You are on page 1of 18

On Some Fragments of Áryabhaṭṭa

Author(s): H. Kern
Source: Journal of the Royal Asiatic Society of Great Britain and Ireland, Vol. 20 (1863), pp.
371-387
Published by: Cambridge University Press
Stable URL: http://www.jstor.org/stable/25581256 .
Accessed: 08/06/2013 04:56

Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at .
http://www.jstor.org/page/info/about/policies/terms.jsp

.
JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of
content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms
of scholarship. For more information about JSTOR, please contact support@jstor.org.

.

Cambridge University Press and Royal Asiatic Society of Great Britain and Ireland are collaborating with
JSTOR to digitize, preserve and extend access to Journal of the Royal Asiatic Society of Great Britain and
Ireland.

http://www.jstor.org

This content downloaded from 122.167.80.236 on Sat, 8 Jun 2013 04:56:59 AM
All use subject to JSTOR Terms and Conditions
371

Art. XV.?On someFragments o/Aryabhatta. By Dr. II. Kern.

[ReadWh March, 1803.J

There arc few names iu the history of Indian science which have
acquired a celebrity eepialling that eif Aryabhatta. lie is?to use
the words eif Ceilebrooke?" the* earliest author known to have

treated of Algebra among the Hindus, anel likely to be, if not
the inventor, the improver, of that analysis; by whom, too, it was

pushed nearly to the whole degree of excellence which it is founel
to have attained them." Rut, the renown
among notwithstanding
of the Indian anel astronomer, not einly among his
algebraist
but also among tho Arabian scholars, his works seemed
countrymen,
to be lost. Neither was Colcbrooke successful, nor was Davis more

so, in their cnelcavours to find work of his. What was known
any
about his doctrine, which iu many points deviated from the prevail
Indian astronomers, was elerived freim
ing opinions among quota
tions iu various mathematical and astronomical writings.
occurring
it is elcscrving of that in Southern India there* are
Now, notice,
extant of works that most lay claim te> being
copies uneepiiveically
the genuine of AryabhaUa. The late Mr. Whish knew
productions
an a treatise on arithmetic anel mathematics, tei
AryabhaUfyam,
which I shall have to revert of this paper. Prof. in the course
Lassen says, iu his Indische Alterthumsknnde,1 that he has received
from Southern India copies of two works ascribed to AryabhaHa,
of the above-mentioned and of the Dasagitaka
viz., Aryabhattfyam
siitra. In an article on the in the Oth veilumc of
Arya-siddluinta
the Journal of the American Oriental Society, Dr. Fitz-Edward Hall
has verified an Arya-siddluinta, of which hepewscssed two imperfect
extracts in the writings of various commen
copies, by occurring
The conclusion he arrives at is, that there were two
tators.

1 Ind. 2nd vol., p. 1136. Prof, Lassen concludes, from the somewhat
Alterth.,
words of Iteinauel, Memoire sur VInde, pp. 321 and 322, that Ilio
equivocal
ib spoken of hy Albfrfml. So far as f can sec, AlJifreinf inti
Aryabhattfyam
could he reehiewl to
mates only that the Indian astronomical Bystcms (mcthodr/t)
the Arynbhatta, anil the Arkanel. R is an inference of
throe, the Siiiflhinel,
that the called is identical with the Aryabbatttyam,
ltcinaud system Aryabbatta
which, to WIiibIi, is not a treatise on astronomy, but on arithmetic and
accoreiing
mathematics.

This content downloaded from 122.167.80.236 on Sat, 8 Jun 2013 04:56:59 AM
All use subject to JSTOR Terms and Conditions
372 ON SOME FRAGMENTS

authors of the name. If the same course were iu regard
adopted
to all the works ascribed to AryabhaUa, or to an AryabhaUa, if their
contents were with tho numerous scattered in
compared fragments
different works, chielly commentaries, one might indulge the heipo
that the question of the authorship of Aryabhavva would be settled
in a manner.
satisfactory
This is a task, however, I cannot assume at present; and I

propeiso to con tribute a small share towards tho
emly solving
question hy giving the following extracts. All of them are taken
from the commentary of Bhav(a -Utpala on the Varahi Sanhitd,
or Brihat-Sanhita, of Varaha-mihira. According to the list of
astronomical writers furnished by the astronomers of Ujjayinf to
Dr. Hunter, and published by Colebrooke (Algebra, Note E), the
commentator flourished about the ?aka-year 890 (a.d. 968-G9); and,
agreeably to this statement, it is said, at the end of the commentary,
that it was finished in the 9aka-ycar 888, or a.d. 9G7.1
I am sorry to say that the copies of Utpala's commentary arc
in such a-condition that I have often been to recur to con
obliged
jectures, most of them, however, being edivious. The uncertainty
about the true reading materially affects the translation; and the
want of any explanation on the part of Utpala docs so in a no less
degree. At the same time, I am of opinion, that the greater or less
degree of correctness of the extracts by no means renders them
less for the purpose of comparing them with the complete
important
works ascribed to the renowned Indian mathematician aud astro
nomer.
I shall begin with tho following extract, which is to be found
in the on the second of the Varahi Sanhitd :
commentary chapter

