Chapter III

REGIONAL CONCEPT OF PUNDRAVARDHANA

The geographical extent of Pundravardhana as a bhukti (province) having political

jurisdiction over the territory and the people can be marked out from the evidence of the
epigraphic records issued by members of the ruling families who held sway in the region
from time to time. The Mauryas and the Guptas, whose main political base was the
middle Gangetic valley, ruled over Bengal and other parts of eastern and northern India
from their capital, Pataliputra. In the post-Gupta period, the Palas of Bengal ruled over
parts of eastern and northern India at certain points of time and were involved in the
tripartite struggle against the Rastrakutas and the Gurjara-Pratiharas for the occupation of
Kanauj. The Candras, Varmans and Senas basically ruled over smaller or larger parts of
Bengal during the period of their hegemony. Of these four dynasties, the Varmans
established their rule over eastern Bengal; and a portion of Pundravardhana formed part
of their domain for a brief period. But the Palas, Candras and Senas not only ruled over
Pundravardhana along with almost the rest of Bengal but even the seat of their authority
(or capital) was sometimes located within Pundravardhana itself Nonetheless, the
territory of Pundravardhana was endowed with the status of a bhukti or province under all
these imperial powers, although the boundary of the province changed from time to time.
The chronological sequence of the development of Pundravardhana is given below on the
basis of epigraphic evidence.

Changing Extent of Pundravardhana-6/rMA://

The Mahasthangarh Fragmentary Stone Inscription of the third century BC' provided the
earliest reference to Pundra as a political division during the Maurya period. In fact, the
Mahasthangarh inscription was the earliest epigraphic source recorded in the early

' For the full text see R.R. Mukherji & S.K. Maity (eds.), Corpus of Bengal Inscriptions, Firma K.L.
Mukhopadhyay, Calcutta, 1967, pp.39-40; Barrie M. Morrison, Political Centers and Cultural Regions in
Early Bengal, Ravvat Publications, Delhi, ]980, pp. 13-14. The details of the landgrants in the
Puiidravardhana-6/;;^A// are given in Table I:. Geographical Extent of Pundravardhana as reflected in the
inscriptions.

54
Mauryan BrahmT character to have been recovered from Mahasthan in the Bagura district
(now in Bangladesh). It recorded that the sadvarika (local administrator or governor) of
the prosperous city of Purjdanagala (i.e., Prakritized form of Pundranagara of the Sanskrit
records) was ordered by the imperial authority to grant paddy, sesamum and mustard
seeds to the SamvamgTyas (i.e., followers of the Sadvargika sect qf Buddhism in Bengal)
during any period of distress or emergency, to be replenished in due course with paddy
and coins.^ The reference to Pundanagala in this inscription confirms the identification of
Pundranagara with Mahasthangarh and indicates that the Pundra region formed an
administrative division within the Mauryan Empire. Two votive inscriptions at Sanchi
(dated the second century BC) referred to Punavadana (i.e., Pundravardhana) while
recording the gifts by two inhabitants from that place (i.e., Punavadana) for the
improvisation of the Sanchi stupa^ Some scholars suggested that 'Punavadana' should
actually be read as Pundravardhana."*

The expansion of the Magadhan Empire during the Gupta period (from the fourth century
to the sixth century) witnessed certain modifications in the territorial organization of the
areas subsumed under the Empire. The region corresponding to Bengal (consisting of
Sumha, Radha, Pundra and Vaiiga or Tamralipta, Radha, Varendra, Vanga, Samatata,
etc.) could not escape the changes brought about by the expanding influence of the
imperial Guptas. The contents of the Mehrauli Pillar Inscription of Candragupta I
(located in Delhi), the Allahabad Pillar Inscription of Samudragupta, and other epigraphic
documents of the Gupta period have been interpreted by scholars to suggest that Vahga,
Samatata, Tamralipta, Pundravardhana and Vardhamana-^rtwA:// were parts of the Gupta
Empire.^ As R.K. Mookherji has shown, Bengal was conquered by Candragupta I. The
Mehrauli Pillar Inscription recorded Candragupia's victory over the coalition of the

^ R.R. Mukherji & S.K. Maity (eds.), op.cit., p.40.

^ A.K. Sur, History and Culture of Bengal, Chuckerverti, Chatterjee & Co., Calcutta, 1963, pp.82-83; B.N.
Choudhury, Buddhist Centres in Ancient India, Sanskrit College, Calcutta, 1969, p. 196; Parmanand Gupta,
Geography in Ancient Indian Inscriptions, D.K. Publishing House, Delhi, 1973, p.93.
"• B.N. Choudhury, op.cit., p.l96.
' Anjali Chattejee, Bengal in the Reign of Aurangzeb (1658-1707), Progressive Publishers, Calcutta, 1967,
p.l. For the Mehrauli Pillar Inscription of Chandragupta 1 see D.C. Sircar (ed.). Select Inscriptions bearing
on Indian History and Civilization, Vol.11, Motilal Banarsidass. Delhi 1983, pp.140-145; and for the
Allahabad Pillar Inscription of Samudragupta, see J.F. Fleet (cd.). Corpus Inscriptionum Indicarum, Vol.Ill,
Indological Book House, New Delhi, 1970, pp. I-8.

55
chiefs of Bengal.^ The Allahabad Pillar Inscription (fourth century) mentioned the rulers
of Samatata, Davaka, Kamarupa, Nepala, Karttrpura, etc., as 'frontier kings' (pratyanta
nrpati) of Samudragupta's empire.^ It may therefore be inferred that Bengal proper (or
Vanga, Tamralipta, Pundravardhana, Vardhamana, etc.) was already a part of the Gupta
Empire during the reign of Samudragupta. Bengal comprised political units such as
bhuktis (provinces), mandalas (divisions) and visaya?, (districts) of the Gupta Empire.
Moreover, even the kingdoms of Kamarupa and Nepala which lay beyond
Pundravardhana were mentioned as 'frontier kingdoms' {pratyanta rajya) in the
inscription. R.K. Mookherji suggested that Bengal continued to be an integral part of the
Gupta Empire even under the successors of Samudragupta, particularly, Candragupta II
and Kumaragupta I. Thus the region of Pundravardhana bounded by the Gahga on the
west and the river Karatoya on the east was under the rule of the Guptas from the second
quarter of the fifth century to about the middle of the sixth century. It included the areas
corresponding to undivided Malda, Dinajpur, Rajshahi, Bagura and Rangpur districts and
parts of Pabna and Murshidabad districts of modern times.^

As stated already, Bengal continued to exist as an integral part of the Gupta Empire under
the successors of Samudragupta upto the reign of Buddhagupta. By the period of
Kumaragupta I, Bengal was administered by his governors ruling over its different
provinces. This fact is borne out by the evidence of a vast range of epigraphic sources,
the important among which were Skandagupta's Bhitari Inscription, and Kumaragupta I's
Damodarpur Copper-plate Inscriptions (AD 443 to AD 448), Baigram Copper-plate
Inscription (AD 447) and the Dhanaidaha grant. The Jagadishpur Copper-plate
Inscription (Gupta Era 128 = AD 447) of Kumaragupta 1, issued from Purnakausika in
Srngavera-vS//r, referred to Pundravardhana, but it did not mention the suffix 'bhuktV.
However, in several other epigraphic records the northern region of Bengal figured as a
regular province of the Gupta Empire under the name Puiidravardhana-ZjAwA//.^ In fact,
Pundravardhana-A/zwA:^/ figured in several inscriptions issued by Kumaragupta I. To cite

* R.K. Mookherji, The Gupta Empire, Delhi, 1925; reprint 1973, Motilal Banarsidass, Delhi, p.24.
' Forthe text of Allahabad Pillar Inscription of Samudragupta see J.F. Fleet (ed.), op.cit, pp.1-8.
' R.K. Mookherji, op.cit., p.82.
' Ibid., p.67; Amiabha Bhattacharyya, Historical Geography of Ancient and Early Medieval Bengal,
Sanskrit Pustak Bhandar, Calcutta, 1977, p.74.

