o

Indian
Historical
Researches
 The Asiatic Society
1, Park Straat, Cslcutta>700 0l6

Book is to ba raturnad on |ha Data Laat Btampad

Data Vouchar No.

 PURCHASED

TWE
BAKHSHALI a\anuscrif>t
in «9ICe</tevai JkCnih^mniic*
 HISTORICAL

THEBAKHSHAU
MANUSCRIPT
Early Hindu /Mathematics
A Stud^ in A^ediaeval Mediernatics,

KAYaR.

vd. 24 (i)

^osMd pgblicatioms
 3: 39
V :

P+ - i-

FirstPubliahed 1933
This series 1987

Published by
KAPOOR (Mrs)
rani
COS/AO PUBLICATIONS
24-B. Ansari Road. Darya Gai^,
New Delhi-1 \0002 (India)
Printed at
tA/S Mehra Offset
New E>elhi
 \
PURCHASED \

CONTENTS.
Preface ........... PAGX.
i

PART I.— INTRODUCTION.

Chapter —
I.

—
....
The history of the manuscript
.... 1.

Chapter
Chapter
Chapter
II.
III.*— Order
—
.........
Description of the manuscript

IV. Contents of the manuscript . .

8
12
.15
Chapter —
V. ^Exposition and method . . . . . .22
—
Chapter
Chapter
Chapter
—
VII. Measures
—
....
VI. Analysis of the mathematical contents .

VIII. ^The sources of the work . . .

38
54
.69
Chapter —
IX. The age of the manuscript and the work . 74

PART II.—THE TEXT,

i.

ii.
The script (with 4 plates)
Tranditeration of the text
.... .
^ .
87
105
iii. Facsimiles of Uie whole text. Pi.ate8 I — X1L.II.
Index.
 PREFACE.

In order to correct an impression that certain passages in this volume might

convey unless distinctly qualified, I must here refer to my indebtedness to the late
Dr. Hoemle. Indeed, a considerable part of the analysis of the MS. is really his

work,* and by his preliminary survey of the manuscript my task was consider-

ably lightened. It was at Dr. Hoernle's special request that I undertook to carry

on the work he had started, and he handed over to me most of the mateiial he had

himself prepared. Had he lived a little longer I should, no doubt, have had the

benefit of further help from him, and this volume might have been issued as our

joint work. Dr. Hoemle’s lamented death prevented that plan being carried out

and unfortunately my views are so often opposed to those that where held by

Dr. Hoemle that it would hardly be proper to make him a participator in them.

I am muoh indebted to Bodley*s Librarian for special facilities that enabled

me to examine the original manuscript under the most favourable conditions to ;

the Oxford University Press for th^ most excellent work in preparing the photo-
graphs of the manuscript and the collotype reproductions of the text ; and to the

Manager, Government of India Press, Calcutta, for the care and skill with

which the transliteration has been printed.

G. R. KAYE.
Baithak,
Attleborough,
Norfolk.

* StciioM a CL R, K ui L tw alaMt Oit work of Dr. Hocralo, who oloo truoitoimted oboat ball of tho Unm of tbo MX
Bafoi«M«otobi«]MikliKM|Mt|«ontboMam^oiionri«aa
 ERRATA
On the pUt£ tmcinig 4 wA *
51 K£cn*€> B* for * 51 v£r&(o B *.

In table 4 (Pai't ii) the hfteenth ligature is sva, not jvy/i.

On p. 41 in the necontl example X|, Xj| and have been a^rongly intorehanfed in the
solution and answer.
In the table cm p. 14 *
folio 55 *
should come before ' foUo 49 That it M9 ia folio 55
and M 10 it folio 49.
p. 20, § 52. Add **
on folio 40 it a phrase ^Pkkt^inr yajamii

p. foot-note if;
: for 2-ff read

p. 29, penultimate line : for read

p. 62, ^ 109. Add ^thetermt pratfUi and kharl occur on fedio 54, hut with doubtful
application/
p. 63. Omit the laat line of the capacity table,

p. 127. Pol. 29v(a). Second line. Rot\d

(/f). Second line. Ke&id dga for tye.
Pol. 29v.(/f). Second line. Read triilga for fritXga,

p. 130. Pol. 33v. Fourth line. Head /tadavgane for pagndkgtinf,

p. 131. Pol. 35r. (<&). Second line. Read gogoigd tor gognm^pP,

p. 13H. Pol. 40v. (rf). Second line. Read pratPamaik to\ prof hnhtP ,

p. 137. Pol. 42v, Third line. Read PimSututm for ritk/fmfoPf

p. 139. Pol. 14v. Second line. Read 6Afr 8 for d. 8,

p. 139. Pol. 45v. First line. Read riXvf for tmga,

p. 142. Pol. 49r. Se<^>nd line. Head dtr Hiisfimwdft fur pddtipas/^‘ft4 kJf*f »

p. 145. Pol. 54r. (A). Read jfikfraik (r) jdiaih,

p. 147. Pol. 56r. Penultimate line. Read ydca/aar/Viyc for gdv/^rfagr,

p. 147. Pol. 67r. Xjast line. Head ekord^i for ekonuriift.