1 own words:
The date in full is given in IJtpala's

werefwi ^1% 3frH f^faw 11
" on a
On the second lunar day of tho elark half of the mouth Phalguna,
Thursday, in the Saka year 888, I finished this commentary."
2
One of tho MSS. belonged formerly to Colebrooke: for convenience Bake, I
shall call it ]1, aud the other A. Tho fragment is introduced, iu 13, by the words

fTOT ^T*WT'> ?* A by cTSlT ^r^FWT*?- Thc varioua readings,

mere blunders being omitted, are: 1) ^J^ffY Vllc\ for *Is[rU ""***
^Tf

This content downloaded from 122.167.80.236 on Sat, 8 Jun 2013 04:56:59 AM
All use subject to JSTOR Terms and Conditions
OF AHYAMIATTA. 373

jpsrerftfwgwT *%*vt*c^4??VfTi: ll

$^lr*?Rinr: WT^Y fWTfTT *rftf%?r:I
t^^r^i *ro T?ore: w4?r> \tt: ii

"
The terrestrial a of earth, water, fire, and
globe, compound air,
entirely round, encompassed by a girdle [the equator], stands in the
air, in the centre of the stellar sphere. Like as the ball formed by
the blossoms of the Nauclea Kadamba is on every side beset with
flowerets, so is the earth-globe with all creatures terrestrial and
Mount Meru, a
aepiatic. measuring only yojana, being wholly
round, and consisting of gems, spreading lustre, and surrounded by
the snowy mountain, is in the midst of Paradise.1 Heaven and
Mount Meru are in the midst of the land; hell and the
Vadavamukha are in the midst of the waters. The gods and tho
demons, of necessity, deem each other to be undermost."
The general of this geographical
features system agree with
that of the Siirya-siddhanta (chapter 12), and other astronomical
works. hell, naraka, is intended the abode of the demons; anel
Ry
Vaelavamukha, in popular belief, as well the submarine
done*ting,
fire as a kind of Charybdis and abyss, is here, as it will appear
more from another passage of AryabhaUa, the supposed
clearly
abode of the dead. We find here, about the division of our planet
into land and water, the same notion that not
nearly prevailed,
in other text-books of Indian astronomy, but also with
only
Greek as Eratosthenes and Strabo, that the
geographers, namely,
northern hemisphere only is a continent and inhabited, and that
the southern half consists of water, and is uninhabitable, at least for
men. A peculiar feature of this piece of AryabhaUa's geographical

1 The worel rendered here hy Paradise, is the well-known
Nandanavana,
of India. It is also called Nandanadyfina in the Kathu-sarit-Hiigani,
garden
laranga 28, vb. 62 (edition of Brockhaue in the Abhandlungenjiir die Kunde des

Morgenlandes, 1802).

This content downloaded from 122.167.80.236 on Sat, 8 Jun 2013 04:56:59 AM
All use subject to JSTOR Terms and Conditions
374 ON SOME FRAGMENTS

system, which distinguishes it from the prevailing opiniem e)f the
Hindus, is his acknowledging only four, neit* live, elements. To
account for this peculiarity by assuming him to have adopted this
doctrine from the Jlauddhas and Jaiuas seems objectionable, be
cause these sects are notorious for their ideas about
preposterous
the heavenly bodies,1 quite different from those of Aryabhavva.
1 would rather suggest that he borrowed this idea from the
Greeks. However, it is but fair to say that he could not have
borrowed the idea from an ancient work eif reputed Greek origin,
viz., the Paiilisa-sieldhanta, the author of which was or, as
Pulisa,
Albmiiif styles him, Paulus the Greek. Feu* we know, accidentally,
from a quotation by BhaYVa Iltpala, that the Paulisa-sielelhanta
acknowledges live elements. It will not be out of [dace to give
the very words:3?

^
rl^T fl<K*Hl<5n*T II
Tligir5T^rTO*ft
"The earth is round like a wheel (eir disk), produced by the
Creator at the end of darkness. It cemsists of five elements. In
the midst of it is the Meru
of the gods."
It will be seen at once, that the shapo of the earth being com
pared to a wheel or disk implies a much lower degree of geogra
phical science than is met with in any Indian book on scientific
astrone)iny.
To
revert to Aryabhavva, I subjoin the original of a passage
already noticed and translated by Colebrooke in his Algebra,
NotclV?

m^T^^Nr^ wii ^^^^ ii
" An
eepial half of the ecliptic, from the beginning of Aries to
1 Sec
Colcbrookc, On the sect of the Jttinas, and As. Rca., xil, p. 228.
a arc to be found iu Cod. A, fol. 29, in B, fol. 33.
They Tho latter MS. has

J
Hero both co])ica concur in reading ff^JT "^T^T^fHT*' intending, very

likely,
^T^IWF2.-A **W^R ?T, 15
TPg^RW0.

