Till<:

Tllll~ ASTILONOMICAL woiu,

IlF

"
ARAliA MIIJlltA.

l'!IE TEXT, Em'I'I':\) WITI! AN OmCUNAL COi\I\\IENTAli.Y IN ~-\AN8h.l:IT

AND AN EN<1LINH Ti:AN~LATl()N AND INT1:oDlf()T10N

,
llAMA HOPJ\DH\' A 8 U I )I! A1\ A H A I) Y I V E D I.
--------------------~~-- ~-----~

PREFACE.

There is some reason to fear that the feeling .of anyone 'who may
examine in detail this edition and translation of Varaha Mihira's astronomi-
cal work will, in the first place, be wonder at the boldness of the editors.
I am indeed fully conscious that on the imperfect materials at our disposal.
an edition in the strict sense of the word cannot be based, and that what we
are able to offer at present deserves no other name but that of a first attempt
to give a general idea of the contents of the Pafichasiddhuntika, It would,
in these circumstances, possibly have been wiser to delay an edition of the
work until more correct Manuscripts have been discovered. Two consider-
ations, however, in the end induced us no longer to keep back the results,
however imperfect, of our long continued endeavours to restore and elucidate
the text of the Panchasiddhantika. In the first place we' were encouraged
by the consideration that texts of purely mathematical or astronomical con-
tents may, without great disadvantages, be submitted to a much rougher and
~.
bolder treatment than texts of other kinds. ~What interests us in these works,
is almost exclusively their matter, not either their general style or the parti-
cular words employed; and the peculiar nature of the subject often enables
us to restore with nearly absolute certainty the general meaning of passage s
the single words of which are past trustworthy emendation. And, in the
second place, we feel convinced that even from that part of the Pancha-
siddhantika which we are able to explain more is to be learned about the
early history of Sanskrit Astronomy than from any other work which has
come down to our time.
Imperfect and fragmentary as text and translation are, we may assert
at any rate that, in our endeavours to overcome the quite unusual obstacles,
which the corrupt and bare text of the Panchasiddhantika opposes to the
interpreter, we have spared no trouble. The time and thought, devoted to
the present volume, would, I may say without exaggeration, have amply
sufficed for the editing and explaining of twenty times the amount of text'
presenting only normal difficulties. This I mention, not of course in order to
extol what we have been able to do, but only as an excuse for what we see
ourselves obliged to leave undone.
Next to the lamentable state of the text as appearing in the two Manu-
scripts at our disposal, the greatest disadvantage under which we laboured
was the absence of a Commentary. Commentaries can be hardly done with-
out in the case of any Sanskrit astronomical work; much less so, when the
text, as that of the Panchaaiddhsntika, describes many mathematical pro . .
VI PREFACE.

cesses more or less divercinc

I:> I:>
from those commonly employed.
v J
Commentaries
probably existed formerly) and possibly exist even now ; but we have failed
to procure an)-. The Commentary published in the present volume is an
entirely original composition by my Collaborator. A mere translation of the
text with notes would, indeed, have sufficed for the European reader; we
:however, wished to make the results of our labour accessible to Pandits also
who understand no English. And a full tiki1 giving full demonstrations
in the ordinary Hindu style will, in many cases, be useful to the European
student also.
The right hand columns of the text give the emended text; the left
hand columns the text of the better one of our t\VO Manuscripts which we
thought advisable to exhibit in extenso. Some remarks on the Manuscripts
and the mode of emendation of the text will be found at the end of the.
Introduction.
As this preface is signed by myself only, I may, I think, here
acknowledce-s-in
I:>
a somewhat more explicit way than the mere association oi
v

names on the title page is capable of doing-the great obligations under

'which I am to my collaborator Pandit Mahamohopadhyaya Sudnakara
Dvivedi. His constant assistance was altogether indispensable to me, and
all the more welcome as among t:le Jyautishas of my acquaintance I know of
no other, fully equal to work of this kind and at the same time equally ready to
devote himself to a task 'which in certain aspects is so entirely unremunerative.
I may express the hope that the .Pandit, who is already so well known for
his efforts to spread a knowledge of modern higher Mathematics among his.
countrymen, will continue to devote a part at least of his learning and talents
to the elucidation of the ancient history of science in this country.
I further wish to express my best thanks to the Bombay Government
and to Professor R. G. Bhandarkar, who with great liberality have allowed
me the use, for lengthened periods. of time, of all those Manuscripts in their
charge which I required for the present edition. N or must I omit to record
my obligations to Professor G. Buehler to whose activity, when in charge of
the search for Sanskrit Manuscripts in parts of the Bombay Presidency) we
are indebted for the discovery of the two Manuscripts on which this edition
is based.
G. THIBA.UT.
ALLiHAB.!.D : }

isu. December, 1888.

 INTRODUCTION.