 THE BAKHSHALI MANUSCRIPT
PART L-INTRODUCTION.
CHAPTER I.

1. In 1881 a mathematical work written on bli^h-bark was found at Bakhsh&U

near Mardftn on the north-west frontier of India. This manuscript was supposed
to be of ffreat age and its discovery aroused considerable interest. Part of it was
examined by Dr. Hoemle, who published a short account together with a transla-
tion of a few of the leaves in 1888. Dr. Hoemle had intended, in due course, to
publish a oompleta edition of the text, but was unable to do so. The present volume
gives the complete text.

BakhshRlT, or Bakhshalai, as it is written in the official mjtps, is a village

S.
of the Yusuf zai sub-division
of the Peshawar district of the North AVest Frontier
Province of .India. It is situated on, or near, the river Mukh&m, which eventually
joins the K&bul river near Nowshera, some twenty miles further south. Six miles
W.N.W. of BakhshSll is Jamalgarhl, twelve nmes to the west is Takht-i-Bhai,.
—
and twenty-five miles W.S.W. is Chftrsada famous for their Indo-Greek art
treasures.

Bakhsh&ll is about 150 miles from Kftbul, 160 from Srinagar, 50 from
Peshawar, 350 from Balkh and 70 from Taxila. It is in the trans-Indus country
and in ancient times was within the Persian boundaries ^in the Arachosian satrapy —
of the Achaemenid kings. It is within that part of the country to which the name
Gandhftra has been given, and was subject to those western influences which are bo>
bountifully illustrated in the so-called Gandhftra art.

3. The only authentic record of the discovery of the manuscript appears to be

contained in the following letter, dated the»5th of July 1881, from the Assistant ‘

Commissioner at Mardftn.
**
In reply to your No. 1306, dated 20tb ultimo, and ite enoloauree, I hUTe the honour to inform you that
the remains of the papyrus MS. referred to were brought to me by the Inspector of Police, Mian An-Wan*
Udin. The fluder, a tenant of the latter, said he had found the manuscript while digging in a ruined stone
enclosure on one of the mounds near Bakhshftli.* These mounds lie on tne west side of the MardSn and
Bakhshsli roads and are evidently the remains of a former village. Close to the same spot the man found
a triangular-shaped * diwa,* a soap-stone pencil, and a large lota of baked clay with a perforated bottom.
I had a further search made but nothing else was found.

**
to the finder’s statement the greater part of the manuscript had been destroyed in taking it
According
up from the place where it lay between etones. The remains when brought to me were like dry tinder, and
there may be about fifty pages left some of which would* be certainly legilne to any one who knew the charac*
tere. The letters on some of the pag^e are very clear and look like some kind of rrakrit^ but it ie most diffi-
cult to separate the pi^g^* without injuring toem. I had intended to forward the manuscript to the Lahore
Mueeum in the hope that it might be tent on thence to some scholar, but I was unabls to have a proper tin
box made for it before I left liftrdAn. I will see to thie on mv
return from leave. The pappus will reouiiw
very tender manipulation. The reault will be intereeting if it enable# ue to judge the age ot the ruins where
the manueoript wae found.**

^ Appeieatly the manuscript was found In Bfay IMl.

* Oeoeral Cunningham In a private letter io Dr. Hoernisb dated Simla, 5th June Itef, sayei ** BakhshSll Is 4 miles north of
ShIhbSeiarhL It la a mound with the village on the top of It. The blroh«bark manusoript was found in a Reid near a well without
Iraee of aay building near the epot, which It outelde the mound village •
 2

This accouat is very unsatisfactory and there are indications that it is not
altogether reliable. It was written, appuently from memory, some month or so
after the discovery of the manuscript. The " ruins,” it appears, were the (Ration
of the writer’s imagination, and the statement generally does not give the impres-
sion of exactitude.

4. In the meantime notices of the discovery had found their way into the
Indian newspapers. Professor Bhhler, who had read of the discovery in the
^
Bombay Gazette communicated the announcement to Professor Weber, who
brought it to the notice of the fifth International Congress of Orientalists then
assembled in Berlin.* In Biihler’s letter to Weber it was stated that the manu-
script had been found ” carefully enclosed in a stone chamber," and it was thought
that the newly discovered manuscript might prove to be " one of the Tripitakas
which Kanishxa ordered to be deposited in Stupas.”

