: chapter tlements and land endowments are much more likely
27; Ludden 1999: 64; but see Ray 1986: chapter 4). The to be recorded in texts and epigraphic sources than are
Gupta/Vakataka period is portrayed as a period of episodes of steady-state agriculture, declining produc-
‘conquest by farming’ involving the foundation and tion and the abandonment of agricultural land. Reality
endowment of temples, the settlement of Brahmins might therefore have been much more complex. This
and the clearing or enhancement of previously uncul- is an important question, but it is one that can only
tivated or under-cultivated areas along with the cul- be resolved through the accumulation of more detailed
tural and political integration of the groups inhabiting archaeobotanical and archaeological evidence against
them. It is generally argued that landholding became which to compare the historical record.
key to social status and political power during this
time and some have argued that agricultural yields
increased markedly (Ludden 1999: 64–65; Randhawa Developments between the pre-Satavahana and
1980: chapter 29; see also Digby 1982; Sharma 1987: Satavahana periods (Period 1 to Period 2)
chapter 10; Shrimali 1987). The Early Medieval period
is seen as a period when emerging localised kingdoms Focussing in on the Paithan region, Ray’s very useful
expanded cultivation as the foundation of their power, review of agriculture in the Satavahana Deccan
especially through the development of irrigation describes an ‘expanding rural economy’ that served
infrastructure, some of which is recorded by inscrip- as the agricultural base for Satavahana rule in the
tions (Randhawa 1980: chapters 31–34; Ludden 1999: Deccan (Ray 1986: 92–104). She acknowledges that
69–76; e.g. Chattopadhyaya 1973[1994]; Kulke 1995b: details of crops are lacking from the historical sources
240). and must be made up by the still very thin evidence
There are some obvious problems with this still very from excavation at sites such as Nevasa, Bhokardan
basic narrative, for example the date of some of the key and Ter, where the presence of crops such as wheat,
sources used, e.g. the Arthashastra for the Mauryan barley, rice, millets, sorghum and Ziziphus is recorded.
period (Willis 2009: 62), the general lack of detail Ray is, however, able to point to texts, inscriptions and
relating to crops and strategies and the fact that the numismatic evidence which testify to the existence of
nature of the sources differs considerably for each irrigation systems, such as water tanks, water wheels
period, as does the information they contain, which and wells. These facilities, she suggests, were privately
obviously makes longue-durée, comparative perspec- controlled even if their construction might have neces-
tives problematic. It is also notable that some periods sitated the patronage of wealthy or royal individuals.
hardly appear in these broad narratives – for example Attempting to filter out those of her conclusions that
the Satavahana period – although such voids probably are based solely on the Arthashastra (which is largely
reflect the relative lack of available historical sources post-Satavahana; see above), she also points to the
rather than the fact that no significant developments in reclaiming of waste land and to the numerous inscrip-
agriculture occurred. Another problem, and one that tions that indicate the private ownership and transfer
is emphasized from a wider review of the literature, of land that, on the basis of Jataka stories, she sug-
is that ‘agrarian expansion’ is a phenomenon that is gests might have been farmed either by the landowner
claimed for almost all periods in Indian history: pre- himself or by waged labour or slaves. Inscriptions also
314   Derek Kennet

record the donation of land and land revenues to mon- zons are likely to have been quite narrow and whose
asteries or to Brahmins. Initially, in the 1st century bc, scope for innovation more limited.
this appears to have been a royal prerogative, but by An important exception to the general lack of
the 1st century ad, increasing numbers of inscriptions archaeological evidence for agricultural practice at
record the donation of land by lay devotees, although this time is Shaw’s recent work around Sanchi and
the majority of these are in the region of Junnar. Vidisha, some 500 km to the north of Paithan but in
Some points from Ray’s analysis are clearly worth a roughly similar geographical setting to the central
emphasizing in relation to the Paithan evidence. The Godavari valley basin. Using a landscape-archaeology
first relates to the advent of private land ownership, approach integrating field survey with environmental
the second to the development of irrigation and a third sampling, Shaw has argued for a marked intensifica-
to the improvement of agricultural methods through tion of agricultural production in this area in the late
land grants to the sangha (Ray 1986: 93, 95, 101). In centuries bc, at the time when Buddhist monasteries
all of these cases, it is possible to see a context for began to expand beyond their original nucleus in the
the changes in agricultural practice that have been Ganges valley. She argues that the intensification of
identified at Paithan between Periods 1 and 2. Firstly, agricultural production was part of a package of cul-
the advent of private land ownership and the rise of tural and economic developments that accompanied
the gahapati as a group of land-owning cultivators the westward spread of Buddhism, urbanization and
(of various scales of wealth), which is suggested by the development of centralized state polities at this
inscriptions at sites such as Junnar, Kanheri and Nasik, time. Key to Shaw’s model are the numerous earthwork
might be linked to the transition towards the increase irrigation dams that, she argues, were constructed to
in household-based processing of crops and also to permit wet-rice cultivation. Aside from being part of
the spread of cultivation into marginal areas as indi- an eastern Indian cultural package imported along-
vidual small land owners sought to maximise the side Buddhism, wet-cultivated rice would have had the
exploitation of their land (see also Chakravarti 1996: advantage of yields up to 11 times greater than might
93). Certainly, the increase in weeds that Fuller has have been obtained from un-irrigated wheat. Shaw
identified appears to be more marked between Periods suggests that it was the resulting increased surplus
1 and 2 than between Periods 2 and 3, suggesting that that effectively underwrote the growth of population,
the transition to household-based processing was towns and Buddhist monasteries in the area (Shaw
most prevalent at this time. Secondly, Ray points to 2007: 233–262).
evidence for the development of irrigation that comes This is a compelling and well-argued model which,
from coins and inscriptions. Although she suggests whilst engaging with previously established models,
that this is likely to have taken place under the control provides a totally new perspective on the study of the
of the king or rich landowners due to the difficulty and Early Historic rural economy of this part of India. As
expense of creating irrigation systems, this nonethe- Shaw admits, however, there are some aspects that are
less provides a context for Fuller’s suggestion, based still open to debate, and one of these is certainly the
on the increased use of winter crops in Period 2, for the significance that is attached to rice. Although the argu-
presence of irrigation infrastructure given the relative ments for rice make perfect sense, Shaw was unable to
lack of rainfall in this part of India. Thirdly and finally, provide any direct evidence for its cultivation, relying
royal land grants to the sangha, which are recorded by instead on economic logic, possible cultural associa-
inscriptions, are suggested by Ray to have been part tions and the fact that the waterlogged environment
of a deliberate policy intended to improve agricultural that was identified through environmental sampling
yields and would presumably have led to the adoption accords closely with what would be expected in an area
of new agricultural techniques brought by the sangha of wet-rice cultivation.
from other parts of India. An example might be new Rice has certainly been identified in Early Historic
cropping strategies that were designed to maximise levels from at least ten excavated sites in India (see
yield and/or profit, which might be reflected in the Table 12.2) but at most of these it is unfortunately not
increase in the diversity of crops and the increased use possible to gain any clear sense of how abundant it was
of the more commercially viable African millets. Such compared to other staple crops. For example, Kajale
innovations were possibly outside the grasp of many of notes that rice was particularly abundant in Sata-
the smaller-scale subsistence cultivators whose hori- vahana levels at Adam, but detailed data are not pre-
sented, making it impossible to assess the reliability
 Chapter 14 Paithan in Its Regional, Historical and Cultural Context    315

of this evidence, which may be affected by taphonomy, have been linked to the expansion of double-cropping
sample size and the presence of wild rice (Kajale 1994: as it was to rice cultivation.
45–46).
In this respect, the Paithan evidence is potentially
very significant. Although Paithan and the upper Developments between the Satavahana and
Godavari are some distance to the south of Sanchi and Vakataka/Rashtrakuta periods (Period 2 to Period 3)
Vidisha, inscriptions make clear that they were closely
connected and were part of the spread of Buddhism From about the mid-4th century ad copper-plate
and Buddhist monasticism by the time the Sanchi dams inscriptions recording land endowments to Brahmins
were being constructed (Ray 1986: 53, 68). Indeed, being and temples began to be issued in significant numbers
an important Satavahana centre, Paithan is likely to across India from Tamil Nadu to Malwa (Willis 2009:
have been fully engaged in the major economic, cul- 81). Such endowments continued throughout the Early
tural and political developments of the time. The low Medieval period, during which time it is generally
relative frequency of rice in the Paithan archaeobot- thought that they were linked to very significant agrar-
anical assemblage indicates, however, that it was not ian expansion and intensification across many parts of
common enough to have had the economic impact that India. They led to virgin, forested land being cleared
Shaw proposes – at least in this area. Indeed, Fuller and brought under cultivation for the first time in areas
has concluded that rice at Paithan must have been a such as Orissa, Bengal and South India as well as parts
high-status or special-use crop alongside the staple of the Deccan and central India, whilst in other areas,
millets and pulses, rather than a staple crop itself. such as Madhya Pradesh, Gujarat and Maharashtra,
This is true for all periods at Paithan, the earliest of including, presumably, the upper Godavari region sur-
which are contemporary with Shaw’s proposed dates rounding Paithan, which had already been urbanized
for the construction of the Sanchi dams (Shaw 2007: and widely cultivated for several centuries, the endow-
240–241). Although the western Deccan traditionally ments encompassed land that was already under culti-
has one of the lowest levels of rice cultivation in India vation and included villages and farmhouses (Kosambi
(Spate and Learmonth 1967: 694), this evidence might 1956 [1975]: 301–302, 321; Sharma 1980: 29, 31–34, 222–3;
nonetheless suggest that it is necessary to reconsider 1987: chapter 10, appendix 1; 2001: 13, 31–32, 108–112,
the importance of rice in the model proposed by Shaw. 291; Shrimali 1987). In some of the latter cases, such
As Fuller has pointed out, the evidence for an donations consisted of an entire village but in others,
increasing reliance on winter pulses in Periods 2 and much smaller units were involved, for example a field,
3 at Paithan suggests an increasing dependence on a well or a disused orchard. It is in these areas that
double-cropping and artificial irrigation through Sharma believes land grants, rather than being linked
those periods. Of course, as was famously reported by to a geographic expansion of agriculture, led instead
Megasthenes, as early as the late 4th/early 3rd century to increased production through better techniques,
bc, double-cropping has long been a feature of South improved methods of animal husbandry, an increase
Asian agriculture (Strabo 1930: 15.I.20; McCrindle 1926: in the varieties of crops grown and the construction of
52–53; e.g. Kajale 1984: xi). However, the key question irrigation systems (Sharma 1987: 172–174; 2001: 108–
is not whether or not double-cropping was practiced 112).
but rather how significant it was to the agricultural Many of Sharma’s points seem to resonate closely
strategies of any particular time. Indeed, Shaw does with changes in the Paithan sequence noted between
consider the possibility that the Sanchi dams were Periods 2 and 3. As far as irrigation is concerned, one
intended to support double-cropping but rejects it on notable point is that the increase in winter pulses that
the basis of the somewhat circular argument that it Fuller has identified appears to be much more marked
would not fit in to the wet-rice cultivation cycle that between Periods 2 and 3 than it is between Periods 1
she proposes (Shaw 2007: 250). Taking into considera- and 2 (e.g. Figs 12.2 and 12.3), and it seems likely that
tion the chronological ambiguities in both the Paithan this reflects an increase in irrigation systems at this
and Sanchi/Vidisha phasing and the potential time lag time. Chattopadhyaya (1973 [1994]) has examined the
and environmental differences between the two areas, development of irrigation systems in Early Medieval
the Paithan evidence suggests that the increased use of Rajasthan and has considered the changes in cropping
irrigation that Shaw has identified might just as easily strategies as well as the implications for the social and
political organization of the communities where these
316   Derek Kennet

systems were introduced. Chattopadhyaya’s study is This is precisely the type of area Paithan and the
based on a different region of India and a slightly later surrounding upper Godavari must have been at this
period. Most significantly, it is based almost entirely time, as is made clear by the mention of Paithan in his-
on evidence from inscriptions, which can perhaps be torical and epigraphic sources and by the density of
expected to emphasize the predominant royal involve- surrounding Satavahana-period settlement (Ray 1986:
ment that he notes (1973 [1994]: 52). There are no such 25, 68–69, 72). In fact the core area of Vakataka land
inscriptions known from the Paithan region which sug- grants is over 200 km to the north-east of Paithan, and
gests that in some areas, the introduction of irrigation there is, at present, no evidence of endowments having
systems may also have taken place within different been made any closer during this period (Shrimali 1987:
social and political spheres. 48, map 2), although, of course, they are known from
The copper plates and the land endowments that later periods, as is recorded by the two Rashtrakuta-
they record have been interpreted in various ways by period copper-plate charters mentioned by Bhandare
historians. D. D. Kosambi and R. S. Sharma, taking a in Chapter 2: one in the late 8th century and one in the
Marxist perspective, have seen the change in land early 10th.
ownership that they record as part of the feudalization Building closely on the so-called ‘processural’ model
of Indian society and economy during the Early Medi- of Early Medieval state formation in which Brahmins are
eval period (Kosambi 1956 [1975]; Sharma 1980; 1987: seen, amongst other things, as conveyors of improved
chapter 10; 2001; contra Sircar 1974: chapter 2; see also agricultural methods (Kulke 1995b: 240), Willis has
Kulke 1995a: 6–18). Other scholars, such as H. Kulke recently explored the issues surrounding the estab-
(1978; 1995b), H. Bakker (1992) and M. Willis (2009), for lishment of Sanskritic worship and the endowment
example, see them as part of a complex socio-political of temples in the Gupta/Vakataka period (2009; see
process that, through the settlement of Brahmins, the also now Bronkhorst 2011). In this work, he makes a
Sanskritisation of ‘tribal’ groups and the foundation number of points and provides a number of examples
of temples, helped to establish, legitimise and expand that might help towards a better understanding of the
the authority of Early Medieval kings and kingdoms Paithan evidence. His analysis makes the same dis-
(Kulke 1995b). tinction between, on the one hand, the establishment
Many of the ideas on which this latter interpreta- of new estates in forested areas that had the potential
tion is based come from fieldwork originally conducted for agricultural development but which had been left
by Eschmann and her colleagues at Jagannath in undeveloped and, on the other hand, areas, such as
Orissa, which focussed particularly on the transfor- that around Vidisha, where intense agricultural devel-
mation and Hinduization of a forest, ‘tribal’ commu- opment had been in place since long before the Guptas
nity (Eschmann et al. 1978). Perhaps for this reason, and in which there was no need to create new estates
the model relates particularly to the transformations (Willis 2009: 159–161). In relation to areas in the latter
that took place in outlying, forested areas, rather than category, Willis explores a Gupta example from the
in the agrarian regions that were already cultivated early 6th century where administrative and priestly
and settled at that time (e.g. Kulke 1995b; Willis 2009: control was imposed on a pre-existing village in order
159). Indeed, as Bakker has pointed out, it seems that to support the construction and maintenance of a new
it was precisely in these outlying areas that most of temple (Willis 2009: 120). In this case, it is actually a
the estates donated by the Vakatakas were located consortium of merchants headed by a priest that came
(Bakker 1992: 91). Bakker also notes that the strategy to control the village and its land. Willis suggests that
and process would have been quite different in the core it is likely to have been the merchants who managed
agrarian areas, the areas he refers to as the ‘political, the endowed land and revenues, thereby instituting
religious and economic fundament’ of the Vakataka changes that would have affected all levels of village
elite (Bakker 1992: 88–90). In such areas, there were life. Such an arrangement is very likely to have led to
few forests to clear or ‘tribal’ groups to settle, instead an increase in agricultural production because people
– from an economic perspective at least – it would such as these, in bringing knowledge and experience
have been a question of improving agricultural yields from other regions, would have been in a position to
through the introduction of new methods and strate- introduce new agricultural techniques and strategies
gies, the enhancement of land that had already been (Willis 2009: 157–162). They may also have brought a
under cultivation for some centuries and the improve- focus on more market-oriented production.
ment of infrastructure.
 Chapter 14 Paithan in Its Regional, Historical and Cultural Context    317

Willis’ example therefore provides a historical It is far from certain that this model is correct,
scenario that potentially allows us to link the changes however. The problem is that many of the agricul-
in the Paithan archaeobotanical and palaeofaunal tural changes identified at Paithan appear to be long-
sequences with the foundation and construction of term and to have continued at a similar rate between
the two temples at the site. Even though the Paithan Periods 1 to 2 and Periods 2 to 3, for example the
temples are relatively small and there is no evidence increases in cotton, African millets and S/E/B millets.
of royal endowment associated with them, they serve This suggests the possibility that some changes were
to link the Paithan sequence with the period of royal driven by longer-term processes (for example popula-
endowments and the settlement of Brahmins that tion growth) and were largely unrelated to an influx of
occurred across India at this time. Even if Paithan was new ideas through land endowments. It suggests that
not directly involved, the innovations and ideas that they might in fact be a sort of Malthusian, longue durée
were transported as part of the movement of Brahmin- backdrop to historical developments rather than a con-
ical groups would soon have become common knowl- sequence of them. This is an important point because
edge and would have been taken up by a wide variety it brings up the question of the relationship between
land owners and cultivators. This scenario therefore the lives of small-scale gahapati and village cultivators
provides a tentative historical context for the changes and the broader historical and political structures of
in agricultural strategy that are visible in the Paithan the time. Unfortunately, it is impossible, at present, to
archaeobotanical sequence. disentangle this issue because land endowments are
At the same time, the Paithan evidence consider- known to have taken place through all three periods
ably enhances the Kulke/Willis scenario by provid- of the Paithan sequence (Bronkhorst 2011: 85–90; Ray
ing precise insights into the nature of the agricultural 1986: 100–104) so the effects of their commencement
changes that were occurring at this time. Here, we cannot necessarily be identified at any particular point
have an indication of how it would have been possi- in the sequence.
ble to have increased agricultural production in what This discussion also raises the problem of how
Kulke refers to as the ecologically favourable riverine changing practice is identified in an archaeologi-
landscapes that had long been under cultivation – this cal sequence such as that at Paithan. Although the
reflects the earliest stages of Kulke’s model, and the sequence is actually made up of a gradual accumu-
stages about which least is known (Kulke 1995b: 234– lation of thin layers, each of which contains a small
242). Elsewhere, and possibly as a later stage, in the amount of archaeobotanical and palaeofaunal mate-
peripheral areas outside the riverine heartlands, agri- rial, these have to be artificially amalgamated into
cultural production was increased by the very different ‘Periods’ in order to create assemblages that are large
method of clearing forest and bringing land under cul- enough to allow differences to be perceptible. This
tivation for the first time (e.g. Hinüber 2007: 192, note amalgamation means that we are able to look at the
38; Kulke 1995b). sequence only through the artificial prism of periodi-
sation that has been imposed on it. One of the draw-
backs is that it is impossible to distinguish between
Longer-term changes slow, steady change that took place gradually over a
long period of time and abrupt, sudden change that
It has been possible to show how Ray’s analysis of evi- took place as the result of a specific historical occur-
dence from Satavahana inscriptions and Willis’ analysis rence. It is also impossible to be certain exactly when
of endowments in the Gupta/Vakataka period might any particular change began to occur. For example, the
provide historical context for the developments in increase in cotton and weeds that Fuller has observed
agricultural practice that have emerged from the between Period 1 and Period 2 may have been the result
Paithan sequence. Even if an increase in agricultural of a gradual change in agricultural practices that
production was not the main intention of land endow- began in Period 1 (or perhaps even earlier), or it may
ments (e.g. Bronkhorst 2011: 91), they might nonethe- have been the result of a more abrupt change that was
less suggest a mechanism for the spread of new agri- implemented as the result of a specific event at some
cultural ideas across India. This is a useful model, time during Period 2. The fact that the increase in cotton
and it has the added attraction of locating the Paithan and weeds both continue between Periods 2 and 3 might
archaeobotanical and palaeo-faunal sequences within suggest that a gradual change in agricultural practices
the broader context of Indian history.
318   Derek Kennet

Fig. 14.4: A histogram of coins from Bhokardan, Maheshwar and Nevasa redated by S. Bandhare showing the numbers of coins deposited
per year (from Bhandare and Kennet forthcoming).

amongst smaller-scale landholders and cultivators is coins that can be expected to reflect actual patterns of
the more likely explanation. ancient coin deposition accurately and it is precisely
for this reason that they need to be more carefully con-
sidered. Of course, when coins retrieved from archaeo-
Coins and monetization logical excavations are used for quantified analysis, it
is necessary to take into consideration the collection
Having identified and discussed shifts in agricultural strategy that was used during the particular excava-
production at Paithan, we will now turn to look at tion – whether it was by hand or whether a sieve was
other aspects of the site’s economy through the same used and, if so, what the mesh size was. This latter
period. One key issue is certainly the numismatic evi- point is crucial, as it has been shown that some base-
dence. Much has been written on possible changes in metal Vakataka issues are quite small, less than 10 mm
the pattern of monetization of the Indian economy in diameter (Shastri 1992), and would regularly escape
between the Gupta/Vakataka and the Medieval periods notice unless excavated earth was consistently sieved
and there is still no clear consensus on the question through a finer mesh. It needs to be stated that whilst
of demonetization (e.g. Chattopadhyaya 1977; Sharma deposition rates calculated from excavated coin assem-
1980: 52–53; 2001: chapter 4; Deyell 1990: 3–8; Ali 2012: blages might provide some indication of the actual
9). Much of the debate has been based on museum circulation rates of low-denomination, base-metal
collections or coin hoards, whilst little emphasis has coinage, they are much less reliable when it comes to
generally been placed on coin assemblages retrieved higher-value coins of gold and silver, that tend to be
from archaeological excavations (e.g. Deyell 1990: 5). much less easily lost (e.g. Harle 1996: 16–17). Nonethe-
There is no doubt that museum collections and coin less, the low-denomination, base-metal aspect of the
hoards have their part to play in understanding pat- monetary system is crucially important, as is demon-
terns of monetization, but these need to be weighed strated by the extremely large numbers of such coins
against excavated evidence. Because of the unsystem- that circulated during certain periods.
atic way in which museum collections are built up and Taking these points into consideration, the present
the sporadic way in which coin hoards generally come author and S. Bandhare recently reviewed the coins
to light, excavated assemblages are the only source of excavated at three sites within the vicinity of Paithan:
 Chapter 14 Paithan in Its Regional, Historical and Cultural Context    319

Bhokardan, Maheshwar and Nevasa, where the report- Medieval India (e.g. Sircar 1974: 18; Shastri 1992: 291–
ing of the coin assemblage is good enough to allow 292). The above figures from the Paithan temple foun-
verification of identification. Unfortunately, no infor- dations are, however, the first to give a precise indica-
mation on artefact retrieval strategy was available for tion of the proportion of ancient coins that might still
these assemblages. Nonetheless, the published coins have been in circulation in the 4th to 7th centuries. The
were re-dated by S. Bhandare and tabulated, taking into figures are actually quite startling: some of the coins
consideration the lengths of the period of time during being used at this time were already as much as 600 or
which they circulated. The results are being prepared for 700 years old – a fact that seems almost incredible. Of
publication (Bhandare and Kennet forthcoming) and are course, some of these coins will have been archaeolog-
shown in Fig. 14.4. They demonstrate a notable increase ically ‘residual’; that is to say that they were deposited
in deposition in the second century bc and a very marked much earlier elsewhere and have simply been acci-
decline in coin deposition – and almost certainly there- dentally redeposited as the earth in which they were
fore also of circulation and minting – by the end of the 1st buried was moved into the temple foundation depos-
century ad at all three sites. its. At the same time, the evidence discussed in the
The pattern of coin deposition as revealed at Appendix to Chapter 9 strongly suggests that a good
Paithan is broadly similar, although the size of the number of these coins were deliberately deposited in
overall Early Historic/Medieval coin assemblage (35 the temple foundations and were therefore still in use.
coins) is too small to be statistically convincing. None- In relation to the debate on the decline of coin cir-
theless, the vast majority of the Paithan coins (32 or culation, or the demonetisation of the Indian economy
91%) are pre-Satavahana or Satavahana, with only in the 4th century ad or later, the Paithan evidence
three coins being post-Satavahana (9%). As stated in does contain some relevant information. Firstly, it is
the Introduction, at Paithan, all excavated earth was clear that the number of deposited coins declined very
sieved through a 5-mm mesh and numerous coins markedly after the 1st century ad in exactly the same
came to light in this way. way as it appears to do at Bhokardan, Maheshwar and
The point has already been made in the Appendix Nevasa based on the analysis shown in Fig. 14.4. Sec-
to Chapter 9 that the coins in the temple foundation ondly, large numbers of ancient coins continued to cir-
deposits (Period 3-temple) appear to have been delib- culate after this time. Thirdly, if we compare the cowrie
erately thrown or placed into the temple foundation shells retrieved during excavation (Chapter 13) to the
deposits whilst the temples were being constructed. coins (see Table 14.2), a clear pattern emerges. Cowrie
This is suggested by the higher ratio of coins to both shells do not occur at the site before Period 3, but in
excavated earth and to pottery sherds in these layers the excavated layers of that period, there are almost
compared to other layers at the site. The point has also equal numbers of cowrie shells and coins. By Period
been made that, whilst the foundation deposits of the 4, cowrie shells outnumber coins by almost 2:1 (24 to
South Temple can be dated to about the 7th century or 14 or 1.7 to 1). It is also notable that no cowrie shells
later on the basis of coin 34, the rest of the coins depos- were recovered from the Period 3-temple deposits,
ited in these layers were already very old, six (35%) despite the large number of coins that were retrieved
of them being Satavahana, three (18%) being pre- from those layers. This point might add credence to the
Satavahana and one (6%) being 4th century ad. It is idea of the coins being deliberately added to the foun-
well known that coins – especially Satavahana coins – dation deposits because of their perceived spiritual or
remained in circulation for long periods in ancient and

Table 14.2: Coins and cowrie shells from the Paithan sequence (summarized from Table 9.1).

Coin date Period 1 Period 2 Period 3 Period 3-temple Period 4 No period Total

Pre-Satavahana 8 2 1 3 14
Satavahana 1 3 6 5 3 18
Post-Satavahana 2 1 3
Total coins 8 5 5 17 14 7 56
Cowrie shells 4 24 28
320   Derek Kennet

superstitious value, a value that cowrie shells presum- tion rates on contemporary sites and therefore provides
ably did not share. a different and perhaps more robust and reliable per-
The use of cowrie shells as low-denomination cur- spective on ancient patterns of monetization and coin
rency is well attested by texts and hoards from the Gupta circulation than other sources of data.
period and into later Medieval times (e.g. Bowrey 1905:
199–200; Fa-Hsien 1923: 21; Sircar 1968: chapter 17; 1974:
18; Deyell 1990: 33–34, 62, 221, 237; Shastri 1992: 291, note Production and consumption
23). However, the Paithan sequence is the first quanti-
fied insight into the degree to which cowries might have Before concluding this section on the economy of the
circulated and how their deposition rates compare to site, it is worth making one further point. Manufac-
those of base-metal coins through the Early Historic ture and distribution were important aspects of the
and Early Medieval periods in this part of India. Early Historic economy as is demonstrated by the large
Whether or not cowrie shells and more ancient numbers of manufactured and traded items that are
base-metal coins were recognized as actual currency found on Early Historic sites. It has been suggested
in the 4th century ad and later (and the Paithan evi- by some scholars that these aspects of the economy
dence suggests very strongly that they were), it seems declined in the later Early Historic and Early Medieval
absolutely clear that after the 1st century ad in the period at a time when, it is argued, there was a disrup-
Paithan area, base-metal coins very largely ceased to tion to urban, mercantile and artisan life (e.g. Sharma
circulate – not entirely but certainly in the same quan- 1987: 153–155, 183; 2001: 285–289; Nath 2001: 22). This
tities that had been minted in the preceding two or is a point of considerable significance and is certainly
three centuries. This indicates that they also ceased to worthy of investigation. The difficulty, however, from
be minted in the same quantity. There are two implica- the archaeologist’s point of view, is finding ways to
tions of this: firstly, the political statement of minting investigate and measure it. It is not simply a case of
coins bearing a king’s name seems to have ceased to be demonstrating that manufactured items were or were
important. As Deyell puts it related to a later period, not present in any particular period. Rather it is neces-
‘coins cease to be used as a message-bearing medium’ sary to be able to make reliable quantified compari-
after about ad 500 (Deyell 1990: 5). Conversely, during sons between periods in order to demonstrate decrease
the Satavahana period, Bhandare has shown that or increase. Nonetheless, few archaeological excava-
coinage was ‘regio-specific’; that is to say it circulated tions record the sort of information that is required to
only within specific and closely defined areas within allow such comparisons to be made.
the Satavahana realm and rarely moved beyond these At Paithan, efforts were made to address this ques-
borders, suggesting that the political message that it tion through the analysis of the density of pottery
carried was understood and was seen as being signifi- sherds by cubic metre of excavated soil (see Chapter 7).
cant (Bhandare 1998: 49–63). Secondly, succeeding Although pottery is only one of a wide range of manu-
dynasties would have had very little control over mon- factured items that were used in the ancient and Medi-
etary policy within their realms. They would not have eval periods, it is by far the most abundant, the most
been able to control the purity of their coinage or in any visible and the least problematic archaeologically
way influence the amount of base-metal coinage in cir- because it has no re-use value and it is cheap and ubiq-
culation. Of course, this is also true of other times and uitous and survives well in buried deposits. For these
places; for example in Han China, minting was carried reasons, the amount of pottery that was deposited in
out privately outside the control of the state (Peng any given period might reasonably be taken as being
1994: 102), but it is nonetheless indicative of a lack of representative of the amount of pottery that was actu-
political control over the lower-denomination parts of ally in use, bearing in mind, of course, factors such as
the monetary economy. variable rates of breakage and repair, neither of which
It is worth stating here that the picture of coin cir- would appear to have been very significant in the
culation presented above does not necessarily fit well present case.
with analyses that are based largely on museum col- The figures given by the analysis mentioned above
lections or hoards. Whilst the reasons behind such suggest that the amount of pottery in use at Paithan
disparities certainly need to be investigated, it must declined by around 20% between Periods 1 and 2, 32%
be remembered that the Paithan, Bhokardan, Mahesh- between Periods 2 and 3 and 54% between Periods 3 and
war and Nevasa evidence is based on actual deposi-
 Chapter 14 Paithan in Its Regional, Historical and Cultural Context    321

4. This is a very marked rate of decline which appears to ufacture. The argument that has been made is based
have accelerated through time. on the general impressions of excavators rather than
Similar data are not available for other manufac- on systematic, quantified analysis. In one case, a pub-
tured materials simply because the numbers involved lished attempt has been made to analyse the quanti-
are too low to allow for statistically reliable compari- ties of manufactured items from Periods IV and V at
sons to be made. In the absence of better information, Nevasa (Gupta et al. 2001). At Nevasa, Period IV pre-
pottery will be taken here to serve as a ‘proxy’ for all dates the occurrence of Roman amphorae at the site,
manufactured items, although it must be remembered whilst Period V contains them. The analysis is based on
that other types of materials might, in fact, have had a simple count of objects made from stone, glass, shell
quite different histories. It is possible, for example, and ivory from the two periods and purports to show an
that the amount of metal utensils increased and that increase in Period V in all types of objects, supposedly
these were partly responsible for a reduction in pottery due to the stimulation provided by Roman trade (Gupta
use. Indeed, it is notable that, at 27, the number of iron et al. 2001: 15). The analysis, however, is fundamentally
objects from Period 3 is much higher than for Periods flawed because the quantity of earth excavated from
1 and 2, where only 14 and 13 were found respectively, each period has not been taken into consideration.
despite the fact that the rim-sherd assemblages from According to a rough calculation by the present author
Periods 1 and 2 were around 3.5 times bigger than based on the published sections of trenches A, F/H, E,
Period 3 (Tables 8.7 and 8.9). Although it is based on a and G in Sankalia et al. (1960: figs 10, 11, 27, 29), almost
small sample, this information might suggest that iron 20 times as much earth was excavated from Period V
containers replaced pottery to some degree, although levels as it was from Period IV. Although such a cal-
it is also notable that the majority of iron objects were culation can only be very approximate, it is accurate
nails and rivets rather than fragments of cauldrons or enough to show that the raw figures cannot be used as
pots. It is also necessary to remember that the trends they have been by Gupta et al. to argue for an increase
identified at Paithan are based on limited samples in manufacturing at Nevasa. Indeed, when corrected
and might simply be due to the specific nature of the by the factor given above, the figures actually indicate
deposits that have been excavated. It is possible that that the number of manufactured items declined in
trenches located elsewhere on the same site might have Period V after the beginning of Roman contact.
given different results. Interestingly enough, this revised conclusion agrees
Nonetheless, the evidence of a marked and contin- with the picture presented by the Paithan evidence,
ued decline in pottery deposition from the pre-Satava- which suggests that the high-point in manufacturing
hana period onwards is potentially highly significant. occurred in the latter centuries bc rather than in the
The most likely explanation is that there was simply first century ad and that it declined quite markedly
less pottery in circulation, meaning that less of it was thereafter. Given that Paithan is an inland site (one of
manufactured and used. the few) that is mentioned in the Periplus, this evi-
There is no general consensus on the patterns of dence puts another large dent in the already heavily
increasing and declining production of manufactured battered theory of Roman economic influence on Early
goods in the Early Historic period, or on the reasons for Historic Indian manufacturing. At the same time, the
them, but a number of scholars have suggested that Paithan evidence does support the argument for a
overseas trade contact from around the 1st century ad decline in production in the later Early Historic/Early
– particularly with the Mediterranean/Roman world Medieval period – but, it must be stressed, only as part
– provided an important external stimulus for many of a much longer trend. As the declining influence of
of the economic developments that took place at Roman trade can now almost certainly be discounted
that time, including manufacture (e.g. Ray 1986: 200; as a possible causal factor in this, it is not clear what
Sharma 1987: 135–138; Gupta et al. 2001: 14–15). It has might have been the cause. All that can be ventured
also been argued that the decline of these same con- here is that the decline appears to have been part of a
tacts around the 4th century ad had commensurately longer process and the explanation therefore needs to
negative effects, leading to a reversal in manufacture be sought at the same historical scale of longue durée
and other areas of the economy (e.g. Sharma 1987: processes. Before this can be done, however, further
135–138). data need to be collected from other archaeological
There is, however, to date no clear archaeological sites in order to corroborate the Paithan evidence.
evidence to demonstrate increasing or declining man-
322   Derek Kennet

Part 3: Conclusion is possible if the necessary evidence is collected and if

it is appropriately analysed.
During the excavations at Paithan, a concerted effort Table 14.3 summarizes the key developments at
was made to collect and analyse quantified evidence Paithan that have been reviewed in the current chapter.
in a way that would allow engagement with key histori- Taken together, these present a complex and sometimes
cal debates about the Early Historic and Early Medieval contradictory picture with different strands of evidence,
economies. The discussion above has shown that this in some cases suggesting developments in apparently

Table 14.3: A summary of the development of key aspects of the Paithan sequence.