The Pallchasiddhtmtild hy Vad,ha Mihira occupies a marked position

(If its own in Indian astrouornical Iitcraturc. As a, rule works treating- of
that branch of science cla.i III either to he eli rcctl y revealed, :18 f i. the S (ll'ya
Siddlulllta, in that form which has conic down to our time; or else to base ill
all essential points on Rome older work of divino <H'i~in, as f '1:. the Sirldhilnt:18
11.1' Hralunagupta alid Bh:l.skn,nkh;b-ya, lxit.h of which arc roproductions,
however greatly anI] t Iifivd and improved, of an old l)aitfLnl:lha, Siddh,l11ta,
One of the COnSC(plCllC(~S (If' tllis is, that these works claim for thcmsolvcs
direct or derived illJ;dlibiJity, propoulld Llicir lloctl'illi's iu a calmly dog'lllatic
tone, awl cithul' pay no :Lttclltion whatever to views divel'gin,!.(' from their own,
or else refer to such ouly o(lcasiona.lly, and most.ly ill the touo of ooutcmp-
tuous doprocintion. Tlw latter attitude iK assullH"d I /. by Bl'alullagl1pt~t
who illdued devotes It spuci:d (:h(l.pt~"r to tho t:tsk of revicwins; thoso astron.nui-
eal ;-;ystulllS whicl: were 0pposl:d to the: tu:whillg' of tho Hraluua Siddhallta,
hut who would havo rOlldon:11 that part; of Jlis WI)J·Ie much lll<)!'(: valuuble alliI
illtel'\~;-;Lillg, lw,d he 1)(~Ull k"s unxiou» to criticiz« awl ridicule t.huu to illlpH,rt
inforuuu.iou. The II,sLl'lllWllli(:a,1 wr itor«, it is true, therein ollly exelllplifY
a g'(;lwnd mental tUlldl'!J<:.y which displays ih(:!l' ill almost u\'\:l'Y departmunt
or Hilldll Litcraturo : llllt inure doglll;~tic axscrt.iou appr::LI'S IlIOI'U thnu
ordiu.uily lllisphc("ll ill nil exact :~eiell('u lib: astrnuoruy, and the H.1>SUllCC of
all appl'l'\:iativu 1'Uf(\I'('ll(:e to Lhu views of pn.lC(ldillg authors is pa.rti("11 1:I1'Jy
vexatious, when we have to do with a l.rnnch of Hiudn Learlling which
shows dear tracca or having bCl'll rClllOdcIkd under the iutlucncc of Greek
teaching.
~eo tho gonoral rule tho :PH.fi{~h:\,Hiddhlllltiki1 formA a Htriking exception.
Al'! far aN we can jtl\lg\) ltt present, Vari\.lm Miliira was the only one among
Hill<hl writers on astronomy who thought it worth while to give all cxpoxition
of a,ll tlw more illljlort,lllt ron!!;.; of astrunoillieal doctrine which wore current
at hiN Limo. Not that be Wa.N uuahle to judgo of tho relative value or tho
::i'yHtcIllH which offered themselves to his cx.uuiuatiou ; for, as we shall see
further on, he knew very well in whali onlor of merit tho Jive Siddh:lutas
whose t(.lil,ehing he summarizes are to be arranged. But he socius ready to
ackuowledgo that even inferior systems deserve a certain amount of attcn-
tion, tLS long aN they continuo to occupy in certain circles a, poxitiou of
authority; and he a,ppears not to be altogether incu.pahle of taking a purdy
intellectual iuterest in examining the various, more or lONH perfect, methods
which may be applied to the solution of scientific problems. At the same
time he seems to have no hesitatiou to acknowledge the eounexion of the
VIII I~TROD1jCTIO~.

modern phase of Hindu astronomy with Greek science. Although

directly stating that the Hindus learned from the Greeks, he at any
mentions certain facts and points of doctrine which sugge:"t the depend,
of Indian astronomv on the science of Alexandria: and, as we know alr-
v •

from his astrological \\Titing::;, he freely employs terms of undoubted G

orurm, The Pancbasiddhantiku thus becomes an invaluable source for
~

who wishes to study Hindu astronomy from the only point of view w
can claim the attention of the modern scholar, ciz. the historical one.
Regarding its form the Pancbesiddhentikf belongs to the class 01
so-called karanagranthas j. e. eouipeudious astronomical treatises whic
not set forth the theorv v
of the subject at comparative lenzth
~ u
as
Siddhantas do, but merely supply a set of concise-and often only app
mately correct-rules which sufiice for the speedy performance of all the
important astronomical calculations. It however contains a few cha
whose contents lie outside the limits of a mere karana and resemble the
responding chapters of the best known SiddMntas; notably the chapter 1,1
describes the general constitution of the universe, and the 15th chapter c
,Tyotishopanishad. And it of course decidedly distinguishes itself fro
ordinary karanas by the fact that it does not base on anyone parti
Siddhanta, but undertakes to reproduce the more important doctrines of
different Siddhantas,
These five Siddhantas, named by Vardha Mihira in the first cha
are the Paitamaha, V~1si;;:htha, ROllaka, Paulisa and Saura Siddha
Varfi.ha Mihira there also states his view as to their order in import:
assigning the first place to the Sllrya Siddhtl.nta, placing next the Romaka
Paulisa Siddhantas as about equally correct, and declaring the two remai
works to be greatly inferior to the three mentioned. In agreement with
estimate very different amounts of space are allotted to the individual
dhantas in the body of the work, the Surya Siddhanta and Paulisa Siddl:
being treated at some le,pgth, next to these the Romaka, and 'very little a
tion being paid to the Paitdmaha Siddhanta, and, although this is a .
somewhat difficult to decide, to the Vasishtha Siddhanta. .
In addition to the general character of the five Siddhantas, this d
ence of treatment is owing to a special cause, mentioned by Varaha M
in the first chapter t,iz. his wish to devote the Pafichasiddhantika chief
the task of setting forth the calculation of solar eclinses, t.hp. mosr rlif1
XVI INTRODUCTION.

check neither by means of the originals nor with the assistance of modern
recasts. There also we must hold Varaha Mihira to have closely followed
the elements and methods of the authors of the Siddhantas, and to have
permitted himself only minor changes, such as facilitate calculation with-
out affecting the fundamental character of the rules. General principles,
e.nabling us to judge with certainty how far those changes may extend, can
however not be laid down; we rather must judge each given case on its own
merits. When we f i. find that the yuga of the Romaka Siddhanta com-
prised, according to Vari1ha Mihira, only 2850 years, we may raise the ques-
tion whether this yuga is the true yuga of the Romaka, or only represents
a subdivision of the true yuga, analogous to the 180000 years of the Silrya
Siddhanta which, as we have seen above, must be considered as the smallest
fraction of the mahityuga with which the calculation of the ahargana can be
effected. But we shall without much hesitation decide in favour of the
former alternative, in the first place because the yuga of the Romaka
Siddhauta is expressly called a yuga of the sun and moon, for the formation of
which a comparatively small number of years was sufficient, and in the second
place because Brahmagupta, in a passage to be quoted later on, testifies that
the Romaka SiddMuta did not conform to the traditional views concerning
the large periods of time. If, again, we find that according to the Paiicha-
siddMntikiL the Paulisa Siddhanta made no use of yugas of any kind to the
end of calculating the ahargal).a and the mean positions of the planets, but
employed for those purposes a peculiar system of its own, we certainly must
conclude that system to have been actually taught in the original Paulisa
Siddhanta, ~nd not constructed, as indeed it might have been, by Varaha
Mihira On the elements of the Paulisa Siddhanta. For why, we must ask
ourselves, should he have transformed in that way the elements of the Paulisa
Siddhanta rather than those of the other SiddM,ntas which without any
difficulty might have been thrown into the same form ~ And, to single out
one further point, if we find that the Paiichasiddhi\ntik agives a rule how to
calculate, accordillg to the Surya Siddhanta, the equation of the centre of sun
and moon for any given anomaly, while it represents the Paulisa and Romaka
Siddhantas as merely stating the amount of those equations for a certain
series of anomalies, without telJ,ching us how to calculate the equations for the
intervening anomalies; we must again SUppose that Var~h!1o Mihira faithfully
renders characteristic features of the original Siddhantas as he found them ;
for if he had held the opinion (which as the writer of 1\ karana he indeed
might have held) that the :pr"ctical astronomer knows enough, if be can
assign the equations of the centre for, let us say, each fifteen degrees of'
anomaly, he would no doubt not have given the general rule from the SCtrya
 ISTRODUOTION. XVII