There is nothing whatever in the record of the find to justify Biihler’s state-
ment, which seems to have originated in a rather strange interpretation of the
words ” while digging in a stone enclosure ” that occur in the letter quoted above,
and which are themselves of doubtful reliability. And Btthler’s views would
hardly have been worth recording here had it not been that their effect was alto-
gether disproportionate to their value. Perhaps the exaggeration of the value of
the find, however, served a useful purpose at the time but now it has become ;

—
embarrassing for refuting it makes the present editor of the manuscript appear
to be decrying his wares.

5. Tl»e manuscript was subsequently sent to the Lieutenant-Governor of the

Punjab, who, on the advice of General Cunningham, directed it to be transmitted
to Dr. Hoernle, then head of the Calcutta Mad rasa, for examination and publica-
tion. In 1882 Dr. Hoernle gave a short description of the manuscript before the
Asiatic Society of Bengal, and this description was published in the Indian Anti-
quary of 1883.* At the seventh oriental conference held at Vienna in 1886 he gave
a fuller account which was publishe<l in the proceedings of the conference,* and also,
with some additions, in the Indian Antiquary of 1888.* In 1902 Dr. Hoernle
^
presented the manuscript to the Bodleian Library.

^ This aooonnt spprsrs in the Bombay Oazkttk of Wednesday, August )3th, 1881, aad is as followa :

The remains very anrient papyrus manuscript have been found near Baskbftll, in the Mardin tahsil, Pediawar District.
of a
On the west side of the MardAn and Baskhkll road are some mounds, believed to be the remains of a former village^ though nothing is
known withany certainty regarding them, and it was while digging in a ruined stone enclosure on one of these mounds ihs disoorery
was made. A triangular- sbape<l stone diwa ', and a soap stone pencil, and a large lotah ol baked day, with a perforated bottom,
*

were found at the same place. Much of the manuscript was destroyed by the ignorant finder in taking it up from the spot where
it lay between the stones and the remains are described as being like dry tinder, in some of the pages. However, the character,
;

which somewhat resembles Prakrit^ is dear, and it is hoped it may be deciphered when it reaches Lahore^ whither ws understand it is
shortly to be sent”

* The official record is given in the report of the Congress (Fart 1, p. 79) as fdlows
Der President (Weber) verlas darsuf aus einem Briefs von Pro! Biihler folgende hochwiohtige Mitthsihing :

£in Penjibi Landmann beim Steinegraben dnen aiten Papvros gsfunden haben, der sorfaHig in einen Steinkammer
soil
eingssohlossen war. Derselbs soli sahr undang^ioh gewssen, dooh VisTss vom Finder duroh Unvordohtlgkeit tsrstort sein. Die sehr
bedeutenden Reste slnd naoh Lahore gebraoht Gam Seiten soilen Isebar sein und die Schrift wit PrAkrit auasehsn* Bskonnie
* *

dies wohl eines der Tripitaka's sein, die Kanishka in Stfipa^s niederlegeii Uess* Ich habe glaloh an Cunningham gssohrsibsn und
um ein Stuck wenigstens in Photographls gebeten.'*

* VoL lii, pp. 89-90.

* Verhandlungen dee vii Intsrnationalen Orientaliiten-Cotigrseeeii Arisohe Section, 197 uq,
» VoL xvii, pp. 33-48 and 275-279.
 3
6.
CHAPTER II.

The manuscript consists of some 70 leaves of birch-bark, but some of these

are mere scraps. The largest leaf measures about 5-75 by 3 5 inches or 14 5 by 8 9
centimetres. The leaves, which are numbered according to the Bodleian Library
arrangement from 1 to 70, may be classified according to their size and condition
as follows :

In fair condition but broken at the edges — size, not less than 5 by 3 inches
(13 by 8 cm.)

1, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 20, 22, 23, 24, 32, 33, 34, 36,
37, 43, 44, 47, 49, 59, 60, 61, 62, 63—
total 35.

Rather more damaged but otherwise in fair condition —not less than 4| by 2
inches (12 by 5 cm.)

2, 25, 26, 42, 45, 46, 48, 50, 51, 52, 55, 56, 57, 58, 64, 69— total 16.

Much damaged.
21, 31, 35, 41, 53, 66, 68—total 7.
Scraps.

27, 28, 29, 30, 35, 38, 39, 40, 54, 68, 70—total 11.

One folio (19) is entirely blank.

Certain folios consist of two leaves stuck together, namely 7, 31, and 65, and
possibly others. It would not be difficult to separate these double leaves without
damaging the manuscript.

7. The eleven folios classified as “ scraps ” at first

appeared completely refrac-
tory but the exercise of considerable patience has led the following to submit to
some arrangement. Here the letters a, ]?, c, etc., refer to the scraps in the order
in which they appear on the right sides of the leaves as illustrated and the revised
arrangement is indicated below :

Refertneee
order. to tbo pUtof.

A1 40* + 39* + 39" + 39* + 38*.