Period Occupation Cultivation Animal Husbandry Temples Coins Pottery

Pre-1 Little known. Nothing known. Nothing known. A few punch-marked No clear
Probably a coins deposited. information
small settlement some- available.
where on the mound.
1 Expands to Well Even mixture of Significant number Highest rates of
cover most of established and cows, pigs and of coins pottery
mound. Wooden archi- mixed. sheep/ deposited. deposition in the
tecture. Predominantly goat. sequence.
millets,
especially Kodo
millet.
2 Continues to Expansion Decline in Slight decline 20% decline
cover most of and increased pigs, increase in deposition. in pottery
mound. Brick and diversity in sheep/ deposition.
tile architecture. of crops, goat possibly
evidence for related to
intensification increased meat
of production. production.
Increased
home
processing.
Increase
in ‘cash-
cropping’?
3 Contracts, Expansion Decline in Construction, Larger numbers Further
possibly to area and increased pigs, increase use and of coins 32% decline
around temples. diversity in sheep/ enlargement. deposited, in pottery
of crops, goat possibly especially deposition.
evidence for related to in temple
intensification increased meat foundations,
of production. production. but many are
Increased old issues.
home Cowrie shells
processing. appear.
Increase
in ‘cash-
cropping’?
4 Re-expands Little known Decline in Abandoned. Large numbers Further
to cover most at the present pigs, increase of coins and 54% decline
of mound and time. in sheep/ cowrie shells. in pottery
surrounding area. goat possibly deposition.
related to
increased meat
production.
 Chapter 14 Paithan in Its Regional, Historical and Cultural Context    323

different directions. For example, the decline in pottery the time, but for the most part, more evidence is still
deposition – which might be taken as indicating a more needed before more can be said on these questions. The
generalised decline in manufacture and consumption – Paithan evidence has taken us forwards by a number of
seems at first glance to be at odds with the archaeo- important steps, it has demonstrated the potential of
botanical and palaeofanual evidence, which suggests a more systematic, quantified approach to archaeo-
a marked movement towards increased production in logical evidence and it has also provided the first real
all aspects of the agricultural sphere. But there is of archaeological insight into the dynamic nature of the
course no reason to expect that the ancient economy agricultural economy of the Early Historic and Early
of Paithan would have been simplistic or monolithic. Medieval periods.
Indeed, the evidence presented here suggests a number A final point that must be made is that there is
of interlocking spheres, each of which demonstrates currently a major failing in Indian archaeology. The
its own distinct trends of development. archaeological methods that were applied during the
The fact that similar approaches to studying exca- excavations at Paithan and during the study of the
vated evidence have not been applied to other sites material have shown their value in addressing ques-
means that we do not yet know how typical (or oth- tions about the nature of the economy and how it
erwise) Paithan was of Early Historic and Early Medi- changed through time. Had other Indian sites been
eval towns in central India more generally. Clearly, analysed in the same way in recent years, there would
comparator evidence is needed from other sites so by now be a corpus of data against which Paithan could
that it is possible to know whether the developments be compared and such comparisons would provide
that occurred at Paithan were isolated and localised, further insights into regional patterns and trends. The
or whether they are representative of regional or methods used at Paithan are now standard amongst
even pan-Indian trends. The Paithan evidence does, archaeologists in almost all countries of the world,
however, make one thing absolutely clear: the ancient but they are still not applied systematically in India.
Indian rural economy was anything but ‘unchang- The reasons for this failure are unclear, but it brings
ing’. On the contrary, there is clear evidence for a a consequential lack of progress in dealing with some
complex, dynamic and constantly changing set of eco- of the key questions that surround the Early Historic
nomic spheres, each with its own trajectory. In some and Medieval periods of one of the greatest and most
cases, it has been possible to suggest ways in which significant countries in world history. It is only to be
these spheres might be linked to each other and to hoped that this failing is soon remedied.
the broader political and cultural developments of
Derek Kennet, Jason Hawkes and Michael Willis

Chapter 15 
Paithan Excavation: Historical, Archaeological,
Geographical and Epigraphical Contexts

The importance of the Paithan excavation goes beyond the principal datasets tend to be different, primarily
the site itself and touches issues central to the history of temple architecture, inscriptions and sculpture. There
India and the profound changes in politics, society and is little communication between history and archaeol-
economic relations during the early centuries AD. That ogy and the disciplines rank their priorities in different
the transition from the Early Historic period to the Medi- ways. Institutional venues also differ, with archaeology
eval was a crucial moment for India has long been noted led primarily by the Archaeological Survey of India and
(e.g. Kosambi 1956; Sharma 1965; Thapar 1968). Impacts state departments of archaeology, while history rests in
were felt not only in established urban centres such as university departments. This is further complicated by
Paithan, but in marginal zones that appear to have been the fact that where archaeology is studied and practiced
transformed by land grants and the foundation of Brah- in universities, it often tends to be within departments of
manical temples (e.g. Kulke 1995b; Willis 2009: 159, fol- history where it is deemed a sub-discipline.
lowing Bakker 1992). Yet while historians have done much We begin with the historical side. Over the last fifty
to develop and modernise the study of India’s transition years, historians have generated a variety of narratives
to the Medieval, archaeologists have made only modest embedded in an historiography that is particular to India.
contributions to the subject (a problem pointed out in, for The literature has been reviewed by Daud Ali (2012, 2014).
eexample, Singh 2011, Hawkes 2014b). Unfortunately, the These articles are essential reading and necessary points
ways archaeology has been practiced in South Asia have of departure for anyone seeking to understand how
not always been conducive to resolving the questions that Indian history—and the understanding of the Medieval in
historians have raised. Archaeology is essential, however, particular—has developed. As Ali (2014) noted, the study
because many changes are outside the concern of Indic of the Medieval “has been vexed by issues of chronologi-
literatures and epigraphic texts and thus not recorded in cal uncertainty, obscurantism, communal distortion and
them. The evidence from Paithan is accordingly crucial. It heavy model building.” The levels of obscurantism are
furnishes not only information about changes at the site usefully introduced by one example: Vishwa Mohan Jha’s
itself, but provides a starting point for new programmes (1996) review of B. D. Chattopadhyaya’s (1990) Aspects
of research that can tackle salient issues at the regional of Rural Settlements and Rural Society in Early Medieval
and pan-regional level. India. This review, chosen from an extensive discourse,
gives insight into the parochial nature of much of the his-
torical scholarship on the Medieval. Jha does not cite the

History and Archaeology book he is reviewing because the readers of the Indian His-
torical Review are expected to know, which of course they
do. But aside from this indicative silence, the patience
Looking back, there are a number of reasons for a lack of
with which Jha has read Chattopadhyaya’s complex argu-
exchange between history and archaeology in the study
ments, understood them and seen links across the script
of South Asia’s ancient and Medieval pasts. Primary
that normal readers would miss and Chattopadhyaya
among them has been inherent differences between the
does not bother to point out and, moreover, the rigorous
disciplines and how they are practiced in the Indian
attention to which Jha has subjected Chattopadhyaya’s
context. Historians tend to focus on political, social, reli-
interpretations and inserted his own carefully worded
gious and economic change, while archaeologists of the
and poignant observations, are as commendable as they
Early Historic period—up to now at least—have focused
are remarkable. Outsiders will be baffled. The perfor-
on tackling sites, establishing chronologies, developing
mance is readily explained, however, by the fact that B. D.
artefact typologies, identifying coins and documenting
Chattopadhyaya was the professor of history at Jawahar-
phases through stratigraphy. For the Medieval period,

https://doi.org/10.1515/9783110653540-015
 Chapter 15 Paithan Excavation: Historical, Archaeological, Geographical and Epigraphical Contexts    325

lal Nehru University until his retirement in 2004 and, fol- stood by historians of the colonial and immediate post-
lowing R. S. Sharma’s retirement from Delhi University in colonial period. During this time, historical scholarship
1985, the father of the ‘Delhi school’ and the leading voice in India borrowed much from the economic, social and
for historical studies in India. From his position of emi- anthropological theories that were emerging in Europe
nence and power, Chattopadhyaya (1998) meticulously and America. Ideas of feudalism and associated socio-
explored what he termed the ‘twin burdens’ of history: economic changes (after Bloch 1933, 1939; Pirenne 1936;
historiography and the written sources. He was not bur- Duby 1952) and social and cultural evolution (Childe 1951)
dened by fieldwork. Throughout his life, Chattopadhyaya all found their way to the writings of historians in India.
described himself to his students as a desk-based scholar Yet over time, as that scholarship became weighed down
with no pretensions to field study (Masahiko Mita, per- by arguments and counter-arguments, it also became
sonal communication, Nagoya, June 2019). more circumscribed. The degree to which the historiog-
This is understandable, up to a point. Chattopad- raphy of the Delhi school (and Calcutta at the formative
hyaya, like Sharma and other luminaries of historical stage) was inward looking and self-referential is shown
scholarship, were historians and not archaeologists. by the ways in which new methodologies were ignored or
Indeed, many were at pains to point this out—explicitly rejected summarily. Scholars like Ronald Inden, Burton
stating the value of archaeological fieldwork and articulat- Stein and Nicholas Dirks, influenced by powerful new
ing what archaeological findings might be useful in their trends in anthropology and sociology (which had contin-
research (see, for example, Sharma 1983, 1987). However, ued to evolve on the international stage), developed inno-
for reasons that we will turn to later, this archaeological vative forms of historical analysis and wrote pieces that
work was not forthcoming and in its absence no attempt are now regarded as classics. Inden (1985), inspired by A.
was made by historians to venture into the landscape to M. Hocart (1950), proposed that the state was an ‘impe-
see what materials might be found to support their inves- rial formation’, constituted as an evolving hierarchy of
tigations. The consequent limitations of this are appar- human and divine lordships (see also Inden 2006). Stein
ent in the slow pace of development in Chattopadhyaya’s (1980) famously drew on the study of acephalous socie-
ideas, although whether a dose of fieldwork would have ties in Africa to propose a ‘segmentary’ model of the Cola
clarified his thinking (and that of others) is moot. The lit- state, while Dirks (1988) created what he termed ‘ethno-
erature is dense and complicated, made more so by the history’ by mapping the changing roles of ‘big kings’ and
vigorous counter-arguments written by Chattopadhyaya’s ‘little kings’ in Tamil Nadu. Daud Ali (2014: 390) paused
academic opponents at each step. From our own side, to reflect on these contributions, saying: “Interestingly,
we would observe only that for every theory or proposi- this literature and its categories, whether as segmentary
tion Chattopadhyaya has put forward, a qualification or polity, ethnohistory or imperial formation, generally did
variant theory seems to appear somewhere in his exten- not articulate clearly with the trends and camps of Medi-
sive writing. To be fair, Chattopadhyaya did not hold a eval historiography well established in India and were
single view, but rather refined his interpretative position largely ignored or refuted, though their contributions
over a long career. With his book The Making of Early have arguably been just as formative for later develop-
Medieval India (Chattopadhyaya 1994) his arguments ments in the field.”
were articulated in mature form. There he accords Early Ali’s detached analysis does him much credit. The
Medieval society, from the time of the Gupta dynasty, inability of the Delhi school to take on ideas from the
with several key socio-economic and political features: outside deserves, however, a harsh rebuke. The problem
an increased clearing and settlement of uncultivated is highlighted by James Heitzman’s (1997) Gifts of Power:
lands (often through the deployment of Brahmins and Lordship in an Early Indian State, a work that involved a
land grants), the growth of networks of rural settlements, reappraisal of temple institutions in their economic, his-
the growth of new political lineages, the transformation torical and political landscapes. This is a book of such
of hinterland and forest societies into ‘state-societies’, innovation and creative scholarship that its lessons are
the peasantisation of cultivators and hunter-gatherers as being assimilated but slowly. Heitzman surpassed much
a part of this process and the concomitant incorporation in the historiography of loosely Marxian affiliation, but his
of non-caste peoples into the caste hierarchy. Ali (2014) cross-cultural comparisons were greeted with a facetious
may be consulted for a guide to the development of these dismissal by R. Champakalakshmi (1998). The impact of
ideas. Heitzman, Inden, Stein and Dirks on the Delhi school
Chattopadhyaya’s thinking emerged from a primary has been minimal, in part because their works were from
focus on the nature and constitution of the state as under- outside the school, but also because most of the publica-
326   Derek Kennet, Jason Hawkes and Michael Willis

tions deal with the Cola state in Tamil Nadu—a foreign concerned with sati, the self-immolation of wives on the
land and cultural anomaly as far as many inhabitants funeral pyres of their husbands. This is a perennially
of Delhi are concerned. The circumscribed geographi- debated topic, especially in the on-going contest between
cal vision shows also in the impact of Upinder Singh’s liberal and conservative thinkers. Spivak explored the
(1994) Kings, Brāhmaṇas and Temples in Orissa, an excel- race and power dynamics involved in the banning of
lent work informed by a careful analysis of inscriptions sati under British rule, noting that what we know of the
and an understanding of the theoretical positions then matter comes only from the accounts written by British
prevailing, especially evident in her cogent critique of and Hindu reformers. We never hear from sati-perform-
Sharma’s theories about Medieval feudalism. If David ing women, a problem that prompted Spivak to reflect on
Lowenthal (1985) is right in believing that The Past Is a whether the subaltern can speak. Looking at the docu-
Foreign Country then, for the Delhi school at least, Orissa mentation of sati in colonial archives, Spivak attempted
and Tamil Nadu are removed both spatially and tempo- to show that western scholars have assumed that their
rally—so doubly foreign. sources are objective and assumed also that the ‘oriental
The impact of anthropology and sociology on histori- other’ is anonymous and mute. However interesting this
cal writing has already been noted. More influential has seems, we will turn the pages of Spivak in vain looking
been the Annales school with its emphasis on long-term for a quotation from a sati pillar inscription, and this
social history and, in more recent manifestations, its despite the fact that these inscriptions survive in great
focus on the histoire des mentalités. Subaltern studies— numbers. To put the matter in plain language: subalterns
an examination of the colonial populations who were are certainly able to speak, but if we want to hear their
(and are) socially, politically and geographically outside voices we will have to listen, and listen in their language.
the hierarchies of power—has a genuine base in Indian More precisely, we will have to read what was written
society and an intellectual reality beyond the abstract on the spot, and look at how the sati was represented in
debates about the nature of the state. Ironically, however, life and death. This subject is not without relevance to
the approach was championed by academics in Europe the present book and its concern with the transition to
and America. The leading figures—Homi K. Bhabha, the Early Medieval. Although inscribed memorial stones
Gayatri Chakravorty Spivak and Ranajit Guha to name of this kind belong mainly to Medieval times, the oldest
the main players—work (or worked) at Harvard, Colum- sati in India is at Eran, on the ancient route to and from
bia, Sussex and the Australian National University. Nor the Deccan—and so also Paithan (Fig. 15.1). It is dated AD
is their style of analysis indigenous, showing as it does 510–11 (Fleet 1888: 91–93; Bhandarkar et al. 1981: 353–54).
an ultimate debt to Antonio Gramsci to whom we owe The text of the inscription has been known for more than
not only term ‘subaltern’ but a general theory of ‘cultural a century, but the pillar was only located and photo-
hegemony’ based (yet again) on Marxist principles (see, graphed recently by Peter Bisschop as part of the Beyond
for example, Gramsci 1995, 1996). Boundaries project. The pillar and its documentation are
Readers of this chapter may think us somewhat available online (Siddham OB00045).
dated in rehearsing the intellectual fashions set by gran-
dees of social and literary theory in the 1990s for a book
completed in 2020, especially since authors like Chat-
terjee (2012) have reflected on how subaltern studies has
given way to a more nuanced cultural history. There is,
however, remarkable deference to authority on the Indian
academic scene and a reticence to engage in robust cri-
tiques. We will not here explore the networks of patron-
age and influence that support the status quo and that
are, to varying degrees, characteristic of elite academia
everywhere. Rather, we will content ourselves with a few
words on approaches to the study of subaltern history as
exemplified by the works of Gayatri Chakravorty Spivak.
Her (1988) essay “Can The Subaltern Speak?” is regarded
by some as a seminal text in subaltern studies. The rami-
fications in the literature is extensive, but a start can be Fig. 15.1: Eran (Dist. Vidisha, Madhya Pradesh). Inscribed memorial
found in Maggio (2007). As is well known, Spivak was stone of Goparāja, dated AD 510–11.
 Chapter 15 Paithan Excavation: Historical, Archaeological, Geographical and Epigraphical Contexts    327

The relative lack of intersection between history, theory

and field archaeology as practiced in India—thrown into
high relief by the Eran pillar—characterises subaltern
studies and the degree to which those engaged in this
mode of historical research have not engaged with the
archaeology of social memory, even when sati itself is the
subject. In a landmark study, S. Settar and Günther-Dietz
Sontheimer (1982) focussed on the Deccan and worked in
a way that showed it is not for historians and ethnogra-
phers to decide if the practice documented by sati memo-
rials is good or bad, but rather to record that it happened
and to understand what these memorials meant and
mean to the people in whose places they are found. Cer-
tainly in our own travels in Central India we found that
the stones are often preserved and respected, the focus
of prayers and offerings (Fig. 15.2). They are taken as
signals that women long past were heroic and that they,
even now, stand as moral exemplars. These memorial
stones fill the countryside and survive in their thousands,
making the landscape a place populated by events and
people. Historians often lament the lack of sources in the
periods they study and rightly note that written texts do
not tells us much about everyday people. But such obser-
vations ignore this vast archive in stone. It is one of the
largest archives in the world and it is, in fact, an archive
without parallel anywhere. If we reflect on the scale of
Fig. 15.2: Salkanpur (Dist. Sehore, Madhya Pradesh). Hero-stone
the corpus, we soon realize that an entire dimension of memorial with a pinnacle in the shape of a temple spire, with a
Indian history and social life is absent from the ways relief sculpture of warriors in battle below, circa 12th century.
historians normally depict the past: it is as if individual
people—and the actions they deemed worthy of record—
it was based on an archaeology that was deemed “unas-
have been banished from the scene. Once we take notice
sailable” (Sharma 1972, 1987). Thanks to this assump-
of these memorials and the memories they embody, we
tion, the urban decline that took place at the end of Early
find that far from lacking local and subaltern histories,
Historic times was, in the words of Chattopadhyaya, “not
India is filled with them. The reservoir is so deep, wide
a matter for speculation” (Kennet 2013: 334, citing Chat-
and varied—and so densely filled with undigested data—
topadhyaya 1974, 1986). However, a close reading of the
that it presents an impossibly huge and daunting task. No
archaeological reports shows that the theory of urban
single person could cope with it, least of all a desk-bound
decline is to some degree an historiographical illusion,
scholar constrained by liberal ideologies that render the
generated by the excavation strategies of the archaeolo-
very subject repugnant.
gists and a selective reading of their publications on the
This excursus brings us back to the central problem
part of R. S. Sharma (Kennet 2013: 334; Hawkes 2014b).
raised in this chapter, namely a lack of connection
The critique is essentially Foucauldian in that the his-
between history and data collection in the field. What is
torical picture created by R. S. Sharma simply rehearses
particularly evident is that history is based on re-cycled
the perceived exclusions of the archaeological, numis-
data and derivative methods. As a result, the analyses
matic and epigraphic archives. So what is the upshot? It
produced are little more than re-configurations of exist-
is simply this: contra Chattopadhyaya, urban decline is
ing information and the historical propositions put
very much a matter for speculation. And while Chatto-
forward mere hypotheses. Kennet (2013), and again in
padhaya is correct to say that archaeology can resolve the
the introduction here, has shown that R. S. Sharma’s
questions he has raised, the way archaeology is practiced
influential theory of urban decline, proposed first in 1972
currently precludes any such a resolution. With this we
and a dominant theme in historical discussion since, has
no substantive basis but has enjoyed influence because
328   Derek Kennet, Jason Hawkes and Michael Willis

come to the discipline of archaeology, the second theme are going home’” (Dass and Willis 2002, citing Colling-
of this chapter. wood 1999: 65). Coming forward in time and back again
to India, we note that the issues have not gone away. For
example, Hawkes (2020) and Hawkes et al. (2020) have

Archaeological Methodology in tried to point to the evidence available, and come to some
evidence-led conclusions, without avoiding the fact that
India much is defective in the way the archaeology has been
done. The problem today is as much methodological as
Jürgen Neuß (2012) has provided a severe critique of it is conceptual. While it is important for archaeology
Indian archaeology and how the nature of its organisa- to be question-driven, the nature of those questions is
tion at the state and national level impedes significant also important. For the last seventy years or so, the vast
progress and innovation. The wider context of this and majority of archaeological activity and research in India
the way that the development of archaeology is also con- has been on pre- and proto-historic periods—as archaeol-
strained by its position as an historic discipline in uni- ogy has developed to fill the gaps where textual history
versities as noted by Chakrabarti (1988, 2003). Dass and cannot reach (Chakrabarti 1988). Yet as Hawkes (2014a,
Willis (2002) raised similar questions based on their pre- 2014b) makes clear, for historical periods such as the first
liminary work at Udayagiri. At that time they suggested— millennium AD, both the questions archaeologists have
naïvely as it turns out—that “future research might involve asked and the interpretations they have made have been
micro-documentation of the visible traces on the site and defined by traditional historical research.
a close study of these features in partnership with his- While we may hope that the publication of excava-
torians of traditional Indian astronomy. As theories and tion reports (as a bare minimum) may trigger improve-
possible lines of investigation developed, this could lead, ments, past experience suggests that they will have little
in the longer term, to some careful excavation work. The immediate impact. It is worth reflecting on why this
degree of interdisciplinary co-ordination and attention to is so. The matter turns, in our view, on the question of
minutiae required would necessitate a type of archaeol- deference noted before and, more importantly, on the
ogy that is not practiced in India today. As there are no way organisations (governmental, academic and non-
signs of such an archaeology developing in the present profit) perpetuate and defend themselves. As civil serv-
circumstances, we can only hope that this essay does not ants know, it is essential to be discreet and impartial,
trigger off some ill-informed escapade of random digging and not to sacrifice one’s career to a cause—be it politi-
similar to that seen at Satdhara and Sanchi over the last cal, intellectual or academic—especially if that cause
few years” (Dass and Willis 2002: 41). This same article has the potential to undermine one’s department and
further cited R. G. Collingwood (1889–1943) who, writing those within it. Criticism is mute. Perhaps “Can The Civil
in the 1930s, was one of the early critics of archaeological Servant Speak?” would have been a more relevant ques-
practice. He wrote: “Once digging has been decided on tion for Spivak to have posed. Certainly it is better not to
excavation becomes a sport for human terriers, tempered speak—and not to publish—and collect a pay cheque and
by the possibility that scientific results—who knows?— pension in the normal way. The alternative is to publish
may be forthcoming. Today, matters have reached a point and be damned. The fraught nature of publishing in the
when it can be said clearly and publicly that no archae- official setting and a reasonable fear of unintended con-
ologist ought even to be allowed to excavate at all unless, sequences naturally encourages deferral. Moreover, all
when he is deciding to work on a certain site, they can organisations develop systems of practice and will per-
answer the question ‘what archaeological or historical petuate those systems unless there are cogent reasons to
problems lead you to that site, why do you think you change. In Indian archaeology there have been numer-
can solve them there, and how exactly do you mean to ous campaigns of reform and re-organisation, the first
go about it?’; unless, every time he or she orders a new undertaken by Alexander Cunningham who established
trench to be opened or even a single shovelful of earth the Archaeological Survey of India and became its first
to be moved, he is prepared to explain, in terms of his- Director General in 1871. At the beginning of the twentieth
torical questions and their possible answers, exactly century vigorous reforms were led by George Curzon who
why he is doing it; unless his record shows that, instead was Viceroy from 1899 to 1905. He took an active inter-
of nibbling away at this site like a small boy with a cake est in cultural heritage and appointed John Marshall as
until nothing is left, he is capable of saying ‘now I have the Director General of the Archaeological Survey in 1902.
answered the question I came here to answer, and we From this point on, Indian nationals were allowed to par-
 Chapter 15 Paithan Excavation: Historical, Archaeological, Geographical and Epigraphical Contexts    329

ticipate in (and direct) excavations in their own country. variant of Wheeler’s deep-sounding system and the exca-
There followed a series of well-known excavations and vation system has remained largely unchanged. There is
restorations at major sites such as Charsadda (1902–3), also a strong resistance to deviation from this method,
Harappa (1920, 1925–34), Mohenjo-daro (1922–27), Sanchi due to the notion that preserving a single system ensures
(1912–19) and Taxila (1913–33). The reports of these exca- a common standard of practice. The level of ownership is
vations continue to book-end much of our archaeological indicated by R. S. Bisht, Deputy Director of the Archaeo-
knowledge (see Marshall 1924, 1931, 1951; Marshall et al. logical Survey at the time of the Paithan excavation, who
1939). Guha-Thakurta (2013) has reflected on the resto- was in regular touch with departmental staff on the site
ration of Sanchi and its publication, The Monuments of to “make sure they [i.e. Kennet et al.] follow our methods
Sanchi, that the books themselves are a sort of monument [i.e. those of Wheeler]”.
in three elephant folio volumes. What is often seen is an absence of a clear research
By the time The Monuments of Sanchi appeared, the framework for many of the excavations that take place,
need to modernise, and move away both academically and a failure to examine the extent of the site from the
and professionally from an Edwardian footing, was felt surface using simple techniques that are available and
clearly. This led to the appointment of Mortimer Wheeler regarded as standard practice in many countries. These
in 1944. In his day, Wheeler was at the forefront of archae- include controlled surface artefact collection, geophysical
ology and museum practice. Influenced by Pitt-Rivers, he examination, aerial photographic examination, detailed
saw that excavation required a more scientific approach mapping and the excavation of a network of test-pits.
central to which was the careful recording of stratigraphy. Once excavation has begun, the focus on Wheeler-style
He was highly critical of the large scale horizontal digs ‘box trenches’ makes it difficult for the archaeologist to
that revealed a single period of a site’s history (as had detect the more subtle large-scale evidence that is easily
characterized excavations carried out under Marshall). In obscured by the excavation baulks that are a part of the
response, he developed a system of excavation that has Wheeler box system. A tendency to phase whilst excavat-
come to be known as the “Wheeler method”, wherein an ing rather than at the post-excavation stage and to record
excavation grid (of any size) is divided into squares or objects by phase only, (making later reinterpretation of
‘box trenches’, which are then excavated so as to leave the phasing and chronology impossible) is still a problem
a baulk of earth between them to facilitate recording of at many sites. But probably the most serious issues are
stratigraphy (Wheeler 1954). This method was applied in the lack of systematic study of the artefacts using robust
India as an expedient means of quickly establishing the methodologies and detailed descriptions, a failure to
chronology of the many hundreds of sites that had been sample systematically for environmental evidence, and
discovered (Wheeler 1955; see further Ray 2008). In doing the almost complete lack of synthetic, regional, compara-
this, Wheeler not only transformed our understanding tive analysis of the data from excavations.
of the archaeological history of India, but also trained Archaeological sites that are obscured by modern set-
what later became the next generation of archaeologists tlement are generally excluded from excavation. Indeed,
in India. the concept of ‘urban archaeology’ which became such
As is well known, world archaeology has moved an important element of European and Scandinavian
ahead in scientific, technical and theoretical ways that archaeology from the 1950s onwards, is practically
would have been inconceivable seventy years ago. We unknown in India. There are two major problems with
now know that the Wheeler-method is not the best (or this. The first is that in only looking at sites that are
indeed the only) method of excavation, and ‘area exca- presently unoccupied we are unable to investigate the
vation’ has now come to be the norm, certainly in the movement of human settlement to locations that are still
west (e.g. Barker 1982). But it has also become clear how occupied today. The second, and real heart of the matter,
fundamentally important it is to take a holistic view of is that while uninhabited sites allow archaeologists to
an archaeological site, before beginning to dig a trench work in the ways they prefer, given the space and lack of
to establish the chronological sequence. The full extent overburden, an exclusive focus on such sites results an
of the site must be understood and the ‘horizontal stra- archaeological profile only of settlements that were aban-
tigraphy’ clearly grasped. If this is not done, the excava- doned at some time—and not those that have continued
tion is likely to provide an unbalanced view of the site’s to be inhabited, or even flourished. This is enforced by
development (Kennet 2013). Institutionally speaking, the archaeological reserve areas that have been set aside
however, the Archaeological Survey of India and the edu- by the Government of India to protect the archaeologi-
cation system that feeds it has perpetuated and retained a cal record from modern development. This is laudable,
330   Derek Kennet, Jason Hawkes and Michael Willis

Fig. 15.3: Maharashtra, region of ancient Vidarbha, showing the distribution of key copper-plate charters of the Vakataka period.

of course, but the reserve areas are often located on the way prescriptive. The remedies posed certainly have
peripheries of sites, in areas that had no modern habi- the potential to address the issues raised by historians
tation when they were set aside—thus compounding the and thus establish common cause between archaeolo-
problem set out above. The combination of Wheeler’s gists and historians on matters relating to the transition
methods with the system of reserves and a reluctance to from the Early Historic period to the Medieval. The rec-
excavate ancient settlements that are still occupied intro- ommendations are as follows: (1) a shift to the excava-
duces important biases into our data collection. We end tion of Medieval rather than only ancient sites, even if
up, in essence, with an archaeology of sites that have this involves excavation in modern urban centres; (2) a
failed over time, not those that endured and prospered. change of focus in the procedure of excavation and explo-
No wonder, then, that India’s greatness is so frequently ration with a more careful and holistic surface evaluation
portrayed as being in the distant past—the wonder that of any site before excavation commences; (3) the routine,
was India—in A. L. Basham’s immortal words (1954). If quantified study of archaeobotanical, zooarchaeological,
we are to believe this notion that India is a place that has ceramic and find sequences; and (4) the documentation
‘had its day’ it is because archaeologists and historians of rural settlement patterns.
have made it so. The present volume sets an example on the first three
points and is, as a result, a milestone marked by innova-
tion in practice, the collection of new empirical data and

New Directions the articulation of interpretative theories to explain that

data. At the same time, the authors of this chapter have
recognised that there is an urgent need to look at wider
The possible solutions to this problem suggested by
settlement patterns around key sites. This strategy allows
Kennet (Chapter 14 this volume) are modest and in no
 Chapter 15 Paithan Excavation: Historical, Archaeological, Geographical and Epigraphical Contexts    331

the exploration of changes that were taking place on a the supposed agrarian innovations that took place during
regional scale, traces of which can be observed at specific the early centuries AD, and the apparent settlement shift
locations like Paithan. from Adam to Mandhal during the mid-first millennium.
To situate Paithan in its regional setting, Hawkes, The results of these surveys and key-hole excavations
Lefranq, Abbas and Willis studied land-granting activi- are being published separately. Here we will restrict our-
ties as documented by copper-plate inscriptions. This selves to some comments on settlement patterns and the
line of analysis was based on the long-established his- distribution of copper-plate charters, the third and final
torical proposition that the appearance of land records theme of this chapter.
marked innovations in land use, agrarian practice, prop-
erty relations and social formation in many parts of India.
The find spots of the copper-plate charters were traced,
from Uttar Pradesh to Karnataka, to test what might be
Vidarbha
concluded from their spatial distribution (Hawkes and
Mapping the find spots of the Vidarbha copper-plates
Abbas 2016). The focus was on the earliest charters—
has shown that of the twenty-nine charters from the
from the fourth, fifth and sixth centuries—and the stone
region, 15 (52%) were found during digging or ploughing
inscriptions of the same period. The number of surviving
in locations that may relate to their original contexts, at
charters is substantial. These data have been compiled
least until evidence emerges to the contrary (for discus-
and published online (see Siddham as well as the aggre-
sion, see Hawkes and Abbas 2016). Others were found
gated data available at www.siddham.network). Having
in the possession of people and museums, having been
plotted the distribution of the copper plates, research
dug up long ago—notably for the present purpose the
then concentrated on an area of eastern Maharashtra—a
copper plates in Pune (Siddham OB00143), and in Patna
region known in ancient times as Vidarbha. Not only does
(Siddham OB00159). It is for this reason that some of the
this area neighbor Paithan (thus facilitating the investi-
Vakataka charters do not appear in the map (Fig. 15.3).
gation of its regional context), it also yielded a number
There are also some plates that are outliers, i.e., they were
of copper plates issued by one one dynasty—the Vaka-
found geographically outside the main corpus and pos-
takas—making it possible to test the effects that these
sibly not in locations that are historically reliable. The
grants had. The results of the findings are shown here in
several charters found in the Satpura hills to the north
Fig. 15.3.
(beyond the area covered by our map) are good exam-
Investigations in Vidarbha made a conscious attempt
ples, with the so-called Siwani copper-plate charter a
to implement the recommendations noted above, but
prime case (Siddham OB00150). This was found in or
adapted them to suit local circumstances. There were three
shortly before 1836 in the possession of a zamindar in the
main aspects to this: (1) in terms of focussing on Medieval
then Seoni District. Before 1888, the plates were still in
sites rather than ancient ones, field surveys showed that
the possession of a zamindar of Pindarai village (Siwani
it was difficult to distinguish reliably between Medieval
Tahsil, Madhya Pradesh). This location is 95 km north of
and ancient sites and that most sites had multi-period
Balaghat, itself the source of a copper plate of uncertain
occupation; consistent abandonment of earlier settle-
provenance. The original find spot of the Siwani plates
ments or movement to new ones seem to have occurred
is not known but, like the other inscriptions recovered
only later in the early second millennium AD; (2) in terms
in the Satpuras, they seem to have been carried into the
of changing procedure, work focused on the systematic
hills and, as a consequence, preserved there. This distri-
reconnaissance and recording of archaeological sites
bution of evidence is not unique to Vidarbha or the Vaka-
in areas selected on the basis of the presence of known
takas. Cultural artefacts deemed valuable or talismanic
epigraphic and monumental remains, coupled with the
are often preserved outside core areas, especially if those
collection of surface material so as to characterize those
areas have experienced a sequence of upheavals and
sites and establish patterns in the regional distribution of
adjacent regions have provided a safe haven.
archaeological material; (3) in order to supplement and
Looking, then, at the distribution of those copper
improve on this regional picture, a programme of limited
plates that were found in or near their original contexts, a
test-pitting was implemented at Adam and Mandhal, sites
picture beings to emerge. Once we discount the ‘noise’ in
known from excavation and for their inscriptions (Nath
the data produced by the highland evidence, we see that
2016; Sali 1998). This was carried out to generate data nec-
all the Vakataka grants are from low-lying geographical
essary to improve pottery typologies, as well as to collect
zones, i.e. at elevations below 500 m. The Masod charter
environmental samples as a step towards understanding
332   Derek Kennet, Jason Hawkes and Michael Willis

Fig. 15.4: Maharashtra, region of ancient Vidarbha, showing the distribution of relevant archaeological sites and coin finds.

seems to be an exception (Fig. 15.3, Siddham OB00171), over the last one-hundred and thirty years (Hawkes and
but this anomaly is explained by the fact that this grant Casile, forthcoming). As would be expected, the consist-
is not an original Vakataka inscription but a forgery ency, dependability and detail of these data varied. For
of Medieval times. With this example set aside, we can instance, the dating of many remains and sites is uncer-
see that the Vakataka charters—and the administrative, tain, with only three sites across the area having been
economic and social practices they reflect—pertain to subject to radiocarbon dating; in addition, there is no
lowland areas, almost all of which were broad, alluvial regional pottery typology and thus little consensus as to
riverine plains. the date of the remains found at any one site (Lefrancq
When we then relate the charters with the distribution et al. 2019; Lefrancq and Hawkes, forthcoming). Addi-
of settlements, further patterns of importance emerge. As tionally, as we know from the accounts, many surveys
noted above, the results of new surveys and excavations and campaigns of exploration were not systematic. Sites
are currently in the process of being finalised for publi- were found through local informants or simply turned
cation. Still, it is still possible to identify certain broad- up along the paths walked by archaeologists, while large
scale patterns in the distribution of known archaeologi- areas of the region have not been surveyed at all. As we
cal remains when they are mapped in space and time (see might expect, areas that are difficult to reach have been
Fig. 15.4). The data in this map were gathered through a surveyed far less intensively than those close to Nagpur
comprehensive survey of the existing archaeological lit- and other towns.
erature (published and unpublished). This resulted in Notwithstanding its obvious idiosyncrasies, the
the compilation of a census of all archaeological sites dataset as a whole is not lacking historical weight and
and remains that have been documented in the region substantive implications. The first of these, and most
 Chapter 15 Paithan Excavation: Historical, Archaeological, Geographical and Epigraphical Contexts    333

important of all, is that the existing settlement data

shows that the entire region was well-settled from at
least the ‘Megalithic’ or early Iron Age (from circa 800
BC in this region). Further, when we look at the distribu-
tion of these sites, we can see that settlements tended
to cluster around water bodies—small tanks developed
and managed by local residents—and along rivers and
streams. In addition, most settlements—like the land
grant inscriptions—are located at elevations below 500
m. The main historical implication of the census data for
the present chapter is this: if we compare land grants and
settlements, we can see that the grants have a settlement
context indicative of the agrarian and geographical envi-
ronment in which the grants operated. It is perhaps better
to rephrase this observation in terms of human agency: if
Fig. 15.5: Mahurjhari (Dist. Nagpur, Maharashtra). Raw material for
settlements are proxies for the people and communities carnelian from ancient quarry.
living on the land and managing its environmental and
agricultural resources, and land grant inscriptions are
proxies for the presence of elites receiving land from the
king, then the find spots of the grants juxtaposed to set-
tlements shows the social and economic matrix in which
beneficiaries of the grants lived and derived their liveli-
hoods. The network of settlements—and their economic
surpluses—also gave rise to monumental remains, such
as temples and monasteries, and the inscriptions that
sometimes survive to document the construction and
endowment of these institutions.