Sicldhlnta, hut calculated from It the amounts whose lmowledge he considered

indi.,pensable, and inserted them ready calculated in his text.
We therefore arrive at the conclusion that Varilha Mihira has in no
'Case obliterated the characteristic features of the Siddhtmtas he had to deal
with. and that whatever clistinglli~hes those works from one another in the
text ~f the PanchasiclclMntikit really distinguished them in their originltl forra.
,Ve may note in conclusion that there is one interesting circumstance which
furnishes a kind of counterproof to this conclusion. According to VII. 1.
and VIII. 9 the Paulisa and Romaka SiddhD,ntas calculated the parallax in
lonzitude at a solar eclipse ill exactly the .sarue manner. Now Varaha Mihira
accentuates this agreement of the two works by stating the rule each time in
exactly the same words. But an author, who is so evidently desirous to mark
the points in which the different authorities 011 which he draws are ltt one,
lllay certainly be supposed to be no less scrupulous in stating the details in
which they diverge.
After having thus cleared the way, I proceed to give short summa-
ries of the doctrines of the five Siddhantas, beginning with that one which,
owing to the existence of a modern recension, is best known, viz. tho Surya
Siddhanta,
According to 1. 14 the Silrya Siddhanta of Vltraha Mihlra taught that S'\"y" Sid,
180000 years contain 66389 intercalary months, and 1045095 omitted lunar
days. The number 180000 is the twenty-fourth part of the years of a maha-
yuga; if we therefore, for comparison's sake, multiply the figures given above
by twenty-four, and deduce from them the number of the savana clays of a
ynga., we obtain 1577917800; while the corresponding figure for the modern
Siddhanta is 15779178:J8. The length of the sidereal year resulting from
these figures is 365 d 6h 12' 36"'56 in the case of the modern, and 365" 6 h 12'
36" in the case of the old Surya Siddhanta, The latter value exactly agrees
with that which, according to BhaHotpala and others, was assigned to the
solar year in the Paulisa SiddMnta.
What the old Surya SiddMnta taught about the mean motions of the
sun and moon, is immediately apparent from the above statement concerning
the nature of the yuga, The number of the moon's sidereal resolutions dur-
ing the yuga is the same as in the modern Siddbanta ; whence it follows that
each rev~lution is a little shorter (the yuga of the old Siddhanta counting
twenty-eIght days fess than that of the modern one), Rules how to calculate
the mean positions o.f the sun and moon are given in chapter IX; they how-
e.ver can for no special remarks, as they follow immediately from the constitu.
tion of the yuga.-The duration of the revolution of the moon's apogee may
xvm IXTRODUOTIOX.

be derived without difficulty from stanzas 3 and 4 of the same chapter. From
stanza 3 it follows that one resolution is performer! in 3231 d 23 h 42' 16"'76;
while the duration resulting from the elements of the modern Siddhilnta
amounts to 3~32d 2h 14' 53"'4. And if, accommodating ourselves to the
general Siddhanta practice, we determine the number of revolutions performed
within one mahayuga, we obtain 48S~19 for Varaha Mihira's Surya Siddhanta;
';hile the modern Siddhunta gives 488203 only. vYe note that according to
Aryabhata also the apogee performs 488219 revolutions within one maMyuga.

From stanza 5 of the same chapter we learn that the old St\rya Sid-
dhanta agreed likewise with Aryabhat~ in reckoning 232226 revolutions of the
moon's node to one mahayuga; while the modern Siddbanta counts 232228.-
In estimating the greatest latitude of the moon at 270 minutes (stanza 6) the
old Surya Siddhanta agreed with the modern one.

According to stanza 7 the old Surya Siddhanta assigned to the sun's

apogee the longitude of eighty degrees. Aryabhata gives 78° only, and a
C

calculation of tho place of the apogee for the epoch of the Paiiehasiddhantika,
based on the elements of the modern Slirya Siddlninta, gives about 77°. The
Panchesiddhantikn says nothing about the revolutions of the apogees of the
sun and planets, and it hence is possible that the old Surya SiddMnta was not
yet acquainted with the theory held, on entirely insufficient ground«, by the
modern treatise, and modern Hindu astronomers in genericl, that the apogees
of the snn and the planets perform a certain number of revolutions within a
inahayugo or kalpa. On the other hand it might be supposed that Var{,ha
Mihira, although acquainted with that doctrine, yet confined himself to stating
the places which the apogees occupied at his time, since so much is sufficient
for the purposes of a karaJ;la-writer.-The rules for finding the true places of
the sun and moon, which are given in stanzas 7 and 8, are analogous to those
of the modern Sitrya Siddhanta, with the one important difference that, while
the latter assumes epicycles of different size for the even and odd qnarters of
the revolution of the two bodies, Varaha Mihira's Surya Siddhunta knows of
one epicycle only for the sun as well as for the moon. The rules for finding
the true motion, etc. given in stanzas 13 and 14 agree with those of the
modern work.
The rules for calculating solarand lunar eclipses agree with the modern
rules as far as general methods are concerned, but at the same time show
many deviation in details; so f i. in the calcnlation of the parallax in solar
eclipses. Some of these rules we have, moreover, not been able to elucidate
to our full satisfaction.
 XIX
D1TRODUCTlOOl.

The mean motions of the planets (apart from sun and moon) are given
in chapter XVI. The following statement shows the numbers of coml~lete
revolutions during one mahdyuga according to the old and moclern Surya
Siddhautas.

Old S{l. Si. Modern Su. Si,

:Mercury 17937000 17937060

"Tenus 7022388 7022376
:Mars 2296824 2296832
Jupiter 364220 364220
Saturn 146564 146568

The two Siddhantas thus agree concerning Jupiter only, and disagree
therein from Aryabha~a, according to whom Jnpiter's revolutions ~mount to
364224 in one mahayuga. The old Surya Siddhauta agrees with Aryabhata
and the Paulisa SiddMnta (according to Bhaetotpala) , as far as Venus, Mars
and Saturn are concerned, while it agrees with the Paulisa Siddhauta only
concerning Mercury and Jupiter.