This content downloaded from 122.167.80.236 on Sat, 8 Jun 2013 04:56:59 AM
All use subject to JSTOR Terms and Conditions
OF ARYAmiATTA. 375

the emelof Virgo, is situated in the* north ; the either half, from the
eif Libra to the e*nel eif Pisces, is iu the seuith."
beginning
The following two passages bear upon the* relative distance of
the the sun and moon, from the earth, the
planets, including sup

posed centre of the celestial sphere, and upon the relative dimen
sions of the plauctury orbits: ?

iwrorg *?*rsn?iT11
*jpwfajcnr
"Relow the fixed stars are, successively, Saturn, Jupiter, Mars,
the Sun, Venus, Mercury, and the Moon; below all tlmse is the
Earth, being in the mielst eif the ether."l
Concerning the relative dimensions eif the orbits, AryabhaUa
has taught thus:?

^rqfwr^^f *rs^ w^cfj ^^Tft il

^TTUT:3n*rT^*N f^THRTW: ^J^P? II
" a small in a short
The Meion, undermost,
completes circle
being
and Saturn, a circle in a time.
time; being uppermost, large long
For a small circle the signs are small, whereas are
upon they great
and minutes are divided
upem a large circle; so, too, the degrees

proportionately to the respective orbits."
for we know, was the lirst, anel remained
AryabhaUa, aught,
almost the astronomer among his countrymen, who aHirmed
sole,
the revolution of the Earth on its own axis. His words, as
daily
quoted by Utpala,2 elilfer from those cited
by Prithiidakaswauiiu,
the commentator eif Rrahmagupta (As. Res., xii, p. 227); but it is
to remark that the same' author hael more than one
superfluous
cither in the same work or iu some other, to repeat his
opportunity,
assertion.

1 The word mcshlbhfUd in the translation,
is omitted since it is devoid of

meaning in this connexion.I surmise that wc ought to read medhfbhtitd. math I
" or barn, to which the cattle
being a post fixed in the centre of a threshing-floor
arc attached, as they turn round it to tread out the corn."?WiUon.

*a has b ^tt^i4^t^.
^T^r^r4*riN.

This content downloaded from 122.167.80.236 on Sat, 8 Jun 2013 04:56:59 AM
All use subject to JSTOR Terms and Conditions
376 ON SOME FRAGMENTS

^TWTfaWfa <T^**wf^*PITft ^nfFEfT1|l?
"
As a person in a vessel, while forwards, sees an im
moving
moveable object moving backwards, in the same manner do tho
stars, however immoveable, seem to move at Lanka (i.e.,
(elaily);
at a situation of no geographical latitude) they go straight to the
west (i.e., in a line that cuts the horizon at or, what
right angles,
is the same, tei the vertical at
parallel prime Lanka)."
The theory of the immobility of the Cxxod stars is not approved
of Bhavva Utpala, who endeavours, to refute it,
hy conseepiently,
partly by reasonings of his own, partly by invoking other autho
rities. One of his arguments is based upon tho supposition that
birds, after having soared to the sky, would be unable to find again
theirnests, *liH (if the Earthmoved) T^RT^T ^ ^TrtpU

^f^RPTOlTEn Further, he appeals to the Paulisa-siddhanta,
which taught:!?

" To its
(the celestial sphere's?) so-called upper pole this wheel
is fastened with cords of air; the (points of) rising and setting of
the stars the wind are
propelled by always revolving."
This is another instance of the inferior epiality of the science
expounded by Pulisa. Another authority appealed to by Utpala is
the Brahma-siddhanta, an extract from which is here subjoined:

^re ^rfY f^j^fcf MT^ft ^ *Wfrt I
WTOHJ^TO ^pfrMrt^l II
^PqT*T
1 The ia conjectural, A **ke
reading exhibiting "U% B TOcfT?$f. *^8 *?

general meaning there can bo no doubt.

This content downloaded from 122.167.80.236 on Sat, 8 Jun 2013 04:56:59 AM
All use subject to JSTOR Terms and Conditions
OF ARYARHATTA. 377

" The
circle of asterisms
(the (ecliptic) connected with the two
poles revolves, beung propelled by the provector wind; tei the gods
(being at the north jieile) it is in the horizon, and moves from left to
right,1 but, to the demons, from right to left. Everywhere else the
sphere is elevated, ami the pole is depressed. For Lanka {i.e., for
a place of no latitude), however, the circle of asterisms goes from
the east point of the horizon to the west point, and there the pole
is in the horizon."
Utpala might have selected more unobjectionable passages than
this, which commits the unwarrantable inaccuracy of confounding
the ecliptic with the equator. A thirel authority opposed to
it is neit a little to learn this?is
Aryabhatta-?anel surprising
Aryabhatfa himself. For Utpala cites from him as follows:?

" The cause of the rising anel setting is, that the circle of
asterisms, with the planets, being always propelled by the pro
vecteir wind, moves to the west."
straight
This may have been one of those on which Rralima
passage
gupta founded his charge of inconsistency against his renowned
(sec Colebr., Algebra, Note G, and As. Res., xii,
predecessor
p. Rut, without more eif the context, wc dare not
213). knowing
assert Aryabhatta to have been really inconsistent: the foregoing
stanza elocs not prove what Utpala supposes it to prove. Nor is it
necessary to assume that this stanza is from another work of

the same author. is by no means It improbable that the author
simply intended, here, to report the opinion of others, or to give a
description how the luminaries appear to move. It is deserving of
remark that Utpala, who could have strengthened his position by
pointing out the inconsistency, does nothing of the kind. This
loeiks as if he were himself aware of his inconclu
quotation's being
sive. I pass from the on the second of the
commentary chapter
Varahf Sanhita to that on the sixteenth, vs. C>. The to
fragment
be found there has been alluded to by Dr. Fitz-Edward nail, in the
Journal of tho Am. Or. Soc, Vol. VI, p. 559.