Agriculture and property relations

The pairing of epigraphic find spots and settlement dis- Fig. 15.6: Mahurjhari (Dist. Nagpur, Maharashtra). Carnelian bead
tribution has significant implications for the historical fragments in fields at village.
theories put forward about the Early Medieval period. Far
from documenting the increased clearing and settlement
of uncultivated lands through the deployment of Brahmins
and land grants, and the growth of networks of rural settle-
ments—as Chattopadhaya would have us believe—matters
are exactly the opposite. The charters document gifts of
land to Brahmins in well developed zones where settle-
ments had been established for at least 1000 years. In one
case, at Mahurjhari, the plates were found at a site where
there was a well-developed industry in carnelian-bead pro-
duction (Mohanty 1999, 2005; Mohanty et al. 2019) (Figs
Fig. 15.7: Mahurjhari (Dist. Nagpur, Maharashtra) (?). Engraved
15.5 and 15.6; for the plates see Siddham OB00178). Some carnelian with a Buddhist inscription in Vakataka-style script
of these stones were engraved and their inscriptions show reading apramāda with corresponding impression. Probably made
they were produced for Buddhist clients and had a wide at Mahurjhari. British Museum, 1892, 1103. 126, see SIDDHAM
distribution (Fig. 15.7). These findings indicate that wealth IN01102.
at Mahurjhari came from a vibrant trade in semi-precious
stones as well as agricultural activity.
334   Derek Kennet, Jason Hawkes and Michael Willis

Fig. 15.8: Chammak (Maharashtra). Copper-plate charter of Pravarasena II with seal of the king and a paper note by J. F. Fleet. British Library
Ind Ch 16, see SIDDHAM OB00149.

Lest we be accused of selecting examples that suit our sena—would necessitate there being adequate resources
interpretation, we can take the Chammak plates as a final to support the people in question. The developed nature
example (Fig. 15.8, Siddham OB00149). These plates were of the wider area is shown by the stupa and associated
found in the ground in the nineteenth century and the settlements at Bhon, slightly to the south on the Purna
inscription on them registers the donation of the village river (Deotare et al. 2007; Naik and Deotare 2016). The
of Carmanka by Pravarasena II at the request of ruler Chammak charter does not represent, therefore, some
named Kondaraja (Mirashi 1963). The village of Chammak sort of pioneering effort in a remote forest area. It is the
perpetuates the ancient name. The donation was made to settlement of a community of learned individuals, their
a large number Brahmins of various Vedic schools and families and attendants in what was already a prosper-
the charter lists 49 of these donees by name. An explora- ous zone. The fact that this site was discovered through
tion of the area under the project led to the discovery of a simple reconnaissance, and that more than a century
large settlement immediately south of Chammak village. stands between the publication of the Sanskrit text on the
This dates from at least the Iron Age to the late Medie- plates in 1888 and the publication of the archaeological
val period (Hawkes et al. 2017, 2020). There can be little context in 2016, underlines the need for the integration of
doubt that this location was the seat of the king named fieldwork and historical studies.
Kondaraja in the inscription. Now the settlement of a
large number of Brahmins in this location at the behest
of a local grandee—and its sanction by king Pravara-
 Chapter 15 Paithan Excavation: Historical, Archaeological, Geographical and Epigraphical Contexts    335

Summary
Returning to the excavations reported in this volume,
we have, it is hoped, demonstrated that some of the
observations made at the regional level correspond with
interpretations that emerge from the site of Paithan. This
provides a methodological basis for placing sites in a
broader context as well as providing insight into the con-
temporary development of two different but key parts of
the region: the Godavari and Vidarbha. As Kennet makes
clear in Chapter 14, the Paithan excavations document
the range and depth of agrarian reform that took place
in a long-settled core area during the Early Medieval.
Kennet suggests these changes may have been stimulated
by the settlement of Brahmins from outside the region,
and by the building of temples whose endowment lands
probably came under new management regimes. These
changes reflect the developments evidenced by the
two land charters discussed above—Mahurjhari and
Chammak—and demonstrate the value of linking land-
scape surveys with in-depth excavations at selected loca-
tions. Moreover, this chapter has shown that a coupling
of landscape studies and excavations can indeed shed
light on the the transition from the Early Historic to the
Medieval, providing a level of insight into processes of
change and their regional variations that are not avail-
able from the epigraphic sources alone.
Reference list
Agrawal, A. 1999. Greener pastures: Politics, markets and Bhandare, S. U. 2006. Political transition in early 5th century
community among a migrant pastoral people. Durham and Gujarat: A numismatic reappraisal based on the silver issues
London. of the Western Ksatrapas, the Guptas and their contem-
Agrawal, V. S. 1942. The geographical contents of the Mahamayuri poraries. Numismatic digest 29–30: 70–108.
text. Journal of the United Provinces Historical Society 15/2: Bhandare, S. and Kennet, D. Forthcoming. A reconsideration of
19–35. Satavahana and later coins from archaeological contexts.
Ali, D. 2002. Vidīśā and its environs in Early Medieval India. Paper Bhandarkar, D. R. 1925. Asoka. Calcutta.
circulated to the Vidisha Research Group. Zenodo http://doi. Bhandarkar, D. R., Chhabra, B. C. and Gai, G. S. 1981. Inscriptions
org/10.5281/zenodo.3623645 of the early Gupta kings. New Delhi.
Ali, D. 2012. The historiography of the Medieval in South Asia. Bhandarkar, R. G. 1895. Early History of the Dekkan. Bombay.
Journal of the Royal Asiatic Society 22(1): 7–12. Binford, L. R. 1981. Bones: Ancient men and modern myths. New
Ali, D. 2014. The idea of the Medieval in the writing of South Asian York.
history: contexts, methods and politics. Social history 39(3): Binford, L. R. 1984. Faunal remains from Klasies river mouth. New
382–407. York.
Allchin, F. R. (ed.) 1995. The Archaeology of Early Historic South Blaikie, P. and Brookfield, H. C. 1987. Defining and debating
Asia: the emergence of cities and states. Cambridge. the problem. In P. Blaikie and C. Brookfield (eds). Land
Ambasta, S. P., Ramachandran, K., Kashyapa, K. and Chand, R. degradation and society. London: 1–26.
(eds). 1986. The Useful plants of India. New Delhi. Bloch, M. 1933. Le problème de l'or au Moyen Age. Annales
Ansari, Z. D. and Dhavalikar, M. K. 1975. Excavations at Kayatha. d'histoire économique et sociale 5: 1–34.
Pune. Bloch, M. 1939. La société féodale 1: la formation des liens de
AnguttaraNikāya = Morris, R. (ed.) 1885–1900. The Anguttara- dépendance. Paris.
nikâya. Pali Text Society/Henry Frowde. 6 vols. London. Boardman, S. 1999. The agricultural foundation of the Aksumite
Arthashastra. Kautilya, the Arthashastra. L. N. Rangarajan (ed., Empire, Ethiopia: An interim report. In M. van der Veem (ed.).
trans. notes). London: Penguin Classics, 1987 [1992]. The Exploitation of plant resources in ancient Africa. London:
Āvaśyakaniryukti, part II (reprint). 1981. Mumbai: Śrī Bherulāl 137–148.
Kanaiyālāl Kothārī Dhārmik Trast, sam 2038. Boessneck, J. 1969. Osteological differences between sheep (Ovis
Bakker, H. 1992. Throne and temple: political power and religious aries Linné) and goats (Capra hircus Linné). In D. Brothwell
prestige in Vidarbha. In H. Bakker (ed.). The Sacred centre as and E. S. Higgs (eds). Science in archaeology. London:
the focus of political interest. Proceedings of a symposium 331–358.
held on the occasion of the 375th anniversary of the University Boland, R., Bedaux, R. and Boser-Sarivaxevanis, R. 1991. Tellem
of Groningen, 5–8 March, 1989. Gronigen: 83–100. textiles: Archaeological finds from burial caves in Mali’s
Bakker, H. 1997. The VākāYakas: an essay in Hindu iconology. Bandiagara. Amsterdam.
Groningen. Bollée, W. B. 1994. Materials for an edition and study of the Pimda-
Barker, P. 1982. Techniques of archaeological excavation. London, and Oha-Nijjuttis of the Śvetāmbara Jaina tradition. Volume II:
2nd edn. Text and glossary. Stuttgart.
Basham, A. L. 1954. The wonder that was India. London. Bopearachchi, O. and Pieper, W. 1998. Ancient Indian coins. Indico-
Bedigian, D. 1998. Early history of sesame cultivation in the Near pleustoi: Archaeologies of the Indian Ocean 2. Turnhout.
East and beyond. In A. B. Damania, J. Valkoun, G. Willcox Bowrey, T. 1905. A Geographical account of countries round the Bay
and C. O. Qualset (eds), The Origins of agriculture and crop of Bengal 1669–1679. R. C. Temple (ed.). London.
domestication. Proceedings of the Harlan Symposium. Aleppo: BrahmaPurāṇa = P. Schreiner and R. Söhnen. Sanskrit indices and
93–101. text of the Brahmapurāna. Purana research publications,
Bedigian, D. 2003. Evolution of sesame revisited: Domestication, Tübingen vol. 1. Wiesbaden, 1987.
diversity and prospects. Genetic resources and crop evolution Brain, C. K. 1981. The Hunters or the hunted? Chicago.
50: 779–787. Brhatkalpabhāsya = Brhatkalpasutra, with the bhāsya of
Bedigian, D. and Harlan, J. R. 1986. Evidence for cultivation of Sanghadāasaganin. Muni Caturave-jaya and Muni Punyavijaya
sesame in the ancient world. Economic botany 40: 137–154. (eds). Bhāvnagar: Jain Atmānand Sabhā 2002, 2nd edn.
.
Bender-Jørgensen, L. and Vogelsang-Eastwood, M. 1991. The Abu Bronkhorst, J. 2011. Buddhism in the shadow of Brahminism.
Sha’ar textiles 1991. Archaeological textiles newsletter 13: Handbook of oriental studies section 2, South Asia, vol. 24.
3–4. Leiden.
Bergman, I. 1975. Late Nubian textiles: Scandinavian joint Brookfield, H. C. 1984. Intensification revisited. Pacific viewpoint
expedition to Nubia, vol. 8. Copenhagen, Oslo, Stockholm. 25(1): 15–44.
Betts, A., van der Borg, K., de Jong, A., McClintock, C. and van Burden, D. 2005. ‘Sesame profile’ in agricultural marketing
Stryndonck, M. 1994. Early cotton in North Arabia. Journal of resource centre: A national information resource for
archaeological science 21: 489–499. value-added agriculture. Website, 11 November 2007: http://
Bhandare, S. U. 1998. Historical analysis of the Satavahana era: www.agmrc.org/agmrc/commodity/grainsoilseeds/sesame/
A study of coins. Unpublished PhD, University of Mumbai, sesameprofile.htm
Ananthacharya Indological Research Institute.

https://doi.org/10.1515/9783110653540-016
 Reference list   337

Casson, L. 1989. Periplus Maris Erythraei: Text with translation, Chowdhury, K. A. and Buth, G. M. 1971. Cotton seeds from the
introduction and commentary. Princeton. neolithic in Egyptian Nubia and the origin of old world cotton.
Casson, L. 1990. New light on maritime loans: P. Vindob. G. 40822. Biological journal of the Linnean Society 111: 303–312.
Zeitschrift für papyrologie und Eepigraphik 84: 195–206. Chowdhury, K. A. and Buth, G. M. 2005. Plant remains from
Casteel, R. W. 1977. Characterization of faunal assemblages and excavation of terraces of the Nile at Afyeh, Nubia and Egypt.
the minimum number of individuals determined from paired Purattatva 35: 154– 159.
elements: Continuing problems in archaeology. Journal of Chumble, P. D. 1991. Rare Satavahana coins. Numismatic digest 15:
archaeological science 4: 125–134. 41–50.
Chakrabarti, D. K. 1988. A history of Indian archaeology from the Clapham, A. and Rowley-Conwy, P. 2007. New discoveries at Qasr
beginning to 1947. New Delhi. Ibrim, Lower Nubia. In R. Cappers (ed.). Fields of change:
Chakrabarti, D. K. 1997. The Archaeology of ancient Indian cities. Proceedings of the 4th international workshop for African
Delhi. archaeobotany. Groningen: 157–164.
Chakrabarti, D. K. 2003. Archaeology in the Third World: a history Clapham, A. J. and Rowley-Conwy, P. 2009. The archaeobotany of
of Indian archaeology since 1947. New Delhi. cotton (Gossypium sp. L.) in Egypt and Nubia with special
Chakravarti, W. 1996. The Social dimensions of early Buddhism. reference to Qasr Ibrim, Egyptian Nubia. In A. S. Fairbairn and
New Delhi. E. Weiss (eds). From foragers to farmers: Papers in honour of
Champakalakshmi, R. 1998. Review of James Heitzman, Gifts Gordon C. Hillman. Oxford: 244–253.
of power: lordship in an early Indian state, Delhi, Oxford Colledge, S. 1998. Identifying pre-domestication cultivation using
University Press, 1997. Studies in history 14: 147–152. multivariate analysis. In A. B. Damania, J. Valkoun, G. Willcox
Chanchala, S. 1992. The fruit and seed remains from ancient and C. O. Qualset (eds). The Origins of agriculture and crop
Hulaskhera, district Lucknow, U.P. (c. 700 BC–500 AD). domestication. ICARDA, Aleppo: 121–131.
Pragdhara: Journal of the Uttar Pradesh State Archaeological Colledge, S. 2001. Plant exploitation on Epipalaeolithic and early
Organisation 2: 65–80. Neolithic sites in the Levant. British Archaeological Reports
Chanchala, S. 1994. Harappan plant economy of Kutch, Gujarat. International Series 986. Oxford.
Geophytology 23(2): 227–233. Collingwood, R. G. 1999. The principles of history and other
Chanchala, S. 1995. Some seed and fruit remains from Kausambi, writings in philosophy of history. W. H. Dray and W. J. van der
district Allahabad, U. P. (ca 600 BC–450 BC). Geophytology 24: Dussen (ed., notes). Oxford.
169– 172. Cooke, M. 2002. Investigating changing agricultural production
Chanchala, S. 2000–2001. Seed and fruit remains from Manjhi. and patterns of subsistence at ancient Hund, in the Vale of
Pragdhara: Journal of the Uttar Pradesh State Archaeological Peshawar, Pakistan. Unpublished BSc Dissertation, Institute
Organisation 11: 143–153. of Archaeology, University College London.
Chanchala, S. 2002. Botanical remains. In D. P. Tewari (ed.). Cooke, M., Fuller, D. Q. and Rajan, K. 2005. Early historic
Excavations at Charda. Lucknow: 166–194. agriculture in southern Tamil Nadu: Archaeobotanical research
Chanchala, S. 2004. Botanical remains. In D. P. Tewari (ed.). at Mangudi, Kodumanal and Perur. In U. Franke-Vogt and J.
Excavations at Saunphari and explorations in Ganga plain. Weisshaar (eds). South Asian archaeology 2003: Proceedings
Lucknow: 189–210. of the European Association for South Asian Archaeology
Chapekar, B. N. 1969. Report on the excavation at Ter 1958. Pune. conference, Bonn, Germany, 7th–11th July 2003. Aachen:
Chatterjee, P. 2012. After Subaltern Studies. Economic and political 329–334.
weekly 47(35): 44–49. Cooke, T. 1903–08. The Flora of the presidency of Bombay. London.
Chattopadhyaya, B. 1973 [1994]. Irrigation in Early Medieval Coomaraswamy, A. K. 1927. History of Indian and Indonesian art.
Rajasthan. In B. Chattopadhyaya. 1994. The Making of Early New York.
Medieval India. Oxford: 38–56. Coomaraswamy, A. K. 1930. Early Indian architecture. Eastern Art 2:
Chattopadhyaya, B. 1974. Trade and urban centres in Early 209–235.
Medieval North India. Indian historical review 1(2): 203–219. Coomaraswamy, A. K. 1992. Early Indian architecture: IV. Huts and
Chattopadhyaya, B. 1990. Aspects of rural settlements and rural related temple types. Essex House Press (Originally published
society in Early Medieval India. Calcutta. 1956. Reprinted in A. K. Coomaraswamy. 1992. Essays in early
Chattopadhyaya, B. 1994. The making of Early Medieval India. Indian architecture. New Delhi: 103–124).
Oxford. Costantini, L. 1979. Plant remains at Pirak. In J. F. Jarrige and M.
Chattopadhyaya, B. 1998. Representing the Other? Sanskrit sources Santoni (eds). Fouilles de Pirak. Paris: 326–333.
and the Muslims. Delhi. Costantini, L. 1983. The beginning of agriculture in the Kachi
Chattopadhyaya, B. D. 1977. Coins and currency systems in south plain: The evidence of Mehrgarh. In B. Allchin (ed.). South
India c. AD 225–1300. New Delhi. Asian archaeology 1981: Papers from the sixth international
Chattopadhyaya, B. D. 1986. Urban centres in Early Medieval India: conference of the Association of South Asian Archaeologists in
An overview. In S. Bhattacharya and R. Thapar (eds). Situating Western Europe held in Cambridge University 5–10 July 1981.
Indian history for Servepalli Gopal. Delhi: 8–33. Cambridge: 29–33.
Chavane, B. 1986. Etude pédologique et palynologique du site Costantini, L. 1987. Appendix B: Vegetal remains. In G. Stacul (ed.).
protohistorique de Ogo (moyenne vallée du Sénégal). Prehistoric and protohistoric Swat, Pakistan. Rome: 155–165.
Archéologie africaine et sciences de la nature appliquées à Costantini, L. 1990. Harrapan agriculture in Pakistan: The evidence
l’archéologie. Symposium international 1. Paris: 355–368. of Nausharo. In M. Taddei (ed.). South Asian archaeology 1987.
Childe, G. 1951. Social evolution. London. Rome: 321–332.
338   Reference list

Costantini, L. and Biasini, L. C. 1985. Agriculture in Baluchistan S.  Ehrhardt. From history to prehistory at Nevasa (1954–56).
between the 7th and 3rd millenium BC. Newsletter of Pune: 531–536.
Baluchistan studies 2: 16–37. Eastwood, G. 1982. Textiles. In D. S. Whitcomb and J. H. Johnson.
Crawford, G. 2006. East Asian plant domestication. In M. T. Stark 1982. Quseir Al-Qadim 1980: Preliminary report. American
(ed.). Archaeology of Asia. Oxford: 77–95. Research Center in Egypt: 285–326.
Cribb, J. 1992. Numismatic evidence for the date of the ‘Periplus’. Erdosy, G. 1995. City states of north India and Pakistan at the
In D. W. MacDowall, S. Sharma and S. Garg (eds). Indian time of the Buddha. In F. R. Allchin (ed.). The Archaeology of
numismatics, history and culture: Essays in honour of Dr early historic South Asia: The emergence of cities and states.
Parmeshwaril Lal Gupta. Delhi: 131–145. Cambridge: 99–122.
Crowfoot, G. M. and Griffiths, F. L. 1934. On the early use of cotton Eschmann, A., Kulke, H. and Tripathi, G. C. 1978. The Cult of
in the Nile Valley. Journal of Egyptian archaeology 20: 5–12. Jagannath and the regional tradition of Orissa. South Asian
D’Andrea, A. C., Kahlheber, S., Logan, A. L. and Watson, D. J. 2007. studies 8. South Asia Institute, New Delhi branch, Heidelberg
Early domesticated cowpea (Vigna unguiculata) from central University. New Delhi.
Ghana. Antiquity 81(313): 686–698. Fa-Hsien. 1923. The Travels of Fa-hsien (399– 414 A.D.), or records
Dales, G. F. 1986. Some fresh approaches to old problems in of the Buddhistic kingdoms. Re-translated by H. A. Giles.
Harappan archaeology. In J. Jacobson (ed.). Studies in the Cambridge.
archaeology of India and Pakistan. Delhi: 117–136. FAO 1995. Sorghum and millets in human nutrition. (FAO food and
Dass, M. and Willis, M. 2002. The lion capital from Udayagiri and nutrition series, No. 27). Food and agriculture organization of
the antiquity of sun worship in Central India. South Asian the United Nations. Rome [available online: http://www.fao.
studies 18(1): 25–45. org/DOCREP/T0818e/T0818E00.htm#Contents].
Davis, S. J. M. 1987. The Archaeology of animals. London. FAO 2001–2003. Food and agriculture organization of the United
De Wet, J. M. J., Prasada Rao, K. E., Mengesha, M. H. and Brink, D. Nations, world census of agriculture (http://www.fao.org/
E. 1983. Diversity in kodo millet, Paspalum scrobiculatum. fileadmin/templates/ess/documents/world_census_of_
Economic botany 37(2): 159–163. agriculture/countries_for_website/INDIA_2001-2003-F.pdf).
Deo, S. B. 1975. Paithan: Ek aitihaseeka samalochan [Marathi]. Fergusson, J. 1868. Tree and serpent worship in India. London.
Marathwada Samshodhana Mandala Warshika. Usmanabad: Fleet, J. F. (ed.) 1888. Inscriptions of the early Gupta kings and their
1–9. successors. Corpus inscriptionum Indicarum, vol. 3. Calcutta.
Deo, S. B. and Dhavalikar, M. K. 1968. Paunar excavations (1967). Flood, G. 2003. Introduction: Establishing the boundaries. In G.
Nagpur. Flood (ed.). The Blackwell companion to Hinduism. Oxford:
Deo, S. B. and Gupte, R. S. 1974. Excavations at Bhorkardan 1–19.
(Bhogavardhana) 1973. Nagpur/Aurangabad. Fukunaga, K., Ichitani, K. and Kawase, M. 2006. Phylogenetic
Deo, S. B. and Joshi, J. P. 1972. Pauni excavations (1969–70). analysis of the rDNA intergenic spacer subrepeats and its
Nagpur. implications for the domestication history of foxtail millet,
Deotare, B., Shete, G., Sawant, R., Kathale, V. and Naik, S. 2007. Setaria italica. Theoretical and applied genetics 113(2):
Discovery of structural stupa at Bhon, District Buldana, 261–269.
Maharashtra. Puratattva 37: 176–185. Fukunaga, K., Wang, Z., Kato, K. and Kawase, M. 2002.
Dey, N. L. 1899. Geographical dictionary of ancient and medieval Geographical variation of nuclear genome RFLPs and genetic
India. Calcutta. differentiation in foxtail millet, Setaria italica (L.) P. Beauv.
Deyell, J. S. 1990. Living without silver: A monetary history of early Genetic resources and crop evolution 49: 95–101.
medieval north India. Delhi. Fuller, D. Q. 1999. The Emergence of agricultural societies in
Dhopate, S. G. 2004. Two Satavahana silver coins with a distinct south India: Botanical and archaeological perspectives.
portrait feature. Studies in South Indian Coins 14: 51–54. Unpublished PhD Dissertation, Cambridge University.
Digby, S. 1982. Northern India under the Sultanate. In T. Fuller, D. Q. 2000. The botanical remains. In T. Insoll (ed.).
Raychaudhuri and I. Habib (eds). The Cambridge economic Urbanism, archaeology and trade: Further observations on the
history of India Vol. 1. c. 1200–c. 1750 AD. Cambridge: 45–47. Gao region (Mali). The 1996 field season results. British
Dikshit, M. G. 1968. Excavations at Kaundinayapura. Bombay. Archaeological Reports International Series 829. Oxford:
Dikshit, M. G. 1973. Excavation at Kausan and Paithan, district 28–35.
Aurangabad. Indian archaeology 1965–66: a review: 28–29. Fuller, D. Q. 2002. Fifty years of archaeobotanical studies in India:
Dirks, N. 1988. The hollow crown: ethnohistory of an Indian Laying a solid foundation. In S. Settar and R. Korisettar (eds).
kingdom. Cambridge. Indian archaeology in retrospect, volume III. Archaeology and
Doniger, W. 2009. Dogs as dalits in Indian literature. In U. Dill and interactive disciplines. Publications of the Indian council for
C. Walde (eds). Antike mythen: Medien, transformationen und historical research. New Delhi: 247–364.
konstruktionen. Berlin/New York: 391–405. Fuller, D. Q. 2003a. Further evidence on the prehistory of sesame.
Driesch, A. E. von den. 1976. A Guide to the measurement of animal Asian Agri-History 7(2): 127–137.
bones from archaeological sites. Peabody Museum bulletin 1. Fuller, D. Q. 2003b. African crops in prehistoric South Asia: A
Cambridge, Massachusetts. critical review. In K. Neumann, A. Butler and S. Kahlheber
Duby, G. 1952. La société aux XIe et XIIe siècles dans la région (eds). Food, fuel and fields. Progress in Africa archaeobotany,
mâconnaise. Paris. Africa Praehistorica 15. Colonge: 239–271.
Eapen, J. 1960. A report on the animal remains of the Nevasa Fuller, D. Q. 2003c. Indus and non-Indus agricultural traditions:
excavations. In H. D. Sankalia, S. B. Deo, Z. D. Ansari and Local developments and crop adoptions on the Indian
 Reference list   339

Peninsula. In S. Weber and W. Belcher (eds). Indus Gautier, A. 1984. How do I count you, let me count the ways?
ethnobiology: New perspectives from the field. Lanham, Problems of archaeozoological quantification. In C. Grigson
Maryland: 343–396. and J. Clutton-Brock (eds). Animals and archaeology:
Fuller, D. Q. 2005. Ceramics, seeds and culinary change in 4. Husbandry in Europe. British Archaeological Reports
prehistoric India. Antiquity 79(306): 761–777. International Series 227. Oxford: 237–251.
Fuller, D. Q. 2006a. Agricultural origins and frontiers in South Asia: Geerlings, E. 2001. Sheep husbandry and ethnoveterinary
A working synthesis. Journal of world prehistory 20: 1–86. knowledge of Raika sheep pastoralists in Rajasthan, India.
Fuller, D. Q. 2006b. The Ganges on the world Neolithic map: The Unpublished MSc dissertation, Department of Environmental
significance of recent research on agriculture origins in Sciences, Wageningen University, Deventer, Netherlands.
northern India. Pragdhara: Journal of the Uttar Pradesh State George, J. C. 1955. Identification of bones. In H. D. Sankalia and
Archaeological Organisation 16: 187–206. S. B. Deo. Report on excavations at Nasik and Jorwe 1950–51.
Fuller, D. Q. 2007. Contrasting patterns in crop domestication and Pune: 142–143.
domestication rates: Recent archaeobotanical insights from Ghosh, A. 1989. An Encyclopaedia of Indian archaeology. 2 vols.
the Old World. Annals of botany 100(5): 903–924. Delhi.
Fuller, D. Q., Boivin, N. and Korisettar, R. 2007. Dating the Neolithic Godbole, S. 2002–2003. Some terracotta sealings from Paithan.
of South India: New radiometric evidence for key economic, Journal of the Numismatic Society of India 64–65: 18–19, pls
social and ritual transformations. Antiquity 81: 755–778. II–III.
Fuller, D. Q. and Edwards, D. N. 2001. Medieval plant economy in Gokhale, S. 1995. New discoveries in the Satavahana coinage.
middle Nubia: Preliminary archaeobotanical evidence from Journal of archaeology – Satavahana Special. Hyderabad.
Nauri. Sudan and Nubia 5: 97–103. Gopal, L. 1980. Aspects of history of agriculture in ancient India.
Fuller, D. Q. and Harvey, E. L. 2006. The archaeobotany of Indian Varanasi.
pulses: Identification, processing and evidence for cultivation. Gopal, L. and Srivastava, V. C. (eds). 2008. History of agriculture
Environmental archaeology 11(2): 219–246. in India (up to c.1200 AD). History of science, philosophy and
Fuller, D. Q., Korisettar, R., Vankatasubbaiah, P. C. and Jones, M. culture in Indian civilization, vol. 5, part 1. General editor D. P.
K. 2004. Early plant domestications in southern India: Some Chattopadhyaya. Delhi.
preliminary archaeobotanical results. Vegetation history Goron, S. and Goenka J. P. 2001. Coins of the Indian sultanates.
archaeobotany 13: 115–129. New Delhi.
Fuller, D. Q., Korisettar, R. and Venkatasubbaiah, P. C. 2001. Govt. of Maharashtra. 1977. Maharashtra state gazeteers:
Southern Neolithic cultivation systems: A reconstruction Aurangabad district (revised edition). Bombay.
based on archaeobotanical evidence. South Asian studies 17: Gramsci, A. 1995. Further selections from the prison notebooks. D.
171–187. Boothman (ed., trans.). London.
Fuller, D. Q. and Madella, M. 2001. Issues in Harappan archae- Gramsci, A. 1996. Prison notebooks, vol. 2. J. A. Buttigieg (ed.,
obotany: Retrospect and prospect. In S. Settar and R. trans.). New York.
Korisettar (eds). Indian archaeology in retrospect, volume II. Grant, A. 1982. The use of tooth wear as a guide to the age of
Protohistory publications of the Indian Council for Historical domestic ungulates. In B. Wilson, C. Grigson and S. Payne
Research. New Delhi: 317–390. (eds). Ageing and sexing animal bones from archaeological
Fuller, D. Q. and Murphy, C. 2018. The Origins and early dispersal of sites. British Archaeological Reports British Series 109.
horsegram (Macrotyloma uniflorum), a major crop of ancient Oxford: 91–108.
India. Genetic resources and crop evolution 65(1): 285–305. Gray, B. 1948–49. Blue and white vessels in Persian miniatures
Fuller, D. Q. and Qin, L. 2009. Water management and labour in of the 14th–15th centuries reexamined. Transactions of the
the origins and dispersal of Asian rice. World archaeology 41: Oriental Ceramic Society, 26: 22–30.
88–111. Grayson, D. K. 1984. Quantitative zooarchaeology. New York.
Fuller, D. Q., Stevens, C. J. and McClatchie, M. 2014. Routine Griffee, P. 2001. Sesamum indicum L. In Ecoport: food and
activities, tertiary refuse and labor organization: Social agriculture organization of the United Nations. Food for all.
inference from everyday archaeobotany. In M. Madella, C. Website 11 November 2007: http://ecoport.org/ep?Plant=
Lancelloti and M. Savard (eds). Ancient plants and people. 1937&entityDisplayCategory=full
Contemporary trends in archaeobotany. Tucson: 174–217. Griffith, F. LL. and Crowfoot, G. M. 1934. On the early use of cotton
Fuller, D. Q. and Weber, S. A. 2005. Formation processes and in the Nile Valley. Journal of Egyptian archaeology 20: 5–12.
paleoethnobotanical interpretation in South Asia. Journal Guha-Thakurta, T. 2013. The production and reproduction of a
of interdisciplinary studies in history and archaeology 2(1): monument: the many lives of the Sanchi stupa. South Asian
91–114. studies 29(1): 77–109.
GacchācārVrtti = Gacchācāraprakīrnaka, with Vijayavimalaganin’s Gulati, A. N. and Turner, A. J. 1929. A note on the early history of
vrtti. Vijayajinendrasūri (ed.). Sāntipurī: Harsapuspāmrta Jain cotton. Journal of the Textile Institute 20: 1–9.
Granthamālā, 1987. Gunawardana, R. A. L. H. 1979. Robe and plough: Monasticism and
Galinato, M., Moody, K. and Piggin, C. 1999. Upland rice weeds economic interest in early medieval Sri Lanka. Tucson.
of South and Southeast Asia. International Rice Research Gupta, P. L. 1960. Punch-marked coins in the Andhra Pradesh
Institute. Manila. Government Museum. Hyderabad.
Gamble, J. S. 1935. Flora of the Presidency of Madras, vol. I. Gupta, P. L. and Hardaker, T. R. 1985. Ancient Indian silver
London. punch-marked coins of the Magadha Maurya Karshapana
series. Nasik.
340   Reference list