The positions of the apogees and the dimensions of the epicycles of the
apsis and the conjllllction are given in XVII, 1-3. If will be observed that,
as regards the numbers indicating the size of the epicycles of the apsis of
Venus and Saturn, the translation diverges from the corrected text given by
us. The manifestly corrupt text was at first emendated on the basis of the
dimensions stated in the modern Surya Siddhanta, the hypothesis of the
agreement of the two Siddhantas in this detail being resorted to in the absence
of evidence decidedly favouring any other assumption. But I afterwards
discovered that such evidence exists. The statements which Brahmagupta in
his Kbal)~lakMdyakakaralfa makes about the places of the apogees and the
dimensions of the epicycles agree with those made in the sixteenth chapter of
the PaiichasiddM.ntilcl, in all those details in which the text of the latter work
needs no emenclation, and it therefore may be presumed that the agreement
extended also to the epicycles of Venus and Saturn. And examining the
traditional text of the Partohasiddhsntika from this point of view, we find that
instead of the 'Suras' of stanza 1 we have to read not 'saris' but 'svaras'
and that the 'trirpsal;!' is correct without any further addition. It is true
that thus the Aryi remains defective; but the word, or words, missing were
most probably expletive rather than essential to the sense. BrahmaO"upta
maintains his katana to be founded on Aryabha~a, or at any rate to gi;e re-
 xx INTRODUCTION.

sults equal to those to be derived from Aryabha~a;* it is then a somewha

curious circumstance-into the discussion of which I cannot enter in this plac
-that the dimensions of the epicycles and the positions of the apogees assume.
in the Khandakhadyaka (as well as in the sixteenth chapter of the Paficha
siddMntika) differ, all of them" more or less from those recorded in the Laghu
..4-...ryabhatiya. t
The method, taug-ht in chapter XVII, of calculating the equations c
the apsis and of the conjunction ,agrees on the whole with that prescribed i
the modern Surya Siddhanta, although there are several divergences in detai ls
Peculiar are the special rule given for Mercury in stanza 10, and the correctio
to be applied to Venus' place according to. stanza 11. The statements as t
the distance from the sun at 'which the planets become visible differ to 80m
extent from those made in the modern Siddhanta; so also the greatest latitude
of the planets given in stanzas 13 and 14.
An omission which might make us suppose that the chapter as give
in our Manuscripts is not complete is that nothing whatever is said about th
places of the planets' nodes.
l>~iUmaha Sid, We next turn to the Paitamaha Siddhanta which. indeed has not corn
dhcntc,
down to our time, but whose teaching throughout agrees with that of a we
known section of Hindu astronomical literature.

Of this Siddhaata there treats only one very shari chapter, of th

Parichasiddhsntikz viz. the twelfth 011e ; but its five stanzas manifestly suffice t
reproduce everything of importance contained in that very primitive treatise
The Paitamaha Siddhanta, known to Varaha Mihira, represents Hindu Astrc
nomy as not yet affected by Greek influences.j and thus belongs to the sam
category as the J yotisha-Vedanga, the Garga Sa111hitli, the Suryaprajriapti an
similar works. From what Vadha Mihira says about its contents, we migl:
almost identify it with the J yotisha Vedal'lga. The yuga on which the calct
lations of the PaitamahaSiddhanta base IS the well known quinquennial on~

'if. Brahmagupta's Khll.I).qa-khil.dyaka begins with the following stanza

lIfl!Itt~ ~:nna:a srTT~fTJf~fn!l'i;itl~R I
~ '"
Cl~Tf;r ~l1G~m;nil"Tun:h~+lG('J;.;tr(Ii~;r II
'" ,
tIt is also worthy of notice that Amasarman , one of the Commentators, of the .Khan.d
-
khadyaka, quotes some stanzas from a Paulisa tantra which make the same statements about'tl
dimensions of the epicycles as the KhaJ,1qakhaclyakil. itself, and, moreover, seems generally to tre
the doctrines of Aryabhil.~a and the Paulisa a's equivalent.
t As already pointed out by me in my paper on the Jyotisha-vedanga, Journal of t]
.A.siat. Soc. of Bengal 1878.
 INTROI) UC,'TION. XXI

which consists of five solar years of 366 days each, and contains sixty solar
months, sixty-two synodical months, and sixty-seven so-calleel nakshatramonths
i. e. sidereal revolutions of the moon. TIle beginning of the yuga is marked
by a conjunction of the sun and moon at the first point of the na~\:shatl'a
Dhanishtba. The duration of the lon.gest day of the yelr amounts to eighteen
muh{lrta~> that of the shortest to twelve nauhurtas ; in the intervening periods
the days increase or decrease by the same daily quantity.-The PaiUmaJw.
Siddhanta refers to two points only which appear not to be mentioned in the
J yotisha Vedanga, as far as I have hitherto succeeded in making out the
meaning of that difficult treatise. It, in the first place, gives a rule for calcu-
lating the so called vyatlpata yogas. (st. 4); and in the second place, fixes a
period from which the quinquennial yug-as are to be counted. In st. 2 VadJw..
Mihira directs us to deduct two from the S'ika date, and to eli vide the remain-
der b}~ five; which implies that a new yuga is supposed to begin with tho
third year of the S'aka Era, or two S'aka elapsed.

Whether this direction is due to Varaha Mihira only, or was already

contained in the Paitamaha Siddhi1nta, may be considered doubtful; the latter
alternative, however, appears to be more probable, as Varaha Mihira, if in any
way adding- to-or rendering more definite-the teaching of the Paitamaha
Siddhanta, would most likely have adapted it to the same initial date as the
other Siddhantas, oiz. 427 S'aka.

The PaiMm.aha (Brahma) Siddhante known to Varitha Mihirahas thus

to be distinguished from the Brahma Siddhanta on which Brahmagupta's
Sphuta Siddhanta is based. That Brahma or Paitamaha Siddhunta is a short
treatise in prose, forming part of the Vishnudharmottara-Punsna, and belong-
ing altogether to the modern phase of Hindu Astronomy. Tho number of
Brahma Siddhentes, known at present, thus amounts to four, viz. the .Paitfi-
maha Siddhanta summarized in the Panchasiddhantilci, the Paitamahc Sid-
dhanta forming part of the Vishnudharmottara, the Sphu~'), Brahmasiddhanta
by Brahmagupta, and that Brahma Siddhanta whose more ordinary name is
S'akalya Siddhanta.