A2 39" + 40* + 38* + 40* + 39*.
A3 40* + 39* + 38*.
A6 51* + 35*.
A9 29* + 29* + 29*.
AlO 27 + 29*.

Of these rearrangements one side of each of Al, A 6 and A9 are shown in the
illustration facing p. 4.These rearrangements were made from the reproductions
— a much more dxmcult task than working with the originals.

Dr. Hoemle’s first estimate of the original size of the leaves was 7 by 81
8.
inches,and this estimate was based upon “ the well-known fact that the old birch-
bark manuscripts were always written on leaves of a squarish size ”, and upon the
 4

obvious incompleteness of folio 17 (xliii),‘ but this estimate he himself corrected

some 1 wolve years later.*

Although no is now complete there are 35 leaves in

leaf of the manuscript
comparutively fair condition, and these
vary in size’ from about 5^ by 3^ inches to
6 by 3i inches. There are some lines of writing quite complete, though not many
an<l these complete lines measure from 5^ to 6 inches in length. There is also
evidence of a small margin on either side, and the actual space occupied is just
over (> inches or 15A centimetres. I estimate that the original length was from
about 6^ to 7 inches.

With regard from 4 to

to the depth I estimate inches. In a few cases we
have the complete set from 10 to 11 to the page) and in some
of lines (generally
other cases the amount missing can be roughly estimated, while the actual maxi-
mum depth that occurs is 33 inches. My estimate of the original size of the leaves
is therefore 64 to 7 inches long by 4 to 44 inches deep or 16-6 to 17-8 centimetres- by
10 2 to 115 centimetres. Dr. Hoernle’s final estimate was 7 by 4 inches.

8 {a). In ordinary Sanskrit manuscripts it is the custom to niimber the leaves

(not the pages) in order, generally in the loft hand margin of the reverse.* No
single leaf of our manuscript has this part intact and there is no evidence of such
numbering.

9. The leaves are now mounted between sheetsof mica and placed within an
album. mica sheets are about 7 4 by 4 6 inches and are fixed together by
'I’he
strips of gummed p.aper at the edges leaving a clear area of 6^ by 33 inches. The
general arrangement is shown in Plate I (Part 11). Some other method of mount
ing (r.fr, between glass plates) might be safer. It should bo possible to separate
the leaves now stuck together, and the possibility of thereby discovering new
material would justify considerable trouble being taken in this matter.

10. Birch-bark is an outer bark of the Silver Birch {Hetula utilis, Betvia
bhoj/Ultra, or the lihurja free, as it is vai’iously called) which flourishes in the
Himalayas from Kashmir to Sikkim. It grows on all the higher ranges of the
Kashmii- hills from a height of about 6,000 feet to 12,000 foot. The forests in the
(Jurais distri<*1 stipjily most of the bhojpatra that is sold in Srinagar. The bark
is used chiefly for the roofing of houses, for wrapping up things, for lining baskets,

etc., and the villagers still use it as a writing material.

To ol)tain the bark from the tree a deep cut is made verticsilly down a clean
piece of bole, and the bark is then peeled otf by the hand. The operation is very
much ihe .same as that employed in girdling ” pine trees except that the upper
“

and lower cuts are not made. The paper bark appears to bo thrown off by an under
red bark atid apparently one layer is produced each year. If all the laminse are
stripped off from a tree, it either dies, or, if it survives, it does not give good bark
a second time. The moat suitable size of tree is from 2 feet to 4^ feet in girth. In

* Indian Antiqimn/, xvii* 1A88, p. 33. The *

wnll-known fact* is not a fact at all.

* JAvSR, UX. p. 126.

* The reproductions arc of tbs same size as actual leaves.

* Thin Is the general practice in Northern India. In the Rtiulh tha number is usually given on tlia obverse
 6

larger trees the paper bark of the bole is rough and lignified and is of no uoe as a
writing material.*

11. Each layer of bark is white or pinky-white on the outer side, but is a red-
dish or yellowish buff on the inner side. The number of layers varies and I have
counted 47 in a strip taken from an old tree. A marked feature of the bark is the
existence of numerous leiiticels (glands) from 1 to about 6 cm. in length and from
about^ 1 to 3 mm. in depth. These lenticels are reddish-brown in colour’ and of
darker shade than the natural reverse of the lamina, and each of them is continued
throughout the several laminae. On the natxiral obverse they appear much more
distinct by contrast with the lighter back-ground, and in the reproductions of
manuscripts this contrast sometimes appears t-o be emphasised. On the bole of
the tree the lenticels are horizontal {i.e., they are .always at right angles to the axis
of the bole or branch). On th- older specimens they are slightly convex on the
obverse and concave on the reverse. Traces of those lenticels can be seen in almost
all the reproductions of our manuscript (sec Fig. 1 and Plate XLI but they show ;

much clearer in some of the Bower manuscript plates). The lenticels are of im-
mrtance from the point of view of the scribe because they are of different structure
i'rom the rest of the bark, and they sometimes break away (see folio 16). There is
a sort of grain running parallel with the lenticels and the bark tears easily in that
direction.