Gupta, S. 1999. Ceramic seriation and chronology of early kingship: archaeology and art in South Asia and along the Silk
Satavahana coins. In A. M. Shastri (ed.). The Age of the Road, 3500 BCE–5th century CE, vol. 1. New Delhi: 43–61.
Satavahanas. New Delhi: 349–359. Heitzman, J. 1997. Gifts of power: lordship in an early Indian state.
Gupta, S., Williams, D. and Peacock, D. 2001. Dressel 2–4 Delhi.
amphorae and Roman trade with India: the evidence from Hernández Fernández, M. and Vrba, E. A. 2005. A complete
Nevasa. South Asian studies 17: 7–18. estimate of the phylogenetic relationships in Ruminantia:
Guy, J. 1986. Oriental trade ceramics in South East Asia, 9th to 16th A dated species-level supertree of the extant ruminants.
Century. Oxford. Biological review 80: 269–302.
Guy, J. 1990. Oriental trade ceramics in South-East Asia, ninth to Hesse, B. and Wapnish, P. 1997. Can pig remains be used for ethnic
sixteenth centuries, with a catalogue of Chinese, Vietnamese diagnosis in the ancient Near East? In N. A. Silberman and
and Thai wares in Australian collections. Oxford. D. Small (eds). The Archaeology of Israel. Constructing the
Habib, I. 1999. The Agrarian system of Mughal India. New Delhi, past, interpreting the present. Journal for the study of the Old
2nd edn. Testament, supplement series 237: 238–270.
Halstead, P. 1987. Man and other animals in later Greek prehistory. Higham, C. F. W. 1967. Stock rearing as a cultural factor in
Annual of the British School of Archaeology at Athens 82: prehistoric Europe. Proceedings of the Prehistoric Society 33:
71–83. 84–106.
Halstead, P. 1998. Mortality models and milking: problems of Hillig, K. W. 2005. Genetic evidence for speciation in Cannabis
uniformitarianism, optimality and equifinality reconsidered. (Cannabaceae). Genetic resources and crop evolution 52(2):
Anthropozoologica 27: 3–20. 161–180.
Halstead, P., Collins, P. and Isaakidou, V. 2002. Sorting the sheep Hillman, G. C. 1984. Interpretation of archaeological plant remains:
from the goats: morphological distinctions between the The application of ethnographic models from Turkey. In W.
mandible and mandibular teeth of adult Ovis and Capra. Van Zeist and W. A. Casparie (eds). Plants and ancient man.
Journal of archaeological science 29: 545–553. Studies in paleoethnobotany. Rotterdam: 1–41.
Hammer, O., Harper, D. A. T. and Ryan, P. D. 2001. PAST: Paleon- Hinüber, O. von. 2007. Review of F. Virkus Politische Strukturen im
tological statistics software package for education and data Guptareich (300–500 n.Chr) by F. Virkus. Indo-Iranian journal
analysis. Palaeontologia electronica 4(1). palaeo-electronica. 50: 183–192.
org/2001_1/past/issue1_01.htm. Hocart, A. M. 1950. Caste: a comparative study. London.
Hardy, A. 1995. Indian temple architecture: Form and Howell, J. 1995. Excavations at Sannathi 1986–1989. Memoirs of the
transformation. New Delhi. Archaeological Survey of India, No. 93. New Delhi.
Harle, K. 1996. Coinage in the Roman economy, 300 B.C. to A.D. Inden, R. 1985. Lordship and caste in Hindu discourse. In A. Cantlie
700. Baltimore. and F. Marglin (eds). Indian religion. London: 159–179.
Harvey, E. and Fuller, D. Q. 2005. Investigating crop processing Inden, R. 2006. Text and practice: essays on South Asian history. D.
through phytolith analysis: The case of rice and millets. Ali (notes). SOAS Studies on South Asia. Oxford.
Journal of archaeological science 32: 739–752. Jain, J. C. 1984. Life in ancient India as depicted in the Jain canon
Harvey, E. L., Fuller, D. Q., Mohanty R. K. and Mohanta, B. 2006. and commentaries: 6th century BC to 17th century AD. Delhi,
Early agriculture in Orissa: Some archaeobotanical results 2nd edn.
and field observations on the Neolithic. Man and environment Jayaswal, V. 2001. Royal temples of Gupta period (excavations at
31(2): 21–32. Bhitari). New Delhi.
Hawkes, J. 2014a. Chronological sequences and the ‘problem’ of Jennings, P. R. 1966. The evolution of plant type in Oryza sativa.
Early Medieval settlement in India. Puratattva 44: 208–228. Economic botany 20: 396–402.
Hawkes, J. 2014b. Finding the Early Medieval in South Asian Jha, V. M. 1996. Settlement, society and polity in Early Medieval
archaeology. Asian perspectives 53(1): 53–96. rural India. Indian historical review 20: 34–65.
Hawkes, J. 2020. The elephant in the room: archaeology and the Jha, A. and Chumble, P. D. 2003–2004. Coin bearing the name
Early Medieval period in India. In E. Myrdal (ed.). South Asian Satavahana/Sadavahana: Some new types. Numismatic
archaeology and art 2014. New Delhi: 169–183. digest 27–28: 47–66.
Hawkes, J. and Abbas, R. 2016. Copperplates in context: Jha, A. and Rajgor, D. 1994. Studies in the coinage of the Western
preliminary investigation of the study and archaeological Ksatrapas. Nasik.
settings of land grant inscriptions. Pratnatattva, 22: 41–71. Jinaprabhasūri = Vividhatīrthakalpa. Jinavijaya (ed.). Śāntiniketan:
Hawkes, J. and Casile, A. 2019 (under review). Back to basics: Singhī Jaina Jñānapītha, 1931.
exploring the importance of returning to the evidence, a case Jones, G. E. M. 1991. Numerical analysis in archaeobotany. In W.
study from South Asia. Asian archaeology. Van Zeist, K. Wasylikowsa and K.–E. Behre (eds). Progress in
Hawkes, J., Abbas, R. and Willis, M. 2016. Report of the archaeo- old world palaeoethnobotany. Rotterdam: 63–80.
logical survey of Chamak and its surroundings, District Kajale, M. D. 1974. Plant economy at Bhokardan. In S. B. Deo and
Amaravati, Maharashtra. Ancient Asia 7: 1–7. R. S. Gupte (eds). Excavations at Bhokardan (Bhoqavardhana)
Hawkes, J., Abbas, R., Shete, G. and Willis, M. 2017. A newly 1973. Nagpur-Auranagabad: 217–224.
discovered Iron Age settlement at Chachondi, District Kajale, M. D. 1977a. On the botanical findings from excavations at
Amravati, Maharashtra. Puratattva 46: 114–125. Daimabad, a Chalcolithic site in Western Maharashtra, India.
Hawkes, J., Abbas, R., Shete, G. and Willis, M. 2020. Iron Age Current science 46: 818–819.
stratigraphy: the sequence at Chachondi, Vidarbha. In L.
Greaves and A. Hardy (eds). Religions, society, trade and
 Reference list   341

Kajale, M. D. 1977b. Ancient plant economy at Nevasa during Khan, A. W. 1963. A Monograph on Yeleswaram excavations.
Satavahana and Indo-Roman period. Bulletin of the Deccan Andhra Pradesh Government Archaeological Series No. 14.
College Research Institute 35: 62–65. Hyderabad.
Kajale, M. D. 1982. First record of ancient grains at Naikund. In S. Kharakwal, J. S., Rawat, Y. S. and Osada, T. 2007. Kanmer: a
B. Deo and A. P. Jamkhedkar (eds). Excavations at Naikund Harappan site in Kachchh, Gujarat, India. In T. Osada (ed.).
1978–1980. Bombay: 60–63. Linguistics, archaeology and the human past. Occasional
Kajale, M. D. 1984. New light on agricultural plant economy during Paper 2. Kyoto: 21–46.
1st millennium BC: Palaeobotanical study of plant remains Kielhorn, F. 1894–95. Epigraphia Indica 3. New Delhi.
from excavations at Veerapuram, district Kurnool, Andhra Kikkawa, Y., Yonekawa, H., Suzuki, H. and Amano, T. 1997. Analysis
Pradesh. In T. V. G. Sastri, M. Katsuri Bai and J. V. P. Rao (eds). of genetic diversity of domestic water buffaloes and anoas
Veerapuram: A type site for cultural study in the Krishna based on variations in the mitochondrial gene for cytochrome
Valley. Hyderabad. Appendix B: 1–15. b. Animal genetics 28: 195–201.
Kajale, M. D. 1988. Plant economy. In M. K. Dhavalikar, H. D. Kimata, M. 1989. Grain crop cookery on the Deccan plateau. In S.
Sankalia and Z. D. Ansari (eds). Excavations at Inamgaon. Sakamoto (ed.). A Preliminary report of the studies on millet
Pune: 727–821. cultivation and its agro-pastoral culture complex in the Indian
Kajale, M. D. 1989. Archaeobotanical investigation on Megalithic subcontinent, II. Research team for the studies on millet
Bhagimohari and its significance. Man and environment 13: cultivation in the Indian subcontinent, Kyoto University. Kyoto:
87–96. 33–50.
Kajale, M. D. 1991. Current status of Indian palaeoethnobotany: Kimata, M., Ashok, E. G. and Seetharam, A. 2000. Domestication,
Introduced and indigenous food plants with a discussion cultivation and utilization of two small millets, Brachiaria
of the historical and evolutionary development of Indian ramosa and Setaria glauca (Poaceae), in South India.
agriculture and agricultural systems in general. In J. Economic botany 54(2): 217–227.
M. Renfrew (ed.). New light on early farming: Recent Kimata, M. and Sakamoto, S. 1992. Utilization of several species
developments in palaeoethnobotany. Edinburgh: 155–189. of millet in Eurasia. Bulletin of Field Studies Institute, Tokyo
Kajale, M. D. 1994. Archaeobotanical investigations on a Gakugei University 3: 1–12.
multicultural site at Adam, Maharashtra, with special Kingwell-Banham, E. and Fuller, D.Q. 2014. Brown top millet:
reference to the development of tropical agriculture in parts of origins and development. In C. Smith (ed.). Encyclopedia of
India. In J. Hather (ed.). Tropical archaeobotany: Applications global archaeology. New York: 1021–1024.
and new developments. London: 34–50. Kirch, P. J. 1994. The Wet and dry: Irrigation and agricultural
Kajale, M. D. 1996a. Plant remains. In M. K. Dhavalikar, M. R. Raval intensification in Polynesia. Chicago.
and Y. M. Chitalwala (eds). Kuntasi: a Harappan emporium on Kirkbride, J. H. Jr., Gunn, C. R. and Dallwitz, M. J. 2006. Family
the west coast. Pune: 285–289. guide for fruits and aeeds, version 1. 0. Website 5 May 2007:
Kajale, M. D. 1996b. Palaeobotanical investigations at Balathal: http://nt.arsgrin.gov/sbmlweb/OnlineResources/frsdfam/
Preliminary results. Man and environment 21: 98–102. Index.cfm.
Kalpasūtra = Kalpa Sutra, Jain Sutras Part I. J. Hermann (ed.). Kobayashi, H. and Kimata, M. 1989. Millets in Maharashtra and
Sacred Books of the East, Vol. 22. Oxford, 1884. Orissa in India. In S. Sakamoto (ed.). A Preliminary report of
Kennet, D. 2004a. Sasanian and Islamic pottery from Ras the studies on millet cultivation and its agro-pastoral culture
al-Khaimah: Classification, chronology and analysis of trade complex in the Indian subcontinent, II. Research team for the
in the western Indian Ocean. Society for Arabian Studies studies on millet cultivation in the Indian subcontinent, Kyoto
Monographs No. 1. BAR International Series 1248. University. Kyoto: 1–9.
Kennet, D. 2004b. The transition from Early Historic to Early Kosambi, D. D. 1956. Introduction to the study of Indian history.
Medieval in the Vakataka realm. In H. Bakker (ed.). The Bombay.
Vakataka heritage: Indian culture at the crossroads. Kosambi, D. D. 1975. An Introduction to the study of Indian history.
Groningen: 11–17. Originally published 1956. Bombay, 2nd edn.
Kennet, D. 2012. Transformations in Early Historic and Early Krahl, R. 1986. Chinese ceramics in the Topkapı Saray Museum. A
Medieval India: Excavations at Paithan, Maharashtra Complete catalogue. 3 vols. London.
1996–1999 [data-set]. York: Archaeology Data Service Krahl, R. 1997. Vietnamese blue-and-white and related wares. In J.
[distributor] (doi:10.5284/1017460). Stevenson and J. Guy (eds). Vietnamese ceramics: A Separate
Kennet, D. 2013. Reconsidering the decline of urbanism in late tradition. Chicago: 146–157.
Early Historic and Early Medieval South Asia. In J. Schiet- Kramrisch, S. 1946. The Hindu temple. 2 vols. Calcutta.
tecatte and C. Robin (eds). Les préludes de l’Islam: ruptures et Kratochvil, Z. 1969. Species criteria on the distal section of the tibia
continuités dans les civilisations du Proche-Orient de l’Afrique in Ovis ammon f. aries and Capra aegagrus f. hircus L. Acta
orientale, de l’Arabie et de l’Inde à la veille de l’Islam. Orient & Veterinaria (Brno) 38: 483–490.
Méditerranée No. 11. Paris: 331–353. Krebs, C. 1985. Ecology: The Experimental analysis of distribution
Kennet, D. and Rao, J. V. P. 2003. Two Early Historic brick temples at and abundance. New York, 3rd edn.
Paithan in Maharashtra. South Asian studies 19: 113–123. Kulke, H. 1978. Royal temple policy and the structure of Medieval
Ketkar, S. V. 1963. Pracheena Maharashtra, volume II [Marathi]. Hindu kingdoms. In A. Eschmann, H. Kulke and G. C. Tripathi
Pune. (eds). The Cult of Jagannath and the regional tradition of
Orissa. South Asian Studies No. VIII. New Delhi: 125–137.
342   Reference list

Kulke, H. 1995a. The study of the state in pre-modern India. In H. Maier, U. 1996. Morphological studies of wheat free-threshing ears
Kulke (ed.). The State in India 1000–1700. Delhi: 1–47. from a Neolithic site in Southwest Germany, and the history
Kulke, H. 1995b. The Early and the imperial kingdom: A Processural of naked wheats. Vegetation history and archaeobotany 5:
model of integrative state formation in Early Medieval India. In 39–55.
H. Kulke (ed.). The State in India 1000–1700. Delhi: 233–262. Majumdar, R. C. 1960. Classical accounts of India. Calcutta.
Kumar, S., Gupta, J., Kumar, N., Dikshit, K., Navani, N., Jain, P. Mallory, J. P. and Mair, V. 2000. Tarim mummies. London.
and Nagarajan, M. 2006. Genetic variation and relationships Marr, C. 1978. West African textiles. African arts 11: 81.
among eight Indian riverine buffalo breeds. Molecular ecology Marshall, F. and Hildebrand, E. 2002. Cattle before crops: The
15: 593–600. beginnings of food production in Africa. Journal of world
Lane, A. 1947. Early Islamic pottery. London. prehistory 16: 99–143.
Laws of Manu. The Laws of Manu. W. Doniger and B. K. Smith Marshall, J. 1924. First light on a long forgotten civilization.
(trans., notes). London: Penguin, 1991. Illustrated London news September 20: 528–532, 548.
Lefrancq, C., Hawkes, J., Jaseera, C. M. and Mohanty, R. K. 2019. Marshall, J. 1931. Mohenjo-daro and the Indus civilization. 3 vols.
A typology of practice: the archaeological ceramics from London.
Mahurjhari. Internet archaeology 52 https://doi.org/10.11141/ Marshall, J., Foucher, A. and Majumdar, N. G. 1939. The monuments
ia.52.9. of Sañchi. 3 vols. Calcutta.
Lefrancq, C. and Hawkes, J. 2019 (in preparation). Accepting Marshall, J. 1951. Taxila: an illustrated account of archaeological
variation and embracing uncertainty: creating a regional excavations carried out at Taxila under the orders of the
pottery typology in South Asian archaeology. Open Government of India between the years 1913 and 1934.
archaeology. Cambridge.
Legge, A. J. 1981. Aspects of cattle husbandry. In R. Mercer (ed.). Mate, M. S. 1957. Trivikrama temple at Ter. Bulletin of Deccan
Farming practice in British prehistory. Edinburgh: 169–181. College Institute XVIII: 1–4.
Legge, A. J. 1992. Excavations at Grimes Graves, Norfolk, Mayer-Thurman, C. C. and Williams, B. B. 1979. Ancient textiles from
1972–1976, 4: Animals, environment and the Bronze Age Nubia. Chicago.
economy. London. McCorriston, J. 1997. The fiber revolution: Textile extensification,
Legge, A. J. and Rowley-Conwy, P. A. 1988. Star Carr revisited. A alienation, and social stratification in ancient Mesopotamia.
Re-analysis of the large mammals. London. Current anthropology 38: 517–549.
Longhurst, A. H. 1917–18. The Kapotesvara temple at Chejerla McCrindle, J. 1885. Ancient India as described by Ptolemy. Calcutta/
in Guntur district. Annual report of the Archaeological London.
Department, Southern Circle. Madras: 33–35. McCrindle, J. W. 1926. Ancient India as described by Megasthenes
Lowenthal, D. 1985. The past is a foreign country. Cambridge. and Arrian. Reprint. New Delhi.
Ludden., D. 1999. An Agrarian history of South Asia. The New McKean, M. B. 1983. The palynology of Balakot, a pre-Harappan
Cambridge history of India 4. 4. Cambridge. and Harappan age site in Las Bela, Pakistan. Unpublished
Lueders, H. 1909–10. A list of Brahmi inscriptions. Appendix to PhD, Southern Methodist University.
Epigraphia Indica, vol. X. Calcutta. Medley, M. 1976. The Chinese potter: A Practical history of Chinese
Lunsingh Scheurleer, D. F. 1974. Chinese export porcelain, ‘Chine de ceramics. Oxford.
commande’. London. Mehta, R. N. and Shah, D. R. 1968. Excavation at Nagara. Baroda.
Madella, M. 2003. Investigating agriculture and environment Mehta, R. N. 1979. Mediaeval archaeology. Delhi.
in South Asia: Present and future considerations of opal Meister, M. W., Dhaky, M. A. and Deva, K. 1988. Encyclopaedia of
phytoliths. In S. A. Weber and W. R. Belcher (eds). Indus Indian temple architecture. Vol. II, part 1, text. Delhi.
ethnobiology. New perspectives from the field. Lanham: Mirashi, V. V. 1963. Corpus inscriptionum Indicarum, vol. 5. New
199–250. Delhi.
Maggio, J. 2007. Can the subaltern be heard? Political theory, Mirashi V. V. 1966. Epigraphia Indica, vol. 36. New Delhi.
translation, representation, and Gayatri Chakravorty Spivak. Mirashi, V. V. 1972. Wategaon hoard of Satavahana coins. Journal of
Alternatives: global, local, political 32(4): 419–443. the Numismatic Society of India 34/II: 205–212, pl. X.
Mahabharata = Sriman Mahabharatam. T. R. Krishnacharya and T. Mirashi, V. V. 1979. Satavahana ani Pashchimi Ksatrapa yancha
R. Vyasacharya (eds). Bombay, 1906–1909. Itihas aani Koreeva Lekha [Marathi]. Bombay.
Maheshwari, K. K. 1977. Some interesting coins from Andhra Misra, V. N. and Mohanty, R. K. 2001. A rare Chalcolithic pottery
Pradesh. Numismatic digest 1/I: 1–13. cache from Balathal, Rajasthan. Man and environment 26(2):
Maheshwari, K. K. 1979. Coins of Kochhiputa Satakarni. 67–74.
Numismatic digest 3/II: 29–30. Mitchiner, M. 1972. Some Kausambi, Rajgir and Mauryan copper
Maheshwari, K. K. and Rath, B. 1996. Fire altar type of coins. Journal of the Numismatic Society of India 34: 157–175.
Skandagupta: Towards a typological and chronological Mitchiner, M. 1973. Local coinage of Taxila and Gandhara. Journal
definition. In K. K. Maheshwari and B. Rath (eds). Numismatic of the Numismatic Society of India 35: 48–72.
panorama: Essays in memory of S. M. Shukla. Delhi: 179–194. Mitchiner, M. 1978. Oriental coins and their values: The ancient and
Maheshwari, K. K. and Rath, B. 1998. Garuda type silver coins of classical world. London.
Kumaragupta reconsidered. In A. K. Jha and S. Garg (eds). Ex Mohanty, R. K. 1999. Significance of a bead manufacturing centre
moneta: Essays in honour of Dr. David W. MacDowall. Delhi: at Mahurjhari, District Nagpur, Maharashtra, India. Man and
293–320. environment 26(2): 79–89.
 Reference list   343

Mohanty, R. K. 2005. Some important observations: excavations Noda, K., Prakongvongs, C. and Chaiwiratnukul, L. 1985.
at Mahurjhari (2001–2004). Man and environment 30(1): Topography of the seeds and leaves of tropical weed. Bangkok.
106–107. Olsen, S. J. 1960. Post-cranial skeletal characters of Bison and Bos.
Mohanty, R. K., Hawkes, J., Lefrancq, C. and Abbas, R. 2019. Report Peabody Museum of Archaeology and Ethnology Papers 45,
on the dating of the Historical Period site at Mahurjhari, no. 4.
Vidarbha. Heritage: journal of multidisciplinary studies in Pandey, S. N. and Chanda, A. 1996. Economic botany. New Delhi.
archaeology 7: 15–28. Payne, S. 1969. A metrical distinction between sheep and goat
Moody, K. 1989. Weeds reported in rice in South and Southeast metacarpals. In P. J. Ucko and G. W. Dimbleby (eds). The
Asia. International Rice Research Institute, Manila. Domestication and exploitation of plants and animals.
Morris, R. (ed.) 1885–1900. The Anguttara-nikâya. Pali Text London: 295–305.
Society/Henry Frowde. 6 vols. London. Payne, S. 1972. Partial recovery and sample bias: The results of
Morrison, K. 1994. The intensification of production: Archaeological some sieving experiments. In E. S. Higgs (ed.). Papers in
approaches. Journal of archaeological method and theory 1(2): economic prehistory. Cambridge: 49–64.
111–159. Payne, S. 1973. Kill-off patterns in sheep and goats: The mandibles
Morwanchikar, R. S. 1985. The City of the saints: Paithan through from Asvan Kale. Anatolian studies 23: 281–303.
the ages. New Delhi. Payne, S. 1975. Partial recovery and sample bias. In A. T. Clason
Moulherat, C., Tengberg, M., Haquet, J.–F. and Mille, B. 2002. First (ed.). Archaeozoological studies. Amsterdam: 7–17.
evidence of cotton at Neolithic Mehrgarh, Pakistan: Analysis Payne, S. 1985. Morphological distinctions between the
of mineralized fibres from a copper bead. Journal of archaeo- mandibular teeth of young sheep, Ovis, and goats, Capra.
logical science 29: 1393–1401. Journal of archaeological science 12: 139–147.
Murray, M. and Nixon, S. 2006. Tadmakka/Essouk: Medieval plant Payne, S. and Bull, G. 1988. Components of variation in
remains from the Malian Sahara. Paper presented at the Fifth measurements of pig bones and teeth, and the use of
International Workshop on African Archaeobotany, University measurements to distinguish wild from domestic pig remains.
College London, London, July 2006 [http://www.homepages. ArchæoZoologia 2: 27–65.
ucl.ac.uk/~tcrndfu/IWAA/AbMurray.htm]. Payne, S. and Munson, P. J. 1985. Ruby and how many squirrels?
Murray, S. S. 2004. Searching for the origins of African rice The destruction of bones by dogs. In N. R. J. Fieller, D. D.
domestication. Antiquity 78(300): Project Gallery: http:// Gilbertson and N. G. A. Ralph (eds). Palaeobiological investi-
antiquity.ac.uk/progall/murray/ gations. British Archaeological Reports International Series
Murray, S. S. 2007. Medieval cotton and wheat finds in the middle 266. Oxford: 31–39.
Niger Delta (Mali). In R. Cappers (ed.), Fields of change. Pearsall, D. M. 2000. Paleoethnobotany. A Handbook of
Progress in African archaeobotany. Groningen: 43–52. procedures. New York, 2nd ed.
Museum of London Archaeology Service. 1994. Archaeology site Pelling, R. 2005. Garamantean agriculture and its significance in
manual. London, 3rd edn. a wider North African context: The evidence of plant remains
Naik, S. S. and Deotare, B. C. 2016. Plant economy at Early Historic from the Fazzan Project. The Journal of North African studies
site of Bhon, District Buldana, Maharashtra. Bulletin of the 10: 397–411.
Deccan College Research Institute 76: 51–68. Peng Xinwei. 1994. A Monetary history of China. E. Kaplan (trans.).
Nath, A. 1993. Taxila coins from Adam: A site in Vidarbha. Journal of Washington.
the Numismatic Society of India 55: 19–22. Penzer, N. M. 1968. The Ocean of story: Being C. H. Tawney’s
Nath, A. 1998. Further excavations at Pauni 1994. New Delhi. translation of Somadeva’s Katha Sarit Sagara. New Delhi.
Nath, A. 2016. Excavations at Adam (1988–1992): a city of Asika Peters, J. 1986. Osteomorphology and osteometry of the
Janapada. Memoirs of the Archaeological Survey of India 109. appendicular skeleton of African buffalo, Syncerus caffer
New Delhi. (Sparrman, 1779) and cattle, Bos primigenius f. taurus,
Nath, V. 2001. From ‘Brahmanism’ to ‘Hinduism’: Negotiating the Bojanus, 1827. Occasional papers, Laboratorium voor Paleon-
myth of the Great Tradition. Social scientist. 29, 3/4: 19–50. tologie, Rijksuniversiteit Gent, no. 1.
Nesbitt, M. N. 1993. Archaeobotanical remains. In G. Herrmann et Pichard, P. 1994. Vingt ans après Tanjavur, Gangaikondacho-
al. The International Merv Project, preliminary report on the lapuram. Publications de l’Ecole française d’Extrême-Orient.
first season (1992). Iran 31: 39–62. Mémoires archéologiques, 20. Paris.
Nesbitt, M. N. 1994. Archaeobotanical research in the Merv Pinto Orton, N. 1991. Red polished ware in Gujarat: A catalogue of
oasis. In G. Herrmann et al. The International Merv Project. twelve sites. In V. Begley and R. D. De Puma (eds). Rome and
Preliminary report on the second season (1993). Iran 32: India: The Ancient sea trade. Madison: 46–81.
53–75. Pirenne, H. 1936. Economic and social history of Medieval Europe.
Neuss, J. 2012. Review of Archaeological Excavations in Central London.
India by R. K. Sharma and O. P. Misra. Journal of the Royal Pliny. Natural History. Vol. II, books III–VII. H. Rackham (ed.). The
Asiatic Society 22/1: 189–191. Loeb classical library, 1942.
Ng, N. Q. 1995. Cowpea. Vigna unguiculata (Leguminosae- Pokharia A. K. 2007. Table 1. Plant remains from Kanmer excavation
Papilionideae). In J. Smartt and N. W. Simmonds (eds). 05–06. In T. Osada (ed.). Linguistics, archaeology and the
Evolution of crop plants. Longman: 326–332 (2nd ed). human past. Occasional paper 2. Kyoto: 30–31.
Niśīthacūrni=Niśītha-Sūtram with bhāsya and Jinadāsa’s cūrni. Pokharia, A. K. 2008. Record of macrobotanical remains from the
Vol. III. Amaramuni and Kanhaiyālāl (eds). Delhi/Varanasi: Aravalli Hill, Ojiyana, Rajasthan: Evidence from agriculture-
Bhāratīya Vidyā Prakāśan, 1982. based subsistence economy. Current science 94/5: 612–617.
344   Reference list

Pokharia, A. K. and Saraswat, K. S. 1999. Plant economy during Rehm, S. and Espig, G. 1991. The Cultivated plants of the tropics
Kushana period (100–300 AD) at ancient Sanghol. Pragdhara: and subtropics: Cultivation, economic value, utilization.
Journal of the Uttar Pradesh State Archaeological Organisation Weikersheim.
9: 75–122. Ritz, L. R., Glowatzki-Mullis, M.-L., MacHugh, D. E. and Gaillard, C.
Possehl, G. L. 1986. African millets in South Asian prehistory. In 2000. Phylogenetic analysis of the tribe Bovini using microsat-
J. Jacobsen (ed.), Studies in the archaeology of India and ellites. Animal genetics 31: 178–185.
Pakistan. New Delhi: 237–256. Robbins, P. 1994. Goats and grasses in Western Rajasthan. Pastoral
Raeside, I. M. P. 1976. The Mahanubhavas. Bulletin of the School of development network paper 36a. Overseas Development
Oriental and African Studies 39/3: 585–600. Institute, London: 6–12.
Raja Reddy, D. and Suryanarayana Reddy, P. 1984. Uninscribed Rowley-Conwy, P. 1994. Dung, dirt and deposits: Site formation
coins of Andhra. Hyderabad. under conditions of near-perfect preservation at Qasr Ibrim,
Rajgor, D. 1991. Recent researches in Kshatrapa numismatics. In D. Egyptian Nubia. In R. Luff and P. Rowley-Conwy (eds). Whither
Handa (ed.). Numismatic studies 1. New Delhi: 33–44. environmental archaeology? Oxford: 25–32.
Ranade, P. V. 1964. A new silver coin of Yajna Sri Satakarni. Journal Rowley-Conwy, P. 1998. Improved separation of Neolithic
of the Numismatic Society of India 26: 98. metapodials of sheep (Ovis) and goats (Capra) from Arene
Randhawa, M. S. 1980. A History of agriculture in India. Vol. I: Candide Cave, Liguria, Italy. Journal of archaeological science
Beginning to 12th century. New Delhi. 25: 251–258.
Rao, J. V. P. and Kennet, D. 1999. Paithan 1999: An Interim report on Rowley-Conwy, P. 2000. Milking caprines, hunting pigs: The
the fourth season of excavations. Unpublished report to the Neolithic economy of Arene Candide in its West Mediterranean
Archaeological Survey of India and the Society for South Asian context. In P. Rowley-Conwy (ed.). Animal bones, human
Studies. societies. Oxford: 124–132.
Rao, J. V. P., Kennet, D. and Howell, J. 1998. Paithan 1996–1998: Saldanha, C. J. 1984. Flora of Karnataka, vol. I. New Delhi.
An Interim report on the first three seasons of excavations. Saldanha, C. J. 1996. Flora of Karnataka, vol. II. New Delhi.
Unpublished report to the Archaeological Survey of India and Sali, C. 1998. Vakataka culture: archaeological perspective.
the Society for South Asian Studies. Sanshodhak 66: 3–38.
Rao, S. R. 1966. Excavations at Amreli: A Kshatrapa-Gupta town. Samuel, D. 2001. Archaeobotanical evidence and analysis.
Museum and Picture Gallery, Baroda. Bulletin Vol. 18. Baroda. In S. Berthier (ed.). Peuplement rural et aménagements
Rao, V. V. 1974. Bone remains from Bhokardan excavations. hydroagricoles dans la moyenne vallée de l’Euphrate fin VIIe–
In S. B. Deo and R. S. Gupte. Excavations at Bhokardan XIXe siècle. Damascus: 343–481.
(Bhoqavardhana) 1973. Nagpur, Aurangabad: 225–233. Sankalia, H. D. and Deo, S. B. 1955. The Excavations at Nasik and
Rapson, E. J. 1908. Catalogue of Indian coins in the British Jorwe. Poona.
Museum: Catalogue of the coins of the Andhra dynasty, the Sankalia, H. D., Deo, S. B., Ansari, Z. D. and Ehrhardt, S. 1960. From
Western Ksatrapas, the Traikutaka dynasty, and the ‘Bodhi’ history to prehistory at Nevasa (1954–56). Pune.
dynasty. London. Sankalia, H. D. and Dikshit, M. G. 1952. Excavations at Brahmapuri
Ray, H. P. 1986. Monastery and guild: Commerce under the (Kolhapur) 1945–46. Poona.
Satavahanas. Delhi. Sankalia, H. D., Subbarao, B. and Deo, S. B. 1958. The Excavations
Ray, H. P. 2008. Colonial archaeology in South Asia: the legacy of at Maheshwar and Navdatoli 1952–53. The Deccan College
Sir Mortimer Wheeler. New Delhi. Research Institute and the Maharaja Sayajirao University
Ray, H. 1989. Early Historical settlement in the Deccan: An publication No. 1. Pune.
ecological perspective. Man and environment 14/1: 103–108. Saraswat, K. S. 1986. Ancient crop remains from Sringeverapura,
Ray, H. P. 2009. The Shrine in early Hinduism: The changing sacred Allahabad, U. P. (ca. 1,050–700 B.C.). Geophytology 16(1):
landscape. The Journal of Hindu studies 2: 76–96. 97–106.
Redding, R. W. 1991. The role of the pig in the subsistence systems Saraswat, K. S. 1992. Archaeobotanical remains in ancient cultural
of ancient Egypt: A Parable on the potential of faunal data. and socio-economical dynamics of the Indian subcontinent.
In P. J. Crabtree and K. Ryan (eds). Animal use and culture Palaeobotanist 40: 514–545.
change. MASCA (Museum Applied Science Center for Saraswat, K. S. 1993. Plant economy of Late Harappans at Hulas.
Archaeology) research papers in science and archaeology, Purattatva 23: 1–12.
supplement to volume 8: 21–30. Saraswat, K. S. 1996–97. Plant economy of Barans at ancient
Reddy, S. N. 1994. Plant usage and subsistence modeling: An Sanghol (ca. 1900–1400 B.C.), Punjab. Pragdhara: Journal
Ethnoarchaeological approach to the Late Harappan of of the Uttar Pradesh State Archaeological Organisation 7:
Northwest India. Ph. D. Dissertation, University of Wisconsin. 97–114.
Reddy, S. N. 1997. If the threshing floor could talk: Integration Saraswat, K. S. 1997. Plant economy of Barans at ancient Sanghol
of agriculture and pastoralism during the Late Harappan in (ca. 1900–1400 B.C.), Punjab. Pragdhara: Journal of the Uttar
Gujarat. Journal of anthropological archaeology 16: 162–187. Pradesh State Archaeological Organisation 7: 97–114.
Reddy, S. N. 2003. Discerning palates of the past: An Ethnoarchae- Saraswat, K. S. 2002. Banawali (29’37’5”N; 75’23’6”E), district
ological study of crop cultivation and plant usage in India. Hissar. Indian archaeology 1996–97: a review: 203.
Ethnoarchaeological series 5, international monographs in Saraswat, K. S. 2004. Plant economy of early farming communities
prehistory. Ann Arbor, Michigan. at Senuwar, Bihar. In B. P. Singh (ed.). Senuwar excavations.
Varanasi: 416–535.
 Reference list   345