There now remain the Romaka, .Paulisa and Vasishtha Siddhflntcs, for
the teaching of none of which we have any other source of importance but
the Panohasiddhantiks, I begin with the first mentioned of these three
treatises.
The fifteenth stanza of the first chapter shortly describes the nature of It"m"ka
the yuga employed by the Romaka Siddbjmta. The yuga is called' one of
the sun and moon' i. e. a luni~olar one, and said to comprise 2850 years,
XXII INTRODUCTION.

which period is further stated to contain 1050 adhimasas and 16547 pra
i. e. tithipralayas, omitted lunar days. The above numbers of yearE
intercalary lunar months allow of being reduced by 150, and we thus find
in the opinion of the author of the Romaka, 19 solar years exactly co
seven intercalary months, or-if we take the entire sum of months-th
solar years comprise 235 synodical lunar months. The yuga of the Eo
is thus evidently based on the so-called Metonic period, named aftei
Athenian astronomer Meton who, about 430 B. C., showed the mea
improving the Greek Calendar of his time bJ the assumption of 19 trc
years comprising 235 synodical months.-That the Romaka Siddhanta, in
of making use of the simple Metonic period, employs its one hundree
fiftieth multiple, has a reason not difficult to discern. The author 0
Rornaka, although manifestly borrowing his fundamental period from the
at the same time wished to accomodate himself to the Indian fashion 0
culating the sum of days which has elapsed from a given epoch (the SO-(
ahargana) by means of a cyclic period comprising integral numbers of
years, lunar months and natural days. N ow the simple Metonic period
not represent an aggregate of the nature required, neither if we-with 1\
himself-estimate the length of the tropical Jear at 365 ~ days, nor if we
ourselves of the more accurate determinations by which later Greek ast
mers improved on the work of Meton, and it therefore becomes requis
employ a multiple. What the multiplying number is to be, of course de"
on the value assigned to the length of the year, and we therefore have
certain the opinion held on this point by the author of the Romaka.
data supplied in stanza 15 enable us to do so without difficulty. For
multiply the 2850 years of the Romaka yuga by 12 (in order to find the
bel' of corresponding solar months), add the 1050 adhimasas (whereb'
obtain the number of synodical lunar months), multiply by 30 (so as tl
the lunar days), and finally deduct the 16547 tithi pralayas, the final
amounts to 1040953 natural days : which being divided by 2850 (the nt
of the years of the yuga), we obtain for the length of one year 365 d 5h 5~
But in order to form an aggregate of years which contains an integral nt
of days and at the same time is divisible by nineteen, 19 x 50 =O! 2850
have to be taken.

'Thence the above determination of the year's length was adopt

the author of the Romaka, there cannot be any doubt. The year (
Romaka is, down to seconds, the tropical year of Hipparchus or, if WI
of Ptolemy who accepted the determination, considerably faultive as i
made by his great predecessor.
"

INTRODUCTION. XXIII

The rule for calculating the ahargana according to the Romaka

(1. 8-10), and so likewise the rules for finding the mean places of the sun
and the moon (VIII. 1. 4) immediately follow from the constitution of the
yuga, and have been elucidated in the notes to the translation. The length
of the periodical month would, according to the Romaka, amount to 27d 7h
43' 6.3".

To the apogee of the sun the longitude of 75° is ascribed in VIII. 2.

-The apogee of the moon and its periods of revolutions are not, in the usual
Indian style, treated apart from the moon's motion; the 8th chapter (stanza 5)
rather contains a rule for calculating the moon's position w ith regard to her
apogee directly i. e. without any preliminary separate calculation of the apo-
gee's place. The kenclra mentioned there is the moon's anomaly, and the rule
implies that the anomaly revolves 110 times within 3031 clays, in other words
that the moon returns to her apogee, or performs one anomalistie revolution,
in 27d 13 h lSI 32". 7.

By deducting the longitude of the sun's apogee from the mean longi-
tude of the sun we find the sun's anomaly, and may then proceed to calculate
his true longitude. For the latter process the Romaka Siddhanta however
does not supply any general rule, enabling us to deduce the required equation
of the centre for any given anomaly; but contents itself with stating the
amounts of the equation from 15 to 15 degrees of anomaly. These amounts
are stated in VIII. 3, and it is of interest to note that they agree very closely
with the corresponding amounts given by Ptolemy. The greatest equation
of the centre, which according to the modern Surya Siddhanta amounts to
2° 10 13 and which in no other Hindu text book known to me greatly differs
lf
,
' 0
from this latter value, according to the Romaka amounts to 2 23' 23 lf , while
Ptolemy assisrns
~ v
to it the value of 2° 23'; and also the equations for the
smaller anomalies show a pretty close agreement, as appears from the follow-
inz
v tabular
. statement

I •
Degrees of Anomaly. 15 30 45 60 75 90

Equation of centre according

to the Romaka. 34' 42 fl
1°8'37" 1 0 3 8 ' 39"1202' 49" 2° 17' 5 ! 2° 23 23"
f f

-
According to Ptolemy.
I 1° 9'
I
2 0 If.
I 2° 23 f

The values quoted from Ptolemy are those given by him for the quadrants of
the apogee. The Romaka Siddhanta apparently makes no distinction of
quadrants, but employs the same equations indiscriminately for all.
XXI\" INTRODUCTION.

In an analogous manner stanza 6 states the moon's equations,

centre from 15 to 15 degrees of anomaly. These equations do not agre,
closely with the corresponding ones of Ptolemy, according to 'whom the
est equation amounts to 5° I/.-The length of the revolution of the I
node amounts, according to VIII. 8, to 6796d 7h , in pretty close agTe
with Ptolemy's determination of the same quantity, ciz. 6796d 14 h etc.-
cerning the greatest latitude of the moon we have two conflicting state
implied in YIII. 11 and VIII. 14, provided the interpretation of
stanzas ginm in the translation be right. According to the former it
amount to 2.10'; according to the latter to 270 /, which is the value ordi
met with in Hinchi astronomical works, Regarding the explanation gi
the translation of stanza 14 I have to remark that it is an attempt on th
of my collaborator to connect the rule with the usual estimation of the 1
greatest latitude, while the fraction ¥-, if its denominator be taken
reduced Radius, would strictly lead back to a greatest latitude of 280/.
difierent values should be ascribed to the same quantity in one and the
book, might prima facie appear inadmissible; but it is by 110 means imp
that in some of the older Siddhantas there were incorporated empirical
borrowed from various sources, the rationale of which was not under

Stanza 13 gives 30/ and 34/ for the mean measure of the diame
sun and moon respectively, and st. 15 gives the ordinary Indian rule fa
ing the true diameters from the meau diameters and the true and
motions.

The greatest parallax is, as in Indian astronomy generally, SUI

to be equal to the mean motion during four naqikas; hence the rule gi
st. 9 for calculating the parallax in longitude, the result being the diff
of the parallaxes of the sun and the moon.

The parallax in latitude is calculated on the same principle (stan

-14), the result however not giving the difference of the solar and the
parallaxes, but merely the latter one, the solar parallax beinv
v nezlectec
v
inaccuracy in the preliminary determination of the zenith distance
nonagesimal is noted
-
in the translation.-The rule for caleulatinz
b
the de
of the eclipse, after the true latitude has been ascertained (st. 16),
usual one.