56
one instance, his Dhanaidaha Copper-plate grant (dated Gupta Era 113 = AD 432-33)
recorded the grant of one kulyavapa of cultivated land, measuring 8x9 reeds, in the
Khadapara-v/Mj^a in favour of a Brahmana. The plate was discovered in Dhanaidaha, a
village in the Natore sub-division of the Rajshahi district. Therefore, it may be concluded
that the Khadapara-vwoya was located within the bhukti of Pundravardhana.'" The
plausibility of this conclusion is further confirmed by the Kalaikuri Copper-plate grant
(Gupta Era 120 = AD 439) and the Sultanpur Copper-plate grant (Gupta Era 120 = AD
439), recovered from the villages Kalaikuri and Sultanpur respectively (both located in
the Bagura district). The common date of the two inscriptions indicates that the plates
belonged to the reign of Kumaragupta I. Apart from valuable information relating to the
local administrative system during the Gupta period, the inscriptions recorded the grant
of land in Smgavera-vrz/iL The editors of the inscriptions have identified Srngavera with
either modern Singra police station in the Natore sub-division of Rajshahi district or
Singahar, twenty miles to the south-west of Mahasthan in the Bagura district. The
identification of Srngavera with Singahar seems to be justified in that the Sultanpur plate
further mentioned that a division called Goh&lx-mandala was in Sxngdi\tx2i-vTthT. A village
called GohalT still exists at a distance of seven miles to the east of Singahar in the Bagura
district. This village was possibly the ancient headquarters of the Go\\2X\-mandala. The
Gohali-mawJfl/a and the Srngavera-vFr/jf were, therefore, within the Pundravardhana-
bhukti}^

The five copper-plate inscriptions discovered in Damodarpur village in the Dinajpur

district are essentially religious sale deeds. They recorded the transactions between the
government and the purchasers. The latter were granted land at reduced price calculated
in dinaras or gold coins of the Gupta rulers. Two of the plates belonged to the reign of
Kumaragupta 1. The first Damodarpur Copper-plate Inscription of Kumaragupta 1 (Gupta
Era 125 = AD 444) mentioned the grant of one kulyavapa of land in the region north-
west of Dofiga for three dinaras (gold coins). The grant further recorded that the gift of

'" R.R. Mukherji & S.K. Maity (eds.), op.cit.. pp.41-43.

R.K. Mookherji, op.cit., pp.S!-82. Surprisingly, these two important inscriptions are not included in the
Corpus of Bengal Inscriptiom compiled by P^.R. Mukherji & S.K. Maity.

57
land was according to nividharma (to oblige the recipient to maintain the endowment in
perpetuity), whereby the land could be enjoyed rent-free but could not be sold or
transferred to anybody.'^ The second plate of the same hoard (Gupta Era 129 = AD 448)
recorded the grant of a plot of land measuring five dronas (eight dronas = one
kulyavapa), lying in the west of the village of 'Airavata-Gorajya'.to a Brahmana for two
dinaras at the established rate of three dinaras for each kulyavapa of land.'^ This grant
was made in accordance with the customary rule of apraksayanivi (the same as
nividharma as mentioned in the earlier inscription).''* Both the inscriptions located the
gifted lands in the Kotivarsa-v/jyayo, which visaya was within the Pundravardhana-
bhukti}^ Therefore, it is clear that like Rajshahi, the modern Dinajpur district was also
within the Pundravardhana-ft/zwto'.

The Baigram Copper-plate Inscription (Gupta Era 128 = AD 448), discovered in 1930 in
the village of Baigram in the modern district of Bagura, was issued from a district office
at PancanagarT, the headquarters of the Vayigrama visaya. The inscription recorded that
three kulyavapas and two dronavapas of land located in Trivrta and SrlgohalT were sold
to two Brahmanas. This was after receiving six dinaras and eight silver coins from them
at the rate of two dinaras for each kulyavapa of land as per nividharma. The village of
Baigram (or Vayigrama) is located in modern Bagura district; hence, it is confirmed that
like Rajshahi and Dinajpur, modern Bagura district was also in the heartland of
Pundravardhana. '^

Another important document on the geographical extent of Pundravardhana is the

Paharpur Copper-plate inscription, which was recovered during the archaeological
excavations from a place called Paharpur in the modern Rajshahi district. The epigraph
(dated to the Gupta Era 159 - AD 479) was recorded during the reign of Buddhagupta.
As stated in the document, a district officer issued the grant from Pundravardhana. It
recorded that a Brahmana named Nathasarman and his wife Rami were granted one and a

'^ R.R. Mukherji & S.K. Maity (eds.), op.cit., p.45.

" Ibid., pp.47-49.
''Ibid.
'^ Ibid, pp. 45-49.
'* Ibid., pp.49-53.

58
half dronavapa of land in VatagohalT; four dronavapas at Prsthimapottaka; four
dronavapas at Gosatapunja; and two dronavapas and two adhavdpas (four adhavapas =
one dronavapa and eight dronavapas = one kulyavdpa) in NitvagohalT, in all measuring
one and a half kulyavapas, after receiving three dinaras at the rate of two dinaras for
each kulyavapa of land. This grant was made for charitable works for the Jaina monks
and according to the principle of perpetual endowment and the custom of nmdharma.
The document also informed that the four villages were situated in the area of Palasatta in
the 'NagiraUa-mandala in the Daksinamsaka-vrr/ji".'^ Since Paharpur (in the Rajshahi
district) is situated in the eastern part of the traditional boundary of the province of
Pundravardhana, it is concluded that the four villages mentioned in the document were
possibly located in the eastern region of the Pundravardhana-MwA:^/.'* Similarly, two of
the five Damodarpur plates are dated to the period AD 476-495, i.e., the regnal years of
Buddhagupta. The first inscription (dated Gupta Era 163 - AD 482) recorded the
purchase of one kulyavapa of land on the receipt of two dinaras by Nabhaka, the
headman of the village of Candagrama, from one Nagadeva for the purpose of settlement
of some Brahmanas. Nabhaka obtained the sanction of the governor of the province of
Pundravardhana. The inscription also located the land thus purchased in the south, west
and east in close proximity to the northern boundary of the village of Vayigrama
(Baigram, near Hili in Bagura district).'^ The date of the second inscription could not be
deciphered. However, considering the fact that the name of Buddhagupta is mentioned in
the text, it is assumed that the inscription belonged to the period of his reign (AD 476-
495). The document registered the purchase of one kulyavdpa of homestead land in the
forest region of Himavacchikhara in the village called Dohgagrama by the guild
merchant, Ribhupala, for three dinaras. The purpose of buying the land was to erect
thereupon two temples and store-rooms for the deities Siva and Visnu. The inscription
located the said forest region of Himavacchikhara within the hhukti of Pundravardhana
and the visaya of Kotivarsa. In both cases, the purchase was according to the prevailing
custom of nmdharma. It is also clear that Candagrama and Dohgagrama were both

'^ Ibid., pp.53-58.

" R.K. Mookherji, op.cit., p.l 14.
" R.R. Mukherji & S.K. Maity (eds.), op.cit., pp.58-61.
^"/ftirf., pp.61-65.

59
within the perimeters of Pundrayardhana-bhukti. The Gunaigarh Copper-plate grant of
Vainyagupta (Gupta Era 188 = AD 507) was found in Tipperah district, in the village of
Gunaigarh situated about eighteen miles to the north-west of the town of Comilla and a
mile and a half to the southwest of the police station of Devilvara in the Tipperah (now,
Comilla) district.^' The fact that the Gupta rule extended to Cornilla during the time of
Vainyagupta in the beginning of the sixth century is further proof that Pundravardhana
was till then under the Guptas and that the geographical extent of this bhukti at that point
of time covered parts of Samatata as well.

In the last of the five copper-plate inscriptions discovered in Damodarpur, dated the
Gupta Era 214 (= AD 543), the name of the issuing ruler is not clearly legible, although
the name undoubtedly ended with the suffix "Gupta'. The document registered the grant
of five kulyavapas of fallow land upon the receipt of fifteen dinaras (at the rate of three
dinaras for one kulyavdpd) according to the custom of mvTdharma. The beneficiary was
one Amritadeva of Ayodhya and the grant was meant for the service of the god Visnu.
The contents of the inscription show that the land so granted was located in the five
villages of Svacchandapataka, Lavangasika, Satuvanasramaka, Paraspatika and
Purnavrndlkahari. This inscription also located Kotivarsa-visaya in Pundravardhana-
bhukti, thus indicating that the five villages in which lands were granted were situated in
the Kol\\arsa-visaya within the Pundravardhana-Z>/2M^//.^^

An important epigraphic source of the post-Gupta period is perhaps the Ragholi

Inscription of Jayavardhana, a ruler of the Saila dynasty. It recorded the victory of his
paternal granduncle over the king of Pundra and its conquest around AD 725. However,
there is no other evidence of the Saila rule in north Bengal or any part of ancient
Pundravardhana that has yet been discovered."^ On the other hand, the extant sources are
very clear about the rule of Sasanka in Bengal, including the Pundra region, during the

^' B.C. Law, Historical Geography ofAncient India, Paris, 1934; reprint 1976, Ess Ess Publications, Delhi,
pp.247-248.
^ R.R. Mukherji & S.K. Maity (eds.), op.cit., pp.70-74.
^^ R.C. Majumdar, History of Ancient Bengal, G. Bharadwaj & Co., Calcutta, 1971, p.74.