Saraswat, K. S. 2005. Agricultural background of the early farming Shastri, A. M. 1991. A terracotta proof piece of the silver coin of
communities in the middle Ganga Plain. Pragdhara: Journal Vasishthiputra Pulumavi. Journal of the Numismatic Society of
of the Uttar Pradesh State Archaeological Organisation 15: India 42: 115–117.
145–177. Shastri, A. M. 1992. Vakataka coins. In A. M. Shastri (ed.). The Age
Saraswat, K. S., Chanchala, S. and Pokharia, A. K. S. 2000. of the Vakatakas. New Delhi: 285–294.
Palaeobotanical and pollen analytical investigations. Indian Shaw, J. 2007. Buddhist landscapes in central India: Sanchi Hill
archaeology 1994–95: a review: 96–7. and archaeologies of religious and social change, c. third
Saraswat, K. S. and Pokharia, A. K. 2003. Palaeoethnobotanical century BC to fifth century AD. London.
investigations at Early Harappan Kunal. Pragdhara: Journal Sherratt, A. 1999. Cash-crops before cash: Organic consumables
of the Uttar Pradesh State Archaeological Organisation 13: and trade. In C. Gosden and J. Hather (eds). The Prehistory of
105–140. food. Appetites for change. London: 13–34.
Saraswat, K. S., Sharma, N. K. and Saini, D. C. 1994. Plant economy Shinnie, P. L. and Anderson J. 2004. Capital of Kush 2 (Meroitica
at ancient Narhan (ca. 1,300 B.C.–300/400 A.D.). In B. P. 20). Wiesbaden.
Singh (ed.). Excavations at Narhan (1984–1989). Varanasi: Shishlina, N. I., Orfinskaya O. V. and Golikov, V. P. 2003. Bronze
255–346. Age textiles from the North Caucasus: New evidence of
Sarkar, H. and Misra, B. N. 1972. Nagarjunakonda. New Delhi, 2nd fourth millennium BC fibres and fabrics. Oxford journal of
edn. archaeology 22(4): 331–344.
Sarma, I. K. 1980. Coinage of the Satavahana empire. Delhi. Shrimali, K. M. 1987. Agrarian structure in central India and
Sarma, I. K. 1982. The Development of early Saiva art and the northern Deccan (c. AD 300–500): A Study in Vakataka
architecture. Delhi. inscriptions. Delhi.
Sarma, I. K. 1993. Alampur temples: Rare evidences on Siddham https://zenodo.org/communities/siddham/
constructional modes and consecrational rites. In A. V. N. Sidebotham, S. E. 1991. Ports of the Red Sea and the Arabia-India
Murthy and I. K. Sarma (eds). Sri Ramachandrika (Professor trade. In V. Begley and R. D. de Puma (eds). Rome and India:
Oruganti Ramachandaiya Festschrift): Essays on Indian The Ancient sea trade. Madison: 12–38.
archaeology, history, epigraphy, numismatics, art and Singh, J., Erenstein, O., Thorpe, W. and Varma, A. 2007.
religion. Delhi: 351–368. Crop-livestock interactions and livelihoods in the Gangetic
Sarma, I. K. 2000. Numismatic researches (critical studies from Plains of Uttar Pradesh, India. Crop-livestock interactions
excavated context). Delhi. scoping study, report 2. Research report 11. Nairobi.
Sastri, T. V. G., Katsuri Bai, M. and Rao, J. V. P. 1984. Veerapuram: A Singh, P., Lakshminarasimhan, N. P., Karthikeyan, S. and Prasanna,
Type site for cultural study in the Krishna Valley. Hyderabad. P. V. 2001. Flora of Maharashtra State, Vol. 2. Botanical survey
Schlingloff, D. 1974. Cotton-manufacture in ancient India. Journal of of India, Calcutta.
the economic and social history of the Orient 17: 81–90. Singh, U. 1994. Kings, Brāhmaṇas and temples in Orissa. Delhi.
Schoff, W. H. 1974. Periplus of the Erythrean Sea. Delhi (reprint). Singh, U. 2011. Introduction. In U. Singh (ed.). Rethinking Early
Scott, H. R. 1908. The Nasik (Jogalthembi) hoard of Nahapana’s Medieval India: a reader. New Delhi: 1–44.
coins. Journal of the Bombay branch of the Royal Asiatic Sircar D. C. 1960. Studies in the geography of ancient and Medieval
Society 22: 223–244. India. New Delhi.
Senart, E. 1905–06. The inscriptions in the caves at Nasik. Sircar, D. C. 1968. Studies in Indian coins. Delhi.
Epigraphia Indica 8: 59–96. Sircar, D. C. 1974. Studies in the political and administrative
Settar, S. and Sontheimer, G.-D. 1982. Memorial stones: a study of systems in ancient and Medieval India. Delhi.
their origin, significance and variety. Dharwad. Small, E. and Cronquist, A. 1976. A practical and natural taxonomy
Shah, U. P. 1975. Beginnings of the superstructure of Indian for Cannabis. Taxon 25(4): 405–435.
temples. In P. Chandra (ed.). Studies in Indian temple Smartt, J. 1990. Grain legumes: Evolution and genetic resources.
architecture: Papers presented at a seminar held in Varanasi, Cambridge.
1967. American Institute of Indian Studies: 80–89. Smith, A. B. and Jacobson, L. 1995. Excavations at Geduld and the
Sharma, R. K. and Mishra, O. P. 2003. Archaeological excavations in appearance of early domestic stock in Namibia. South African
central India (Madhya Pradesh and Chhatttisgarh). New Delhi. archaeological bulletin 50: 3–14.
Sharma, R. S. 1965. Indian feudalism, c. 300–1200. Calcutta. Soustiel, J. 1985. La Céramique Islamique. Fribourg.
Sharma, R. S. 1972. Decay of Gangetic towns in Gupta and Spate, O. H. K. and Learmonth, A. T. A. 1967. India and Pakistan: a
post-Gupta times. In S. Chandra (ed.). Proceedings of the general and regional geography. London, 3rd ed.
Indian History Congress, 33rd session. Muzaffarpur: 92–104. Spivak, G. C. 1988. Can the subaltern speak? In C. Nelson and L.
Sharma, R. S. 1980. Indian feudalism c. AD 300–1200. Madras, 2nd Grossberg (eds). Marxism and the interpretation of culture.
edn. Urbana, IL: 271–313.
Sharma, R. S. 1983. Material culture and social formations in Staubwasser, M., Sirocko, F., Grootes, P. M. and Segl, M. 2003.
Ancient India. Delhi. Climate change at the 4. 2 ka BP termination of the Indus
Sharma, R. S. 1987. Urban decay in India (c. 300–c. 1000). New Valley civilization and Holocene South Asian monsoon
Delhi. variability. Geophysical research letters 30: 1425–1429.
Sharma, R. S. 2001. Early Medieval Indian society: A Study in Stein, B. 1980. Peasant state and society in Medieval South India.
feudalisation. London. Oxford.
346   Reference list

Stevens, C. J. 2003. An investigation of agricultural consumption Viot, C. 2019. Domestication and varietal diversification of Old
and production models for prehistoric and Roman Britain. World cultivated cottons (Gossypium sp.) in the Antiquity.
Environmental archaeology 8(1): 61–76. Revue d’ethnoécologie [Website, posted 30-6-2019, consulted
Strabo. 1930. Geography. The Geography of Strabo, vol. vii. The 2-1-2020]. URL : http://journals.openedition.org/ethnoe-
Loeb classical library. T. E. Page, E. Capps and W. H. D. Rouse cologie/4404 ; DOI : 10.4000/ethnoecologie.4404.
(eds). London. Vishnu-Mittre. 1961. Plant economy in ancient Navdatoli-
Strouhal E. 1984. Wadi Qitna and Kalabsha-South. Prague. Maheshwar. In Technical report on archaeological remains.
Subbarao, B. 1953. Baroda through the ages. Baroda. Pune: 13–52.
Sutta Nipata = K. R. Norman. The Group of Discourses. Pali Text Vishnu-Mittre. 1968. Kaundinyapura plant economy in pre-historic
Society. Oxford, 2001, 2nd edn. and historic times. In M. G. Dikshit (ed.). Excavations at
Tadulingam, C. and Venatanarayana, G. 1985. A Handbook of some Kaundinyapura. Bombay: 140–147.
South Indian weeds. Delhi. Vishnu-Mittre. 1971. Ancient plant economy at Hallur. In M.
Tengberg, M. 1999. Crop husbandry at Miri Qalat, Makran, SW S. Nagaraja Rao (ed.), The Protohistoric cultures of the
Pakistan (4000–2000 B.C.). Vegetation history and archae- Tungabhadra Valley. Dharwad University: 125–130.
obotany 8: 3–12. Vishnu-Mittre. 1990. Plant remains. In J. P. Joshi (ed.), Excavations
Tewari, R., Srivastava, R. K., Saraswat, K. S. and Singh, K. K. 2000. at Surkotada. Memoirs of the Archaeological Survey of India.
Excavations at Malhar, district Chandauli (U. P.) 1999: A New Delhi: 388–392.
Preliminary report. Pragdhara: Journal of the Uttar Pradesh Vishnu-Mittre and Gupta, H. P. 1968a. Ancient plant economy at
State Archaeological Organisation 10: 69–98. Paunar, Maharashtra. In S. B. Deo and M. K. Dhavalikar (eds).
Thanheiser, U. 2002. Roman agriculture and gardening in Egypt as Paunar excavations 1967. Nagpur: 128–134.
seen from Kellis. In C. A. Hope and G. E. Bowen (eds). Dakhleh Vishnu-Mittre and Gupta, H. P. 1968b. Plant remains from Ancient
Oasis Project: Preliminary reports on the 1994–1995 to Bhatkuli, district Amraoti, Maharashtra. Puratattva 2: 21–22.
1998–1999 field seasons. Dakhleh Oasis Project: monograph Vishnu-Mittre, Prakash, U. and Awasthi, N. 1971. Ancient plant
11. Oxford and Oakville: 299–310. economy at Ter, Maharashtra. Geophytology 1(2): 170–177.
Thapar, B. K. 1967. Prakash 1955: A Chalcolithic site in the Tapti Vishnu-Mittre and Savithri, R. 1979. Palaeobotanical and pollen
Valley. Ancient India 20/21: 4–177. analytical investigations. Indian archaeology 1974–1975: a
Thapar, R. 1968. Interpretations of ancient Indian history. History review: 78–81.
and theory 7(3): 318–335. Vishnu-Mittre, Sharma, A. and Chanchala, S. 1984. Palaeobotanical
Thapur, R. 1984. From Lineage to state: Social formations in the and pollen analytical investigations. Indian Archaeology
mid-first millennium B.C. in the Ganga Valley. Bombay. 1981–82: a review: 105–106.
Reprinted in R. Thapar. 2000. History and beyond. Oxford. Vishnu-Mittre, Sharma, A. and Chanchala, S. 1986. Ancient
Tisdell, C. and Gali, J. 1999. Trends and developments in India’s plant economy at Daimabad. In S. A. Sali (ed.). Daimabad
livestock industry. University of Queensland School of 1976–1979. New Delhi: 588–626.
Economics, economics, ecology and environment working Wakankar, V.S. 1967. Kayatha excavation number. The Vikram:
paper 31. Journal of Vikram University, Ujjain. Ujjain.
Todywalla/Hurmuz Kaus. 2004. Todywalla Auctions catalogue Walshaw, S. C. 2005. Swahili urbanization, trade, and food
No. 8: Coins from the Hurmuz Kaus Collection, Hyderabad. production: botanical perspectives from Pemba Island,
Mumbai. Tanzania, A. D. 600–1500. Unpublished PhD Dissertation,
Tomber, R. 2007. Rome and Mesopotamia: Importers into India in Washington University, St. Louis.
the first millennium AD. Antiquity 81, 972–988. Watson, J. 1979. The estimation of the relative frequencies of
Tomooka, N., Vaughan, D. A., Moss, H. and Maxted, N. 2003. The mammalian species. Journal of archaeological science 6:
Asian Vigna: Genus Vigna subgenus Ceratotropis genetic 127–137.
resources. The Netherlands. Watt, G. 1889–1893. A Dictionary of the economic products of India.
Vaillancourt, R. E. and Weeden, N. F. 1992. Chloroplast DNA 7 vols. London.
polymorphism suggests Nigerian center of domestication Weber, S. A. 1991. Plants and Harappan subsistence. An Example of
for the cowpea, Vigna unguiculata, Leguminosae. American stability and change from Rojdi. New Delhi.
journal of botany 79: 1194–1199. Weber, S. A. 1998. Out of Africa: The impact of millets in South
van Schrötter, F. 1930. Wörterbuch der Münzkunde. Berlin. Asia. Current anthropology 39: 267–274.
Vasudeva Hindi = Vasudevhiṇḍi: a narrative text in archaic Prakrit. Weber, S. A. 1999. Seeds of urbanism: Paleoethnobotany and the
M. Caturvijaya and M. Punyavijaya (eds). Bhavnagar, 1930–31. Indus civilization. Antiquity 73: 813–826.
Vats, M. S. 1940. Excavations at Harappa. I. Calcutta. Weber, S. A. 2001. Seeds of urbanism revisited. Antiquity 75:
Venkateshaiah, S. V., Varaprasada Rao, J. and Visweswara, M. 413–414.
V. 2006. Constructional modes and consecrational rites of Weber, S. A. 2003. Archaeobotany at Harappa: Indications
a Rashtrakuta temple, Pattadakal, Karnataka. In P. Chenna for change. In S. A. Weber and W. R. Belcher (eds). Indus
Reddy (ed.). Mahasenasiri: Riches of Indian archaeological ethnobiology. New perspectives from the field. Lanham:
and cultural studies (a felicitation volume in honour of Dr. I. K. 175–198.
Sarma). 2 vols. Delhi: 335–344. Weber, S. A. and Fuller, D. Q. 2008. Millets and their role in early
Verdcourt, B. 1971. Phaeseoleae. In E. Milne-Redhead and R. M. agriculture, Pragdhara: Journal of the Uttar Pradesh State
Polhill (eds). Flora of tropical East Africa, Legumonisae (Part Archaeological Organisation 18: 69–90.
4), Papilionidae (2). London.
 Reference list   347

Wendel, J. F. 1995. Cotton. In J. Smartt and N. W. Simmonds (eds). Willcox, G. 1992. Some differences between crops of Near Eastern
Evolution of crop plants. Longman: 358–366 (2nd ed). origin and those from the tropics. In C. Jarrige (ed.). South
Wheeler, R. E. M. 1954. Archaeology from the Earth. Oxford. Asian archaeology 1989. Madison: 291–299.
Wheeler, R. E. M. 1955. Still digging. London. Willis, M. 2009. The Archaeology of Hindu ritual: Temples and the
Wheeler, R. E. M. 1962. Charsada: A Metropolis of the north-west establishment of the Gods. Cambridge.
frontier. Oxford. Wilson, L. M. 1933. Ancient textiles from Egypt in the University of
Whitehouse, D. and Williamson, A. 1973. Sasanian maritime trade. Michigan collection. Ann Arbor, Michigan.
Iran 11: 29–49. Yazdani, G. 1960. Early history of the Deccan. Oxford.
Wild, J. P. and Wild, F. C. 1998. The textiles. In S. E. Sidebotham and Yusuf, S. 1938. Paithan excavations. Indian art and letters XII, 2:
W. Z. Wendrich (eds). Berenike 96: Report of the excavations at 1–7.
Berenike (Egyptian Red Sea coast) and the survey of the Eastern Yusuf, S. 1939. Appendix C: Paithan excavations. Annual report of
Desert. Leiden: 221–236. the archaeological department of His Exalted Highness the
Wild, J. P. and Wild, F. C. 2001. Sails from the Roman port at Nizam’s dominions. Calcutta: 39–42, pls XV–XX.
Berenike, Egypt. The International journal of nautical Zeder, M. A. 1996. The role of pigs in Near Eastern subsistence: A
archaeology 30: 211–220. view from the southern Levant. In J. D. Seger (ed.). Retrieving
Wild, J. P. and Wild, F. C. 2005. Rome and India: early Indian cotton the past: Essays on archaeological research and methodology
textiles from Berenike, Red Sea coast of Egypt. In R. Barnes in honor of Gus W. Van Beek. Winona Lake: 297–312.
(ed.), Textiles in Indian Ocean studies. London: 11–16. Zhou, Z. and Bathelomew, B. 2003. Cannabaceae. In Zhengyi Wu,
Wild, J. P. and Wild, F. 2008. Early Indian cotton textiles from P. H. Raven and Deyuan Hong (eds). Flora of China 5. St. Louis:
Berenike. In E. M. Raven (ed.). South Asian Archaeology 1999. 74–75.
Groningen: 229–233. Zohary, D. and Hopf, M. 2000. Domestication of plants in the Old
Wilkinson, K. and Stevens, C. 2003. Environmental archaeology: World. Oxford, 3rd edn.
Approaches, techniques and applications. Stroud.
Appendix I
List of recorded contexts

Context Trench Type Description Date Trench Period Bone Sherds

phase (approx)

0 All Virtual context used for surface finds and other 10/04/00
unstratified finds.
1 C Topsoil Pottery and tile inclusions. 17/03/96
2 C Layer Lime mortar deposit from ring-well construction, 17/03/96
and possible cap of context 3.
3 C Layer Re-deposited erosional slump deposit. 17/03/96
4 C Layer Lime mortar deposit, similar to context 2. 16/03/96
5 C Fill Fill of cut 7. Very fine, with occasional large frag- 17/03/96
ments of tile and stone limited to sides and very
bottom of pit.
6 C Fill Fill of animal burrows. No cut number. 22/02/96
7 C Cut Cut for fill 5. Large sub-circular pit, steep sided 17/03/96
with flat bottom.
8 C Layer Fill of cut 19. 18/03/96
9 C Fill Fill of cut 22. Occasional medium pebble and tile 19/03/96
fragments in base.
10 C Layer Fill of cut 21. Upper fill, representing disuse phase 18/03/96
of feature.
11 C Fill Fill of ring-well. 19/03/96
12 C Fill Fill of cut 23. 19/03/96
13 C Fill Fill of cut 24. 19/03/96
14 C Fill Fill of cut 18. Possible post-hole, very similar to 19/03/96
Fill 10.
15 C Fill Fill of construction trench of ring-well. Disconti- 19/03/96
nuous in places, but located between mortar of
context 2 and wall of context 17.
16 C Layer Inclusions of pottery and tile. 19/03/96
17 C Wall Brick wall of a ring-well. 19/03/96
18 C Cut Cut for fill 14. Sub-circular, steep-sided cut from 19/03/96
the surface of Context 16.
19 C Cut Cut for fill 8. SE portion of a steep-sided sub-circu- 19/03/96
lar cut. Too deep to excavate completely.
20 C Fill Post-hole base and packing of cut 21. Large frag- 19/03/96
ments of brick and stone packed around a central
circular boulder.
21 C Cut Post-hole cut with two fills. Upper fill 10 repre- 19/03/96
sents disuse of feature. Lower fill 20 is foundation
of post.
22 C Cut Shallow cut of fill 9. Not fully exposed. 19/03/96
23 C Fill Lower silt in-washing fill of cut 24. 19/03/96
24 C Cut Cut of shallow pit. Not fully exposed in plan, but 19/03/96
similar to cuts 7 and 28.

https://doi.org/10.1515/9783110653540-017
 List of recorded contexts   349

Context Trench Type Description Date Trench Period Bone Sherds

phase (approx)

25 C Layer Ash lens in layer 16. 21/03/96

26 C Fill Fill of cut 27. Possible organic dump, including lots 22/03/96
of charcoal soot and flecks.
27 C Cut Very shallow cut, almost on top of context 16, 22/03/96
rather than into it. Extensive site horizon below.
28 C Cut Cut of fill 13. Cut into context 16, but complete 22/03/96
plan unknown.
29 C Layer Mid-brown fine/medium sandy silt. Frequent 23/03/96
small/medium river pebbles.
30 C Fill Fills of cut 33. Mixed contexts of ash and clayey- 25/03/96
silt.
31 C Layer Rammed earth floor. One of several occupation/ 24/03/96
dereliction deposits in connection with context 32.
32 C Wall Single course of broken brick fragments, aligned 24/03/96
on a slight curve East-West. Faced on southern
side.
33 C Cut Cut of fill 30. Cuts into context 34. 25/03/96
34 C Layer Compact silt/clay with frequent large pebbles, 25/03/96
brick, tile and flecks of lime or burnt shell. Not
homogenous distribution.
35 C Drain Brick drain. Two parallel lines of bricks on edge, 25/03/96
with single line lying flat to form base, packed in
by tile and brick rubble.
36 C Drain? Two and a half bricks, roughly aligned with drain 25/03/96
35. May represent cap or channel bottom of drain.
37 C Layer Sandy silt deposit. Small stone flecks, occasional 25/03/96
burnt bone/lime, charcoal and brick rubble.
38 C Layer Sandy silt deposit. Frequent brick rubble, stone 26/03/96
fragments and charcoal, but no homogeneity of
distribution.
40 C Layer Sandy silt, of varying composition, compaction, 27/03/96
and inclusions, but no clear differentiation.
41 C Not used Last context number from PTN96.
42 C Topsoil Soil wash accumulation across low level depres- 05/03/97
sion. Baulk sample.
43 C Layer Latest make-up soil layer. Possible last packed 05/03/97
earth floor, but very disturbed by mixing and wash.
Baulk sample.
44 C Layer Compact silt soil (baulk sample). 06/03/97
45 C Layer Compact silt soil (baulk sample). 06/03/97
46 C Topsoil Baulk sample. 06/03/97
47 C Layer Compact silt soil. 06/03/97
48 C Layer Compact silt soil. 06/03/97
49 C Topsoil No record kept. 08/03/97
50 C Layer No record kept. 08/03/97
51 C Layer No record kept. 08/03/97
52 C Layer No record kept. 08/03/97
53 C Layer No record kept. 09/03/97
350   Appendix I

Context Trench Type Description Date Trench Period Bone Sherds

phase (approx)

54 C Topsoil Equal to context 56. 09/03/97

55 C Layer Compact silt layer. 09/03/97
56 C Topsoil Equal to context 54. 09/03/97
57 C Layer Equal to context 58. 10/03/97
58 C Layer Equal to context 57. 10/03/97
59 C Layer Soil layer in baulk. Compact and silty, with pebble 10/03/97
inclusions and sparse cultural material.
60 C Fill Pit fill. Coarse chalky inclusions, stones and 11/03/97
pottery.
61 C Fill Rubble pit fill. 11/03/97
62 C Layer Equal to context 63. 11/03/97
63 C Layer Equal to context 62. 11/03/97
64 C Fill Pit (top fill). 11/03/97
65 C Layer Uncertain. Layer next to wall (?), north. 12/03/97
66 C Layer Uncertain. Layer next to wall (?), south. Equal to 12/03/97
context 71.
67 C ? No record kept. 12/03/97
68 C Topsoil Topsoil. 12/03/97
69 C Fill Lower pit fill. 12/03/97
70 C Layer No record kept. 12/03/97
71 C Layer Equal to context 66. 13/03/97
72 C Layer Compact silt soil. 13/03/97
73 C Fill Fill of possible post-hole 75. 13/03/97
74 C Fill Fill, and layer of lower pit. 13/03/97
75 C Cut Cut of possible post-hole. 13/03/97
76 C Fill Lower pit, possibly of a ditch. 13/03/97
77 C Fill Fill of cut 78. 13/03/97
78 C Cut Cut of fill 77. Pit or post-hole. 13/03/97
79 C Layer No record kept. 15/03/97
80 C Layer Rubbly layer. 15/03/97
81 C Layer Layer/wall? 15/03/97
82 C Layer Layer/wall? 15/03/97
83 C Layer No record kept. 16/03/97
84 C Layer No record kept. 16/03/97
85 C Fill Fill/layer. 16/03/97
86 C Layer No record kept. 16/03/97
87 C Layer Lowest ‘definite archaeology’ layer. 17/03/97
88 C Layer No record kept. 17/03/97
89 C Layer No record kept. 17/03/97
90 C Layer No record kept. 17/03/97
91 C Layer No record kept. 17/03/97
92 C Layer No record kept. 17/03/97
93 C Layer Semi-natural layer including ceramics. 17/03/97
94 C Layer No record kept. 18/03/97
95 C Layer Silt-like soil, mixed with ash deposits. 18/03/97

96 C Layer Soil mixed with silt. 18/03/97

 List of recorded contexts   351

Context Trench Type Description Date Trench Period Bone Sherds

phase (approx)

97 C Layer No record kept. 18/03/97

98 C Layer Lowest. 18/03/97
99 C Topsoil No record kept. 18/03/97
100 C Layer Mixed rubble layer. 18/03/97
101 C Layer Compact silt. 19/03/97
102 C Fill Fill of post-hole. Infrequent charcoal, moderate 19/03/97
organic material, pebbles and grit.
103 C Layer Natural? 19/03/97
104 C Wall Stone wall, aligned East-West. 20/03/97
105 C Layer No record kept. 20/03/97
106 C Fill Upper fill of collapsed ring-well. May represent 20/03/97
same context as 100.
107 C Layer Red burnt clay. 20/03/97
108 C Fill Fill of pit (against east-facing section). 21/03/97
109 C Ring- Terracotta ring-well pieces. 21/03/97
well
110 C Fill Fill of cut 111. Upper fill of pit. 21/03/97
111 C Cut Cut of narrow, deep pit. Squarish in plan, with 21/03/97
rounded corners. Extends under baulk and not fully
excavated. Possibly related to ring-well construc-
tion.
112 C Fill Rubbly top fill. 21/03/97
113 C Fill Fill of cut 114. 21/03/97
114 C Cut Cut of 113. Cut into 117. 21/03/97
115 C Fill Top fill of cut 119. 21/03/97
116 C Layer Grey soil mixed with ash and flecks of charcoal. 21/03/97
Large number of potsherds.
117 C Layer Grey/brown compact sandy silt. Occasional lime/ 21/03/97
mortar and pebbles.
118 C Fill Middle fill of cut 119. Soft, red sandy silt. Occa- 22/03/97
sional mortar, charcoal and shell flecks, brick
fragments and pebbles.
119 C Cut Cut of fills 115, 118 and 121. Deep pit with lip at 22/03/97
top. Cut into natural.
120 C Layer Compact sandy, silty clay. Pale brown, with occa- 22/03/97
sional pebbles, small stones and very occasional
mica.
121 C Fill Bottom fill of cut 119. Soft red/brown sandy silt, 22/03/97
occasional pebbles, stone flecks, grit, mortar and
shell flecks.
122 C Layer Dark brown. 22/03/97
123 C Layer No record kept. 22/03/97
124 C Cut Construction cut for terracotta ring-well. Oval plan, 22/03/97
due to collapse of ring-well. May have been cut
from later layer.
125 C Layer Cut by 124. 23/03/97
126 C Layer Compact soil, mixed with ash, charcoal, brick bats 23/03/97
and pot-sherds.
352   Appendix I

Context Trench Type Description Date Trench Period Bone Sherds

phase (approx)

127 C Layer Compact yellow. 23/03/97

128 C Layer Lowest north of terracotta ring-well. 23/03/97
129 C Fill Mixed pit fill of cut 136. 23/03/97
130 C Fill Lower fill of cut 111. Pottery dump, with seepage 23/03/97
of clayey silt into the deposit.
131 C Layer South of terracotta ring-well and 111. 24/03/97
132 C Fill Fill of construction cut for terracotta ring-well. 24/03/97
133 C Fill Lower fill of ring-well. Possibly the original fill, 24/03/97
before collapse of the upper levels of ring-well.
134 C Layer South of ring-well (lowest). 24/03/97
135 C Fill Lower fill of cut 136. 25/03/97
136 C Cut Cut of fills 129 and 136. 25/03/97
137 C Cut Cut of fills 60 and 69 ‘pit’? 25/03/97
138 C Cut Cut of fills 61 and 64. 25/03/97
139 C Cut Cut of fill 108. 25/03/97
140 C Cut Cut for fill 85. 25/03/97
141 C Layer Number allocated from section.
142 C Layer Number allocated from section.
143 C Layer Number allocated from section.
144 C Layer Number allocated from section.
145 C Layer Number allocated from section.
146 C Layer Number allocated from section.
147 C Layer Number allocated from section.
148 C Layer Number allocated from section.
149 C Layer Number allocated from section.
150 C Layer Number allocated from section.
151 C Fill Number allocated from section.
152 C Cut Number allocated from section.
153 C Layer Number allocated from section.
154 C Layer Number allocated from section.
155 C Fill Number allocated from section.
156 C Fill Number allocated from section.
157 C Fill Number allocated from section.
158 C Fill Number allocated from section.
159 C Fill Number allocated from section.
160 C Fill Number allocated from section.
161 C Cut Number allocated from section.
162 C Layer Number allocated from section.
163 C Fill Number allocated from section.
164 C Cut Number allocated from section.
165 C Layer Number allocated from section.
166 C Layer Number allocated from section.
167 C Layer Number allocated from section.
168 C Layer Number allocated from section.
169 C Layer Number allocated from section.
 List of recorded contexts   353

Context Trench Type Description Date Trench Period Bone Sherds

phase (approx)

170 C Fill Number allocated from section.

171 C Fill Number allocated from section.
172 C Fill Number allocated from section.
173 C Fill Number allocated from section.
174 C Cut Number allocated from section.
175 C Layer Number allocated from section.
176 C Floor Number allocated from section.
177 C Fill/layer Number allocated from section.
178 C Fill/layer Number allocated from section.
179 C Fill/layer Number allocated from section.
180 C Fill/layer Number allocated from section.
181 C Fill/layer Number allocated from section.
182 C Cut Number allocated from section.
183 C Layer Number allocated from section.
184 C Layer Number allocated from section.
185 C Floor Number allocated from section.
186 C Layer Number allocated from section. Y
187 C Layer Number allocated from section.
188 C Layer Number allocated from section.
189 C Layer Number allocated from section.
190 C Layer Number allocated from section.
191 C Layer Number allocated from section.
192 C Layer Number allocated from section.
193 C Layer Number allocated from section.
194 C Layer Number allocated from section.
195 C Fill Number allocated from section.
196 C Fill Number allocated from section.
197 C Cut Number allocated from section.
198 C Layer Number allocated from section.
199 C Floor Number allocated from section.
200 C Layer Number allocated from section.
201 C Floor Number allocated from section.
202 C Layer Number allocated from section.
203 C Layer Number allocated from section. Probably = 179
(same layer).
204-299 Not Numbers not used.
used.
300 A:TP1 Layer Trample/backfill in bottom of Yusuf’s trench. 23/02/98 4 Y 258
301 A:TP1 Layer Layer of brick rubble and silt below 300. 24/02/98 4 Y 579
302 A:S of S Rubble Brick jelly patches abutting the southern wall of 24/02/98 4 Y 219
Temple the South Temple.
303 A:S of S Rubble Silt and rubble to south of the South Temple. 24/02/98 4 Y 2,088
Temple
304 A:TP1 Rubble Bricks, possible in situ wall collapse. 24/02/98 4
305 A:TP1 Layer Trample under wall collapse 304. 24/02/98 3 Y 583
306 A:TP1 Layer Deposit next to west wall of the North Temple; 24/02/98 3-temple 125
possibly top fill of foundation trench.
354   Appendix I

Context Trench Type Description Date Trench Period Bone Sherds

phase (approx)