1 For the sense of the terms savyaga and apamvyaga, sec Sdryn-siddhfinta,
ch. 12, vs. 66, auel the valuable translation of it by tbe Hov.Mr. Burgees and Prof.
Whitney.

This content downloaded from 122.167.80.236 on Sat, 8 Jun 2013 04:56:59 AM
All use subject to JSTOR Terms and Conditions
378 ON SOME FRAGMENTS

^reirert ri ^t%i ^fg^^^^TO^iw: ii1
"The gods dwelling on Mount Meru, in the north, sec half
of the asterisms going from left to right; the same half is seen by
the dead in Vadavainukha, in the south, mewing from right to left."
As tei Meru and Vadaviiniukha, 1 refer hi the extract near the

beginning of this paper.
The last extract I have to offer is, in more than one respect, of
as has been mentioned Davis,
high iin|K)i*tauce. Aryabhavva, by
in the As. Res., iii, p. 215, and afterwards by Colebrooke, in tho
As. Res., xii, p. 244, and Algebra, Ne)te 1, stated the revolutions of

Jupiter in a Mahayuga tei be 301,221. The original is given by
in his comment on ch. 8, vs. I, of the Varahi Sanhitd:2
Utpala

*t^t*tot TTft^WT ^TFq^pirgT^kH^t: ii
^W
^^vr^TWT^t II
f^^^^^^cJW
"The revolutions of Jupiter, multiplied by the number of
the signs (12), are the years of Jupiter, called Aswayuja, &c; his
revolutions are equal to the number of the Jiuas, a couple, the
Vedas, the seasons, the fires (i.e., 3(14,224)."
There are, iu this stanza, two circumstance's that call for par
ticular attention. In the first [dace, it is clear that Aryabhavva
most distinctly states the first of the twelve years in the revolution
of Jupiter to be Aswayuja, whereas, generally, up to the present
day, the first year is called Kdrttiku. If we were sure that this
fact implied another, viz., that Aryabhavva considered the lunar
mansion Aswinf to be the first of the series, instead of Kriftikd, we

1 Uoth MSS. have W*U which here is devoid of sense, and more
^irSf?
ls *s
over eliaturha tho metre. Further, A has ^Tp(jR^> B
^e^^TO^.

* B has
^TSJT for ^J^SfT* and 3<^ I4!- The latter ia unobjectionable,

it ho taken for two worels, not for a compound.?A
provldeel l%^^?*Icl<^*rv?

Uoth havo fT^JTti *"iei have added, in numerals,
<^8^8?

This content downloaded from 122.167.80.236 on Sat, 8 Jun 2013 04:56:59 AM
All use subject to JSTOR Terms and Conditions
OF ARYARIIATTA. 379

should have one datum mow for determining his
approximately age.
Before upon the we have to bear iu mind
pronouncing question,
that the rule*, according to which the twelve years of Jupiter
received their names, is not uniform. Varalia-mihira1 proscribes,
"
that the name eif the ye*ar of Jupiter is determined from the
asterisrn with which Jupiter rises hediaeally, and that the years
follow iu the same order with the months." Other astronomers or

of far more ancient elate, and Rishi
astrologora asCarga, Kasyapa,
rules of the same In order to avoid needless
putra, give purport.
prolixity, I refer to the copious treatment of Ihis subject by Davis,
in the third volume of the As. Res., pp. 217, sqq. If is obviems that
the' rule laid elown in Varalia-mihira, or rather repeals! by him,
was framed at a time when the series of asterisms was reckoned
from Krittika, and not yet from Aswini. It is, further, obvious that
the same rule could not be followed after Aswini's beginning the
unless the fust year its name. Now, wc are
ecliptic, changed
directed by the Surya-sielelhanta (ch. M, vs. 17) to determine the
name of a }rear of Jupiter from that astcristn in which the Sun and
Moon will be in conjunction at the end of the month iu which
the heliacal rising aud setting take place. The first month being
called Vaisakha,?so-called because originally full moon occurred in

the asterisrn Visakha,?it feillows that the conjunction of the Sun
and Moon, at the end of this month, will fall inKrittika; and, agree
to the rule, the first year of Jupiter is to receive the name
ably
of Karttika. In this manner the order of the years is preserved;
but it could be so only if the Siirya-siddhauta, which considers
Aswini to be the first asterisrn, changed the rule of Varsilia-mihira,
or rather of his predecessors. Very likely the rule of the Surya
sielelhanta is au innovation by which itwas designed to find a means
of leaving unaltered the traelitional order eif the years. It is

sure, that AryabhaHa, whether the lirst asterisrn with him
pretty
was Krittika or ABwiuf, could not have determined the names eif

the in the same manner with the For, on
years Surya-sielelhanta.
the former supposition, the first year would have been called
in the latter, it would receive the name of Karttika.
Margasirsha;
we cannot assert so whether he did en
Unfortunately, confidently
did not make use of the rule with Varalia-mihira. If hereally
diel, then the necessary consequence woulel be*, that iu his time
the first of the series of asterisms was Aswini'. Rut it will be
permitted tei affirm that the fragment quoted affords no sufficient
data for deciding this epieslioit satisfacteirily.
1 Var.
Sanh., ch. 8, vs. 1.