60
seventh-eighth centuries.^'' Conventional history of course spoke of a period of political
disintegration and disorder in the region after the death of Sasahka. The eighth century
saw the rise of petty chieftaincies and then the eventual rise of the powerfiil Pala dynasty.
Some writers therefore thought that Jayavardhana established his rule in a part of
Pundravardhana for a short period during that phase of political instability but that he
could not extend his authority over the entire area.^^

When the Palas became prominent rulers of Bengal, Pundravardhana seems to have
continued to be suffixed with the term bhukti. This was despite the fact that there were no
overarching imperial systems emanating from any part of northern India. The Khalimpur
Copper-plate Inscription of Dharmapala (discovered at Khalimpur in Malda district)
dated to AD 802 recorded the grant of land made by Dharmapala to SrT Narayanavarman
for the worship and maintenance of the venerable Nanna-Narayana. The gifted land
included the four villages of (a) Crancasvabhra in the Mahantaprakasa-vwoyo, which was
attached to the Byaghratati-wa«Ja/a; (b) Madhasalmali; (c) Palitaka; and (d) Goppali, all
three in the Sthalikatta-v/jflya within the Amvasandika-wan</a/a. The donated land taken
together was bounded on the east by the western boundary of the Udragrama-wa«(/a/o.
The mention of Pundravardhana-^/zz/^/ in the document indicates that the Vyaghratati-
mandala, Amvasandika-7«an(/a/a and Udragrama-.wa«</a/a, along with the
aforementioned visayas and the villages within those mandalas, were included within the
Vundravasdhana-bhukti. The mandalas, visayas and gramas named in the inscription have
not yet been identified beyond doubt, but it has been tentatively concluded that they fell
under the area corresponding to the modern Malda area of north Bengal.^^ The Nalanda
Copper-plate Inscription of Devapala (dated tenth century) also mentioned the grant of
land by Devapala in the village of Vyaghralati within the Pundravardhana-i/?M^//.^^ In
this inscription, Vyaghratati was mentioned only as a village (grama). Similarly, the
Amgachi Copper-plate Inscription of Vigrahapala III (dated to AD 1055), recovered from
a place called Amgachi in the Dinajpur district, recorded the grant of a plot of land in the

^'' R.R. Mukherji & S.K. Maity (eds.), op.cil., p.70; Amitabha Bhattacharyya, op.cit., pp.52-53.
^^ Supra, 23n.
^* R.R. Mukherji & S.K. Maity (cds.), op.cit., pp.95-110.
^^ Amitabha Bhattacharyya, op.cit., p.75.
Kotivarsa-visaya within the Pundravardhana-iAw^// to a Brahmana by the king. The
contents of the document implied that Pundravardhana-hhukti included Kotlvarsa-vwoya
within its territorial limits.^^ The Banagada Copper-plate Inscription of Mahipala 1 (AD
988-1023), which was recovered from the famous remnants of Bangarh in the Dinajpur
district, recorded the gift of a plot of land in the village of Kuratapallika by Mahipala I to
a Brahmana named Bhattaputra Krsnaditya, a resident of the village of Cavati. The grant
was made for religious purpose, presumably for the worship of the Buddha. The land was
donated according to the principles of non-transferable rent-free perpetual endowment.
As inferred from the text of the document, the village of Kuratapallika was situated
within GokaWka-mandala, Kotivarsa-visaya and ?mdrawardhana-bhukti. Further, the
royal proclamation was issued by king Vigrahapala III from the victory camp at
Vilasapura, possibly situated in the northern division of Bengal.^^ Again, the Manhali
Copper-plate grant (recovered from Manhali in Dinajpur district) of Madanapala (first
half of twelfth century), announced the gift of a village named Kautogiri to a Brahmana
named Sri Batesvara-Svamisarman, as remuneration for reciting slokas from the
Mahabharata. The grant was made at the request of Citramatika, the chief queen of the
Pala king Madanapala. The grant further recorded that Madanapala executed the royal
mandate from the victory camp situated in the precincts of the city of Ramavati, while the
village Kautogiri was situated in the Halavarta-/?7a«(/o/a and the Kotivarsa-v/j?aya lying
TA

within the jurisdiction of the Pundravardhana-^/zMfe/."

The Bangladesh National Museum Stone Inscription of Pahila (dated about the ninth
century) has been recovered recently from Nimgacchi in Pabna district (now renamed as
Serajganj district) of present-day Bangladesh. It recorded the gifting of a matha (cloister)
in the Bhattasala-wan</a/fl to the Vaisnava recluses by one Pahila having noticed the
steadfastness of dharma and the transitoriness of wealth, life and youth. The inscription
informed that Pahila was the local administrator of the Bhattasala-mandala under
Devapala of the Pala dynasty, who ruled over Gauda-Magadha. It has been suggested that
the administrative division called Bhattasala-mandala was included within the

^* R.R. Mukherji & S.K. Maity (eds.), op.at., pp. 192-197.

^' Ibid., p.208.
^° Ibid., pp.209-2\9.

62
Pundravardhana-6/2wfe/ and that ihe mandala was situated in the modern Pabna
(Serajganj) district.-" Another recently discovered inscription, the Jagjivanpur Copper-
plate of Mahendrapala (dated the eleventh century, found at Tulabhita in the Jagjivanpur
Mouza of the Habibpur Police Station in the Malda district of West Bengal) recorded that
the king granted an area of land called Nandandighikodranga in the Kuddalakhataka-
visaya in Pundravardhana-6/zwA;// from his victory camp at Kuddalakhataka. The donation
was made at the request of the commander of forces, Vajradeva, for the purpose of
building a monastery {vihara). The grant was issued in perpetuity according to the
bhumicchidranyaya with the approval of the officials present. The grant stipulated that
the neighbouring cultivators would supply necessary cash and ration and that all future
kings would uphold the terms and conditions of the grant.''^

The rise of the Candras in the Harikela-Vanga region (i.e., southeast and east Bengal)
witnessed a significant development in the history of Bengal during the tenth-eleventh
centuries. Founded by one Purnacandra, the Candra dynasty grew in strength during the
reign of its fourth ruler Srfcandra who expanded the boundary of the kingdom when the
Palas were on the decline. SrTcandra conquered Pundravardhana or at least a part of it, as
is evident from some of the inscriptions. The Kampala Copper-plate Inscription (ascribed
to tenth century) recorded the gift o^ one pdtaka of land in Nehakasthi in Nanya-wo«£/o/o
of Pundravardhana-6/zM^// by Sricandradeva, king of Harikela, and son of
Trailokyacandra. The donee was one santivdrika PTtavasaguptasarman. He received the
grant for performing one crore sacrifices. The royal proclamation was issued from the
victory camp at Vikramapura. The copper-plate was discovered at the village of Kampala
in the Munshiganj subdivision of the Dacca district in east Bengal. As known from
various sources, the heartland of the Harikela kingdom was southeast Bengal, precisely
from SrThatta (Sylhet) to Cattograma (ChitJagong); under the Candras this kingdom
expanded to include parts of Vanga (East Bengal) and the CandradvTpa area.^'' This

'' Gouriswar Bhattacharya, "Bangladesh National Museum Prasasti of Pahila (9* century AD)" in Journal
of Bengal Art, Fo/.2, w.4-10, Dhaka, 1997, pp.111-116.
•'^ Gouriswar Bhattacharya, "The New Pala Ruler Mahendrapala" in Pratna Samiksha, Vol.1, Calcutta,
1998, pp. 165-170.
" R.R. Mukherji & S.K. Maity (eds.), op.cit., pp.22I-229.
''' R.C. Majumdar (ed.), History of Bengal, Vol.1, Dacca University, Dacca. 1943, pp.9-11.