307 A:TP1 Layer Soapy trampled surface under rubble and brick 25/02/98 3 Y 263
collapse.
308 A:TP1 Layer Silty layer, may be contaminated by fill of construc- 25/02/98 TP1-4 3 Y 702
tion trench.
309 A:TP2 Layer Silty layer between two temples. 28/02/98 344
310 A:TP1 Fill Fill of construction trench 311. 28/02/98 3-temple Y 56
311 A:TP1 Cut Cut of construction trench for the South Temple 28/02/98 3-temple
(west end).
312 A:TP1 Layer Layer cut by construction trench, same as 308 but 28/02/98 TP1-4 3 Y 576
no contamination.
313 A:TP1 Fill Fill of construction trench 311. 01/03/98 3-temple 17
314 A:TP2 Layer Loose grey silty clay; no finds. 01/03/98
315 A:TP2 Layer Brick rich layer, may be back-fill post 1967 conser- 01/03/98 Y 359
vation.
316 A:TP1 Layer Whitish compact layer, possible floor in north-west 02/03/98 3 Y 240
corner of TP1.
317 A:TP2 Layer Deposit of brick bats, a few pot-sherds and brick 02/03/98 Y 451
jelly.
318 A:TP1 Layer Soft ashy layer. Not excavated 1998. 02/03/98 3
319 A:TP2 Layer Compact, dense grey clay layer – mandapa 02/03/98 99
surface?
320 A:TP9 Layer Silt accumulation in bottom of Yusuf’s trench. 04/03/98 4 0
321 A:W Layer Silt accumulation in Yusuf’s trench to west of both 05/03/98 4 17
Area temples.
322 A:N Rubble Rubble or crude wall on top of east wall, North 05/03/98
Temple Temple.
323 A:TP2 Fill Loose, granular, blackish fill adjacent to south wall 08/03/98 Y 2,512
of the North Temple filling 388.
324 A:NWNT Layer Dark brown, stoney, clayey deposit north of the 07/03/98 4?
North Temple.
325 A:NWNT Rubble Rubble and stone linear deposit, probably post- 07/03/98 4?
Yusuf.
326 A:NWNT, Fill Upper silty fill of robber trench 329. 07/03/98 4 366
TP9
327 A:NWNT Fill Fill of robber trench 377. 07/03/98 4? Y 394
328 A:NWNT, Fill Lower fill of robber trench 329 (west half). 07/03/98 4 Y 1,002
TP9
329 A:NWNT, Cut Cut of robber trench of north wall of the North 07/03/98 4
TP9 Temple (fills 326, 328 and 332).
330 A:TP5 Rubble Rubble and brick pavement within the South 08/03/98 4 Y 939
Temple.
331 A:NWNT Fill Fill of robber trench 341 between remaining walls. 08/03/98 4? Y 1,264
332 A:NWNT Fill Fill of robber trench 396. 08/03/98 4? 137
333 A:W Fill Fill of cut 393 within 323, with compact pale grey 09/03/98
Area earth and brick bats.
334 A:TP5 Fill Fill of pit 335 cut into north-west corner of the 09/03/98 4 Y 650
South Temple.
335 A:TP5 Cut Cut of pit in north-west corner of the South Temple 09/03/98 4
(fill 334).
336 A:NWNT Fill Fill of robber trench 337. 09/03/98 4? 280
 List of recorded contexts   355

Context Trench Type Description Date Trench Period Bone Sherds

phase (approx)

337 A:NWNT Cut Cut of robber trench in north-east corner of the 09/03/98 4?
North Temple (fill 336).
338 A:W Fill Brick jelly fill overlying 323 and cut 339. 09/03/98 3 59
Area
339 A:W Cut Cut filled by 338 and 640. 09/03/98 3
Area,
TP3
340 A:N Founda- Phase I foundation of the North Temple. 10/03/98 5
Temple tion
341 A:NWNT Cut Cut of robber trench in north wall of the North 10/03/98 4?
Temple (fill 331).
342 A:TP6 Rubble Compact rubble packing along the south of the 10/03/98 3-temple Y 189
South Temple.
343 A:TP6 Fill Fill of pit 369 cut below 320. 11/03/98 Y 431
344 A:S Wall Phase III east wall of the South Temple. 11/03/98
Temple
345 A:S Wall Phase IV extension to east end of north wall of the 11/03/98
Temple South Temple.
346 A:S Surface Surface to east of the South Temple (not exca- 11/03/98
Temple vated).
347 A:S Cut Foundation cut for wall 352, South Temple. 11/03/98
Temple
348 A:EEST Fill Packing of foundation cut 347 after construction 11/03/98 3 Y 1,599
of wall 352.
349 A:S Wall Short wall abutting wall 345 to east of the South 11/03/98
Temple Temple (N side).
350 A:EEST Fill Fill of robber trench 364. 11/03/98 3 Y 406
351 A:S Wall Number originally given to part of wall 560. Equals 11/03/98
Temple 560.
352 A:N Wall Phase IV and V east wall of the South Temple. 11/03/98
Temple
353 A:S Floor Brick floor to east of wall 352. 11/03/98
Temple
354 A:S Wall Small, crude brick alignment abutting east end of 11/03/98
Temple wall 349.
355 A:EEST Stone Unexcavated stones to east of the South Temple – 11/03/98 3
may be later walling or flooring.
356 A:EEST Fill Fill under 355, not excavated. 11/03/98 3
357 A:EEST Fill Fill of small post-hole 358, east of the South 11/03/98 3 Y 251
Temple.
358 A:EEST Cut Cut of small post-hole, east of the South Temple. 11/03/98 3
359 A:S Wall Phase IV extension to east of south wall of the 11/03/98
Temple South Temple.
360 A:S Wall Crude stone wall on top of wall 352, east of the 11/03/98
Temple South Temple.
361 A:S Wall Number originally given to part of wall 560. Equals 11/03/98
Temple 560.
362 A:EEST Layer Layer under wall 354 (not excavated). 11/03/98 3
363 A:EEST Layer Layer under 353 (not excavated). 11/03/98 3
364 A:EEST Cut Robber cut in north end of wall 352 of the South 11/03/98 3
Temple (fill 350).
356   Appendix I

Context Trench Type Description Date Trench Period Bone Sherds

phase (approx)

365 A:NWNT Layer Whitish/light grey silt layer, north-west corner 11/03/98 4? Y 91
Trench A.
366 A:S Layer Layer beneath floor 330 inside the South Temple. 12/03/98 Y 1,631
Temple
367 A:NWNT Fill Foundation cut fill for brick pedestal 375 (with 12/03/98 4? Y 107
368).
368 A:NWNT Cut Foundation cut fill for brick pedestal 375 (with 12/03/98 4?
367).
369 A:TP6 Cut Cut of a pit below 320 (fill 343). 12/03/98
370 A:TP6 Layer Hard greyish deposit along the south wall of the 14/03/98 3 Y 1,836
South Temple, fills cut 371.
371 A:TP6 Cut Cut filled by 370. 14/03/98
372 A:TP6 Cut Cut or re-cut of foundation trench of south wall of 14/03/98 3-temple
the South Temple.
373 A:TP6 Surface Hard surface south of the South Temple. 14/03/98 3 Y 1,349
374 A:EEST Layer Silt deposit associated with floor 353. 14/03/98 3
375 A:NWNT Wall Brick pedestal in the north wall of the North 14/03/98 4?
Temple with mortar pointing; possibly 1967 recon-
struction.
376 A:NWNT Fill Fill of cut 377. 14/03/98 4 Y 3,544
377 A:NWNT, Cut Robber trench cut in north wall of the North 14/03/98 4
TP9 Temple.
378 A:NWNT Layer Brick jelly patch, north of the North Temple. 14/03/98 4?
379 A:NWNT Layer Brick jelly and rubble layer north of cut 377. 14/03/98 4? Y 280
380 A:NWNT Cut Cut filled by deposit 365. 14/03/98 4?
381 A:NWNT Fill Fill of cut 382, compact dark brown. 14/03/98 4? Y 194
382 A:NWNT Cut Foundation cut for wall 514. Same as 515. Cuts 14/03/98 4?
through eastern end of 398.
383 A:N Wall Thin Phase V partition wall in west of mandapa, 14/03/98
Temple North Temple (equals 544).
384 A:N Wall Internal cross wall above 383, North Temple. Pos- 14/03/98
Temple sibly a 1967 reconstruction of a late internal wall.
385 A:NWNT Fill Loose fill, probably robber fill (similar to 331). 14/03/98 4?
386 A:NWNT Cut Robber cut, continuation of cut 377. 14/03/98 4? Y 971
387 A:NWNT Rubble Medium-sized brick/rubble deposit within 379. 14/03/98 4?
388 A:TP2 Cut Cut filled by 323. 14/03/98
389 A:TP2 Surface Hard surface, cut by 388. 14/03/98
390 A:TP2 Fill Fill of foundation trench 391 of south wall of the 14/03/98
North Temple.
391 A:TP2 Cut Cut of foundation trench of south wall of the North 14/03/98
Temple (filled by 390).
392 A:TP2 Layer Brick jelly deposit below fill 323. 14/03/98
393 A:TP2 Cut Cut filled with 333. 14/03/98
394 A:NWNT Layer Layer under brick jelly rubble, north-west corner of 14/03/98 4? 278
Trench A.
395 A:TP9 Fill Loose fill below walls 383/384. May be fill deposi- 15/03/98 4 0
ted during 1967 reconstruction.
 List of recorded contexts   357

Context Trench Type Description Date Trench Period Bone Sherds

phase (approx)

396 A:NWNT Cut Cut of robber trench with fill 332 east of line of 15/03/98 4?
bricks.
397 A:N Wall Upper rubble wall on the south wall of the North 16/03/98
Temple Temple – probably resulting from 1967 restora-
tion.
398 A:TP9 Layer Top layer in interior of the North Temple; re-num- 16/03/98 62
bered 519.
399 A:NWNT Fill Foundation packing for wall 502. 16/03/98 4? 217
400 B Layer Surface. 12/03/98 B-58 4 Y 2,018
401 B Wall Late structure on eastern side of trench. 12/03/98 B-56 4
402 B Layer Dump material deposit against 401. 12/03/98 B-57 4 Y 3,626
403 B Pillar Pillar base – south. 14/03/98 B-55 4
404 B Pillar Pillar base – north. 14/03/98 B-55 4
405 B Fill Fill of 406. 15/03/98 B-57 4
406 B Cut Cut of pit for pot filled by 405. 16/03/98 B-57 4 120
407 B Pillar Fill of 408. 16/03/98 B-55 4
408 B Cut Cut of northern side pillar base foundation (filled 16/03/98 B-55 4 34
407).
409 B Cut Cut for southern side pillar base foundation (filled 16/03/98 B-55 4
410).
410 B Fill Fill of 409. 16/03/98 B-55 4 0
411 B Cut Cut for north-eastern side pillar base along wall. 16/03/98 B-55 4
412 B Fill Fill of 411. 16/03/98 B-55 4
413 B Floor Floor level into which pillar bases 403, 404 are cut. 16/03/98 B-54 4 Y 625
414 B Layer Deposit inside structure 401. 16/03/98 B-57 4 Y 315
415 B Fill Fill of 416. 17/03/98 B-53 4 Y 159
416 B Cut Cut along northern section of trench, filled by 415. 17/03/98 B-53 4
417 B Floor Floor level into which pit 416 is cut. 17/03/98 B-52 4 Y 470
418 B Fill Fill of pit 419. 17/03/98 B-51 4 Y 171
419 B Cut Cut of pit below floor level 417 (filled by 418). 17/03/98 B-51 4
420 B Fill Fill of pit 421. 17/03/98 B-51 4 Y 89
421 B Cut Cut of pit in north-western corner of trench (filled 17/03/98 B-51 4
by 420).
422 B Fill Fill of pit 423. 17/03/98 B-55 4 24
423 B Cut Cut of pit in south-west corner (filled by 422). 17/03/98 B-55 4
424 B Fill Fill of 425. 17/03/98 B-55 4
425 B Cut Foundation cut for south-eastern pillar base (filled 17/03/98 B-55 4
by 424).
426 B Fill Loose fill, north-eastern corner of trench. 18/03/98 B-53 4 15
427 B Floor Floor in which pits 419 and 421 are cut. 18/03/98 B-50 4 Y 378
428 B Layer Ash deposit on western side of trench. 18/03/98 B-49 4 Y 32
429 B Floor Compact floor deposit covering whole trench. 18/03/98 B-47 4 Y 510
430 B Fill Fill of pit 431. 18/03/98 B-48 4 Y 188
358   Appendix I

Context Trench Type Description Date Trench Period Bone Sherds

phase (approx)

431 B Cut Pit cut into 429 (filled by 430). 18/03/98 B-48 4
432 B Fill Fill of 433. 18/03/98 B-48 4 Y 385
433 B Cut Cut of pit into floor 429 (filled by 432). 18/03/98 B-48 4
434 B Layer Stone deposit overlying pit 436. 19/03/98 B-48 4
435 B Fill Fill of pit 436. 19/03/98 B-48 4 Y 4,224
436 B Cut Pit cut into floor 429, filled by 435. 21/01/98 B-48 4
437 B Fill Compact brown deposit, north-eastern corner of 21/03/98 B-43 4 Y 202
trench.
438 B Layer Thick clay deposit with few sherds. 22/03/98 B-46 4 Y 148
439 B Floor Floor level with white plaster patches. 22/03/98 B-44 4 Y 1,370
440 B Layer Loose ashy deposit with a lot of sherds and under- 23/03/98 B-45 4 Y 729
lying red burnt deposit.
441 B Pot Large pot covered with 442. 24/03/98 B-41 2 Y 1,318
442 B Fill Mud lining of pot 441, burnt to a high temperature. 24/02/98 B-42 2 Y
443 B Layer Black burnt deposit in south-east corner of trench. 24/02/98 B-40 2 Y 182
444 B Pot Half pot in southern section, related to 441. 24/02/98 B-39 2 Y 175
445 B Pot Small pot in south-west corner next to 441. 24/03/98 B-39 2
446 B Pot Pot on the western side of pit 436. 25/03/98 B-39 2
447 B Fill Mud plaster lining of pot 446. 25/03/98 B-39 2 40
448 B Fill Fill of 449. 16/02/99 B-39 2 Y 607
449 B Cut Channel cut running from south section to cut of 17/02/99 B-39 2
pit 436 (filled by 448).
450 B Floor Floor level, loose blackish brown below 443. 17/02/99 B-38 2 Y 142
451 B Floor Compact brown yellow floor level below 450. 17/02/99 B-37 2 Y 182
452 B Floor Compact brown floor level. 18/02/99 B-36 1 645
453 B Floor Semi-compact brownish black floor level. 18/02/99 B-35 1 Y 588
454 B Fill Fill of 455. 18/02/99 B-34 1
455 B Cut Pit cut (filled by 454). 18/02/99 B-34 1
456 B Layer Compact yellow deposit in north-west corner of 20/02/99 B-33 1 69
trench.
457 B Floor Compact floor level does not extend to western 20/02/99 B-32 1 Y 196
side of trench.
458 B Floor Semi-compact floor level on western side of 20/02/99 B-30 1 Y 470
trench.
459 B Fill Fill of post-hole 460. 21/02/99 B-31 1
460 B Cut Post-hole cut, filled by 459, cutting 458. 21/02/99 B-31 1
461 B Fill Fill of post-hole 462. 21/02/99 B-31 1
462 B Cut Post-hole cut, filled by 461, cutting 458. 21/02/99 B-31 1
463 B Fill Fill of post-hole 464. 21/02/99 B-31 1
464 B Cut Post-hole cut, filled by 463, cutting 458. 21/02/99 B-31 1 Y
465 B Floor Compact floor with brownish-black patches and 22/02/99 B-29 1 Y 501
burnt clay.
 List of recorded contexts   359

Context Trench Type Description Date Trench Period Bone Sherds

phase (approx)

466 B Floor White floor level on north side of trench. 23/02/99 B-28 1
467 B Floor Floor with burnt clay lumps and charcoal. 23/02/99 B-27 1 Y 930
468 B Floor Floor level with white patches. 24/02/99 B-26 1 Y 255
469 B Fill Fill of pit 470, with a lot of pottery. 24/02/99 B-25 1 Y 2,305
470 B Cut Pit cut, filled by 469, cutting 468. 24/02/99 B-25 1
471 B Floor Floor with patches of white lime. 24/02/99 B-24 1 154
472 B Floor Floor with white lime patches. 24/02/99 B-23 1 Y 109
473 B Floor Floor level. 25/02/99 B-22 1 Y 165
474 B Floor Very compact floor level. 25/02/99 B-19 1 Y 375
475 B Fill Fill of pit 476. 25/02/99 B-21 1
476 B Cut Pit cut, filled by 475. 25/02/99 B-20 1
477 B Floor Floor level with clay and silt. 25/02/99 B-18 1 Y 312
478 B Floor Floor level into which some post-holes have been 27/02/99 B-16 1 Y 310
cut.
479 B Fill Fill of post-hole 480. 27/02/99 B-17 1
480 B Cut Post-hole cut, filled by 479, cutting floor 478. 27/02/99 B-17 1
481 B Fill Fill of pit 482. 27/02/99 B-17 1 Y 382
482 B Cut Pit cut into floor 478 (or 477?), filled by 481. 27/02/99 B-17 1
483 B Floor Semi-compact floor deposit with charcoal and 27/02/99 B-15 1 Y 272
pottery.
484 B Floor Floor with white and red patches. 28/02/99 B-14 1 Y 117
485 B Floor Thick floor deposit with some fire activity. 28/02/99 B-13 1 Y 130
486 B Floor Floor level with evidence of fire-related activity. 01/03/99 B-12 1 Y 145
487 B Floor Floor level into which two post-holes have been 01/03/99 B-10 1 Y 133
cut.
488 B Fill Fill of post-hole 489. 01/03/99 B-11 1
489 B Cut Post-hole cut into floor 487, filled by 488. 01/03/99 B-11 1
490 B Fill Fill of post-hole 491. 01/03/99 B-11 1
491 B Cut Post-hole cut into floor 487, filled by 490. 01/03/99 B-11 1
492 B Floor Floor with hearth on western side. 01/03/99 B-09 1 Y 104
493 B Floor Floor with hearth and pit on north-east corner. 04/03/99 B-08 1 Y 217
494 B Floor Compact floor, black with white plaster. 04/03/99 B-06 1 Y 301
495 B Fill Fill of pit 496. 04/03/99 B-07 1 15
496 B Cut Pit cut, filled by 495, cut into 493. 04/03/99 B-07 1
497 B Fill Charcoal-rich fill of pit 498. 04/03/99 B-07 1 Y 200
498 B Cut Pit cut into floor 494, filled with 497. 04/03/99 B-07 1
499 B Floor Compact, black floor packing on northern side of 06/03/99 B-04 1 Y 68
trench.
500 B Floor Yellow floor deposit on southern side of trench. 06/03/99 B-04 1 13
501 A:NWNT Cut Foundation cut for wall 502. 16/03/98 4?
360   Appendix I

Context Trench Type Description Date Trench Period Bone Sherds

phase (approx)

502 A:NWNT Wall Single line brick ‘wall’ within robber cut 377, 16/03/98 4?
lowest two courses probably remaining bricks of
the North Temple wall 557.
503 A:NWNT Cut Linear East-West cut, equals 396. 16/03/98 4?
504 A:NWNT Fill Fill of 503. 16/03/98 4? Y 457
505 A:NWNT Fill Lower brick jelly/rubble fill in west of cut 377 16/03/98 4? Y 401
under 376.
506 A:NWNT Fill Lower brick jelly/rubble fill in east of cut 377 under 16/03/98 4?
376.
507 A:NWNT Fill Lower robber fill of 377 in north-east corner. 16/03/98 4? Y 578
508 A:S Wall Crude brick-bat repair to gap in south wall, South 16/03/98
Temple Temple.
509 A:TP5 Wall Single course thickening of interior of south wall of 16/03/98 3
the South Temple.
510 A:TP5 Wall Interior wall in the South Temple mandapa. 16/03/98 3
511 A:TP5 Wall Interior wall in the South Temple mandapa. 16/03/98 3
512 A:TP2 Fill Fill abutting 397. 16/03/98 Y 596
513 A:South Surface Hard surface with brick bats and clay, south of the 16/03/98 Y 293
Area South Temple, below 300.
514 A:N Wall Remaining portion of original northern wall abut- 16/03/98
Temple ting shrine, North Temple.
515 A:NWNT Fill Foundation fill of cut 516 for wall 514. 16/03/98 4?
516 A:NWNT Cut Foundation trench cut for wall 514 (fill 515). 16/03/98 4?
517 A:EEST Fill Lower foundation fill for wall 352, fill of cut 347. 17/03/98 3 57
518 A:TP9 Cut Cut in interior of the North Temple (fill 519). 17/03/98 4
519 A:TP9 Fill Fill of 518. 17/03/98 4 Y 506
520 A:EEST Fill Fill of cut 521. 17/03/98 3
521 A:EEST Cut Cut against wall 345/1240, north side of the South 17/03/98 3
Temple.
522 A:EEST Fill Fill of post-hole 523. 17/03/98 3
523 A:EEST Cut Cut of post-hole against eastern side of Phase III 17/03/98 3
South Temple.
524 A:EEST Fill Fill of post-hole 525. 17/03/98 3 Y 139
525 A:EEST Cut Cut of post-hole against eastern side of Phase III 17/03/98 3
South Temple.
526 A:EEST Fill Fill of foundation cut 527. 17/03/98 3
527 A:EEST Cut Cut against wall 359/1247 south side of the South 17/03/98 3
Temple.
528 A:TP5 Wall Thickening to interior face of north wall, South 16/03/98 3
Temple.
529 A:NWNT Fill Lower fill robber trench 377 to west. 18/03/98 4? Y 1,062
530 A:S Cut Cut into north-west corner of the South Temple, 18/03/98
Temple cutting 334.
531 A:S Fill Fill of cut 530. 18/03/98
Temple
532 A:NWNT Fill Fill of circular pit 533. 18/03/98 4? Y 515
 List of recorded contexts   361

Context Trench Type Description Date Trench Period Bone Sherds

phase (approx)

533 A:NWNT Cut Cut of pit (fill 532), similar to 541. 18/03/98 4?
534 A:S Rubble Brick-bat rubble overlying south-west corner of the 18/03/98
Temple South Temple, appears to have been deliberately
placed.
535 A:NWNT Fill Lower robber fill of 377 (equal to 506). 18/03/98 4? Y 537
536 A:N Fill Fill of cut 537. 18/03/98 Y 98
Temple
537 A:N Cut Cut of north-south trench in interior of the North 18/03/98
Temple Temple, perhaps by Yusuf or 1967 restoration.
538 A:TP2 Rubble Rubble packing on the southern wall of the North 19/03/98 Y 631
Temple. May result from 1967 restoration.
539 A:N Layer Surface or compact layer in interior of the North 19/03/98
Temple Temple; may represent limit of later disturbance.
540 A:NWNT Fill Soft ashy pit fill of cut 541. 19/03/98 4? 263
541 A:NWNT Cut Truncated pit cut filled by 540, east of north-east 19/03/98 4?
corner, North Temple.
542 A:TP9 Wall Crude tile alignment in south aisle of the North 19/03/98 4
Temple. May result from 1967 restoration.
543 A:TP9 Layer Deposit under tile wall 543, south aisle, North 19/03/98 4 255
Temple.
544 A:TP9 Wall Thin Phase V partition wall in south aisle of 19/03/98
mandapa, North Temple (equals 383).
545 A:TP9 Layer Black cotton soil behind wall 544, south aisle, 19/03/98 255
North Temple (equals 551).
546 A:NWNT Fill Lower fill of cut 541. 19/03/98 4? Y 733
547 ? ? Not used. This number was added in Sept 2008 88
as 88 sherds were recorded under it in the pottery
book, probably in error.
549 A:NWNT Fill Lower brick jelly fill of cut 377. 21/03/98 4? 69
550 A:NWNT Surface Interior surface of the North Temple, probably 21/03/98 4?
originally abutting wall 514.
551 A:TP9 Layer Black cotton soil in western end of North Temple 23/03/98 3 80
mandapa.
552 A:NWNT Fill Ashy fill of pit 553. 23/03/98 4?
553 A:NWNT Cut Pit cut in north-east corner of Trench A. 23/03/98 4?
554 A:N Wall Phase I wall, North Temple. 23/03/98 3-temple
Temple
555 A:N Wall Phase I platform, North Temple. 23/03/98 3-temple
Temple,
TP9, GG
North
556 A:N Wall Phase II structure with offsets, North Temple. 23/03/98 3-temple
Temple,
GG
North
557 A:N Wall Phase III mandapa wall, North Temple. 23/03/98 3-temple
Temple,
TP9
362   Appendix I

Context Trench Type Description Date Trench Period Bone Sherds

phase (approx)

558 A:N Wall Phase IV interior wall, North Temple. 23/03/98

Temple
559 A:N Wall Rebuilding of outer face of wall 514. 23/03/98
Temple
560 A:S Wall Phase II wall of the South Temple. 23/03/98 3
Temple,
TP3,
TP5
561 A:S Wall Rebuilding of north wall, South Temple. 23/03/98 3
Temple,
TP5
562 A:S Wall Thickening of inner face of south wall, South 24/03/98
Temple Temple.
563 A:N Wall Northern lime-mortar mud levelling, ‘ghost wall’ 01/04/98
Temple under 514.
564 A:N Wall Southern lime-mortar mud levelling, ‘ghost wall’ 01/04/98
Temple under 514.
565 A:N Layer Foundation levelling under wall 514 + 557. 01/04/98
Temple
566 A:N Layer Brick jelly north of robber cut 377 (west). 02/04/98
Temple
567 A:N Layer Brick jelly north of robber cut 377 (east). 02/04/98
Temple
568 A:South Layer General post-Yusuf accumulation. 25/01/99 4 Y 743
Area
569 A:S Wall Cut Robber or excavation cut alongside south wall of 28/01/99 4
of N the North Temple.
570 A:TP5 Layer Rubble and brick bat foundation of possible later 28/01/99 4 Y 382
wall.
571 A:S Wall Fill Fill of 569. 28/01/99 4 Y 342
of N
572 A:S Wall Layer Layer cut by 569 and 585, next to north wall of the 28/01/99
of N South Temple.
573 A:S Wall Fill Fill of 569, beneath 571. 30/01/99 4 Y 684
of N
574 A:South Wall Possible wall below 568 or collapse of 578. 30/01/99 4
Area
575 A:S Wall Fill Fill of cut 576. 30/01/99 4 Y 207
of N
576 A:S Wall Cut Cut filled by 575 next to south wall of the North 30/01/99 4 Y
of N Temple.
577 A:South Cut Robber trench cut to rob wall foundation 699. 30/01/99 3
Area
578 A:South Wall Wall foundation in cut 577. 30/01/99 3 Y 669
Area
579 A:TP5 Founda- Packed stone foundation under 570. 30/01/99 4
tion
580 A:TP6 Fill Stone rubble fill of 581 with brick fragments, tiles, 31/01/99 4 Y 772
pottery.
581 A:TP6 Cut Filled by 580. May represent post-Yusuf activity. 31/01/99 4
582 A:TP5 Fill Silt packing in foundation cut 595 at southern end. 31/01/99 4 Y 266
583 A:South Fill Fill of pit/pot dump filling cut 587. 31/01/99 3 370
Area
584 A:S Wall Fill Fill of possible robber trench 585. 31/01/99 4 Y 197
of N
 List of recorded contexts   363

Context Trench Type Description Date Trench Period Bone Sherds

phase (approx)

585 A:S Wall Cut Possible robber trench cut. 31/01/99 4 Y

of N
586 A:South Fill Fill of pit/pot dump 588. 31/01/99 3 Y 223
Area
587 A:South Fill Shallow pit/scoop cutting wall 578. 01/02/99 3
Area
588 A:South Cut Shallow cut/scoop cutting wall 578. 01/02/99 3
Area
589 A:South Layer Cleaning layer, equal to 568. 01/02/99 3 Y 852
Area
590 A:TP5 Fill Foundation fill between and under stone packing 01/02/99 4 Y 283
579.
591 A:South Fill Fill of robber cut 577 below 578, remains of wall. 02/02/99 3 Y 1,570
Area
592 A:South Layer Laminated slope wash. 02/02/99 4 Y 267
Area
593 A:TP5 Fill Northern end of foundation fill 330. 02/02/99 4 64
594 A:TP5 Cut Foundation trench cut for lost Medieval wall (with 02/02/99 4
fill 330; 593; 604).
595 A:TP5 Cut Foundation trench cut for lost Medieval wall (with 02/02/99 4
fill 570; 582; 579; 590).
596 A:TP6 Fill Fill of cut 597. 02/02/99 4 Y 151
597 A:TP6 Cut Cut, filled by 596. May be post-Yusuf. 02/02/99 4
598 A:TP5 Layer Mud mortar levelling course of wall 511. 02/02/99 3
599 A:TP5 Fill Foundation fill of cut 600. 02/02/99 3 Y 149
600 A:TP5 Cut Foundation trench cut for wall 511. 02/02/99 3
601 A:TP5 Layer Mud mortar levelling course for wall 510. 02/02/99 3
602 A:TP5 Fill Foundation fill of cut 603. 02/02/99 3 246
603 A:TP5 Cut Foundation cut for wall 510. 02/02/99 3
604 A:TP5 Fill Stone foundation packing for lost Medieval wall, 03/02/99 4
north end of cut 594.
605 A:South Rubble Rubble tumble, slope wash below 592. 04/02/99 4 Y 417
Area
606 A:South Fill Fill of pit 607. 04/02/99 3 Y 378
Area
607 A:South Cut Pit cut, filled by 606. 04/02/99 3
Area
608 A:South Fill Lowest fill of robber cut 577, below 591. 04/02/99 3 Y 569
Area
609 A:TP5 Layer Mud levelling under wall 528. 04/02/99 3
610 A:TP5 Founda- Foundation for wall 528. 04/02/99 3 Y 138
tion
611 A Fill Fill of last phase of interior structure of the North 04/02/99
Temple, post-Yusuf.
612 A:W Layer Layer immediately to south-west of the North 06/02/99 4 Y 158
Area Temple (equals 644).
613 A:TP6, Layer Slope wash/tumble below rubble 605. 06/02/99 4 Y 393
South
Area
614 A:TP5 Founda- Possible foundation fill put down during re-build of 06/02/99 3
tion north wall of the South Temple (561).
364   Appendix I

Context Trench Type Description Date Trench Period Bone Sherds

phase (approx)

615 A:TP5 Layer Silt and stone packing under wall 589. 06/02/99 3 38
616 A:W Fill Fill of 617. 06/02/99 4 Y 518
Area
617 A:W Cut Cut for possible boundary wall. 06/02/99 4
Area
618 A:W Layer Homogenous layer below 612, to west and south- 07/02/99 4 Y 802
Area west of the North Temple.
619 A:South Layer Patch of rubble tumble/slope wash south-east of 07/02/99 4 Y
Area robbing 607.
620 A:TP5 Cut Possible foundation cut for wall 509. 07/02/99 3
621 A:TP5 Cut Cut made during rebuild of north wall of the South 07/02/99 3
Temple (fill 614).
622 A:TP5 Layer Latest surviving floor foundation inside the South 07/02/99 3 Y 2,392
Temple, associated with walls 509 and 528.
623 A:South Cut Shallow pit/scoop filled with 624, south of robber 07/02/99 3
Area cut 577.
624 A:South Fill Fill of 623. 07/02/99 3 Y 75
Area
625 A:TP5 Fill Lower foundation fill of cut 620. 08/02/99 3 Y 114
626 A:TP5 Layer Made-up construction layer for later internal 08/02/99 3-temple
surface of the South Temple.
627 A:W Layer Layer or fill almost completely removed by Yusuf. 08/02/99 4 Y 290
Area
628 A:South Cut Cut of pit in western section. 08/02/99 3
Area
629 A:South Fill Uppermost fill of pit 628. 08/02/99 3 Y 241
Area
630 A:South Fill Ashy lens fill of pit 628, below 629. 08/02/99 3
Area
631 A:South Cut Cut of pit. 08/02/99 3
Area
632 A:South Fill Uppermost fill of pit 631. 08/02/99 3 Y 474
Area
633 A:TP5 Fill Fill of cut 335. 08/02/99 4 Y 456
634 A:TP5 Floor Floor surface inside the South Temple. 08/02/99 3-temple 99
635 A:South Layer Slope wash below 613. 08/02/99 4 Y 843
Area,
TP6
636 A:W Layer Rubbly layer with silty matrix. 09/02/99 4 120
Area
637 A:TP9 Surface Trample surface in layer 551 in the North Temple. 09/02/99 3
638 A:TP5 Layer Foundation and make-up layer under floor 634. 09/02/99 3-temple Y 954
639 A:TP9 Layer Black cotton soil in the North Temple below 637. 09/02/99 3 Y 75
640 A:W Rubble Brick rubble layer against wall 357, fill of 339. 09/02/99 3 55
Area
641 A:TP9 Layer Thin layer in west end of south aisle, North Temple. 09/02/99 4 Y 283
642 A:TP5 Layer Silt layer under 638. 10/02/99 3-temple Y 1,301
643 A:W Layer Compacted silt and brick layer around west end of 10/02/99 4 194
Area the North Temple.
 List of recorded contexts   365

Context Trench Type Description Date Trench Period Bone Sherds

phase (approx)

644 A:W Layer Equals 612. 10/02/99 4 Y 109

Агeа
645 A:South Rubble Rubble patch, possibly equal to 619. 10/02/99 4 Y 425
Area
646 A:GG Wall Rough rubble wall in garbhagriha of the North 11/02/99 3-temple
North Temple.
647 A: South Layer Ashy dump. 11/02/99 3 Y 419
Area
648 A:TF5 Layer Stones in garbhagriha of the South Temple. 11/02/99 3-temple
649 А:TР6 Layer Deposit underlying 635. 11/02/99 4 Y 634
650 A:W Floor Hard surface (floor level) below 649. 11/02/99 3 Y 157
Area,
TF6
651 A: South Layer Silt deposit from trial sounding. Contaminated. 11/02/99 3 Y 358
Area, W
Area
652 A:GG Layer Black cotton soil fill of garbhagriha, North Temple. 11/02/99 3-temple
North, W
Area
653 A:TP5 Layer Silt layer under stones 648 in garbhagriha, South 11/02/99 3-temple Y 272
Temple.
654 A:TP5 Fill Regularly spaced stone packing within fill 642. 11/02/99 3-temple
655 A:W Fill Fill of 656. 11/02/99 4 Y 24
Area
656 A:W Cut Post-hole cut(?) (fill 655). 11/02/99 4
Area
657 A:W Surface Trample surface to west of 617. 11/02/99 4 Y 616
Area
658 A:GG Layer Layer of stones and black cotton soil in garbhag- 11/02/99 3-temple
North riha, North Temple.
659 A:W Fill Fill of 660. 13/02/99 4 Y 149
Area
660 A:W Cut Cut, truncated by drain made by Yusuf. 13/02/99 4
Area
661 A:TP5 Layer Brick jelly layer under 654, interior of the South 13/02/99 3-temple Y 231
Temple.
662 A:TP5 Layer Clayey deposit against internal sides of South 13/02/99 3-temple
Temple walls.
663 A:TP5, Fill Uppermost fill of ditch 688, silt and rubble. 13/02/99 4 Y 748
South
Area
664 A:GG Rubble Hard rubble surface in garbhagriha, North Temple. 13/02/99
North
665 A:TP5 Deposit Lower stone packing in garbhagriha, South Temple. 14/02/99 3-temple
666 A:TP5 Fill Silt fill under 665, garbhagriha of the South 14/02/99 3-temple Y 251
Temple.
667 A:TP5 Layer Compact brick-rubble jelly under 661. 14/02/99 3-temple Y 1,459
668 A: South Fill Fill of pit 669. 14/02/99 3 Y 297
Area
669 A: South Cut Cut of pit (fill 668). 14/02/99 3
Area
670 A: South Layer Gravel rich silt fill of 688. 14/02/99 3 Y 170
Area
671 A: South Cut Linear North-South cut along west section. Similar 14/02/99 3
Area to 688.
366   Appendix I