1.

* notm by Mr. E. Raddiffc in the Foresttr (xxviii, pp. 25 *27). T ntn much imlnbted to Lt. Col. A.
Director of the Botanical Survey of India, Mr. !*• H. Chitterbuck, Inepcctor-Ccnrral of Forrj»t«, and Mr. B. Coventry, (.V)niervaior
of Forests, Kashmir, for information very kindly supplied by them.
The following particulars of specimen strips of birch-bark taken from trees of various ages, and kindly sent l-o me by Mr. B. f 'ovontry,
are perhaps worthy of record.

Siza or Strip.

Qirth of trcM^. XiimbiT of laminflp.

l^rngth.* Breadth.*

Ft. Inches. Ft. I nches.

Ft. Inchea.

6 2 6 20
2 10 10 19
1
3 4 i t5 29
2 o
H 4 7 ... Shows previous stripping.
o 2 19
3 3 7
2 10 2 6 27 Bottom of t/unk woody.
4
3 6 4 33 ,, ., ,,
5
3 7 1 10 47 Taken from an tipper branch.
6

* Here length means vertioal length measured at right angles to the lenticels and breadth means the measurement paroVtil to
the lentioela
• On the outside of the tree they are eometimes black, and on very old trete they form woody ezorescenoea.

R 2
 6

12. The art of preparing the bark for writing upon appears to be lost, but
AibIrunI tells us that the stripe were rubbed with ( ?) oil and polished. The inanu>
scripts preserve no evidence of either of these processes. All that they tell us is
something of the process of sub-division, and arrangement. Each leaf of the
fiakhshall manuscript appears to consist of half of the original thickness of the
strip, i.e., the original strip was divided into layers each of which consisted of some
six laminae. In the. Bower manuscript, the number of laminae to a leaf varies from
two to six while one leaf consists of at least twelve liminae. The Kashmirian
Arthava Veda exhibits rather more elaboration the process of sub-division is
;

carried to the extreme limits and each leaf consists of two single laminae pasted
together.

13. Possibly the original strip of birch-bark from which the leaves of the
BakhshS.ll manuscript were taken was roughly of
the shape of the annexed diagram and was cut up
A
D
1
— £
C

F
into the oblongs indicated. If A, B, C, etc.,

represent the upper layer, and A', B‘, C‘, etc., the G H I

lower layer, then, according to the evidence of the

J K L
leaves themselves, they were arranged for purposes
of writing upon in the order A, A* B, B’ C, C’
M ! N 0
; ; 1

etc., or A, A* D, D' etc. P « R

; ;
1 1

14.

Fig. t.

Birch-bark, even after preparation for writing upon, retains the natural
marks wood showing the lenticels and occasionally knots. Such knots are like
of the
thumb marks in their individuality, so that, if a knot recurs, it can generally be
identified. —
Returning to our diagram it is obvious that if a knot occurs on, say,
D its doublet will occur on D‘. In the Bower manuscript three knots that occur on
folio II, 7 recur on II, 8 the knot
; on III, 8 recurs on III, 9 several knots on II, 23
;

recur on II, 24 one on

;
II, 31 recurs on II, 32 and so on and in no case does any
; ;

particular knot occur on more than two leaves. In the Bakhshail manuscript the
order of the leaves was uncertain but it was noticed that folios 12 and 13 had a
common knot, and so had folios 32 and 36, folios 44 and 49, folios 51 and 52, folios
53 and 66. If our scheme of the arrangement of leaves also holds good here, then
each of these pairs of leaves should consist of consecutive leaves, and the final order,
based on other considerations, places them thus

Bodleian Final
order. order.

12 G3
13 G4
32 M3
36 M4
44 MIO
49 M9
51 A8, 9

52 AlO

53 K2

66 El

15. It should be noted also that where the reproductions are of double leaves
stuck together the two sides should have no common natural markings fsee, for
example, folio 65) and, conversely, when the two aides have no natural marks in
;

common there is ground for suspicion (e.g., folios 46 and 3). In this matter too
much reliance should not be placed on the reproductions.

Format.

16. Of the very few early birch-bark manuscripts the following are perhaps
the best known ; — *

() The Kharoshthi Dhammapada from Khotan. Size 8 by 36* inches,

(Ratio 0-2)*. Period 2nd century A.D.

() The Bower manuscript from Khotgarh. Size 9 by 2 to 11^ by 2^ inches,

(Ratio 4-5). Period 5th century A.D.

(c) The Bakhshftll manuscript (Oxford). Size 7 by 4 inches, (Ratio 17)

Period ?