What remains unexplained of the Romaka Siddhanba are, princ

the different kshepa-quantities met with in the rules for finding the ah
(Chapter I), and the mean places of sun, moon. etc. (Chapter VIII).
 INTRODUCTION. xxv

of course, are intended to enable us to start in our calculation from the epoch
of the Panchasiddhantika (or of the, or some, Romaka-Siddhanta, about which
see below), and their elucidation would probably lead to some interesting
results. It will be observed that the rule for calculating the ahargan a
professes to be adapted to the meridian of Yavanapura, while the rules for
finding the places of the sun, moon etc. refer to the meridian of Ujjayini."
The difference in longitude of those two places is stated by Vadhamihira-=-
following the Paulisa Siddhanta as it appears-in III. 13.-A further
reference to the Romaka which has remained obscure to us seems to be made
in III. 73.-Whether any of the rules concerning the planets which are
given in the last chapter base on the Romaka Siddhanta, is doubtful.
From this short summary of the contents of the Romaka Siddhdnta
I pass on to the consideration of its authorship and time of composition,
coupling therewith-for reasons which will appear later on-an enquiry as to
the date of the Panchasiddhttntika itself
Hitherto it has been generally held, 011 the authority of Colebrooke
and Bhau D~jl, that the original Romaka Siddhanta was composed by Srishena ;
an opinion which I myself, when writing my paper on the Pafichasiddha,ntik;t
(Journ. Asiat. Soc. of Bengal) was not prepared to abandon entirely, although
then already certain considerations led Inc to suggest that S'dsheJ)a's work
might after all have been a mere recast of an older treatise of the b:1.1nO narue.
This latter view I now feel inclined to set forth as the only true one.
The authorities for Colebrooke's and Bhnu D£ljl's opinion were
Brahmagupta and his commentator Prithudaka Svamin, Brahrnagupta, in
a considerable number of passages of his Sphuta Siddhunta, refers to tS'rishOJ.Hl,
by name, and in connexion with those passages his commentator repeatedly
remarks that S'risho1)a was the author of the Romaka Siddh:lnta. And in.
one passage at least Brabmagupta himself mentions Sdshel)a in eonnexion
with the Romaka Siddhanta. That passage which is found in the Tantra-
parikshadhyaya (the 11th chapter of the Sphuta Siddhanta) was discussed by
me in the paper referred to above (pp. 290 ff.), but o\ving to the very corrupt
form in which the Manuscripts of the Sphuta Siddhanta exhibit its text I
did not at that time fully understand it, so that the meaning of just its most

-lI- The truth of this remark of course depends, in the first place, on the correctness of

the emendation in VIII. 5 owing to which we have subatituted 5 t:ollii: S Cl;;:<;lT~ (read 80 in the
text, instead of 51;('llllf6ot'l1TJ!) for the Q{(HTlfCla:lT of the Manuscript; and in the second place, OIl
the assumption that the clause "at sunset, at Avanti" has to be connected generally with the
rules given in stanzas 1-5. But both this assumption and the emendation appear to me
well founded.
 7 "PIW

IXTROD'CCTIOX.

important clause escaped me, R-S it seems to have escaped Cole brooke an
Bhau Daii. The text of the passilO'e.
O,.! 0
as appearing
I 0
in Colebrooke's manuscrij
~

(no\,> in the India Office Library), runs as follows:-

......... , . . . . ~ ~.....

~.n':l:r-TiTao:~p!l?:~l';m':;.;i'c;;;m;:n<;Y;i"l'lI Q

;;r!F!nTc;TE!'~ClTC;Tf1 'ifflfc;a~ f<3'~<:i'iiiRl~ I ~

~i7i1Hil-i~HI1Tfo:j 11~ci; ~1Jtll1fo:j ihS';;nr., I ~
~l:q'TI!l'inti'lt ElHT.,T'aC;'::!nI;; a~rf;r I I:l
't ~ ...... t"'.... e-
"'>ITlJT~l:l'jj'l"r"l'il 'l:l'I:ZlH:::r~~?;,,!fiT :a I \I.

~~l:r:iil~"ll!~fuf<3fl:;JTI1<3'rTl'.j;;;;n"'l <=l"'l::C11 I f!
BT[!llRCl1h.:ml:m;:qTf;;te;Tf;:q~;;r;:ffG:Z;R'J13;T1 ; s

.
if?;]:a tlf~f"<.'ltlTfHo:av.sT'l'iH!J1i":HlJ"GTR. I ':
!;jttiii!., 'lWRiT <:~r~n;rn:nl'i';!iR'l'i<1: I ~
.. ~ r ""' l"'" .. -
<J:RFiEi 'lTI lRiI EiH<;iror TCI~'5'~lIT I 'l.0

The other Manuscripts of the Sphuta Siddhanta known to me (or

belonging to the Bombay Government; one, a modern copy, in the library (
the Benares College; and one in the Royal Library of Berlin) have sam
important different readings. They all read in line 1 ~TC:° instead of ~T;;I° an
in line 5 i;lTc;n:r;J
G.
0 instead of <>!H.h;:;:r~ In line 7 the Ben. MS. reads l:l1f~lf7Si(
~
0.
~~.

the Berlin :MS. has CnIW~l~Sltl;:jfc;~fltfBF,!; and the Born. MS. 6lr~Y~~Fni
i'iii'H!ToTf!. In line 8 the Born. and Ben. =\188. read tfT<:fU11H'rl::1:rr,c:T1ili<m1aJ<=lTo. Lir
9 runs in the Berlin nIS. <J:Fli~C7 'li'ti'EiT :as:::~ra:l!ilm1ili: roR: <ii'<l'T. The Ben. :r.n
reads 'litRiT {F.ir~~"i":li'i!'iR: el'i'<J1'., and the Bom. 1'.18. ;zitR.lT t~l'atiul=lEfiTf1 ~i'l': oiil
I~l line 10, instead of cnfBJ:i5r the Ben. =\1:8. has Tar:rl'lZT, the Born. MS. f61fl!~T'
(not to mention less important differences).
The general purport of this passage is clear. It is meant as a eriticisi
of the performance of S'rishel~a, who in composing his astronomical text boo
borrowed rules and processes from various sources, and combined them int
an incongruous whole. Leaving aside for the present the second half (
line ~,andline IG, we may-emendating the text as given above with tl
help of the varietas Iectjonis-s-render the passage as follows.
'From the fact that 8'1'i8he1).3., Vishnuchandra, Pradyumna, Aryabhat
Lata, and Simha contradict one another regarding eclipses and similar topic
their ignorance is proved daily. The criticisms whieh I (in the preceding pa:
of the chapter) have passed on Aryabha~a are, with the requisite modification
to be applied to the doctrines of each of those teachers as well. I will howev.
make some further critical remarks on Srishena and others.
SHshel).a took from Lata the rules concerning the mean motions of tt
sun and moon, the moon's apogee and her node) and the mean motions,
 IXTRODt"CTIO::\. XXVII