63
particular inscription further showed that the area corresponding to northern Bengal also
came under the Candra suzerainty during the period of Sricandra. It can be inferred from
this Kampala Copper-plate Inscription that ?undrawardhana-bhukti included within its
territorial limits 'Manya-mandala within which was situated Nehakasthigrama.^^ The
Candras probably initiated the expansion of their empire from the heartland of Harikela
(i.e., southeast Bengal or the Srlhatta-Mainamati-Cattagrama region) and gradually
spread to northern Bengal (the heartland of Pundravardhana) through the Meghna valley.
The PascTmbhag Copper-plate grant of Sricandra (dated to about the tenth century, which
was discovered in a village called Pasclmbhag in the Maulavibazar subdivision of the
Sylhet district in East Pakistan in 1958) recorded that the Srihatta area (geographically
northeast Bengal) was also included in the Pundravardhana-Z)/2MA:^? during the reign of
Sricandra. The royal charter was executed from the victory camp at Vikramapura. It
recorded that maharaja Sricandra made over land in the three visayas (districts) of
Candrapura, Garala and Pagara situated in Snhana-mandala within the jurisdiction of the
Pundravardhana-Mw/t//. The grant was for charitable purposes. The gifted land extended
over all the three visayas within the Srlhatta-/wa«<;/a/a. In Candrapura-vwoya alone, four
hundred patakas of land were donated to six thousand Brahmanas associated with the
eight mathas (monasteries). The land was granted according to the principle of
bhumicchidra and was to be enjoyed in perpetuity free of rent and any kind of
interference or fear of confiscation." It is clear from the text of the Pasclmbhag
Inscription that Srihatta was a mandala within the ?\xndvavaxd\\ax\a-bhukti, and that
Srlhatta-wa«(/a/a included three districts, namely, Candrapura, Garala and Pagara. Thus,
the V\xndravavdi\\ana-bhukti at the time of Sricandra included the Srihatta region, which
geographically formed the extreme northeastern corner of Bengal (Isvan-Vanga or
northeast Bengal). In this context it may be noted that Bhaskaravarman, king of
Kamarupa, renewed the grant of land to a large number of Brahmanas in the Candrapura-
visaya by the Nidhanpur Copper-plate (dated the seventh century). This grant had
originally been issued by his great grandfather Bhutivarman. It is interesting to note that

^^ Supra, 33n.
^* K.K. Gupta (ed.), Copper-plates of Sylhet, VoLl (f -ll"' Century AD), Self Published, Sylhet, 1967,
pp.81-118.

64
the fmdspot of this inscription was the village Nidhanpur located in the Sylhet district.
The boundaries of the Candrapuri-vwoya of the Nidhanpur grant and those of the
Candrapura-v/5aya of the PascTmbhag Inscription, as described in the respective
inscriptions, are almost identical. On the basis of the data provided by both the
inscriptions, K.K. Gupta prepared a map of the donated villages in the Candrapura-
visaya, which shov/ed that this visaya included parts of the modern Sylhet district (now
Sylhet Division consisting of four districts) in Bangladesh, the three Barak valley districts
of Cachar, Hailakandi and Karimganj in Assam, and the Kailasahar-Dharmanagar area of
modern Tripura.''^ The evidence of this inscription therefore confirmed that
Pundravardhana-i/?i/^// not only covered northern Bengal and parts of eastern, western
and southern Bengal but also included a large portion of southeast Bengal up to the
borders of Tripura. The extent of this bhukti must have been very large under Srlcandra.
His Dhulla Copper-plate grant (dated tenth century) further indicated that
Pundravardhana-^AwA:?/ included Yola-mandala of Ikkadasi-v/j'aya in which the donated
villages of Pakkadimunda and Bahupatra were located. It also included the Ballimunda-
khandala of Khadiravilli-vwoya in which the villages of Loniyajoda, Jivanavilli and
Durbapatra were situated.^* The evidence points to the fact that Pundravardhana-i/iwA://
during the reign of the Candras covered not only northern Bengal but also extended as far
as the Barak-Surma valley and the Tripura plains in the northeast and southeast regions of
Bengal.

Most likely the Candras reorganized the bhukti to extend their jurisdiction over the whole
of southeast Bengal. The Mainamati Copper-plate Inscription (eleventh century) of
Ladahacandra, the grandson of Srlcandra and son of Kalyancandra (recovered from
Mainamati in Comilla district) recorded the grant of land situated in the village of
Suravoraka in Peranatana-vwa>'a within Samatata-mandala that was under
Pundravardhana-bhukti?'^ Amitabha Bhattacharyya identified the place Campavanl
mentioned in the inscription with Campaknagar in the Comilla district and added that the
Mainamati inscription effectively showed that Samatata, which was coterminous with the

"/A/a'.,pp.l26-127.
Amitabha Bhattacharj'va, op cit., pp.75-76.
^' K.K. Gupta (ed.), op.cii., p.80.

65
Tippera-Noakhali area, formed a mandala within Pundravardhana-6/?Mfe/ at the time of
Ladahacandra.'*" In this connection it may be noted that the Meher Copper-plate of
Damodaradeva (dated AD 1234 or Saka Era 1156) also made reference to Meharagrama
(located in the Candrapur subdivision of Comilla district) situated in Vayisagrama-
khandala in the Paratayi-vwoyo of Samatata-wctwJa/a, . which was within
Pundravardhana-6/zM^?z. The Pundravardhana-A/zwA:// therefore sometimes extended up to
Samatata.'*'

The Belava Copper-plate Inscription (twelfth century) of Bhojavarman (discovered in

village Belava in the Narayanganj sub-division of the Dacca district) recorded the grant
of one pataka and 9% dronas of land in the village Upayalika within KausambT-
Astagachchhakhandaia situated in Adhahpattana-mawafa/a of Pundravardhana-AZ/w/t//. The
grant mentioned the name of the donee as Ramadevasarman, a priest and an inhabitant of
the village of Siddhala in the Uttara-Radha. The grant was issued from the victory camp
at Vikramapura. The contents of the document are indicate that Pundravardhana-^Z/wA:;/
included within its jurisdiction the village of Upayalika located within KausambT-
Astagachchhakhandala lying within the Adhahpattana -mandala."^^

Pundravardhana maintained its identity as a bhukti under the Sena kings of Bengal also.
The Senas had their political base at Lakhnauti (in modern Malda area). The
Madhainagar Copper-plate Inscription of l.aksmanasena (dated to about the twelfth
century) and the Deopara Stone Inscription of Vijayasena (dated the twelfth century)
claimed that the Sena rulers had subjugated Kamarupa.''^ It bears noting however that so
far there is no evidence of the Sena kings having issued landgrants in Kamarupa,
although such instances were plentiful in the case of the region of Pundravardhana. The
Barrackpur Inscription of Vijayasena (dated to about the twelfth century), who ascended
the throne in AD 1095, referred to a grant of land in Ghasasambhogabhattavida in KhadT-
visaya within the Pundravardhana-^/iwW/, while his Deopara Inscription informed that

""^ Amitabha Bhatlacharyya, op.cit., p.ll.

"' D.C. Sircar (ed.), op.cit., pp. 140-145; Amitabha Bhattacharyya, opcil., p.77.
"^ R.R. Mukherji & S.K. Maity (eds.), op.cit., pp.234-244.
"' Ibid., pp.32-38, 244-258, 277-289.

66
Varendra was a large region within Pundravardhana.'''' The twelfth century Madhainagar
Copper-plate Inscription of Laksmanasena, son of Vallalasena (which inscription was
recovered from Madhainagar in the Pabna district) registered the donation of a village
named Dapamya-pataka, situated on Ravana lake in Kantapura-vr/// of Varendrl within
the boundary of Puridravardhana-6/zMfo/. The donated land rneasured one hundred
bhukhadis and sixty-one khddikas and yielded an annual income of one hundred/?wra«as
and sixty-eight kapardakas. The donation was in favour of Govindadevasarman, officer-
in-charge of the room of propitiatory rites for religious purposes. The contents of the
inscription also indicated that the grant was made in perpetuity free from all dues or
threats of interference or confiscation. The document further mentioned that the donation
was made according to bhiimicchidranyaya. The grant was issued from the victory camp
in the vicinity o{ Dhaxyya-grama. Thus, Dapamya-pataka in Kantapura-vr^/7 of VarendrT
was included within the jurisdiction of the Pundravardhana-ft/jwA:^/.'*^ The twelfth century
Sunderban Copper-plate Inscription of Laksmanasena (which was discovered in the
village Bakultala to the south of Kosinagar in the Diamond Harbour sub-division of the
24-Parganas district of West Bengal) recorded that he granted a plot of land along with a
homestead in Mandalagrama, altogether measuring three bhudronas, one khddlka,
twenty-three unmanas and two and a \\a\^kakinis^^ and yielding an annual income of fifty
puranas. The donated land lay within the jurisdiction of Kantallapura-cfl/wraA»,
belonging to KhadT-wan</a/fl of the Pundravardhana-/?/zwA://. The grant was made in
favour of Krsnadharadevasarman, the priest-in-charge of the room where propitiatory
rites were performed. It was to be enjoyed in perpetuity free from all dues and any kind
of interference or confiscation in accordance with the principle of bhumicchidra. It was
issued from the victory camp at Vikramapura. As inferred from the contents of the
document, Mandalagrama in YJ^nXaWapnra-caturaka, lying within KhaAi-mandala was
situated within the bhukti of Pundravardhana.'*'' Similarly, the twelfth century
TarpandTghi Copper-plate grant of Laksmanasena (recovered from the northern part of
TarpandTghi under Gangarampur Police Station in the Balurghat subdivision in Dinajpur