(d) The Kashmirian Arthava Veda. Size 9 8 by 7 8 inches, (Ratio 125).

Period 15th century A.D.

(e) The Deccan Oollege Rig Veda (Poona). Size lOJ by 9i inches, (Ratio
108). Period ?

» The siwi kerf* IflTWi are, of ooum, only approiimfttionn to the tverage Rices. See (o) Joumai
pp. ff
by A. F. R. Hoibklb. (*f) The Katihmirian AHhnw Veda, BrxioirmLD k Oarbk.
(h) Thi Bowtir Cat M88. of th^
Poona, Vol. I, p. 1.
 8

The shapes and relative sizes of these manuscripts are represented in the accom-
panying diagram

Bower MS. I

Bakhshsli
Kharo8t4hi MS.

MS.

Khotan. Kashmir

Arthava

Veda

Rig Veda

Poona

17. The format of a book is generally determined by several semi-independent

factors, e.g., material, economy, fashion, convenience, contents. Our first example
of a birch-bark book, the Khotan manuscript, seems to have been shaped in imita-
tion of the western papyrus roll' the second example, the Bower manuscript, is
;

probably in imitation of the palm leaf manuscripts of India while the later ex- ;

amples are in imitation of paper books. The BakhshAll manuscript appears to

stand in an intermediate position and there is, of course, a possibility that its format
may indicate its age within certain limits.
18. In birch-bark manuscripts the writing is parallel to the lenticels, which
all
on the bole of the tree are horizontal ; and the reason for this direction of the writ-
ing is that the bark tends to split in the same direction. In the Bower manuscript
there are one or two exceptional examples of writing inclined to the lenticMS
(Plates I and II) but the angle of inclination is small. It should also be borne in
mind that the length of the strip of bark (parallel to the writing) is limited to the
circumference of the tree, and the depth of the strip is also practically limited,'
and that the format was, at least, partly determined by the necessity for dividing
these strips economically.

19. But fashion was probably a more powerful factor than material in the deter-
mination of the format of birch-bark books, and it is possible that the method of
preparing papyrus for writing upon to some extent influenced those who prepared
birch-banc for the same purpose. The stem of the papyrus plant was cut into

‘ NymbM of oaob rolli wore diionTered hy Sir M. A. Stein In Control Aeio.

See fooiooto on poge
longitudinal atripa which were laid on a board aide by aide to the required width,
and across this layer another of shorter strips was laid at right angles. After
soaking, the combined sheet was hammered and dried and pohshed with ivory or
a smooth shell. Some twenty such sheets were pasted together to form a roll.
Later it was the practice to back one sheet with a second in order to obtain a more
suitable paper. The depth of the rolls varied from 4 to 12 inches. The later limit
20.
for the use of papyrus as a writing material was about the 9th century ' A. D.

The palm leaf manuscripts of India are made from the leaves of Corypha
vmhraculi/eTa or Borassus flabelHfer. The former is indigenous in south India
but the latter was probably introduced from Africa. The leaves of both of these
trees are long and tapering, with central ribs. All the earlier palm leaf manu-
scripts are made from the leaves of Corypha. The following are examples :

Reference. Locality. Leng^th. Depth. Ratio. Date.

Horiusi .... W. India . 11 in. 2 in. h-.n ? 5£0 A.D.

B«iidall Add. 1049 n tt 2 „ 80 fha „

„ „ 1083 Nepal . 21 2 „ 10-5 1089 „

j

„ „ um Bihar . 22 „ 2 ..
110 1054

Kielhoin No. . W. India 18 „ 21 „ 5*8 1183 „

Petereon No. ii*20 . i

„ 21 i;>’0 1405 „
i1
1 1

From measuiements of a large number of (.Corypha manuscripts it appears that

the most usual depth was about two inches and that the lengths varied from about
10 to 34 inches. The chief mode was from 10 -16 inches but there was a second
mode from 30 34 inches. It appears that the whole length of the prepared leaf
was occasionally used, but that more often it was cut into thirds, or sometimes into
halves.

About the middle of the fifteenth century the use of palm leaf as a writing
material entirely ceased in western India, but in eastern India it continued to be
used side by side with paper until much later.

21. paper making appears to have been practised by the Chinese at

The art of
a very early date. From the Chinese the Muslims learnt the process in the eighth
century and they introduced it into Europe and also India about the twelfth
century. Among the earliest Muslim paper manuscripts now preserved, one
was written in A.D. 866, others in 974, 980, 990, etc. The earliest Indian paper
manuscripts known were written in A.D. 1231 and 1343. The earlier one measures
6 by 4 inches (ratio 15) and the other 13^ by 5 inches (ratio 2 7).