Mars, Mercury's S'lghra, Jupiter, Venus' S\ghra and Saturn; from - - the
elapsed years and revolutions of the yuga; from Aryabhat,a the rules
concerning the apogees, epicycles and nodes, and those referring to the
true motions of the planets; and thus - - - '
Here we are confronted by the latter half of line 9, which seems to
state that thus the Romaka (Siddhanta] ',',-208 composed (krital.l) by Srishena.
But this would leave unexplained the last word of the line which three
L

Manuscripts give in the form 'kanthfL.' Keeping therefore this latter reading,
and substituting (with the Berlin and Born. 1\188.), 'ratnochchayo' for the
four aksharas preceding' Romakal),' I translate' and thus the Romaka
[Siddhanta) which 'was (or 'is ') a heap of jewels (as it were) has, by
Srishena, been made into a patched rag (as it were).'
In other words: S'rlShei)a incorporated into the old genuine Romaka
Siddhanta elements borrowed from various heterogeneous sources, and thereby
spoilt it, making it look like a piece of cloth, 01' dress, made up of various
patches.
The Homaka Siddhanta going under S'dsheI).8,'s name was thus not the
original on8, but merely a recast of it, into which new matter borrowed from
different astronomical writers had been introduced. This is neither improb-
able in itself, nor altogether destitute of collateral proof. For if we compare
the information concerning Srisheua's Eomaka, Siddhanta, given by Brahma-
gupta, with what we now know about the Romaka Siddhanta epitomized by
Varaha Mihira, certain differences between the doctrines of the two works
present themselves at once. I here confine myself to two points, the
consideration of which does not necessitate a reference to any other passage
from the Sphuta Brabma Siddhanta but the one quoted above. Tbe first
point of disagreement is that Srlshe1].a, according to Bralunagupta, borrowed
his rules for the spashtikarana i. e. for the calculation of the true places of
the planets, from Aryabhata, Now Aryabhata's rules are known to 1.13 frorn
the Laghv-aryabhatlya, and we observe that they agTee in an essential points
with the corresponding rules of the Suryn Siddlu111ta: specifying, as the latter
'Work does, the dimensions of the paridhi-epicycle of each planet, and
teaching how the equation of the centre is to be calculated trigonometrically
.c .
lor any gIven anoma1y. "y,
varana
1
11l.1mra
"Th.'f'l· 'sRomaza
1 Siddha
1 an t a on the other
hand, as we have seen above, makes no mention of epicycles, does not in fact
give any generally applicable rule for calculating the equation of the centre,
but merely states in a tabular form the equations, howsoever calculated, for
each fifteenth degree of the anomalies of sun and moon. That Romaka Sid-
dhanta therefore manifestly had not borrowed its rules from Aryabhata, and
we .i

XXVIII INTRODUCTION.

hence cannot be identified with S'rishe:r;ta's work. On the other hand it is

quite intelligible that Srlshe:r;ta, who appears to have followed the old Romaka
Siddhanta as far as the mean motions of the planets are concerned, should
have borrowed the rules for. calculating the true places-which his principal
authority was unable to supply-from the work of A.ryabhata. A second
firgument may be drawn from what, in line 7 of the extract quoted above
from Brahmagupta, is said about S'risheI).a having borrowed from some other
work (apparently some Vasishtha Siddhanta) his theory as to the elapsed
years and revolutions of the yuga. Judging from the expressions made use
of in that place and from the context in which it stands, Srishena's views
about the yuga must have been akin to those generally held in the Siddhantas
on that point, the yuga being a vast period of time comprising integral
numbers of complete revolutions of all the planets. But as we have seen
above, the yuga employed in the old Romaka Siddaanta was an altogether
different one, of a strictly Iunisolar character and hence consisting of a com-
paratively moderate number of years. When, therefore, Brahmagupta, in
the first chapter of the Sphuta Siddhanta, animadverts on the non-traditional
character of the Romaka Siddhanta,* he manifestly does not refer to the
recast by S'rishe:r;ta in whose hands the Romaka Siddhanta had assumed Q
more orthodox form, but to the genuine Siddhanta, which at Brahmagupta'a
time was no doubt still in existence and duly distinguished from Srtshena's
treatise.
We next have to consider the bearings of a date which, in the first
chapter of the Panehasiddhantiks, is mentioned in connexion with the Romaka- .
Siddhanta, Stanzas 8-10 which give a rule for calculating the ahargana
(i. e. the sum of civil days which have elapsed from an initial epoch up to a
given date) direct us first to deduct 427 from the number of the current Saka
year, which means that the initial epoch of the calculation is 427 Saka, It
then proceeds to explain the details of the calculation of the ahargana, and
closes with the words 'this is the ahargana in (or, according to) the Romaka
Siddhsnta,' •
That this date-427 Saka-ismentioned in the PanchasiadMntik!,
has been known to scholars since a considerable time. The astronomers, of
Ujjayint who furnished to Dr. William Hunter the list of astronomers with
their dates, published by Colebrooke (Algebra p. XXXIII), gave 427 S'aka as
the time of' (their· second) V~a Mihira. Albe, full refers to it as the date
 INTRODUCTION. XXIX