Context Trench Type Description Date Trench Period Bone Sherds

phase (approx)

672 A:South Fill Fill of 671. 14/02/99 3

Area
673 A:South Cut Linear cut along wall/boulder fill 591. 14/02/99 3
Area
674 A:South Fill Fill of 673. 14/02/99 3
Area
675 A:South Fill Gravely stony fill of 688. 14/02/99 3 Y 186
Area
676 A:GG Layer Stone layer in garbhagriha, North Temple. 14/02/99 3-temple 42
North
677 A:GG Layer Pure black cotton soil fill below 676 in garbha- 14/02/99 3-temple 38
North griha, North Temple.
678 A:South Layer Laminated slope wash (equals 635). 14/02/99 4 Y 275
Area
679 A:South Layer Thin calcareous wash in ditch 688. 14/02/99 3
Area
680 A:W Fill Fill of 681. 14/02/99 4 Y 40
Area
681 A:W Cut Possible grave cuts in north of trench. 14/02/99 4
Area
682 A:TP3 Fill Black cotton soil fill of 683. 14/02/99 3 Y 386
683 A:TP3 Cut Linear cut. 14/02/99 3
684 A:TP9 Layer Layer in south aisle, North Temple, building under 15/02/99 4 Y 933
641.
685 A:GG Layer Stone layer below 677 in garbhagriha, North 15/02/99 3-temple Y
North Temple.
686 A:TP5 Layer Stone packing layer below 666, garbhagriha, 15/02/99 3-temple
South Temple.
687 A:GG Layer Black cotton soil fill below 685 in garbhagriha, 15/02/99 3-temple 15
North North Temple.
688 A:South Cut Shallow U-shaped ditch following line of wall 691. 15/02/99 3
Area
689 A:TP6 Deposit Deposit under 635. 15/02/99 4 Y 587
690 A:TP6 Layer Dark brown earth below 689. 15/02/99 4
691 A:South Wall Wall of large irregular igneous boulders. 16/02/99 3
Area
692 A:South Cut Cut; possible edge of Yusuf’s excavation trench. 16/02/99 4
Area
693 A:South Layer Remnant ‘finger’ of in situ stratigraphy between 16/02/99 3
Area cuts 692 and 577.
694 A:GG Surface Trample surface in front of garbhagriha, North 16/02/99 16
North Temple.
695 A:GG Step Crude brick step in west entrance to garbhagriha, 16/02/99
North North Temple.
696 A:W Layer Soil covering a compact layer in north-west corner 16/02/99 4 Y 367
Area of Trench A.
697 A:GG Surface Trample surface over walls and hearting in west 16/02/99
North entrance, garbhagriha, North Temple.
698 A:GG Layer Black cotton soil and brick flecks in garbhagriha, 16/02/99 3-temple
North North Temple.
699 A:TP6, Foundation Possible wall foundation of broken bricks cut by 16/02/99 3
GG 577.
North,
South
Area
 List of recorded contexts   367

Context Trench Type Description Date Trench Period Bone Sherds

phase (approx)

700 A:TP5 Layer Construction layer under 667 inside the South 16/02/99 3-temple Y 2,573
Temple.
701 D2 Layer Loose compact dump deposit. 31/01/99 D2-24 2 Y 18
702 D2 Layer Semi-compact yellow deposit. 01/02/99 D2-23 2 Y 252
703 D2 Layer Compact black brown deposit. 02/02/99 D2-22 2 Y 652
704 D2 Floor Compact white floor. 03/02/99 D2-21 2 73
705 D2 Layer Compact yellow brown deposit, packing for floor 04/02/99 D2-20 2 Y 303
704?
706 D2 Layer Loose brownish deposit. 04/02/99 D2-19 2 Y 323
707 D2 Floor Compact brown floor deposit. 05/02/99 D2-18 2 Y 147
708 D2 Cut Pit cut, filled by 709. 05/02/99 D2-17 2
709 D2 Fill Fill of 708. 05/02/99 D2-17 2 282
710 D2 Floor Floor sloping towards the north. 06/02/99 D2-16 2 Y 488
711 D2 Floor Compact white floor deposit. 06/02/99 D2-15 2 Y 443
712 D2 Floor Compact brownish floor. 09/02/99 D2-14 2 Y 1,401
713 D2 Fill Fill of 714. 10/02/99 D2-13 1 38
714 D2 Cut Cut of a root hole, filled by 713. 10/02/99 D2-13 1
715 D2 Floor Compact floor. 10/02/99 D2-12 1 Y 1,228
716 D2 Fill Fill of 717. 11/02/99 D2-11 1 14
717 D2 Cut Circular post-hole cut, filled by 716, cuts 718. 11/02/99 D2-11 1
718 D2 Floor Compact blackish brown floor covered with 11/02/99 D2-10 1 Y 2,567
pottery.
719 D2 Fill Ashy fill of 720. 11/02/99 D2-09 1 Y 109
720 D2 Cut Pit cut, filled by 719, cutting floor 721. 11/02/99 D2-09 1
721 D2 Floor Compact floor level. 13/02/99 D2-08 1 Y 687
722 D2 Floor Compact brownish black floor. 14/02/99 D2-06 1 Y 885
723 D2 Fill Fill of 724. 15/02/99 D2-07 1 94
724 D2 Cut Pit cut, filled by 723, cutting 725. 15/02/99 D2-07 1
725 D2 Floor Floor with ashy patches and pit cut. 15/02/99 D2-05 1 Y 194
726 D2 Floor Less compact floor with fire place in south-west 16/02/99 D2-04 1 Y 1,143
and south-east.
727 D2 Fill Ash and charcoal fill of 728. 17/02/99 D2-03 1 Y 297
728 D2 Cut Shallow pit, filled by 727, cutting floor 729. 17/02/99 D2-03 1 333
729 D2 Floor Less compact blackish brown floor. 17/02/99 D2-02 1 Y 69
730 D2 Layer Natural black cotton soil. 17/02/99 D2-01 1
731 B Fill Fill of 732. 06/03/99 B-05 1 1
732 B Cut Pit cut along eastern section of trench, cut into 06/03/99 B-05 1
735, filled by 731.
733 B Fill Fill of 734. 06/03/99 B-05 1
734 B Cut Pit cut along western section, cut into 735, filled 06/03/99 B-05 1
by 733.
735 B Floor Black soil floor with pottery and charcoal. 07/03/99 B-02 1 Y 809
736 B Floor Compact, black floor. 08/03/99 B-02 1 Y 311
737 B Fill Fill of 738. 08/03/99 B-03 1
738 B Cut Post-hole cut, filled by 737. 08/03/99 B-03 1
739 B Fill Fill of 740. 08/03/99 B-03 1
368   Appendix I

Context Trench Type Description Date Trench Period Bone Sherds

phase (approx)

740 B Cut Post-hole cut, filled by 739. 08/03/99 B-03 1

741 B Fill Fill of 742. 08/03/99 B-03 1

742 B Cut Post-hole cut, filled by 741. 08/03/99 B-03 1

743 B Natural Natural soil. 09/03/99 B-01 1
751 D1 Layer Loose wash material. 01/02/99 D1-13 356
752 D1 Floor Semi-compact disturbed lime floor. 02/02/99 D1-12 18
753 D1 Layer Black earth with very few sherds. 03/02/99 D1-10 1
754 D1 Layer Pottery dump. 03/02/99 D1-10 1 Y 2,236
755 D1 Fill Fill of pit 756. 06/02/99 D1-11 1
756 D1 Cut Cut of pit, cutting 752 and filled with 755. 06/02/99 D1-11 1
757 D1 Fill Ashy fill of pit 758. 07/02/99 D1-09 1
758 D1 Cut Pit cutting floor 759, filled with 757. 07/02/99 D1-09 1
759 D1 Floor Mud and pot-sherd floor. 07/02/99 D1-08 1 Y 180
760 D1 Fill Fill of pit 761 containing large sherds of a storage 09/02/99 D1-07 1 416
jar.
761 D1 Cut Pit cut to hold storage jar, filled by 760, cutting 09/02/99 D1-07 1
floor 766.
762 D1 Fill Fill of two post-hole cuts 763. 09/02/99 D1-07 1 Y 642
763 D1 Cut Cut of two post-holes, cutting 766, filled by 762. 09/02/99 D1-07 1
764 D1 Fill Fill of 765. 09/02/99 D1-07 1 49
765 D1 Cut Cut of a small pit or post-hole, cutting 767, filled 09/02/99 D1-07 1
with 764.
766 D1 Floor Compact mud floor. 10/02/99 D1-06 1 Y 495
767 D1 Floor Semi-compact mud floor beneath 766. 11/02/99 D1-05 1 Y 2,288
768 D1 Fill Fill of pit 769. 11/02/99 D1-07 1 20
769 D1 Cut Cut of pit, cutting floor 767, filled with 768. 11/02/99 D1-07 1
770 D1 Floor Compact black grey floor level. 13/02/99 D1-03 1 Y 138
771 D1 Fill Fill of 772. 14/02/99 D1-04 1 Y 51
772 D1 Cut Pit cut, cutting floor 770, filled with 771. 14/02/99 D1-04 1
773 D1 Fill Fill of 774. 14/02/99 D1-04 1 112
774 D1 Cut Cut of post-hole, cutting 767, filled with 773. 14/02/99 D1-04 1
775 D1 Fill Fill of 776. 14/02/99 D1-02 1 87
776 D1 Cut Irregular cut of pit, cutting 777, filled with 775. 14/02/99 D1-02 1
777 D1 Layer Natural soil. 15/02/99 D1-01 1 Y 58
778 F Layer Surface layer, thick washed material and dump. 20/02/99 F-16 3 Y 1,286
784 F Floor Mud floor. 23/02/99 F-14 2 Y 329
785 F Hearth Fill of hearth 786. 24/02/99 F-15 2 34
786 F Hearth Shallow cut of hearth, cutting floor 787, filled with 24/02/99 F-15 2
785.
787 F Floor Compact mud floor. 24/02/99 F-13 2 Y 732
788 F Floor Compact blackish grey mud floor. 25/02/99 F-12 2 Y 1,749
789 F Floor Mud floor. 27/02/99 F-11 2 Y 228
790 F Floor Brick floor. 27/02/99 F-09 2
791 F Fill Fill of post-hole cut 792. 28/02/99 F-10 2
792 F Cut Cut of post-hole, cutting 790, filled with 791. 28/02/99 F-10 2
 List of recorded contexts   369

Context Trench Type Description Date Trench Period Bone Sherds

phase (approx)

793 F Fill Fill of post-hole cut 794. 28/02/99 F-10 2

794 F Cut Cut of post-hole, cutting 790, filled with 793. 28/02/99 F-10 2
795 F Fill Fill of post-hole cut 796. 28/02/99 F-10 2
796 F Cut Cut of post-hole, cutting 790, filled with 795. 28/02/99 F-10 2
797 F Layer Hard surface beneath brick floor 790. 28/02/99 F-08 2 Y 246
798 F Layer Gravel deposit beneath 797. 28/02/99 F-07 2 Y 92
799 F Hearth Fill of hearth 800. 01/03/99 F-06 2
800 F Hearth Hearth cut into floor 846, filled with 799. 01/03/99 F-06 2
801 E Surface Surface wash. 10/02/99 Y 74
802 E Wall Stone wall extending from east section. 10/02/99
803 E Wall Stone wall extending from north section. 10/02/99
804 E Wall Stone wall. 10/02/99
805 E Fill Fill of 806. 10/02/99 Y 105
806 E Cut Foundation cut for stone wall 804. 10/02/99 2
807 E Fill Fill of 808. 10/02/99
808 E Cut Rectangular cut near north section, filled by 807. 10/02/99
809 E Fill Fill of 810. 10/02/99 Y 45
810 E Cut Pit, filled by 809. 10/02/99
811 E Surface Compact floor surface. 13/02/99
812 E Fill Fill of 813. 14/02/99 Y 50
813 E Cut Wavy cut feature. 14/02/99
814 E Cut Foundation cut for wall 804, equals 806. 16/02/99
815 E Floor Plastered floor. 14/02/99 Y 408
816 E Fill Fill of 817. 17/02/99 Y 700
817 E Cut Pit cut, filled by 816. 17/02/99
818 E Floor Rubble packing below floor 815. 20/02/99 Y 96
819 E Fill Fill of 820. 20/02/99 24
820 E Cut Shallow cut. 20/02/99
821 E Layer Rubble and silt deposit. 21/02/99
822 E Layer Loose yellow deposit abutting stone packing 823. 21/02/99 128
823 E Layer Stone packing. 21/02/99
824 E Layer Semi-compact yellow deposit below stones 823. 21/02/99 Y 60
825 E Layer Brick bat packing in pit 817. 22/02/99 96
826 E Layer Semi-compact yellow silt deposit on north side. 22/02/99 181
827 E Layer Brick deposit on line of pit 817. 23/02/99 101
828 E Layer Compact deposit below 827. 24/02/99 35
829 E Fill Fill of 830. 27/02/99 Y 441
830 E Cut Cut of pit in north section, filled by 829. 27/02/99 105
831 E Fill Brick lining to circular pit 832. 28/02/99 Y 882
832 E Cut Circular pit or well. 01/03/99 261
370   Appendix I

Context Trench Type Description Date Trench Period Bone Sherds

phase (approx)

833 E Floor Compact rammed floor. 01/03/99 65

834 E Floor Floor with traces of fire. 01/03/99 70
835 E Hearth Hearth or circular fireplace. 01/03/99 Y 33
836 E Cut Cut for hearth 835. 01/03/99
837 E Fill Fill of hearth 838. 03/03/99
838 E Cut Cut of circular hearth or ash pit. 03/03/99
839 E Layer Silt deposit. 03/03/99 Y 30
840 E Floor Semi-compact floor deposit. 04/03/99
841 E Fill Ash deposit, fill of 842. 04/03/99 Y 69
842 E Cut Irregular cut, filled by 841. 04/03/99
843 E Layer Loose brown clayey silt deposit. 07/03/99 Y 248
844 E Wall Bottom course of mud and stone wall. 07/03/99 Y 480
845 E Cut Foundation cut for wall 844. 07/03/99
846 F Floor Mud floor beneath 798. 01/03/99 F-05 2 Y 3,009
847 F Fill Fill of hearth 848. 01/03/99 F-06 2 4
848 F Cut Cut of hearth, cutting 846, filled by 847. 01/03/99 F-06 2
849 F Fill Fill of pit 850 below 846. 04/03/99 F-04 1 Y 74
850 F Cut Round pit cut, cutting 851, filled with 849. 04/03/99 F-04 1
851 F Floor Hard clay surface below 846. 04/03/99 F-03 1 Y 5,409
852 F Floor Hard yellow floor. 07/03/99 F-02 1 Y 250
853 F Natural Natural black cotton soil. 09/03/99 F-01 1 54
901 A:TP5 Layer Thin brick dust layer in garbhagriha, South 16/02/99 3-temple
Temple.
902 A:TP5 Layer Clay layer under 901. 16/02/99 3-temple Y 166
903 A:W Fill Five large stones set in cut 617, possible remains 16/02/99 4
Area of boundary wall.
904 A:W Rubble Silty layer with brick rubble, fill of 909. 16/02/99 4 Y 372
Area
905 A:TP6 Fill Fill of 906. 16/02/99 4 Y 218
906 A:TP6 Cut Cut, filled by 905. 16/02/99 4
907 A:GG Fill Third deposit of stones in garbhagriha, North 16/02/99 3-temple
North Temple.
908 A:TP1 Layer Layer of silt and ash, equal to 312. 17/02/99 TP1-4 3 Y 309
909 A:TP3, Cut Cut around west end of the North Temple (equals 17/02/99 4
TP10 939).
910 A:GG Layer Black cotton soil below 907, garbhagriha, North 17/02/99 3-temple 10
North Temple.
911 A:GG Fill Brick hearting within wall 555, North Temple. 17/02/99 3-temple Y 80
North
912 A:TP1 Layer Water-derived silt layer below 908. 17/02/99 TP1-4 3 Y 99
913 A:TP1 Layer Ashy silt layer below 912. 17/02/99 TP1-4 3 Y 576
914 A:TP6 Fill Fill of cut 915. 17/02/99 4 Y 133
915 A:TP6 Cut Cut, filled by 914. 17/02/99 4
916 Number changed to 940. 17/02/99
917 A:GG Fill First stone fill in base of garbhagriha, North 17/02/99 3-temple
North Temple.
918 A:TP5 Layer Silt layer underlying 700, South Temple interior. 18/02/99 3-temple Y 2,205
 List of recorded contexts   371

Context Trench Type Description Date Trench Period Bone Sherds

phase (approx)

919 A:TP5 Layer Silt layer under 907, garbhagriha, South Temple. 18/02/99 3-temple Y 369
920 A:W Fill Rubble layer filling cuts 909 and 939. 18/02/99 4 955
Area
921 A:TP1 Layer Ash and brick rich silt below 922. 18/02/99 TP1-4 3 Y 1,081
922 A:TP1 Layer Clay rich silt almost black cotton soil below 913. 18/02/99 TP1-4 3 Y 77
923 A:TP9, Floor Brick powder floor, garbhagriha, North Temple 18/02/99 3-temple 1
GG under 917.
North
924 A:TP6 Fill Fill of cut 925. 18/02/99 3 Y 765
925 A:TP6 Cut Cut, filled by 924, 930. 18/02/99 3
926 A:TP6 Layer Rammed floor of sandy silt with course of fine sand 18/02/99 3 Y 434
and specks of brick jelly.
927 A:TP9, Surface Hard clay upper foundation fill, garbhagriha, 18/02/99 3-temple Y 616
GG North Temple.
North
928 A:TP10 Fill Silty fill with some brick rubble in cut 909 (equals 18/02/99 4 Y 425
934).
929 A:TP6 Fill Early fill of 925. 18/02/99 3 18
930 A:TP1 Layer Ashy layer below 921. 20/02/99 TP1-4 3 Y 829
931 A:South Fill Fill of robber cut 577. Equals 608. 20/02/99 3 297
Area
932 A:GG Surface Brick surface or inner construction inside wall 969. 20/02/99 3-temple
North
933 A:TP1 Layer Mix clayey loam layer below 930. 20/02/99 TP1-4 3 Y 928
934 A:W Fill Silty fill of 909 (equals 928). 20/02/99 4 Y 100
Area,
TP3
935 A:TP6 Feature Rectangular stump of clay silt associated with floor 20/02/99
926.
936 A:TP6 Fill Fill of 937. 20/02/99 3 45
937 A:TP6 Cut Circular cut filled by 936. 20/02/99 3
938 A:GG Layer Below 927, layer of foundation fill for Phase I 20/02/99 3-temple Y 111
North shrine, North Temple.
939 A:TP3 Cut Cut running alongside west wall of the North 20/02/99 4
Temple (equals 909).
940 A:TP9 Surface Trodden earth floor above 916 in Phase III of the 20/02/99 3 15
North Temple.
941 A:GG Rubble Rubble hearting under 932. 20/02/99 3-temple
North
942 A:TP3 Fill Brick jelly fill of 683. 21/02/99 3 65
943 A:TP10 Fill Brick rubble fill of 909. 21/02/99 4 22
944 A:TP1 Rubble Accumulation of brick and topsoil forming an 21/02/99 TP1-3 2 Y 981
occupation horizon.
945 A:TP1 Fill Ash fill of pit 946. 21/02/99 TP1-3 2 Y 81
946 A:TP1 Cut Pit cut into 944. 21/02/99 TP1-3 2
947 A:TP3 Layer Silty layer, possibly accumulated since Yusuf’s 21/02/99 3 365
excavations.
948 A:TP1 Deposit Ash dump on 944. 21/02/99 TP1-3 2 Y 35
949 A:GG Layer Foundation fill for Phase I shrine, beneath 938, 21/02/99 3-temple Y 243
North garbhagriha, North Temple.
950 A:TP10 Layer Surface remnant, excavation not continued below. 21/02/99 4 121
372   Appendix I

Context Trench Type Description Date Trench Period Bone Sherds

phase (approx)

951 A:TP5 Rubble Small brick rubble dump on north end of 953. 21/02/99 3-temple
952 A:TP3, Deposit Stone packing laid on 955. 21/02/99 3-temple
TP5
953 A:TP5 Floor Floor surface layer in garbhagriha, South Temple 21/02/99 3-temple
(equals 955?).
954 A:TP5 Packing Small stone packing laid on floor 953. 21/02/99 3-temple
955 A:TP5 Floor Mud floor in the South Temple (equals 953?). 21/02/99 3-temple Y 148
956 A:TP3 Layer May be fill of cut 982, but truncated by drain cut 22/02/99 3-temple Y 450
by Yusuf.
957 A:TP3 Layer Tumble from a brick structure. 22/02/99 3 Y 317
958 A:TP5 Fill Uppermost fill of foundation trench 963. 22/02/99 3-temple Y 16
959 A:TP3, Layer Thin spread of material. 22/02/99 3 298
W Area
960 A:TP5 Fill Sandy silt fill of 963 below 958. 22/02/99 3-temple Y 17
961 A:N Cut Small cut into top of wall 554. 22/02/99
Temple
962 A:N Fill Fill of cut 961. 22/02/99 20
Temple
963 A:TP5 Cut Foundation cut for the South Temple. 22/02/99 3-temple
964 A:TP5 Layer Clay surface in South Temple mandapa (equals 22/02/99 3-temple Y 275
965?).
965 A:TP5 Layer Clay surface in South Temple garbhagriha (equals 22/02/99 3-temple Y 50
964?).
966 A:TP1 Layer Soft silty loam below 944. 22/02/99 TP1-3 2 Y 1,081
967 A:GG Fill Hearting of wall 556, North Temple. 22/02/99 3-temple
North
968 A:GG Wall Internal thickening of wall 554 of garbhagriha, 22/02/99 3-temple
North North Temple.
969 A:GG Wall Possible earlier phase of wall 555, North Temple. 22/02/99 3-temple
North
970 A:GG Wall First two courses of wall 554, North Temple. 22/02/99 3-temple
North
971 A:GG Wall Rebuilds to wall 556, North Temple using purple 22/02/99 3-temple 1
North bricks.
972 A:TP3 Layer Compact layer with gravel inclusions. 23/02/99 3 Y
973 A:TP3 Surface Compact, well-defined clay surface. 23/02/99 3 Y 286
974 A:TP3 Fill Fill of 975, probably equals 981. 23/02/99 3 Y 606
975 A:TP3 Cut Cut of pit, filled by 974. 23/02/99 3
976 A:TP9 Deposit Hard clay feature at western end of North Temple 23/02/99 3 4
interior.
977 A:TP9 Deposit Loose crumbly material filling cut 980. 23/02/99 3
978 A:TP5 Layer Silt layer under 964 in South Temple mandapa 22/02/99 3-temple Y 203
(equals 979?).
979 A:TP5 Layer Silt layer under 963 in South Temple garbhagriha 22/02/99 3-temple Y 106
(equals 978?).
980 A:TP9 Cut Cut into 976 at western end of North Temple 22/02/99 3
interior.
981 A:TP3 Fill Fill of 982. 23/02/99 3-temple Y 5,877
982 A:TP5, Cut Foundation cut for the South Temple. 23/02/99 3-temple
TP3
 List of recorded contexts   373

Context Trench Type Description Date Trench Period Bone Sherds

phase (approx)

983 A:TP9 Struc- Brick altar base? Interior of North Temple 23/02/99 3 4
ture mandapa.
984 A:TP1 Cut Pit cut. 24/02/99 TP1-3 2
985 A:TP1 Fill Upper fill of pit 984. 24/02/99 TP1-3 2 Y 79
986 A:TP1 Fill Ash fill of pit 984 below 985. 24/02/99 TP1-3 2 Y 106
987 A:TP6 Surface Hard surface strewn with brick fragments and 24/02/99 3 Y 1,737
pot-sherds.
988 A:TP1 Fill Ash and sand fill of pit 984, below 986. 24/02/99 TP1-3 2 Y 197
989 A:TP1 Cut Small post-hole cut. 24/02/99 TP1-3 2
990 A:TP1 Fill Fill of post-hole. 24/02/99 TP1-3 2 Y 42
991 A:TP3 Rubble Brick rubble layer. 24/02/99 3-temple
992 A:TP9 Surface Surface in mandapa of the North Temple, associa- 24/02/99 3
ted with 983.
993 A:TP3 Fill Fill of feature 994. 24/02/99 3 Y 503
994 A:TP3 Cut Shallow pit. 24/02/99 3 Y
995 A:TP1 Cut Pit cut. 24/02/99 TP1-3 2
996 A:TP1 Fill Upper fill of pit 995. 24/02/99 TP1-3 2 Y 177
997 A:TP1 Fill Ash fill of pit 995, below 996. 24/02/99 TP1-3 2 Y 159
998 A:TP9 Deposit Deliberate sandy deposit under 992, mandapa of 24/02/99 3
the North Temple.
999 A:TP5 Deposit Gravely construction layer deposit overlying north 25/02/99 3-temple 11
end of 1002.
1000 A:TP5 Packing Stone packing in South Temple mandapa, overly- 25/02/99 3-temple
ing 1002 (equals 1001).
1001 A:TP5 Packing Stone packing in South Temple garbhagriha 25/02/99 3-temple
(equals 1000).
1002 A:TP5 Layer Compact clayey silt layer under 1000 in mandapa 25/02/99 3-temple Y 496
of the South Temple (equals 1003).
1003 A:TP5 Layer Compact clayey silt layer under 1001 in garbha- 25/02/99 3-temple Y 82
griha of the South Temple (equals 1002).
1004 A:TP3 Fill Fill of 1005. 25/02/99 3 94
1005 A:TP3 Cut Truncated feature. 25/02/99 3
1006 A:TP3 Fill Fill of 1011. 25/02/99 3
1007 A:TP9 Cut Cut into mandapa floor, west end of the North 25/02/99 3
Temple.
1008 A:TP9 Fill Fill of 1007. 25/02/99 3
1009 A:TP1 Rubble Rubble spread in sand matrix below 966. 25/02/99 TP1-3 2 Y 1,053
1010 A:TP9 Floor Mandapa floor of the North Temple (equals 1100, 25/02/99 3-temple 474
1080, 550?, 319?).
1011 A:TP3 Cut Cut of shallow feature. 25/02/99 3 1
1012 A:TP3 Layer Homogenous layer. 25/02/99 3 Y 1,083
1013 A:TP1 Layer Compact clay and rubble spread below 1009. 25/02/99 TP1-3 2 Y 1,341
1014 A:TP1 Deposit Pot and rubble dump filling 1029. 25/02/99 TP1-3 2 Y 1,060
1015 A:TP3 Fill Fill of post-hole 1016. 25/02/99 3
1016 A:TP3 Cut Post-hole cut, cutting 1030. 25/02/99 3
1017 A:TP3 Fill Fill of post-hole 1018. 25/02/99 3
1018 A:TP3 Cut Post-hole cut, cutting 1030. 25/02/99 3
1019 A:TP3 Fill Fill of post-hole 1020. 25/02/99 3
374   Appendix I

Context Trench Type Description Date Trench Period Bone Sherds

phase (approx)

1020 A:TP3 Cut Post-hole cut, cutting 1030. 25/02/99 3

1021 A:TP3 Fill Fill of 1022. 25/02/99 3 Y 603
1022 A:TP3 Cut Possible post-hole cut, cutting 1030. 25/02/99 3
1023 A:TP3 Fill Fill of 1024. 25/02/99 3
1024 A:TP3 Cut Cut of uncertain feature, cutting 1030. 25/02/99 3
1025 A:TP3 Fill Fill of post-hole 1026. 25/02/99 3 123
1026 A:TP3 Cut Post-hole cut, cutting 1030. 25/02/99 3 1
1027 A:TP3 Fill Fill of post-hole 1028. 25/02/99 3
1028 A:TP3 Cut Possible post-hole cut, cutting 1030. 25/02/99 3
1029 A:TP1 Cut Irregular cut filled by 1014. 27/02/99 TP1-3 2
1030 A:TP3 Layer Occupation layer cut by a series of post-holes and 27/02/99 3 Y 742
other features.
1031 A:TP6 Fill Fill of cut 372, compact clayey deposit. 27/02/99 3-temple 180
1032 A:TP9 Fill Loose fill below floor 1010 interior of the North 27/02/99 3-temple 66
Temple.
1033 A:TP6 Packing Stone packing filling foundation cut 1246. 27/02/99 3-temple
1034 A:TP6 Deposit Deposit below stone packing 1033. 27/02/99 3-temple 26
1035 A:TP6 Deposit Deposit below the compact floor 987. 27/02/99 3 Y 1,700
1036 A:TP3 Fill Ashy lenses and silty clay matrix. 27/02/99 3 Y 204
1039 A:TP3, Fill Compact layer, fill of cut 982. 27/02/99 3-temple 52
TP5
1040 A:TP1 Layer Olive green clay and rubble layer below 1013. 27/02/99 TP1-3 2 Y 1,953
1041 A:TP3 Fill Silty loam, fill of cut 982. 27/02/99 3-temple
1042 A:TP9 Surface Hard surface under 1010 in mandapa of the North 28/02/99 3-temple
Temple.
1043 A:TP1 Fill Fill of pit 1044. 28/02/99 TP1-3 2 Y 53
1044 A:TP1 Cut Shallow irregular cut. 28/02/99 TP1-3 2
1045 A:TP5 Fill Third foundation deposit of stones and sand filling 28/02/99 3-temple Y
cut 963.
1046 A:TP5, Fill Stone layer, fill of 982. 28/02/99 3-temple
TP3
1047 A:TP5, Fill Compact clay layer, fill of 982. 28/02/99 3-temple 153
TP3
1048 A:TP5 Fill Compact clay fill of foundation cut 963 above 28/02/99 3-temple
1045.
1049 A:TP9 Cut Small cut, filled by 1032 in interior of the North 28/02/99 3-temple
Temple. Possible shrine.
1050 A:TP9 Layer Loose silt deposit in mandapa, North Temple. 28/02/99 3-temple Y 256
1051 A:TP6 Layer Deposit below 1035. 28/02/99 3 Y 1,589
1052 A:TP1 Rubble Rubble layer below 1040. Contaminated. 28/02/99 TP1-3 2 Y 1,332
1053 A:TP5, Fill Silty fill of foundation cut 982. 28/02/99 3-temple 94
TP3
1054 A:TP6 Layer Compact deposit below silt 1034 and 1033. 28/02/99 3-temple
1055 A:TP9 Cut Cut against walls 555, 556. 28/02/99 3-temple 1
1056 A:TP9 Fill Fill of cut 1055. 28/02/99 3-temple
 List of recorded contexts   375

Context Trench Type Description Date Trench Period Bone Sherds

phase (approx)

1057 A:TP1 Cut Pit cut, filled by 1058. 01/03/99 TP1-3 2

1058 A:TP1 Fill Fill of pit 1057. 01/03/99 TP1-3 2 Y 310
1059 A:TP6 Packing Stone packing below 1054. 01/03/99 3-temple
1060 A:TP6 Layer Deposit beneath 1059. 01/03/99 3-temple 143
1061 A:TP3, Fill Stone layer in fill of foundation trench 982. 01/03/99 3-temple
TP5
1062 A:TP3, Fill Layer beneath 1061 in foundation trench 01/03/99 3-temple Y 161
TP5 982.
1063 A:TP1 Layer Rubble deposit, equals 1052 but unconta- 01/03/99 TP1-3 2 Y 1,206
minated.
1064 A:TP5 Fill Clay fill of foundation trench 963, below 01/03/99 3-temple Y
1045.
1065 A:TP9 Layer Stone layer, interior of the North Temple 05/03/99 3-temple
(equals 1115).
1066 A:TP9 Layer Surface in interior of the North Temple 01/03/99 3-temple
mandapa.
1067 A:TP9 Layer Crumbly brick and green sand layer; inte- 01/03/99 3-temple
rior of mandapa of the North Temple.
1068 A:TP5, Fill Silty layer, fill of 982. 01/03/99 3-temple Y 202
TP3
1069 A:TP5, Fill Stone layer, fill of 982. 01/03/99 3-temple
TP3
1070 A:TP5 Packing Stone packing under the South Temple 01/03/99 3-temple
(equals 1071).
1071 A:TP5 Packing Stone packing under the South Temple 01/03/99 3-temple
(equals 1070).
1072 A:TP5 Fill Compact clay layer, fill of South Temple 01/03/99 3-temple Y 78
foundation trench, under 1070.
1073 A:TP3 Fill Compact black cotton soil, fill of 1171 01/03/99 3-temple Y 893
(equals 927).
1074 A:TP1 Fill Fill of pit 1075. 01/03/99 TP1-3 2 Y 168
1075 A:TP1 Cut Pit cut. 01/03/99 TP1-3 2
1076 A:TP6 Layer Deposit below 1060, less compact clayey 03/03/99 3-temple
silt.
1077 A:TP1 Rubble Rubble layer below 1063. 03/03/99 TP1-3 2 Y 1,365
1078 A:TP9 Cut Cut related to late rebuilding, western end, 03/03/99 3
north aisle, North Temple.
1079 A:TP9 Fill Fill of 1078. 03/03/99 3
1080 A:TP9 Surface Surface of mandapa, North Temple (equals 03/03/99 3-temple 108
1010 and 1100).
1081 A:TP9 Layer Green sand deposit, south aisle, North 03/03/99 3
Temple (equals 1107).
1082 A:TP9 Cut Cut (by Yusuf?) into south aisle, North 03/03/99 4
Temple.
1083 A:TP3 Fill Fill in cut 1171. 03/03/99 3-temple Y 1,463
1084 A:TP9 Cut Post-hole cut into 1089. 03/03/99 3
1085 A:TP9 Fill Fill of cut 1084. 03/03/99 3
1086 A:TP1 Fill Stone and green silt fill of foundation cut 03/03/99 3-temple Y
963 below 1064.
1087 A:TP9 Cut Cut against west wall, south aisle, North 03/03/99 3
Temple.
1088 A:TP9 Fill Fill of 1087. 03/03/99 3
1089 A:TP9 Layer Brick jelly layer on floor of south aisle of 03/03/99 3
mandapa, North Temple.
376   Appendix I