22. Another very common writing material early India was copper. in
Btthler states that the size of the copper plate used was largely determined by the
common writing material of the locality, e.gr., the palm leaf, birch-bark, etc. This
statement, however, is not fully supported by actual, evidence, for the more usual
ratios of lengths to depths of copper plates vary from* about 1’4 to 2 6 as against 5
to 15 for palm leaf manuscripts.*

There are rere ezemplie of copper pUtec of rolio over 4 Iwo TasHa plelee of redo 4*7).
 10’

23. Before making definite deductions from the formats of our birch-bark
nianuecripts, a great deal more investigation is obviously required, and in parti-
cular the formats of the western tablets and codices, the early Chinese and Muslim
paper books should be studied. Uutil the introduction of paper into India there
does not seem to have been much consistency in the format of birch-bark manu-
scripts. The Khotan manuscript may have been in imitation of the papyrus rolls
of the west. The width is about 8 inches (20 cm.), the length is unknown, but the
total length is about 4 feet (1 m. 23) and the longest piece of the Paris portion is
about 24 inches (61 cm.). The scapus was made by joining pieces of unknown
depth together* and at each side a fine cord was threaded through the bark about
one cm. from each edge. Most probably it was never intended that the manuscript
should be rolled up possibly it was to be hung on a wall. The writing is parallel
:

to the short side, and in this matter differs entirely from the pap3rri, on which the
writing is generally parallel to the len^h of the roU, and in columns of from two to
three and a half inches wide. The writing is on one side only of most of the frag-
ments of the Khotan manuscript, but on one long piece (C) both sides are written
upon. This piece (C) consists of several layers of bark (it is, of course, hnpossible
to count these layers without examining the original). The lenticels are identical
on the two sides and the verso side is the lighter side of the bark. The manuscript
appears to have been folded up to the size of 20 by 5 cm., but obviously this was not
the original intention, for the folds are independent of the position of the lines of
writing, and folding birch-bark manuscripts was never a wise procedure.

That the Bower manuscript format was determined by the palm leaf pothi is
probably true, although should be noted that 9x2 inches is a very uncommon
it
size for an early palm leaf manuscript. (See Dr. Hoernle’s lists in the JASB, 1900,
pp. 99ff.)

24. format of the Bakhshall manuscript (ratio 16) as a cri-

Br:^^a.rding the
terion oL age, I can come
to no positive conclusion from the evidence before me.
Further investigation is required and might be profitably undertaken if time
allowed. Dr. Hoernle, however, writes as follows :

“ It is noteworthy that the two oldest (Indian paper) manuscripts known to

25. us point to their having been made in imitation of such a birch-bark
prototype as the Bakhshall manuscript.”

It is not easy to accept this argument, for it would be quite as reasonable to con-
clude that the Bakhshall format was determined by the paper manuscript formats,
and that it is of later date than the introduction into India of paper as a writing
material and this would place the Bakhshall manuscript aoout the twelfth
;

century of our era at the earliest. However I only give this hypothesis as a set-off
to Dr. Hoernle’s unjustifiable deduction.

The script.

The Bakhshall text is written in the Sarada script, which flourished on the
north-west borders of India from about the ninth century until within recent times.
Its distribution in space is fairly definitely limited to a comparatively small area
lying between longitudes 72 and 78 east of Greenwich and latitudes 32 and 36 nqrth.
Dr. Vogel distinguishes between Sarada proper, of which the latest examples are
of the early thirteenth century, and modern ftarada.

Rtit I rjm no eridenoe of thit In th« illusiraUoni given in the Jaumal A9i4Xt%qu€ (1808* p. 308
 11

The writing of the Bakhshall manuscript is of the earlier period and is gene-
rally yety good writing indeed. was wntten by at least two scribes. In the
It
table on p. 97 the styles of writing are indicated by the letters a and of
y8 pertains whoUy to the
which “ ” section. MStyle a is divided into four sub-
sections which possibly belong to the work of four separate scribes, although it is
not easy to point out any fundamental differences between these styles. Folio 6.5
ossibly exhibits the writings of two separate scribes on the two sides, which do not
E elong to the same original leaf. Compare also folios 29 and 17.

Possible points of differentiation between the sub-divisions a„ az, a, and a, are

the length of medial the length of the final up-stroke of initial u, the length of
the vir&ma mark, the occurrence of the sickle shaped i ^folios 1, 35, 52, 60), of the
" clubbed ” ai and e (folios 16 and 60), of the “ curvM ” medial e (folios 3, 18, 5), and
the different methods of forming medial e, ai and o.

The style (jS) of section M " is distinguished by its boldness (contrast plates
“

II and XXV), by “ tails ” or flourishes (including very long virUma marks), the
methods of writing medial ai and o, the looped six, etc., etc.

In Part II the script is examined in detail and further peculiarities of the

different sections are given, e.g., section “ M
” has no examples of and contains M
practically all the examples of d‘ Sections L (a,) and (a^) show marked G
mfferencea in the methods of writing medial e, while section F is, in this matter,
very much like section L. Section C (ot,) has no example of either the
jihvdmiUiya or upadhmaniya, and so on ; but it must be borne in mind that these
statistics are only of value in the mass.