of the Pafichaaiddhantika. Bhau Daji quotes the stanza from the Paneha-
siddhantikd as furnishing the epoch of the Romaka Siddhanta, adopted by
Varsha Mihira also. (J ourn, Royal Asiat. Soc. New Series Vol. I). Dr.
Kern is inclined to look upon 427 Saka as marking the year of the birth of
Varaha Mihira who, as appears from a passage quoted by Bhau Daji, died in
S'aka 509. .
All these views clearly have no further foundation than the passage of
the Panchasiddhgntika about the calculation of the ahargana. The view that
427 Saka is the year of Varaha .Mihira's birth we may set aside without hesita-
tion. Dr. Kern was led to that hypothesis partly by the consideration that the
Pafiohasiddhantikft, which in one place refers to Aryabhata's views, could
hardly have been composed in 505 A. D. when Aryabhata-born in 476
A. D.-was only 29 years old. We now know-from Dr. Kern's edition of
the Aryabhatiya-that Aryabhatx composed his work in 499 A. D. already,
so that he might very well have been quoted in a book writtten in 505 A. D.
The other argument brought for.ward by Dr. Kern, viz. that V araha Mihira
died in 587, certainly goes some way to prove that the Pafichasiddhantikd
was not written in 505, but not that Varaha Mihira was born in the latter
year. The text of the Patiohasiddhftntikg enables us at present to judge of
the position of Varaha Mihira with regard to the date 427 Saka. From the
chapters on the Surya Siddhanta it appears that Varaha Mihira considers
that year to be the epoch of his karanagrantha from which all astronomical
calculations have to start; for all the kshepa quantities involved in the
different rules, given in those chapters for finding the mean places of sun,
moon, and planets, can be accounted for satisfactorily on that basis. I
have no doubt that also the kshepa quantities stated in the Romaka and
Paulisa Chapters admit of being explained on the same supposition, but
unfortunately we have so far not succeeded in finding the clue' to their right
understanding. N ow it would certainly be most satisfactory, if we could
assume that the Parichasiddhantika was composed in the very year which it
selects for its astronomical epoch, or at any rate within a few years of that
year; for as nearness of the epoch tends to facilitate all astronomical calcula-
tions and, at the same time, to minimize the inaccuracies resulting from the
fact that karana rules are often only approximatively correct, it is the interest
and the practice of karana writers to choose for their epoch a year, as little
remote as may be from the time of the composition of their treatises. The
positive statement, however, made by .A.maraja (as quoted by Bhau DijJ)
about the date of Varaha Mihira's death does not favour such an assumption;
and we mor~over find that the deduction of 427 forms part of a rule which
in the end is' saidto be 'in' or 'according to' the Romaka Siddhanta, This
- ---- ------------------------------------------

XXVIII rXTRODUCTION.

hence cannot be identified with S'rishelJ-a's work. On the other hand

quite intelligible that SrishelJ-a, who appears to have followed the old Ro
Siddhanta as far as the mean motions of the planets are concerned, s:
have borrowed the rules for. calculating the true places-which his prir
authority was unable to supply-from the work of Aryabhafa. A s
Mgument may be drawn from 'what, in line 7 of the extract quoted:
from Brahmagupta, is said about S'dsheJ).a having borrowed from some
work (apparently some Vasishtha Siddhanta) his theory as to the el
years and revolutions of the yuga. Judging from the expressions mad,
of in that place and from the context in which it stands, Srishena's
about the yuga must have been akin to those generally held in the Siddb
on that point, the yuga being a vast period of time comprising ini
numbers of complete revolutions of all the planets. But as we have
above, the yuga employed in the old Romaka Siddaanta was an altog
different one, of a strictly lunisolar character and hence consisting of a
paratively moderate number of years. When, therefore, Brahmagup
the first chapter of the Sphuta Siddhjmta, animadverts on the non-tradi
character of the Romaka Siddhanta, * he manifestly does not refer t(
recast by tMsheI).a in whose hands the Romaka Siddhanta had assui
more orthodox form, but to the genuine Siddhanta, which at Brahmagi
time was no doubt still in existence and duly distinguished from Srtsl
treatise.

We next have to consider the bearings of a date which, in the

chapter of the Panchasiddhantika, is mentioned in connexion with the ROl
Siddhsnta, Stanzas 8-10 which give a rule for calculating the aha
(i. e. the sum of civil days 'which have elapsed from an initial epoch uJ
given date) direct us first to deduct 427 from the number of the current
year, which means that the initial epoch of the calculation is 427 SakI
then proceeds to explain the details of the calculation of the ahargana
closes with the words 'this is the ahargana in (or, according to) the Rc
Siddhanta.' •

That this date-427 Saka-is mentioned in the Panchasiddhs

has been known to scholars since a considerable time. The astronome
Ujjayini who furnished to Dr. William Hunter the list of astronomers
their dates, published by Colebrooke (Algebra p. XXXIII), gave 427 S
the time of (their second) Varaha .Mihira, Albe. Unl refers to it as the

* 'Il:m~Ofi~T:
~
~qf~OfiT: ~nt'CffiiT: I
~

'Ilan;;t ~n{ii a mfi'lGtlih l1+IEti'E.aam~ II

 4

INTRODUCTION. XXIX

of the Paachasiddhantika, Bhau Daji quotes the stanza from the Pancha-
eiddhantika as furnishing the epoch of the Romaka Siddhanta, adopted by
Varaha Mihira also. (J ourn. Royal Asiat. Soc. New Series VoL I). Dr.
Kern is inclined to look upon 427 S'aka as marking the year of the birth of
Val'aha Mihira who, as appears from a passage quoted by Bhau Daji, died in
S'aka 509.
All these views clearly have no further foundation than the passage of
the Panchasiddhautika about the calculation of the ahargana, The view that
427 S'aka is the year of Varaha Mihira's birth ,'i 8 may set aside without hesita-
7

tion. Dr. Kern was led to that hypothesis partly by the consideration that the
Panchasiddhantika, which in one place refers to .Aryabhata's views, could
hardly have been composed in 505 A. D. when Aryabhata-born in 476
A. D.-was only 29 years old. 'Ve now know-from Dr. Kern's edition of
the Aryabhatiya-that .Aryabhata composed his work in 499 A. D. already,
so that he might very well have been quoted in a book writtten in 505 A. D.
The other argument brought forward by Dr. Kern, viz. that Vafilha Mihira
died in 587, certainly goes some way to prove that the Patichasiddhantikg
was not written in 505, but not that Varaha Mihira was born in the latter
year. The text of the Patichasiddhantika enables us at present to judge of
the position of Val'aha Mihira with regard to the date 427 S'aka. From the
chapters on the Surya Siddhanta it appears that Varaha Mihira considers
that year to be the epoch of his karanagrantha from which all astronomical
calculations have to start; for all the kshepa quantities involved in the
different rules, given in those chapters for finding the mean places of sun,
moon, and planets, can be accounted for satisfactorily on that basis. I
have no doubt that also the kshepa quantities stated in the Romaka and
Paulisa Chapters admit of being explained on the same supposition, but
unfortunately we have so far not succeeded in finding the clue to their right
understanding. Now it would certainly be most satisfactory, if we could
assume that the Panchasiddhantika was composed in the very year which it
selects for its astronomical epoch, or at any rate within a few years of that
year; for as nearness of the epoch tends to facilitate all astronomical calcula-
tions and, at the same time, to minimize the inaccuracies resulting from the
fact that karana rules are often only approximatively correct, it is the interest
and the practice of karana writers to choose for their epoch a year, as little
remote as may be from the time of the composition of their treatises. The
positive statement, however, made by Amaraja (as quoted by Bhau Daj:l)
about the date of Varaha Mihira's death does not favour such an assumption;
1 ~ 1 L 1_ 1 J 1 1 1 1 ~ ( l . , ..... .... J10