^ Ibid., pp.244-258; Amitabha Bhattacharyya, op.cit., p.76; B.C. Law, op.cit., p.247.
"' R.R. Mukherji & S.K. Malty (eds.), op cit., pp.277-286.
"* Ibid., pp.290-294.
'Ubid

67
district) recorded the donation of a plot of land situated within Velahisthi in VarendrT
belonging to the Pundrayardhana-bhukti. It measured one hundred and twenty adhavapas
and five unmanas and yielded an annual income of one hundred and fifty kapardaka-
puranas. It was gifted by way of remuneration to Isvaradevasarman, the priest who
performed the ceremony of giving away a golden horse and chariot. The document
ftirther mentioned that the grant was made according to bhiimicchidranyaya and the
donee was to be the beneficiary of the custom of perpetual endowment free of any rent
and any kind of interference or danger of confiscation. The copper-plate was issued from
the victory camp at Vikramapura. This shows that the bhukti of Pundravardhana included
within its jurisdiction the Velahisthi area in VarendrT.''^ The Bakultala Copper-plate
Inscription of Laksmanasena designated KhadT as a mandala within the jurisdiction of
Pundravardhana-Zj/zwA://. KhadT is identified with the region presently known as KharT
located in the Diamond Harbour area of the 24-Parganas of West Bengal.'*^ Another
inscription of Laksmanasena, the Anulia Copper-plate grant, was recovered from the
village Anulia near Ranaghat in the Nadia district (West Bengal). It recorded the transfer
of a plot of land in the village of Matharandiya, measuring one pataka, nine drorias, one
adhdvapa, thirty-seven unmanas, and one kakinika and yielding one hundred kapardaka-
puranas as an annual income in favour of a scholar named Raghudevasarman. The grant
was made according to the principle of perpetual endowment, with provision for enjoying
freedom from payment of any rent and immunity from interference or confiscation as per
the principle of bhumicchidra. The document further stated that the village Matharandiya
was situated within the jurisdiction of Vyaghratati which belonged to the
Pundravardhana-Z)/zwA;//.^° It may be mentioned in this connection that according to the
Khalimpur Copper-plate grant of Dharmapala and the Nalanda Copper-plate Inscription
of Devapala, Vyaghratati was a mandala within the bhukti of Pundravardhana.^' It is
interesting to note that while the TarpandTghi Copper-plate Inscription of Laksmansena
recorded the grant of a plot of land within the traditional boundary of Puiidravardhana
(i.e., in the Dinajpur area of north Bengal), the Madhainagar charter issued by the same

•^ Ibid., pp.295-302.
"" Amitabha Bhattacharyya, op.cit., p.76.
^° R.R. Mukherji & S.K. Maity (eds.), op.cit., pp.302-3] 2.
" Ibid., pp. 95-110, 312; Amitabha Bhattacharyya, op.cil., p.75.

68
ruler recorded the grant in the Pabna district, the Sunderban Copper-plate Inscription
recorded a grant in the 24-Parganas and the fourth epigraph, Anulia, recorded another
grant in the Nadia district. The latter three inscriptions effectively showed that during the
Sena period the boundary of Pundravardhana-6/?M^// included not only the northern
districts of Bengal but had also been extended further eastwards (Pabna) and included
parts of the modem 24-Parganas and Nadia districts (western and southern Bengal).

At least two of Laksmanasena's successors issued land grants in east Bengal within
Pundravardhana. The tv/elfth century Madanapada Copper-plate charter of Visvarupasena
(found in Madanapada in Faridpur district in East Bengal) was executed from the victory
camp at Vikramapura (modern Vikrampur in East Bengal, i.e., Bangladesh). This
inscription recorded the grant of the village of Pinjokasti situated in Vikramapura within
the jurisdiction of the Pundravardhana-^Awto'. The grant was made in favour of a
Brahmana named Visvarupadevasarman, who was a reciter of the moral texts. The village
thus gifted was, according to the document, divided into two plots — one situated in the
locality called Kandarpasankara and yielding an annual income of five hundred puranas;
the other situated in the area called Kandarpasankarasa belonging to the locality
Narandapa and yielding an annual income of one hundred and twenty-seven purams.
Thus both together fetched a total annual income of six hundred and twenty-seven
purams. As the inscription indicated, the grant was made in accordance with the
principle of bhumicchidra to be enjoyed by the donee in perpetuity free of all dues and
any kind of interference or threat of confiscation.^^ The twelfth century Calcutta Sahitya-
Parisat Copper-plate Inscription of Visvarupasena (preserved in the Sahitya Parisat
Museum, Calcutta, after it was recovered from the neighbourhood of Dacca) recorded the
grant of land within the jurisdiction of Pundravardhana-^AwM. The land measured sixty-
seven and three-fourth udanas in two plots in Ramasiddhi village situated in the
navigable region; twenty-five udanas in the village of Vinayatilaka; one hundred and
sixty-five udanas in Ajikula in Navasamgraha-c'a/MraA:a within MadhuksTraka-avr////;
twenty-five udanas and ten udanas respectively in two plots in DeuldhastT in Lauhanda-

52
R.R. Mukherji & S.K. Maity (eds.), op.cit., pp,312-321.

69
caturaka in Vikramapura; twelve and three-fourth udanas in Ghagharakatti in Ura-
caturaka within CandradvTpa; and twenty-four udanas in PatiladivTka in CandradvTpa.
The donee was a Brahmana named Avapallika Pandita Halayudhasarmana. He received a
total area measuring three hundred and thirty-six and a half unmanas and yielding an
annual income amounting to five hundred purdnas. He was to enjoy in perpetuity
freedom from payment of rent and any kind of interference or danger of confiscation as
per the principle of bhumicchidra. The grant was executed from the victory camp at
Vikramapura.^^ Thus, the villages of Ramasiddhi, Vinayatilaka, Ajikula in
Navasamgraha-carwra/to within MadhuksTraka-avr////; DeuldhastT in Lauhanda-caturaka
in Vikramapura; Ghagharakatti in {Jra-caturaka within CandradvTpa; and PatiladivTka in
CandradvTpa were all within the jurisdiction of the Pundravardhana-6/2M^//. The evidence
of this inscription confirmed that the jurisdiction of the bhukti called Pundravardhana
included for sometime at least the heartland of eastern Bengal like m.odern Dacca-
Vikrampur-Faridpur area up to CandradvTpa. Further, the twelfth century Edilpur Copper-
plate grant of Kesavasena, (executed from the victory camp at Phalyugrama and
recovered from Edilpur in the Faridpur district) recorded the perpetual grant of the village
of Talapada, situated in Vikramapura, lying within the Jurisdiction of Pundravardhana-
bhukti. The donated land yielded an annual income of two hundred drdmmas. The grant
was made in favour of a Brahmana named Isvaradevasarman, the reciter of moral texts.
The land was gifted as per the bhUmicchidranydya.^'^ Thus, Pundravardhana-ft/7MA:n'
included within its jurisdiction Vikramapura to which belonged the village of Talapada.
The status of Vikramapura, which included a number of gifted villages as mentioned in
the inscriptions of Laksmanasena, Visvarupasena and Kesavasena, is not known for
certain. However, it is possible that Vikramapura was either a visaya or a martdala or
some other revenue division, since the inscriptions located it within the province of
Pundravardhana. The Sena kings issued these grants from their various victory camps in
eastern, western and southern Bengal, although they were the kings of Gauda, with their
capital at Lakhnauti in north Bengal. Therefore, it is possible that from the north they
expanded to other parts of Bengal through wars and conquests and some of those newly

"/6W., pp.321-332.
^'' Ibid., pp.333-349.