This content downloaded from 122.167.80.236 on Sat, 8 Jun 2013 04:56:59 AM
All use subject to JSTOR Terms and Conditions
380 ON SOME FRAGMENTS

We are in a more favourable position as to the second fact worthy
of notice, which I alluded to. Feir it may be sherwn that the way in
which Aryabhavva gives the number of Jupiter's revolutions implies
that he has made use of our arithmetical notation. This fact,
although by no means surprising of itself, has acquired importance
since the discovery, by Whish,1 of au alphabetical notation in the
AiyabhaHiyani. This eliscovery has induced Prof. Brockhaus3
to assert that was with the decimal
Aryubhavva unacquainted
notation; and with him agrees Rcinaud, to whom we arc indebted
for an Memoire sur VInde. The latter scholar expresses
interesting
his opinion in the following words:8 "Apparemment, le proceehs
par Aryabhatta (viz., the alphabetical notation) etait tombe
empleiyei
en desuetude au Neanmoins, Icb traites scien
temps d'Albyrouny.
au 7? sicclo de notre ere, et
tifie[iies comjKiscB parBrahmagupta,
les ecrivaius ne en
par posteirieurs, supposeut pas, general, Tusage
deschiffres; les nombres semt par des mots susccptibles
exprimes
d'etre rattaches a uno Now, it is very true,
quality quelconque."
as well as known, that, iu Sanskrit works, numbers
generally
may bo denoted by other words than the usual names of the
numerals. This, however, has nothing whatever to do with tho
question of a particular notation, but is a matter merely of prosody.
It is quite immaterial whether one chooses to say tri or agni, as,
iu English, twenty or a score; the essential point is to know whether
the name of a unit, whatever the sound of that name may be, may
occur at another place than the first right-hand place. When it is
unquestionablo?and so it is in the above-cited stanza?that, feir
instance, the sixth numeral from the right, denoted by a word
for three, has the value of three hundred thousand, it is clear that
the notation must have been a decimal one. Every passage in any

Sanskrit book where wo meet with the same method of expressing
the numerals the namcB of considered to present them
by things,
selves in a certain quantity, tells us that it originated at a time
when the decimal notation was known iu India, with this restric
tion, that the numeral must cemsist of at least twei figures. The

shows, not the views of Prof. Brockhaus
fragment quoted only
anel M. Rcinaud to be inadmissible, but also, I believe, those of
Prof. Lassen. This most scholar argues, from tho
distinguished
existence of an inseriptiem of the Samvat-ye&r 380 or 323, which
exhibits an alphabetical notation, that, in tho time of AryabhaMa,
1 Transactions 1827.
of the Lit. Soc. of Madras,
9
Ze'UscItrift fur die KumU des Morgenlandes, vol. Iv, p. 81.
* sur VInde, p. 299.
M6moirc

This content downloaded from 122.167.80.236 on Sat, 8 Jun 2013 04:56:59 AM
All use subject to JSTOR Terms and Conditions
OF ARYABHATTA. 381

the decimal system was not yet fully developed; although he rightly
urges, ou the other hand, that the discovery of an alphabetical
notation does not involve an entire ignorance of the ciphers (see
lud. Alterth., 2nd veil., p. 1131), aud Thomas, iu his edition of
Prinscp's Indian Antiquities, 2nd vol., pp. 80-84). Assuming the
Samvat-yviiv to mean the year of the Vikramaelifya-era, it remains
still to be* proved, first, that AryabhaHa lived before or about JJ80
or 82') of the and, that the use of a
Vikramaelitya-cra, secouell^,
certain notation iu one monument involves the currency
general
eif the same notatiein, for every use, the whole of India.
throughout
Thus we may safely conclude, 1 think, that the great algebraist
has made use, in his calculations, of the system handed down to us

by the Arabians, and to these by the Hindus. Tei the unpractical
alphabetical notation no regard lias been paid, even by those
Indian mathematicians who derived their science mainly from the
works eif the reputed founder of algebra. Whish, I. c, p. 50, tells
us: "I am not, indeed, aware that it has since been iu use among

mathematicians, never having observed it in their works. It must,
however, have been understood by them ; for the three books,
Lilavati, and Maha-Bhuskaiiyam, of Bhas
Laghu-Bliaskariyam,
are founded upon the principles of the Aryabhatiyani;
kara-acharya
the Lilavati, in particular, is composed from the second chapter of