Context Trench Type Description Date Trench Period Bone Sherds

phase (approx)

1090 A:TP9 Cut Post-hole cut, north aisle of mandapa, North 03/03/99 3-temple
Temple.
1091 A:TP9 Fill Fill of 1090. 03/03/99 3-temple
1092 A:TP5 Fill Silt layer/fill around stones 1093, South Temple 03/03/99 3-temple Y
foundations (equals 1130).
1093 A:TP5 Packing Stone packing in 1092. 03/03/99 3-temple
1094 A:TP3 Fill Stone layer within cut 1171. 03/03/99 3-temple
1095 A:TP3 Fill Compact clay and silt layer in cut 1171. 03/03/99 3-temple Y 669
1096 A:TP9 Surface Surface below 1080, northern aisle of mandapa, 03/03/99 3-temple Y
North Temple.
1097 A:TP9 Layer Green silt (dung?) on floor, south aisle of 04/03/99 3
mandapa, North Temple.
1098 A:TP6 Packing Thick stone packing. 04/03/99 3-temple
1099 A:TP1 Layer Rubble/clay layer filling pit 1114, below 1077. 04/03/99 TP1-3 2 Y 1,684
1100 A:TP9 Floor Floor in south aisle of mandapa, North Temple 04/03/99 3-temple
(equals 1010 and 1080).
1101 A:TP9 Cut Re-cut against wall 556, south aisle of mandapa, 04/03/99 3
North Temple.
1102 A:TP9 Fill Foundation fill of 1101. 04/03/99 3
1103 A:TP3 Fill Foundation fill of 1171. 04/03/99 3-temple Y 386
1104 A:TP3 Fill Foundation fill of 1171. 04/03/99 3-temple
1105 A:TP1 Fill Fill of pit 1057 below 1058. 04/03/99 TP1-3 2 Y 166
1106 A:TP6 Layer Fine sand and silt with pot sherds. 04/03/99
1107 A:TP9 Layer Green sand deposit, south aisle, North Temple 04/03/99 3
(equals 1081).
1108 A:TP9 Surface Second surface, south aisle of mandapa, North 06/03/99 3-temple 515
Temple.
1109 A:TP3 Fill Foundation fill of 1171. 06/03/99 3-temple Y
1110 A:TP1 Rubble Rubble below 1099, fill of 1114. 06/03/99 TP1-3 2 Y 489
1111 A:TP1 Fill Gravel patch below 1099, fill of pit 1113. 06/03/99 TP1-3 2 Y 283
1112 A:TP3 Fill Foundation fill of 1171. 06/03/99 3-temple Y 390
1113 A:TP1 Cut Cut of irregular and disturbed pit. 06/03/99 TP1-3 2
1114 A:TP1 Cut Shallow pit/sump filled with 1110 and 1099. 06/03/99 TP1-3 2
1115 A:TP9 Layer Stone packing below 1108, south aisle, mandapa, 06/03/99 3-temple
North Temple.
1116 A:TP5 Layer Compact clay foundation layer of the South 06/03/99 3-temple Y
Temple, under 1093.
1117 A:GG Layer Fill of Phase I foundations, North Temple. 06/03/99 3-temple Y 591
North
1118 A:TP5 Deposit Silt matrix around stones 1119. 06/03/99 3-temple Y
1119 A:TP5 Packing Stone packing in foundations of the South Temple. 06/03/99 3-temple
1120 A:TP1 Layer Compact, clay rich layer below 1110, cut by 1114. 07/03/99 TP1-2 2 Y 823
1121 A:TP9 Deposit Silt in between stones, foundation layer of 07/03/99 3-temple Y 268
mandapa, North Temple.
1122 A:TP3 Fill Stone layer in foundation fill of 1171. 07/03/99 3-temple
1123 A:TP5 Layer Compact clay layer under 1071, South Temple 07/03/99 3-temple 33
foundations.
 List of recorded contexts   377

Context Trench Type Description Date Trench Period Bone Sherds

phase (approx)

1124 A:TP5, Layer Compact layer in pre-temple deposits. 07/03/99 3 Y 1,217

TP3
1125 A:TP1 Fill Clay fill of foundation cut 963, below 1086. 07/03/99 3-temple Y
1126 A:TP1 Fill Deposit below 1120, fill of pit 1131. 07/03/99 TP1-2 2 Y 798
1127 A:TP1 Layer Compact clay deposit below 1120. 07/03/99 TP1-2 2 Y 494
1128 A:TP5 Layer Compact clay foundation layer under 1119 in 07/03/99 3-temple Y
South Temple foundations.
1129 A:TP3 Layer Pre-temple deposit consisting of thin ash lenses. 07/03/99 3 612
1130 A:TP5 Layer Soft silt under 1123 in South Temple foundations. 07/03/99 3-temple 718
1131 A:TP1 Cut Pit cut, filled with 1126. 07/03/99 TP1-2 2
1132 A:TP9 Layer Red brick jelly under 1115, mandapa, North 07/03/99 3-temple
Temple.
1133 A:TP9 Layer Yellow brick jelly under 1132, mandapa, North 07/03/99 3-temple
Temple.
1134 A:TP3 Layer Ashy layer in pre-temple deposits. 07/03/99 3
1135 A:TP5 Deposit Soft silt matrix around stones 1136, South Temple 07/03/99 3-temple Y 151
foundation.
1136 A:TP5 Packing Stone packing under 1128/1135, South Temple 07/03/99 3-temple
foundations.
1137 A:TP9 Struc- Possible traces of shrine under 1132, mandapa, 07/03/99 3-temple
ture North Temple.
1138 A:TP3 Fill Fill of 1139. 07/03/99 3
1139 A:TP3 Cut Post-hole, cutting 1134. 07/03/99 3
1140 A:TP3 Layer Silty clay layer in pre-temple deposits. 08/03/99 3 808
1141 A:TP1 Layer Small patch of burning on 1127. 08/03/99 TP1-2 2
1142 A:TP9 Surface Surface of mandapa, North Temple. 08/03/99 3-temple
1143 A:TP9 Layer Brick jelly under 1142, mandapa, North Temple. 08/03/99 3-temple Y
1144 A:TP1 Fill Fill of pit 1145. 08/03/99 TP1-2 1 Y 1,238
1145 A:TP1 Cut Pit cut (fill 1144). 08/03/99 TP1-2 1
1146 A:GG Layer Stone layer under garbhagriha, North Temple. 08/03/99 3-temple
North
1147 A:GG Layer Fill of foundation of Phase I shrine, beneath 1146, 08/03/99 3-temple Y 329
North North Temple.
1148 A:TP3 Layer Pre-temple layer with remnant of brick surface. 08/03/99 2 Y 1,906
1149 A:TP9 Layer Crumbly silt under possible ‘altar’, mandapa, 08/03/99 3-temple
North Temple.
1150 A:TP1 Fill Second fill of pit 1145, below 1144. 08/03/99 TP1-2 1 Y 2,215
1151 A:TP9 Layer Hard clay fill of mandapa foundations, North 08/03/99 3-temple Y 349
Temple.
1152 A:GG Layer Fill in Phase I foundations, North Temple. 08/03/99 3-temple 314
North
1153 A:TP5 Packing Stone packing under garbhagriha, South Temple 09/03/99 3-temple
(under 1130).
1154 A:TP3 Layer Lowest level of pre-temple deposits reached 1999. 09/03/99 2
1155 A:TP3 Layer Lowest level of pre-temple deposits reached 1999. 09/03/99 2 Y
378   Appendix I

Context Trench Type Description Date Trench Period Bone Sherds

phase (approx)

1156 A:TP9 Layer Slightly clayey yellow silt in mandapa foundations, 09/03/99 3-temple
North Temple.
1157 A:TP9 Layer Lowest stone layer in mandapa foundations, North 09/03/99 3-temple 389
Temple.
1158 A:TP1 Fill Primary fill of pit 1145. 08/03/99 TP1-2 1 Y 43
1159 A:TP1 Layer Subsoil above natural. 08/03/99 TP1-1 1 4
1160 A:TP1 Natural Natural soil. 09/03/99 TP1-1 1
1161 A:TP5 Layer Compact clay under stones 1153 in South Temple 08/03/99 3-temple
foundations.
1162 A:GG Layer Stone layer in garbhagriha foundations, North 09/03/99 3-temple
North Temple.
1163 A:GG Layer Hard layer below 1162 in garbhagriha founda- 09/03/99 3-temple Y 218
North, tions, North Temple.
TP5
1164 A:GG Layer Looser layer below 1163 in garbhagriha founda- 09/03/99 3-temple Y
North tions, North Temple.
1165 A:TP9 Layer Lens of brick jelly under 1168, mandapa of the 09/03/99 3-temple
North Temple.
1166 A:TP9 Cut Cut into foundation deposits of mandapa, North 09/03/99 3-temple
Temple.
1167 A:TP5 Fill Silt matrix surrounding stones 1174 under 1161, 09/03/99 3-temple
South Temple foundations.
1168 A:TP9 Layer Hard surface under walls of mandapa, North 09/03/99 3-temple
Temple.
1169 A:TP9, Surface Brick dust surface under Phase I shrine, North 09/03/99 3-temple
GG Temple.
North
1170 A:TP9 Layer Hard layer in bottom of foundation trench of 10/03/99 3-temple 244
mandapa wall, North Temple.
1171 A:TP3, Cut Foundation cut for garbhagriha, North Temple. 10/03/99 3-temple
GG
North
1172 A:GG Layer Stone layer below 1164, in Phase I foundations, 10/03/99 3-temple
North North Temple.
1173 A:GG Layer Bottom foundation fill of 1171, Phase I North 10/03/99 3-temple
North Temple.
1174 A:TP5 Layer Stone layer in 1167, under garbhagriha, South 10/03/99 3-temple
Temple.
1175 A:GG Natural Natural soil. 10/03/99 3-temple
North
1176 A:TP5 Layer Compact clay under 1174, garbhagriha, South 10/03/99 3-temple
Temple.
1177 A:TP7 Fill Fill of foundation cut 1178. 10/03/99 Y 209
1178 A:TP7 Cut Foundation cut of southern mandapa wall, South 10/03/99
Temple (fill 1177).
1179 A:TP5 Layer Silt layer under 1176, garbhagriha foundations, 11/03/99 3-temple
South Temple.
1180 A:TP5 Layer Compact silt under 1179, garbhagriha founda- 11/03/99 3-temple
tions, South Temple.
1181 A:TP5 Cut Bottom of foundation cut under the South Temple. 11/03/99 3-temple
1182 A:TP8 Layer Hard surface layer on north eastern corner of the 11/03/99 242
South Temple.
 List of recorded contexts   379

Context Trench Type Description Date Trench Period Bone Sherds

phase (approx)

1183 A:TP1 Fill Clean silt fill of foundation cut for the South 11/03/99 3-temple
Temple observed in section only.
1184 A:TP8 Layer Silt layer underlying 1182. 11/03/99 Y 135
1185 A:TP8 Layer Less compact layer under 1184. 12/03/99 177
1186 A:TP5, Surface Top of Phase I foundation fill, North Temple (equals 11/03/99 3-temple 479
TP9, GG 927).
North
1187 A:TP7 Fill Fill of cut 1188. 11/03/99
1188 A:TP7 Cut Small rectangular cut filled with 1187. 11/03/99
1189 A:TP7 Fill Fill underlying 342. 11/03/99 102
1190 A:TP7 Packing Rammed floor with rubble packing cut by 372. 11/03/99 128
1191 A:TP7 Surface Compact surface. 11/03/99 121
1192 A:TP3, Fill Compact layer in which stones 1069 are set, fill of 11/03/99 3-temple
TP5 982.
1193 A:TP5 Fill Thin mud, uppermost fill of foundation for the 11/03/99 3-temple
South Temple.
1194 A:TP5 Fill Green silt and stones below 1193, part of upper- 11/03/99 3-temple
most fill of foundation for the South Temple.
1195 A:TP5 Fill Green grey silt, fill of foundation cut 963 for the 11/03/99 3-temple
South Temple.
1196 A:TP9 Layer Pre-foundation of the North Temple mandapa 11/03/99 3-temple Y 524
deposit.
1197 A:TP7 Layer Yellowish brown silt layer overlying 1190. 11/03/99 Y
1198 A:TP7 Fill Fill of cut 1199. 11/03/99
1199 A:TP7 Cut Circular cut, filled by 1198. 11/03/99
1200 A:TP7 Layer Thin yellowish silt deposit underlying 1197. 11/03/99 67
1201 E Fill Loose grey fill of cut 1202. 07/03/99
1202 E Cut Circular cut, possibly a post-hole. 07/03/99
1203 E Layer Compact ashy deposit. 07/03/99 Y 207
1204 E Fill Fill of post-hole 1205. 08/03/99
1205 E Cut Cut of post-hole cut, filled by 1204. 08/03/99
1206 E Layer Loose brown clayey silt deposit, possibly a 08/03/99 Y 1,364
surface.
1207 E Layer Rubble packing. 09/03/99 581
1208 E Layer Compact yellowish brown silt deposit. 09/03/99 Y 408
1209 E Layer Brownish clayey deposit. 09/03/99 190
1210 E Layer Semi-compact brownish deposit. 09/03/99 153
1211 A:TP7 Layer Ashy patch underlying 1200. 13/03/99
1212 A:TP5 Layer Lowest stone packing in construction levels under 13/03/99 3-temple
the South Temple.
1213 A:N Wall Moulding added to north face of wall 514. 13/03/99
Temple
1214 A:TP6 Fill Fill overlying 373, south-west of the South Temple. 13/03/99 147
1215 A:TP8 Cut Cut. 13/03/99
1216 A:TP8 Layer Floor deposit underlying 1182. 13/03/99 98
380   Appendix I

Context Trench Type Description Date Trench Period Bone Sherds

phase (approx)

1217 A:N Feature Striations on upper surface of wall 555 of the 14/03/99
Temple North Temple.
1218 A:N Wall Ledge on west side of garbhagriha, North Temple. 14/03/99
Temple
1219 A:TP8 Packing Stone packing underlying 1245. 14/03/99 76
1220 A:TP8 Floor Compact pinkish silt clay floor abutting eastern 14/03/99 448
wall of temple and underlying 1219.
1221 A:TP7 Fill Fill within the foundation core make-up of the 14/03/99 59
South Temple.
1222 A:TP8 Fill Rubble fill of foundation cut 1226. 14/03/99 88
1223 A:TP8 Packing Stone packing fill underlying 1222. 15/03/99 33
1224 A:S Wall Foundation wall for Phase I of the South Temple. 15/03/99 3-temple
Temple
1225 A:TP7 Fill Fill within the foundation core make-up of the 15/03/99 103
South Temple.
1226 A:TP8 Cut Possible cut of foundation trench for the South 15/03/99
Temple, filled with 1222 and 1223.
1227 A:TP8 Surface Hard surface below floor 1220. 15/03/99 183
1228 A:TP8 Fill Hard surface below 1227, may be foundation fill. 16/03/99
1229 A:TP7 Packing Stone packing part of core filling of foundation of 16/03/99
South Temple mandapa.
1230 A:TP7 Cut Foundation cut of the South Temple in south-east 16/03/99
corner.
1231 A:TP7 Layer Deposit underlying 1191 and cut by 1230. 16/03/99
1240 A:S Wall South Temple Phase II wall, north side. 16/03/99 3-temple
Temple
1241 A:S Wall Phase V? rebuild of south wall, South Temple 16/03/99 3
Temple (equals 1242).
1242 A:S Wall Phase V? rebuild of south wall, South Temple 18/03/99 3
Temple (equals 1241).
1243 A:TP9 Cut Foundation cut for mandapa, North Temple. 16/03/99 3-temple
1245 A:TP8 Fill Yellowish silt, top fill of foundation (was 1218A). 14/03/99
1246 A:TP6 Cut Foundation cut of the South Temple on south side. 16/03/99 3-temple
1247 A:S Wall South Temple Phase II wall, south side. 17/03/99 3-temple
Temple
1248 A:N Wall Late rebuild of mandapa wall 557, south side, 17/03/99
Temple North Temple.
1249 A:TP5 Cut Foundation cut of wall thickening 528 inside the 17/03/99 3
South Temple.
1250 A:TP5 Cut Foundation cut for the South Temple foundation 18/03/99 3-temple
under 1212.
1251 A:S Wall Phase II wall of garbhagriha of the South Temple. 29/10/06 3-temple
Temple
1252 A:TP9 Layer Layer numbered ‘551 cleaning’ in 1998 season. 21/03/98
General Author Index
Ali, Daud 4, 324–325 Mirashi, V. V. 15–17
Bakker, Hans 4, 316 Morwanchikar, R. S. 11–14, 16, 234, 242, 307
Bhandarkar, R. G. 11, 17 Ray, H. P. 300, 313–315, 317
Chattopadhyaya, B. D. 3–5, 315–316, 324–325, 327 Sharma, R. S. 315–316, 325–327
Collingwood, R. G. 328 Singh, Upinder 326
Dikshit, M. G. 18–20, 57, 86, 98, 299–303 Sircar, D. C. 3, 12
Dirks, Nicholas 325 Spivak, Gayatri Chakravorty 326, 328
Eschmann, A. 4, 308, 316 Stein, Burton 325
Gramsci, Antonio 326 Thosar, H. S. 16
Heitzman, James 325 Wheeler, Mortimer 329, 330
Inden, Ronald 325 Willis, Michael 316–317, 328, 331
Jain, J. C. 12 Yazdani, G. 12
Jha, Vishwa Mohan 324–325 Yusuf, Syed 13, 18, 19, 22, 24, 25, 27, 28, 31, 39–41, 43, 44, 53, 54,
Kosambi, D. D. 316 56, 60, 64–66, 68, 70, 72, 73, 76, 77, 86, 226, 300–303, 307
Kulke, H. 4, 316–317

https://doi.org/10.1515/9783110653540-018
General Subject Index
absolute chronology 59, 60, 304 BrahmaPurāṇa 12, 13
Adam 226, 254, 264, 314, 331 LingaPurāṇa 12
agrarian expansion 4, 313, 315 MatsyaPurāṇa 12
agriculture, see also crops; livestock PadmaPurāṇa 12
agricultural development 316 Bharuch 12, 13
agricultural economy 308–318 Bhitargaon 36, 61, 306
agricultural labour 277, 308–309 Bhitari 58, 61, 304
agricultural methods 314, 316 Bhojakas 15
agricultural production 275, 280, 309, 314, 316–317 Bhokardan 18, 57, 101, 112, 115, 116, 118, 119, 125, 129, 139, 227,
commercialisation 308 242, 254, 264, 284, 313, 319, 320
diversification 252, 258, 277, 280, 308 Bhumara 61
household-based processing 277, 308–309 , 314 Bhuvaneshvara 62
intensification 247, 258, 277, 280, 308–311, 315 Bodhgaya 61, 306
subsistence 310 bone 71, 87, 183, 194, 233, 242, 281–295
Ajanta Caves 61 bone objects 87, 183, 194, 233, 242
Alampur 57 faunal remains 281–295, 309
Amaravati 17, 234, 305 age at death 289
animals, see also bone; livestock Minimum Animal Units 286–287, 288, 290–291
bird 233, 283, 288 Minimum Number of Individuals 286
blackbuck 283 Number of Identified Specimens 282–283, 285–287
cattle 283, 285–289, 291, 292–295, 311 human remains 71
crocodile 283, 289 borers 185, 188
deer 283 bowls 105, 112, 115, 118, 119, 122, 127, 129, 131, 133, 136, 185, 194,
dog 281–282, 283, 289, 309, 311 196–198, 199
fish 283, 285, 288 box trenches 10, 329
goat 281–290, 292, 293, 309–312 Brahmagiri 57, 93
horse 283 Brahmapuri 14, 18, 93, 106, 107, 115, 118, 124, 125, 139, 164
jackal 283 Brahminism 3, 4, 16, 57–58, 305, 313–317, 333–334, 335
monkey 283 brass 184
pig 281–282, 284–295, 309–312 brick drain 19, 300–302
sheep 281–290, 292, 293, 309–312 British period 3, 97, 225, 303
turtle 283 bronze 184–185
water buffalo 283–284 buckles 185, 193
zebu 283 Buddhism 314, 315
Annales school 326 Buddhist texts 11–12
Archaeological Survey of India 6, 8, 10, 324, 328, 329 Anguttara Nikāya 11, 12
area excavation 10, 62, 69, 329 Jatakas 11, 313
Arrian 13 Assaka Jataka 11
arrowheads 87, 185–187, 233 Baveru Jataka 11
Arthashastra 311, 313 Chulla Kalinga Jataka 11
Asaf Jahi period 303 Mahabharata 11, 12
Aśmaka 11–12, 15, 16 Mahāmayurī 12
Asoka 15 Sutta Nipāta 11, 12
Assaka 11 butchery 288
axles 193
C14 dates 95, 96, 106
Bahmani Sultanate 228 cereals 246–247, 252, 253–263, 275, 280
Ahmed Shah II 228 barley 246, 247, 253, 254, 256, 257, 258–259, 313
balls 193, 196, 198, 201 emmer 258
bangles 84, 158, 171–179, 232, 233, 242 millet 247, 248–249, 253, 255, 256–257, 259–263, 280, 308,
Bāvari 11 309, 313, 314, 315, 317
beads 19, 84, 87, 158, 160–171, 232, 242, 333 rice 246, 247, 253, 254, 255–258, 280, 313, 314–315
Berenike 273, 274, 280 sorghum 252, 253, 254, 256, 259–261, 280, 313
Bhadrabahu 12 wheat 6, 246, 247, 253, 254, 258–259, 313, 314
Brahmanic texts 12–13 Chakradhara 14
Mahabharata 11, 12 Chalcolithic 158, 234, 245, 247, 255, 258, 260, 262, 265, 266, 267,
Purāṇas 12, 13 276, 277, 278, 279, 309

https://doi.org/10.1515/9783110653540-019
 General Subject Index   383

Chalukyan architecture 57 flotation 5, 10, 244, 245–246, 255, 280

Chammak 334 fodder 255, 270, 276, 278
Charsadda 329 folklore 11, 14–15, 17
chisels 185, 187–188, 233
Chejerla 60, 61 gahapati 314, 317
coins, see also hoards; monetization glass 165–167, 174–176, 180, 182, 194, 213, 214, 232, 233, 242, 321
British coins 19, 225 Godavari river 4–6, 8, 9, 11–14, 17, 19, 24, 244, 280, 314–316, 335
coin circulation 3, 91, 96, 97, 318–320 gold 182
coin deposition 96, 97, 229–231, 302, 308, 318–320 Gunāđhya 13
‘elephant’-type 219, 221, 227, 235, 237 Gupta period 3, 228, 313, 316, 317, 318, 320, 325
‘hollow cross’-type 96, 215–217, 225, 226, 235, 300 Gupta architecture 58, 61
Indo-Sasanian coins 228, 238 Gupta coins 228, 238
Islamic coins 225, 228 Kumara Gupta 228, 238
Kshatrapa coins 221, 227, 228, 231, 232, 238
Maitraka coins 62, 228, 231 Harappa 263, 269, 270–272, 276, 329
post-Satavahana coins 97, 221, 225, 228, 231, 319 Harappan civilization 112, 256, 258, 264–265, 267, 268, 269, 270,
pre-Satavahana coins 95, 96, 215–217, 226–227, 231, 232, 300, 275, 276
319 hemp 275–276
punch-marked coins 96, 226, 234–235, 300 Hinduization 308, 316
Sarvva-Bhattaraka coins 62, 228, 231 historiography 3, 312, 324–328
Satavahana coins 19, 20, 91, 95, 96, 219–221, 223, 227–228, hoards 318, 320
231, 232, 237–238, 239, 319 Amaravati hoard 234
‘tree’-type 217, 226–227 Jogalthembi hoard 228
copper 19, 171–174, 179–180, 182, 184–185 Wategaon hoard 227
copper-plate inscriptions 3, 15–16, 315, 316, 331–334
cotton 6, 13, 270–274, 280, 308, 309, 317 Iron Age 254–257, 261, 264–266, 268, 277, 279, 333, 334
cowrie shells 285, 288, 319–320 iron objects 19, 185–193, 196, 233, 321
crops, see also agriculture; cereals; cotton; flax; pulses iron slag 213
cash-cropping 274, 275, 308, 310 irrigation 6, 247, 258, 274, 308, 313–316
crop-livestock interaction 309, 312 Isanasivagurudevapaddhati 57–58
cropping strategies 280, 308, 309, 312, 314, 315 ivory 19, 176–177, 242, 321
crop-processing 246, 260, 276, 277, 309
crop residues 309, 310 Jaggayyapeta 305
double-cropping 247, 260, 280, 308, 310, 315 Jaina texts 12
fallow periods 247, 280, 308 Āvaśyakaniryukti 12
summer crops 247, 253, 257, 268 Bṛhatkalpabhāṣya 12
surplus 274, 308, 310, 313, 314 GacchācārṾrtti 12
winter crops 246, 247, 253, 258, 263, 267–268, 275–276, 280, Kālakāchārya Kathānaka 12
308, 314, 315 Kalpasūtra 12
Niśīthacūrṇi 12
dabbers 199 Piṇḍadaniryukti 12
dairy products 289 Tiloya-Paṇṇati 12
Darra 61 Vasudeva Hindi 12
Daulatabad 14, 228 Junnar 314
Delhi school 325, 326
Deogarh 61, 306 Kanheri 15, 314
Dhanyakataka 17 kaolin 19, 56, 205, 208, 233, 239, 242, 307
Dnyaneshwara 14 knives 187, 233, 283
kohlsticks 19, 84, 182, 242
ear-studs 180, 181, 233, 242 Kshatrapas 14, 15, 221, 227, 228, 231, 232, 238
eggplant 268 Chashtana 14
Eknath 14 Nahapana 12, 227, 228, 238
Elephanta Caves 61 Kuda 14, 17
Eran 326, 327 Kunda 61

Fa-Hsien 311 lamps 198, 199

fasteners 184, 191 land ownership 314, 316
feudalism 3, 316, 325, 326 endowments 313, 315–317, 333, 335
figurines 15, 19, 56, 108, 204, 205–213, 239, 241, 242, 307 grants 3, 4, 314–316, 324, 331–333
flax 274–275 Laws of Manu 309, 311
384   General Subject Index

lead objects 193 Pliny 13

Limbārāmikā 16 population growth 280, 313, 314, 317
lime mortar 28, 43, 44, 56, 67, 70, 71 porcelain 92, 97, 194, 204, 213
livestock, see also animals; bone pottery 4, 19–20, 32, 57, 60, 62–65, 70, 73, 81, 84, 86, 87, 90–93,
age at death 289 95, 97, 101–157, 199, 233, 234, 320–321, 323
cattle 283, 285–289, 291, 292–295, 311 Black and Red ware (B&R) 91–92, 95, 101–102, 104–105,
crop-livestock interaction 309, 312 112–113, 119, 131, 201, 204, 300
goat 281–290, 292, 293, 309–312 Black Burnished ware (BBW) 112, 119, 127, 131, 137
pig 281–282, 284–295, 309–312 Blue and White Frit (PERS) 93, 97, 108–110, 111, 149
sheep 281–290, 292, 293, 309–312 Chinese Blue and White porcelain (CBW) 93, 108, 110
transhumant migration 310 Jorwe ware 92, 96, 105–106, 119, 234, 300
Northern Black Polished ware (NBP) 19, 91–93, 95, 106–107,
Mahānubhāva literature 14 113, 149, 300
Leelā-Charitra 14 Red Polished ware (RPW) 84, 91–93, 95, 107–108, 110, 113, 149,
Smṛtisthala 14 234
Sthānpothī 14 Slip-painted ware (SLIP) 93, 110, 111, 149
Maheshwar 3, 93, 101, 107, 111–113, 115, 116, 118, 119, 125, 129, 131, Thick Grey ware (GREY) 73, 84, 85, 91, 92, 95, 97, 104, 111–113,
136, 139, 141, 226, 319, 320 122, 125, 133, 139
Mahurjhari 333, 335 White and Red Painted (WARP) 105, 110, 111, 136, 149
Mandhal 331 Prakash 18, 101, 105, 107, 108, 111–113, 115, 116, 118, 119, 133, 139,
Marathi literature 14 141, 226
Marxist perspectives 3, 316, 325–326 Prakrit literature 13
Mauryan period 3, 6, 15, 107, 300, 313 Bṛhatkathā 13
Megalithic 256, 271, 333 Bṛhatkathā-Manjirī 13
microliths 92, 158, 159, 300 Gāthā Saptaśati 13
millet 247, 248–249, 253, 255, 256–257, 259–263, 277, 278, 279, Kathā-Saritsāgara 13, 16
280, 308, 309, 313, 314, 315, 317 Ptolemy 11, 13–14, 16–17
African millets 247, 255, 259–261, 308, 309, 314, 317 pulses 6, 246–247, 252, 257, 258, 263–268, 280, 308, 315
broomcorn millet 261–262 chickpea 263, 267, 268, 308
browntop millet 262–263 cowpea 266
Brachiaria ramosa 247, 248, 252, 254–257, 260–262, 280 grasspea 263, 267, 308
Chinese millets 261–262 horsegram 266–267
foxtail millet 261–262, 280 hyacinth bean 267
finger millet 259–261 lentil 263, 266–268, 308
Indian small millets 262–263 moth bean 265–266
kodo millet 247, 254–257, 262, 263, 277, 280, 308 mung bean 264–265
pearl millet 253, 256–257, 259–261, 280 pigeonpea 267
sawa millet 262–263 urd bean 264–265
Mohenjo-daro 270, 271, 329 winter pulses 246–247, 257, 258, 267–268, 280, 308, 315
monetization 318–320
Mughal period 19, 84, 87, 225, 228, 303, 309 querns 19, 87, 198, 242
Aurangzeb 225, 228 Qutb Shahi Sultanate 225, 228
Mūlaka 11, 12, 15, 16
Murundas 12 Raika 310
Rajim 62
Nachana 61 Ramtek 33, 60, 62
Naga 12–15 Rashtrakutas 16, 19–20, 57, 302, 304, 315, 316
Nagara 18, 107, 304 Govind III 16
Nagarjunakonda 305 Indra III 16
Nasik 15–17, 18, 93, 105, 107, 115, 116, 118, 119, 122, 125, 133, 139, Rathikas 15
141, 163, 167, 182, 194, 281, 284, 304, 314 rings 179–180, 191, 242
Neolithic 245, 247, 253, 255, 256, 258, 259, 262–268, 271, 277–279 finger-rings 179–180, 242
Nevasa 18, 58, 90, 93, 95, 101, 105–108, 111–113, 115, 116, 118, 119, toe-rings 180
122, 124, 125, 129, 131, 139, 177, 179, 182, 194, 219, 227, 234, ring-wells 84–85, 301
235, 237, 254, 257, 264, 271, 275, 281, 284, 304, 313, 319 321 rivets 184, 191
Roman trade 274, 280, 321
Pattadakal 57 Roman amphorae 95, 108, 152, 321
Periplus Maris Erythraei 11, 13, 274, 309, 321 rural economy 312–314, 323
Petenikas 15
Pitalkhora Caves 15
 General Subject Index   385

safflower 275 kumuda 61

Sakor 61 mandapa 62, 304, 306–308
Sanchi 18, 314, 315, 328, 329 mulaprasada 62, 305–306
sangha 314 shikhara 61, 306, 307
Satavahanas 6, 12–17, 19–20, 88–89, 91, 95, 96, 219–221, shrine 305–307
227–228, 231, 242, 237–238, 280, 302, 304, 311–312, standard temple plan 304–305
313–316, 320 triratha 58, 61, 306
Gautamiputra Yajnasri Satakarni 237–239 economic and political role 3–4, 313, 315–317, 324, 325–326,
Gotamiputa Siri Satakani 15–17, 96, 227–228, 238 333, 335
Hāla 13
Kosikiputa Siri Satakani 219, 227 Ter 13, 18, 62, 107, 115, 116, 118, 119, 122, 125, 129, 139, 141, 205,
Sālivāhana 12, 14–16 242, 274, 309, 313
Vasithiputa Siri Pulumavi 13, 15, 227–228, 237 terracotta 14–15, 19, 56, 84–85, 160, 169–171, 179, 180–182,
Satdhara 328 198–201, 205–214, 232–233, 234, 238–242, 307
sati 326–327 textile production 270–274, 280, 308–309
Satpura hills 331 Tigawa 61
sculpture 56, 213, 224, 305, 307, 334 toys 14–15, 201–205, 233
seals and sealings 16, 84, 232, 242 typology 102, 113, 139, 149
sesame 268–269
shell 168, 177–179, 180, 213, 232, 242, 283, 288, 319–320 Ujjain 14, 15, 274
Sirpur 62 urban archaeology 98, 329
skin rubbers 84, 199, 233, 242 urbanization 314
Strabo 13
subaltern studies 326–327 Varkari 14
Sufism 14, 228 Vakatakas 15, 95, 304, 313, 315, 318, 331–334
Pravarasena II 15, 334
Taxila 93, 226, 329 vases 198, 199, 234
temples 23–80, 304–308 Veerapuram 254, 256, 264, 305
Buddhist cave temples 15 Vidarbha 14, 17, 226, 331–335
brick temples at Paithan 5–6, 23–80, 92–93, 95–97, 149, Vidisha 314–316
304–308 Vikrama 14, 15
1937 excavation 19, 24, 27, 41
1965 excavation and restoration 19, 27 weeds 246, 252, 253, 262–263, 276–279, 309, 314, 317
coin deposition 229–231 Wheeler method 329
faunal remains 288–289 wire 185, 193
early Hindu temple architecture 23, 307–308
adhisthana mouldings 33, 61–62, 306, 307 Yadava period 14
foundations 56–58, 305 Yelleshvarum 305
garbhagriha 61, 304