70
acquired areas were made parts of Pundravardhana-/).'?^^://. R.R. Mukherji and S.K. Maity
identified Isvaradevasarman of the lildilpur Copper-plate Inscription and
Visvarupadevasarman of Madanapada Copper-piate Inscription as brothers.^^ It is
therefore not unlikely that the Sena kings facilitated the settlement of some of the
Gaudiya Brahmana [i.e., Brahmanas from Gauda (north Bengal)], in the newly acquired
regions through land grants.

The epigraphic and other sources depicted Pragjyotisa-Kamarupa and Pundravardhana as

immediate neighbours, but the political boundary between the two political formations —
or the geographical extent of the two regions — changed at different points of time.
Based on the evidence of the Doobi Copper-place Inscription of Bhaskaravarrnan (dated
to seventh century), some scholars suggested that parts of Pundravardhana were then
included within KamarOpa.^'' In fact, the Doobi grant renewed a gift of land, originally
made by Bhutivannan, to some Brahmanas.^' P.C. Choudhury claimed that the
information contained in the sixth century Badgahga Rock Inscription of the time of
Bhutivarman was suggestive of the conquest of Pundravardhana by Bhutivarman around
AD 550, the period which marked the end of the Gupta hold over Bengal.^* However,
there is no direct mention of any conquest in the said inscription.^' Similarly,
Choudhury's interpretation of the Tezpur grant of Vanamala (ninth century) that
Pundravardhana was included within Vanamala's kingdom''" is not corroborated by the
text of the inscription.^' It is also doubtful that Vanamala had ever conquered
Pundravardhana (in north Bengal) despite Choudhury's location of Hensiva, the donated
area in Dijjina-vwoya to the west of Teesta or Karatoya in Pundravardhana, which was
made over to a Brahmana named Shrutidhara. This is because the Nowgong grant (on
which Choudhury had based his interpretation)''" did not mention either Teesta or

"lbid..p32\.
^ P.C. Choudhury, History of the Civilisation of the People of Assam to the Twelfth century AD, DHAS,
Gauhati, 1959, pp.158-160.
' ' M.M. Sharma (ed.), Inscriptions of Ancient Assam, Gauhati Universily, Gauhati, 1978, p.29.
^* P.C. Choudhury, op.cit., p. 154.
^' M.M. Sharma (ed.), op.cit., p.49.
^ P.C. Choudhury, op.cit., p.43.
*' M.M. Sharma (ed.), op.cit., pp.95-105.
" P.C. Choudhury, op.cit., p.223.

71
Karatoya.^^ The twelfth century Subhankarapataka grant of Dharmapala of Kamarupa
also recorded the donation of land in Dijjina-v/>fl>'a to a Brahmana who hailed from the
village of Krosaiija in Sravasti-v;'.vaya.^'* D.C. Sircar identified Sravasti with the Hili-
Balurghat area of north Bengal and concluded that the village of Krosanja and the visayas
of Sravasti and Dijjina mentioned in the document were situated vvithin the jurisdiction of
Pundravardhana-6/zw^?/.^^ M.M. Sharma, however, felt that Dijjina probably
corresponded to the Dijjina-vwoya of the Nowgong grant of Vanamalavarman III and was
located it in the Daksinakuta or the south bank of the Brahmaputra.^^

D.C. Sircar examined in detail the possibility of a portion of Pundravardhana having been
included in Pragjyotisa-Kamarupa for sometime. He refuted the possibility on the ground
that many of the copper-plate grants of the Palas of Bengal and Bihar, ranging from about
eighth to the twelfth century speak of the bhukti or province of Pundravardhana as an
integral pan of the empire. Besides, the Pala records as well as the copper-plate grants of
the age of the Imperial Gupta, issued in the fifth and sixth centuries, often locate
Kotivarsa-v/i'aya, i.e., modern Bangarh near Balurghat in the West Dinajpur District,
West Bengal. DevTkotta and Banapura, which still exist under their ancient names, are
mentioned in some records as within Kot'warsa-visaya. Moreover, an inscription of about
the third century BC (i.e., of the Maurya age) has conclusively proved that Pundranagara,
which was originally the headquarters of the Pundra people and afterwards that of the
territory named after them, has to be identified with modern Mahasthan in the Bagura
district of Bangladesh. Finally, Sircar opined that since the earliest Gupta inscriptions
recording transactions in the Pundravardhana-bhukti belong to the reign of Kumaragupta
I (AD 413-455) and the latest of them bears a date in AD 543., it is impossible to accept
the theory regarding the inclusion of the northern areas of Bengal or the eastern part of

" M.M. Sharma (ed.), op.cit., pp.142-151.

^ Ibid., p.247.
*^ D.C. Sircar, Studies in the Geography of Ancient and Medieval India, Motilal Banarsidass, Delhi, 1970,
p.298.
^ M.M. Sharma (ed.), op.cit., p.249.
^'' D.C. Sircar, "Pragjyotisha Kamarupa", in H.K. Barpujari (ed.), The Comprehensive History of Assam,
Vol.1, Assam Publication Board, Guwahati, 1990, p.65.

72
Bihar in the Pragjyotisa-Kamarupa during the reign of king Bhutivarman (dated AD 518-
542).^^

Be that as it may, it is not unlikely that at certain points of time during the early period,
some areas of modern northern Bengal were included within Kamarupa, because, as
already discussed, some literary sources suggested that the territorial boundaries of
Kamarupa extended up to the river Karatoya.^'^

The eleventh century Silimpur Stone-slab Inscription (found at Silimpur in the Bagura
district of Bangladesh) issued by Jayapala, the king of Kamarupa, recorded the grant of
nine hundred gold coins and a gift of a landed property to a Brahmana named Prahasa on
the occasion of the tiildpurusadana. The donee was credited with the construction of a
temple, the installation of the deities Amaranatha and Trivikrama, the foundation of an
alms-house, the creation of a garden and the excavation of a tank. According to the
contents of the charter, the donated land was in the Balagrama village in Sravasti-i'/j'<a'>'i3
in Pundravardhana-6/2wA://. The land yielded one thousand measures of paddy. However,
the donee refused to accept the gift even though persistently pressed.^° As observed by
D.C. Sircar, the name of Jayapala was only casually mentioned in the inscription.^' It is
not clear whether the inscription was issued by Jayapala or the refusal of Prahasa to
accept the offer of the tulapurusadana is mentioned as an example of the greatness of the
Brahmana named Prahasa. After all, the inscription is a prasasti (eulogy) in honour of
Prahasa. As the facts indicate, the award of the tulapurusadana was not limited only to
Brahmanas within the kingdom of Kamarupa; it could be conferred on learned and
righteous Brahmanas from other regions as well. The Guwakuchi Copper-plate grant of
Indrapala (dated AD 1071) recorded the grant of land to a Brahmaiia named Vasudeva, in

** Ibid., p.66.
® K.L. Barua, Early History of Kamarupa, Shiilong, 1933, reprint 1966, Lawyers Book Stall, Gauhati,
p.41; P.C. Choudhury, op at, p.219
° Epigraphia Indica, VolXJJ, p.283, D C Sircan'Tolnical Hi.story", in H.K. Barpujari (ed.), op.cil, p.l60;
R.C. Majumdar, History of Ancient Bengal, loc at, p.l 3; M R.Singh, A Critical Study of the Geographical
Data in the Early Purdnas, Punthi Pustak Publication, Calcutta, 1972, p.232; Ichhamuddin Sarkar, Aspects
of Historical Geography of Pragjyottsa-Kamarupa, iVaya Prafcash, Calcutta, 1992, p. 126; D.C Sircar,
op.cit., '^.191; P.C.Choudhury, op at, pp 242, 384.
"" D.C. Sircar, "Political History", in H.K Barpujari (ed.), op cU , p.160.