the work."
If Bhaskara nowhere mentions the alphabetical notation,
Albfriini knew nothing of it (see Memoire sur VInde,
decidedly
p. 299); which would be rather strange, if, by the third of the three
astronomical systems, Siiidhiud, Arkaud, and AryabhaHa* the
really was intended. Rut I have already stated
Aryabhatiyam
my grounds fen*not believing this work to be identical with the
system of AryabhaHa.
The above of AryabhaHa are not of a nature, at
fragments
least to my idea, tei afford new data for fixing his age. The
last extract, exhibiting Aswayuja as the name of the first year in
revolution, makes, indeed, the impression that he considered
Jupiter's
Aswini to be the first asterisiu; but it is no more than an hypo

so as it is neit corroborated by other evidence. And
thesis, long
it may be observed that, if the hypothesis, after further investiga
tion, should prove fo be true, it still does neit admit of the inference
that, in his time, the very beginning of Aswini coincided with the
vernal Anel every eif difference in
cquiimx. de*gree longitude
involves a difference of about years in the time of preces
seventy
sion, Therefore I shall content myself with referring to tho dis
vol. xx. 2D

This content downloaded from 122.167.80.236 on Sat, 8 Jun 2013 04:56:59 AM
All use subject to JSTOR Terms and Conditions
382 ON SOME FRAGMENTS

cussion of the subject by a scholar who with most profound and
vast combined so much cautiousness and conscientiousness,
learning
?by Colebrooke. He has made out that Aryabhatta lived before
Varaha-mihira, who, to external evidence, lived in the 6th
according
century of our era (Algebra, Notes I and K; cf. Memoire sur VInde,
p. 337). He was not aware of Vardha-inihira's distinctly mentioning
Aryabhavva; and, certainly, neither in the Sanhitd, nor in the Brihad
Jataka, nor in the is the name to be met with.
Laghu-Jataka,
Nevertheless, I believe I have found a passage, quoted by^Utpala,
and perhaps taken from the Panchasiddhdntika, inwhich the name
occurs. The passage itself must have been known to Colebrooke;
but the MS. he possessed is so badly corrupted, that it seems to
have his notice that the passage was from Varaha-mihira.
escaped
In order that others may judge for themselves, I subjoin the
whole passage, which is a little but, for
long, many reasons,
curious.

Bhavva Utpala, after giving the rules for finding the lord of
the day, month, year, &c., to intimate that learned
proceeds
authorities are at variance the commencement of the
touching days
of the week (and consequently about that of the creation);
some from others from others from
reckoning sunrise, midday,
sunset, and others, again, from midnight.

" The lord of the day of the week begins, with some, at sunrise,
with others at with others at with a fourth
midday, again sunset,
school at midnight. So says also the Achdrya." The title dchcirya,
by itself, invariably means, with Utpala, and the
Varaha-mihira;
following extract exhibits, in every the famous
respect, astrologer's
style, being defective in precision and clearness;?

1
B, tbe MS. formerly in the of omits
possession Colebrooke,

^ I*i | *^?Tl^I3iL which may account for his knowing only three schools,
S **!
instead of four. See Algebra, Dissert., p. viii. As.
Cf, however, Davis, Ites., ii,
p. 2C1.

This content downloaded from 122.167.80.236 on Sat, 8 Jun 2013 04:56:59 AM
All use subject to JSTOR Terms and Conditions
OF AUYABHATTA. 383

^TWTf^i3nrfim*T*!rrsfa f%^r^rrcnfrrs: i
^TZT^T^Tt^ *RTgTS *4t^PT **T II'

*njfn?rrTO*ra f^iTT^f? anrr^ ^ref*r?: i

^^ ^r *r>^nyr*if irrs^Hre: ^fagr* f%^3^
i

? A has
cTT^pgT^UT.-B cTT^R

? B omits
^nf.-A cf^WT:.
3 A out of the question.
"^T^T^r^Tl"** winch is wholly The reading of

B, TTpEJ^Tir!, however, is also against the metre ; and therefore I venture to

read "^T^RTZ"!, & form not infrequent in MSS., and in favour of which might

be the fact that the Arabian transcription sounds Arjabhar. A has
urged

*a B ^rm ^ rr%r[ i
TRrrfw5! 1^ ^^rN?
^TrI^Tf%l^ ^T^lt rTTTt^.
2 1) 2

This content downloaded from 122.167.80.236 on Sat, 8 Jun 2013 04:56:59 AM
All use subject to JSTOR Terms and Conditions
384 ON SOME FRAGMENTS

^^T
^ ^T^ ^T^^4 f^^frl^^ II

^
cTW^ftfaFT^TTTTfblrTrrt: IttfH II
^frqr%i tit^V f^^K ^Rq^: ^tits^-t i

" The from the sum of elays
day of the week is to be determined
the creation); the sum of stands in connexion with
(viz., from days
situation and daytime. Lavdcluirya says that the days are to be
reckoned from sunset in the of the Yavanas. Sinhdeharya
city
states the sum of days (to begin) from sunrise at Lanka; and, in
adopting this view, they must begin, with the Yavanas, at the time
when ten muhurtas of the are has
night past. Aryabhava
allirmed that the day begins (i.e., is considered by some to begin)
at midnight at Lanka; but the same says, further, that the eluy
commences to his own from sunrise at
(viz., according opinion)
Lanka, if, after taking into account the* different eif a
longitiule
country, the rc-sult eleies not agree with the in this
actuality
country, yet the correspondence of time has been sfateel the
by
same authorities, on scientific as billows :2 The
principles, sun,
while rising in Inelia, at the same moment cause;s in the
midday
region of the Bhadrdswas, sunset iu that of the Kurus, midnight in
Ketuniald. At the time of the sun's at Lanka, he is setting
rising
iu the city ot* the Perfected; it is midday in Yamakovi, and
midnight in The the omitted
Roinaka-country. intercalary months,
lunar the
days, elays of the planets,' the lunar days, the elays, the

1A
Tjfr^i?.
5
The meaning of this obscure stanza seems to he, that there is no general
rule for determining the commencement of tho days of the week, this depending
not only upon longitude, but also on the particular custom of a country.
What, can be brought under scientific
nevertheless, rules, is the correspondence
of time. In other words, the rules are
only theoretic, yathdulstnnn.
3 IJ'nariitia and its synonym Kinardtra are wanting in the dictionaries.
"With the aid of we can make out that tho niyht, or nearly
etymology, they mean

This content downloaded from 122.167.80.236 on Sat, 8 Jun 2013 04:56:59 AM
All use subject to JSTOR Terms and Conditions
OK AIIVAIU1ATTA. 38-0

Ram, the Moon, the Sun, the half-years, the seasons, the daily
motion of the stars, the start at must be irekenied
nights, {i.e.,
the same moment in the commeiie*.einent of the ymja. It
from)
is not the same if you count the freim Romaka
longitude
anel from nor is it alike, if you
country Yavana-city; begin
from at Lanka and from sunrise there. Neither elo wc,
midnight
when determining freim sunset the lend of the day feir every situa
have oral evhlonoo, nor is there any menus for
tion, trustweirthy
inference. In one place it is twilight, elsewhere day, or night; in
short, for every small distance there is a different lord of the elay.
the lord of the day is an entangled The
So, ascertaining epiestiem.
case stands the same for the return eif the hords; for the hard is

the first (hour) e>fthe lord of the elay: the latter not being fully
how can the lord of the hard be sei? Generally, poe>ple
ascertained,
here their concerns on a eif the week, withemt con
regulate day
such the learned preinounce it to be right,
sidering questions;
when one takes the day tei coincide with the limits of the
not the mean) lunar day."
apparent {i.e.,
Besides the name of AryabhaUa, or Aryahhata, we are indebted
to this for one more into the* geographical system
quotation glimpse
of this astronomer and two others, La is toll a ry a and Sinhae-harya.
It appears, from the worels of the text, that flu'se three agreed
with each other. their with the doctrine of
Comparing eipiniems
the vs. 37-40 and vs. we find the
Siirya-siddhanta (ch. 12, 70),
samo As Albirunf seems to give a somewhat different
agreement.
as ami
account, 1 will quote*
his words, translated by M. Reiuauel,
examine the weight tei be attached to them:1 "11 est dit elans
le Paulisa-Siddhunta, Panics le (Jroc, que la terrc et la
compete; par
incr sont une nominee Tilknschu,* o'ost-a-dire,
ceiupt'*es par lignc

ligne sous laepielle il n'y a pas de latitude. Cette ligne rcpoml a
ce nous la Aux ipiatro peiints
que nppelons ligne eepiinoxiale.
cardinaux sout villes considerables, a savoir: Yama-lvota,
quatrc
au midi; a Pouest; et au
al'oricnt; Lanka, Romaka, Siddhapour,

the synod icai lunar month of thirty tithis ib less than thirty solar
so, by which
elays. The number
of tithis continually gaining upon that of the polar elays, tho
will amount, in a little more than two mouths, to a whole lithi, which
difference
lie ex itmirdtra is rendered by tithinaya; it properly
must pun gee!. Therefore,
being the cause of the latter becoming necessary.?The elays of tho planets mean
here, I think, the elays of the week.
1Mtm. snr Vlndc, p. 341.
* This worel is certainly wrong. For 'j one may read j,and pronounce

nilakslia, the Sanskrit niroksha.
closely approaching

This content downloaded from 122.167.80.236 on Sat, 8 Jun 2013 04:56:59 AM
All use subject to JSTOR Terms and Conditions
38G ON SOME FRAGMENTS

a
nord. Quand lc sedeil se leve sur la ligne epii va de Lanka
Mcrou, il est midi it Yama-Kola, minuit a Romaka et six heures
du soir a Siddhapoiir." Rcinaud proceeds thus: " Albyrouny ajoute
que la mcmc avait etc* emise elans les Siddhfuitas compels
opinion
par Aryabhatta, Vasishtha ed Lat; main il fait observer avee raison
que, d'aprcs les tonnes iiiciucs, il ne* pent lades epiatre points
s'agir
eardinaux, niais eh*s ejuaelraus ele 1 cannot mis
IVepialour." help
trusting here the ae'euraey of the Arabian astremeuiier. The
doctrine eif the Paulisa-Siehlhanta, absurd as it is, is consistent
with the false assmnptieni eif the Earth's the of a wheed
having shape
eir disk, as we have seen abeive. Tluu-cfore this Siddhanta must
have inteueh*d the* feiur cardinal On the other hami,
really points.
it is quite imjKissihle that Aryabhatta, or any one else who supposed
the Earth to be a took the same view eif the course eif the
ghdie,
Sun. The weirds of Varaha-mihira, as well as those e)f the Siirya
siddhanta, leave nei doubt as tei the author's of
meaning quaelrants
the equator, and not the cardinal wo know
points. Moreover, that,
according to Aryabhatta, at the* north pole was situated Mount
Meru, and, at the south Hell. 1 fear that Albi'runf has been
pole,
too in passing his criticisms..
hasty
I shall conclude with a remark the occurrence of
suggested by
the names of Romaka and
country Yavana-city.
It is intimated, that, iu the country eif the Yavanas, the time
te) sunrise at Lanka, that in the morn
corresponding is, six o'clock