73
Vainamagrama in the SavaXh'i-visaya. The Savathi-vwayo (identified as Sravasti) has been
located by some scholars in Pundravardhana like Sravasti of the Subhankarapataka grant
of Dharmapala.'^ As already discussed, the Ramacariiam by Sandhyakaranandin, besides
some other sources made it clear that during the last quarter of the eleventh century, the
region of Pundravardhana came to be known as Gauda and the city of VarendrT within
Pundravardhana was deemed to be 'the ornament' of the region. The Ramacaritam also
mentioned Ramapala's conquest of Kamarupa. ' This view finds corroboration in the
Kamauli plates of Vaidyadeva.'''*

On the other hand, the evidence of the Kamauli grant of Vaidyadeva (dated AD 1142)
showed that the region of Pundravardhana o^ VarendrT included some parts of modern
Lower Assam under the later Palas. The document recorded the donation of land located
in the Vada-vwoya which was within the Kamarupa-mcwc/a/o to a Brahmana family that
presumably came from VarendrT.''^ The text of this copper-plate grant of Vaidyadeva
clearly pronounced that maharajadhiraja paramesvara paramabhattaraka illustrious
Vaidyadeva was the great devotee of Siva ana the great devotee of Visnu, that while in a
perfect state of health he greeted, informed and commanded the cultivators and other
citizens beginning with the catta, the bhatta and the visayillaka about the grant of land
belonging to the villages of Santibada and Mandara, possessed by Gangadhara Bhatta and
situated in the visaya of Vada, in the mandala of Kamarupa, within the bhukti of
Pragjyotisa, and that the charter was issued from the illustrious victorious royal
headquarters situated at Hamsakonci.^^ This inscription also mentioned that Vaidyadeva
was a minister under the Pala king, Kumarapala of Gauda and that this widely celebrated
Sri Vaidyadeva was appointed as a ruler by the Lord of Gauda, in place of Tingyadeva."

''^ P.C. Choudhury, op.cit, p.242.

^^ Haraprasad Sastri (ed.), Ramacaritam, Canto I, v.38; Canto III, v\'.i-29, pp.26, 6\-l\\ Kaviprasasti, v.l,
p.98.
^ Kamauli Copper-plate grant in M.M. Sharma (ed.), op cil., pp.273-290; Amitabha Bhattacharyya, op.cit.,
p.44. Also see, R.R. Mukherji & S.K. Maity (eds.j, op.cit, pp..370-386.
^^ Kamauli Copper-plate grant in M.M. Sharma (cds.), op en , pp.273-290. .Also see, R.R. Mukherji & S.K.
Maity (eds.), op.cit., pp.370-386.
'* M.M. Sharma (ed.), op.cit., pp.273-290.
" Ibid., pp.283-284.

74
The evidence of the charter suggested that Kamarupa was at that point of time a mandala
and Pragjyotisa a bhukti within Cauda (earlier known as Pundravardhana and then as
VarendrT). Tingyadeva had probably been a local governor or ruler in Kamarupa under
the later Palas of Bengal. Vaidyadeva may have replaced him due to the latter's hostility
to Kumarapala of Cauda. Perhaps this had happened after the fall, of the Pala dynasty of
Kamarupa. In this context, M.M. Sharma seems to be convinced that Ramapala of Bengal
earned authority over Kamarupa, and more probably over the western part of it, through
the military expedition of one of his allies, that Tingyadeva ruled over the eastern region
of the Pala kingdom as a feudal ruler during the reign of Ramapala's son Kumarapala,
that Vaidyadeva came to the eastern region himself when Tingyadeva started a rebellion,
that Pragjyotisa was a bhukti and Kamarnpa was a mandala in his kingdom, and that his
seat of government at Hamsakonci was decidedly situated within the area of the modern
Kamrup district.^* Therefore, during the period of the Pala kings of Bengal, Pragjyotisa
was a bhukti within the Kamarupa mandala and Vaidyadeva was a feudatory ruler of the
eastern region of the Pala kingdom.

Heartland of Pundravardhana

The available sources suggest ihat the geographical extent of Pundravardhana underwent
continuous change during the period starting from the third century BC to the twelfth
century AD. Some scholars formulated opinions about the geographical extent of the
region of Pundravardhana on the basis of the information available to them through other
scholars' researches. In fact, the development of the geographical concept of
Pundravardhana makes an interesting study. As the researchers brought out more and
more literary, epigraphic and other archaeological sources to light, opinions on the
subject also underwent significant changes. Apparently, Alexander Cunningham had
mistakenly identified Pundravardhana with 'Pubiia' (Pabna district). It appears from his
own statement that he was wrongly guided by the idea that the kingdom of Pun-na-fa-
tan-na (later on identified as Pundravardhana), as recorded by the Chinese pilgrim Hiuen

' ' Ibid., p.0.36, with reference toriarapias-adGastri led.), op cit.. New Introduction, p.xxxi; Canto HI.
w.39-40, pp.37-39.

75
Tsang, would have been called 'Ponbardhan' or 'Pobadhan' from which the name Pabna
came. According to Hiuen Tsang. the circuit of Pnn-na-fa-tan-na measured about four
thousand // or six hundred sixty-seven miles.''^ Alexander Cunningham was further
prompted by the information that Pubna was situated just a hundred miles from 'Kankjoi'
(Kajangala) on the opposite bank of the Ganga, notwithstanding the fact that its location
was almost southeasterly rather than towards the east. He found Pun-na-fa-tan-na,
Puridravardhana and Pabna sounding very similar to one another.*^ Nevertheless,
Cunningham's effort inspired a number of scholars to direct their attention towards the
proper estimation of the location of Pundravardhana. Therefore, it is obvious, as later on
pointed out by scholars like B.C. Law and Amitabha Bhattacharyya, that Cunningham
was initially mistaken in identifying the location of Pundravardhana, but that later on he
had revised his opinion in subsequent writings and located the ancient city of
Pundravardhana in the present ruins of Mahasthan or Mahasthangarh which lay
seven miles north of the modern town of Bagura.*' Writing at about the same time of
Cunningham, Sir M. Monier-Williams identified 'Pundra' as a country which included
parts of Bengal and south Bihar.^"

B.C. Law not only identified the city of Pundravardhana Vi'ith Mahasthan in the Bagura
district but also discussed the geographical extent of Pundravardhana-6/?M^;, as gleaned
from the information contained in several literary, epigraphic and other archaeological
sources. He concluded that the territorial limits of the bhukti changed at different points
of time, and from time to time its boundary extended, besides northern Bengal, to include
parts of western, southern and eastern Bengal during the period of study. According to
Law, Pundravardhana-^/jwA;// in its heydays included the districts of Dinajpur, Bagura,
Rajshahi, Malda, Rangpur, Sunderbans and 24-Parganas and parts of the Dacca division;
but for much of the ancient period it was confined to northern Bengal only corresponding

' ' Samual Beal & Shaman Hwui Li, The Life of Hiueii-Tsmn^. London, 1884: reprint 1973, Academica
Asiatica, New Delhi, pp. 130-131.
*° Alexander Cunningham, op.cit., pp.404-4G5.
*' B.C. Law, op.cit., p.248; Amitabha Bhattacharyya, op.cit.. p.94.
'^ Sir M. Monier-Williams, A San.skrit-Er1gli.9h Diclioncry, London, 1899; reprint 1974, Indological Book
House, Delhi, pp.632. 650.

76
to Dinajpur, Bagura, Rajshahi, Malda and Rangpur districts.^^ Law did not mention
anything about the SrThatta-Samatata region having formed a part of Pundravardhana.
This was perhaps because the copper-plate grants pertaining to the SrThatta-Samatata
region were not known during his time.

R.C. Majumdar, who dealt in detail with the geographical extent of Pundravardhana,
sourced a vast range of literary and epigraphic materia! in particular, those dated to the
ancient and the early medieval periods. According to him, Pundravardhana mainly
corresponded to the region of northern Bengal. However, he suggested that at various
points of time it included eastern, southeastern and southern Bengal, thus stretching its
boundary up to the sea-coast of the Bay of Bengal. The evidence of ancient and medieval
literature and inscriptions cited by Majumdar proved that Pundravardhana included major
portions of the present districts of Bagura, Rajshahi, Rangpur and Dinajpur in
Bangladesh, besides the north Bengal region of the present .state of West Bengal India.^"*
Majumdar's opinion is more or less corroborated by the findings of M.R. Singh. Basing
his research on the study of the Purarias, Singh stated that Pundravardhana in the Puranic
period included Varendra or Varendrl, which covered the region roughly corresponding
to northern Bengal.^^

D.C. Sircar studied the geographical extent of Pundravardhana in the specific context of
the geographical regions of ancient and medieval India. On the basis of a vast range of
literary, epigraphic and archaeological sources dated to the ancient and medieval periods
and strictly conforming to the available data, Sircar concluded that the city of Pundra
(Pundranagara / Pundravardhana-7Wgara) should be identified with Mahasthangarh in the
modern district of Bagura, while the geographical extent of the province
(Pundravardhana-6//wA:/0 included a large area covering parts of northern, eastern and
western Bengal. Sircar argued his identification on the grounds that,

* ' B . C . Law, o/;.a/,p.247.