ing, is ten mulnirtas in the night. It is not distinctly saiel at what
time the night is to begin; but it can hardly be at another hour than
six o'clock in the ten muhurtas later two o'cleick
evening. Now, gives
in the morning, from which we derive the that
fact, Yavana-city
was to have a west from or the meridian
presumed longitude Lanka,
?f Ujjayhif, tei GO degrees. The meridian over Yavana
amounting
city a eif 30 east from
having longitude de?grees
(ceuiseepieutly,
we arrive at the conclusion
that is
Romaka-eity), by Yavana-pura
meant Alexandria. The crreu* iu the hingitiiele of Reiiue
ilele*rmining
from the inerieliau of is but neit sei very
Ujjayini conspicuous,
astonishing, if we*, remember even
that estimates of
Ptedi'iny's
longitudes exceeel the* truth. Sei, fei take* eine* iu.stauce1,
Kanogiza,
or is maele him tei have a
Kaneij, by huigitueh*, east from the
lnsuhe Portmiutu*, of Mfi the erreir te>
degroe.'s; niiuiiuitiug nearly
50 degrees. The llinelnerrors of peisitiem, with
carefully compared
the errors iu the writings of (Jreek
gveigraphers, successively,
would load tei results for the times
perhaps important ascertaining
of intercourse between Greek ami Indian science. It is, however,

This content downloaded from 122.167.80.236 on Sat, 8 Jun 2013 04:56:59 AM
All use subject to JSTOR Terms and Conditions
OF AttYABHATTA. 387

not the object of the present paper to do this; and 1 fear 1 have
made too a
already long digression.

ADDITIONAL NOTE.
Since the reaeling of the foregoing paper, 1 have been enabled,
by the kindness eif Prof. Weber, tei ascertain, that all the quotations
of with the eif one half-stanza, oe:cur in a MS. eif
Utpala, exception
the Berlin Library. In the ceileiphon of that weirk the title is given
as feillows: "Ity Aryabhata-siddhauta-vyakhyane Bhutapradipe
The stanza is ceiinmu
Dasagiti-bhashyain sunuiptam." conolueling
nicateel in the Jourii. Ainer. Or. Soc, fil.h Vol., p. 5(11, hy IVof.
u
with the translation : Bhula-Vishnu hath
Whitney, feilleiwing (?)
lcuriie*d it by the* favour of
thusconqireheiisively explained?having
his teacher?the text-boeik e?fwry obscure
Dusagiti (Dasagiti-sutram),
llluitta." It is evielent that
meaning, formerly promulgated by
must have considered this Dasagiti-bhashyain either as a
Utpala
eir as an authoritative; para
composition by AryabhaUa himself,
and substitute. It is somewhat that, at the end of
phrase puzzling
"
the* 3rd eif the Berlin we finel the words: Iti emnael
chapter MS.,
siddluinta golapmlas chalurthah."
acharyaryabhatavirachite (r.-te)
"
In the sccoriel the author his own elate*:
chapter gives shustynh
danain cha iadhika vinsatifr
shashtir yaela vyatitas trayas yugopadnh
tad ilia mama The metre eif the second
a)bdas jannialei'li'tali."
half-stanza an emendation : we arc to reael tadeha,
reepiires perhaps
in lien of tad tha. But still there remains a dilliculty as to the
"
of adhikd vinsatir. The stanza may mean : Whe*n three
meaning
eif the four and are then there are
Ages, 3,000 years, past, past
more than freim iny birth." If (his be the* sense,
twenty years1
the date wouhl be the y<*ur 478 from the beginning of the
Christian era. it is not clear whether the date of
Unfortunately,
is meant, eir that of llhutavislmu. However it may be,
A'ryabhaHa
certain it is, that one stanza in the large extract which I have
ascribed to Varnha-niihira, is to be found in the Dasagfti-bhashyam,

T^f &c- Hence this work must have been
viz., ^^"SjY "fcft^flf
anterior tei his time.

1As a date I should surmise that adhikd is a misread
requires precision,

the characters ^ and liable to he confounded. The
tryadhikd; "3J" being
" the date, a.d. 475.
rendering, in this case, would be, twenty-three yearn," and

This content downloaded from 122.167.80.236 on Sat, 8 Jun 2013 04:56:59 AM
All use subject to JSTOR Terms and Conditions