*" R.C. Majumdar, op.cit., pp.13. 29,
* ' M . R . Singh, o/7.c/;.,p.225.
'^ D.C. Sircar, op at., p.?7.

77
 (a) according to Divyavadana. the city of Pundravardhana lay beyond Kajangala,
the eastern border of the land which was the early sphere of Buddhism (i.e.
Bihar and eastern Ultar Pradesh);
(b) the Mahasthan inscription of the third century B.C. not only proves that
Pundravardhana was also called Pundranagara, 'the city of the Pundras' but
also supports its identification with Mahasthan;
(c) the Damodarpur plates of the Gupta age shov/ that Kotivarsa (the area around
modern Bangarh in the West Dinajpur district. West Bengal) was a visaya or
district of the bhukti or province of Pundravardhana;
(d) Hiuen Tsang, the Chinese pilgrim of the seventh century, travelled from
Kajangala (modern Rajmahal in Bihar) to Pundravardhana and thence to the
farther east to Kamarijpa (Pragjyolisa);
(e) the epigraphic records of the early medieval period applied the name
Pundravardhana-6/2wto' to north Bengal, its chief city being identified
with modern Mahasthan in the Bagura district (Bangladesh).*' '

Among subsequent writers, Parmanand Gupta suggested that the region of northern
Bengal (consisting of north Bengal in the present state of West Bengal and the northern
districts of Bangladesh) remained a constant part of Pundravardhana, while it sometimes
also corresponded to the modern districts of Malda. Dinajpur, Rajshahi and portions of
Bagura and Rangpur districts situated to the west of the Karatoya. He further concluded
that Pundravardhana was bounded by the main stream of the Ganga on the west and the
Karatoya on the east.*^ However, the most detailed study so far on the historical
geography of ancient Bengal is by Amitabha Bhattacharyya. He observed that although
the geographical extent of Pundravardhana experienced a continuous change the region
or the bhukti was always identical with the region of northern Bengal which, besides the
present north Bengal districts in West Bengal, included Dinajpur, Bagura, Rajshahi and
Rangpur districts in northern Bangladesh.^^ Geographically, all these areas together
constituted the northern geographical division of undi^'!ded Bengal. Therefore, the recent

''/Wc?., pp. 159-160.

*^ Parmanand Gupta, op cil, p.24.
*^ Amitabha Bhattacharyya, op.cit., pp.6-7.

78
writers also seem to be in agreement with B.C. Law, who had observed that
Pundravardhana as a bhukti at times included the districts of Dinajpur, Bagura, Rajshahi,
Malda, Rangpur, Sunderbans and 24-Parganas and parts of the Dacca division of his time,
but the Dinajpur, Bagura, Rajshahi, Malda and Rangpur areas were always included
within that bhukti. These districts have been further bifurcated in recent years into more
districts both in India and Bangladesh.

B.N. Mukherjee is also of the opinion that the ancient Pundravardhana was situated
around Pundanagala (Pundranagara), identified with Mahasthan in the Bagura district
(now in Bangladesh), while as an administrative division, called bhukti, it included in the
age of the Imperial Guptas substantial parts of the territories now included in the northern
sections of Bangladesh and West Bengal. Several area^ to rhe east of the Bhagirathi were
brought within the limits of the bhukti of Pundravardhana in the Pala-Sena age. The
extension of its jurisdiction to the east of the Brahmaputra is suggested by the
PaschTmbhag copper-plate inscription of Srlcandra, which included SrThatta-w7a«<ia/a in
Pundravardhana-6/zw^//. The mandala of Srihatta certainly included the modern Sylhet
area to the east of the Brahmaputra (Lauhitya). The name Pundravardhana thus gradually
became associated with a large territory outside its normal geographical limits in the
Pala-Sena age.^' In the same vein. Mandira Bhattacharyya also observed that ancient
Pundravardhana roughly corresponded to the undivided districts of Malda, Bagura,
Rajshahi, Rangpur and Dinajpur of undivided Bengal. On the east it was separated from
the Kingdom of Pragjyotisa by the river Karatoya and in the west from Kajangala and
Karnasuvarna by the river Ganges, and this tract of land was slightly higher than the
alluvial plains of the rest of Bengal. In about the ninth century, this area came to be
known as Varendra or VarendrT and formed an important administrative unit of the Pala
Kingdom.^^

" B.N. Mukherjee, "Pundravardhana and the Sculptural Art of Kamarupa" in B.N. Mukherjee & P.K.
Bhattacharyya (eds.), Early Historical Perspective of North Bengal, North Bengal University, Siligiiri,
1987, p. 9S,
'^ Mandira Bhattacharyya. "Religion in Pundravardhdn.i" in B.N. Mukherjee & P.K. Bhattacharyya (eds.),
op.cit., p.64.

79
The survey of the available source-materials confirms that the scholars like B.C. Law,
R.C. Majumdar, D.C. Sircar, Amitabha Bhattacharyya and others were more or less
justified in their observation chat Pundravardhana was generally identifiable with
northern Bengal. These scholars, however, overlooked a few facts, perhaps mainly
because some of the documents and inscriptions of Kamarupa and SrThatta had not been
available to them. Consequently, some of the important facts regarding the changes in the
geographical extent of Pundravardhana are not clearly known from their works, as
brought out below:

(a) that Pundravardhana or north Bengal came to be better known as Varendra or

VarendrT between the tenth and twelfth centuries, and under the later Palas and the
Senas the same area was recognized as 'Cauda' (indeed, a revival of an earlier
name), and under the later Senas it was called LaksmanavatT, which the Persian
writers mentioned as Lakrauti (Laksm.anavatT) or Barind (Varendra);
(b) that the later Palas conqueica and ruled over a part of Kamarupa (more precisely
Lower Assam), as borne cut by the evidence of the Kamauli grant of Vaidyadeva
and the Ramacaritam of Sandhyakaranandin;
(c) that the PascTmbhag grant of SrTcandra clearly mentioned that SrThatta was a.-
mandala within the Pundravardhana-^^/jwA:// and there were three visayas within
the SnhaUa-mandala {viz., Garala, Pagara and Candrapura);
(d) that other copper-plates of SrTcandra and his successors also recorded grants of
land in different visayas v.ithin the Pundravardhana-bhukti, while the find-spots
and the villages named in the inscriptions were spread over Dacca, Vikrampur,
Tripura and other districts of eastern, northeastern and southeastern Bengal;
(e) that the territorial limits of Pundravardhana-/)/jMM from time to time extended to
other regions of Bengal also, although for most of the time it might have been
identifiable with northern Bengal only.

It may thus be concluded that the geographical extent of the Pundra region or
Putidravardhana as a bhukti changed fiom time to time during the period under study
since the Pundras were noticed in the literary sources as inhabitants of the area in later

80
Vedic times. The area was known as Pundra and Pundravardhana, and then as Varendra
or VarendrT till about the twelfth century. The area denoted as Pundravardhana was first
limited to modern northern Bengal, then it expanded to cover the neighbouring areas, and
finally, it again shrunk to northern Bengal in the eleventh-twelfth centuries, when it more
popularly came to be known as Varendra."^ Pundranagara (modern Mahasthangarh)
continued to be its main political centre. However, Pundravardhana, as a bhukti or
province, denoted a much larger area than what was originally known as Pundra or the
land of the Pundras. Although for most of the times Pundravardhana was limited to
northern Bengal, under the Candras and the Palas, Pundravardhana-^/jwM included parts
of western, eastern and northeastern Bengal also, besides northern Bengal. Nonetheless,
for much of its history the Pundra region coincided with modern northern Bengal, the
part of which is now included in the state of West Bengal in India and another part is in
Bangladesh. Like the geographical extent, the ethnic composition of the area had also
undergone changes. Although the Pundras were first referred to as the inhabitants of a
part of the region and the word 'Pundra' sometimes denoted land and sometimes people,
at a point of time in historv the Pundras and the Kiratas were seen as people occupying
areas adjacent to one another, while at an earlier point of tim^e the entire area was referred
to as Kirata (both as a land and the people who inhabited that land). Social elements from
the western direction spread into the area at a later stage and introduced the Brahmanical
culture principally manifested in the idea of the vamasrama-dharma.

'^ See Map 2: Heartland of Pundravaidhana.