The language-

26. The language of the text may be described as an irregular Sanskrit.

Nearly all the words used are Sanskrit, and the rules of Sanskrit grammar and
prosody are followed with some laxity. The peculiarities of spelling, sandhi,
Igramnutr, etc., that occur in the text are exceedingly common in the inscriptions
I
of the eleventh and twelfth centuries found in the north-west of India. Dr.
Hoemle, however, implies that the language is much older. He states that the
text “ is written in the so-called Gatha dialect, or in that literary form of the North
Western Pr&krit which preceded the employment, in secular composition, of the
classical Sanskrit.” He also states that this dialect " appears to have been in
general use, in North Western India, for literary purposes, till' about the end of the
3rd century A.D.'"

The subject of language is discussed to some extent in Chapter viii and will
be more fully dealt with in Part III of this work.

^ Ind. Ant. «vii, 188K, pp. 37 Sk 38.

 12

CHAPTER III.

Oroeb.

27. of the folios, which was definitely fixed by

The Bodleian Library order
circumstances not within the control of the editor, is necessarily followed in this
volume in the arrangement of the facsimiles and the first ti’ansliteration of the
text. But a more satisfactory arrangement was considered desirable, and a very
considerable amount of labour has been involved in the attempt to achieve an
approximately correct order.

The leaves were disarranged to some extent before they reached Dr. Hoernle,
but unfortunately he did not leave any proper record of the order in which the
leaves reached him.'’^
. if I

A
detniled analysis of all the leaves showed that Dr. Hoernle’s rearrange-
28.
ment was faulty,* that the “ find order ", as far as known, was better, but not alto-
gether reliable Further rearrangement was necessary and had to be attempted ;

but leaves were missing, .some were in fragments placed anyhow, not a single leaf
was complete, and the connecting portions of the text were all wanting.

29. Dr. Hoernle attempted to arrange the leaves on the basis of the numbered
sutras; but the numberetl sufras were too few and too unevenly distributed to
serve this pui-pose, and in some ways they were even misleading. The chief
criterion of order is, of course, the nature of the contents of the leaves, but I have
seized upon any means available that has offered any help towards the solution of
the problem of order. Had all the leaves been extant, oven in fragments, the
problem could have been completely solved but some loaves are completely miss- ;

ing and many fragments have disappeared altogether, so that the problem is only
partially soluble. I have already explained how the knots in the birch-bark were
of assistance, and besides this natural aid there was the accidental one of the effects
of the method of storage. Possibly for some hundreds of years the bundle of leaves
was subject to a certain amount of pressure, and was exposed, particularly at the
edges, to chemical and other disintegrating actions. Some oi the leaves stuck
together, the edges of all became frayed and certain leaves became so frail as to
break up into scraps when handled. On the principle that contiguous leaves would
be affected approximately to the same extent, we might, if no disintegrating effects
had taken place since the find, rebuild in layers the original bundle. But we know
that fuT-ther disintegration has taken place, and is still not altogether eliminated ;

nevertheless, similarity in size and shape and mechanical makings were distinctly
helpful in rearranging the leaves.

30. The script was helpful in a general way indeed it led me to differentiate :

the " M
” section from the rest of the text. The language employed was also help-
ful, particularly with reference to the use of special technical terms while such ;

examples as the following were noted after some order had been restored ghna :

*'
multiplied bv ” occurs only on folios 7, 45, 46, 57 and 65, all of which belong to
the same section (C) pratyaya, yuta and rUpa occur very often but never in the
;

" M " section, etc., etc.

31. The chief criterion —

that of the subject matter could not always bo —
utilised, owing to the absence of all the actual connecting portions and to the

* In ih® following not®® by And order I m®nn th® order in wbioh the lenre® renohed Dr. Hoernle.
* ’
Thie order ie not oom-
pletely known bat Dr. Hoernle merked th® order of certain of th® leeree.
It would perhepe b® more oorreot to eej that it wee inoomplete. He pleoed leoTee 10 to IS end 00 to 63 in proper leqnenee.
 13

numerous lax^ei* gaps, but occasionally it was definitely decisive e.g,^ one example :

goes from folio 65 to folio 66 and then to 64, another from 66 to 67, and another
from 29 to 27, etc. We also find the same topic spread over a number of leaves,
and here it was often possible to reconstruct a logical sequence that must be in
some correspondence with the original.

Finally the numbered sutras were useful as a check and also as indicating to
some extent the relative positions of certain sections.

32. The problem of rearrangement, however, had not only to do with the order
of the leaves but with the order of the pages and also the order of the fragments.
In a number of cases the Bodleian arrangement places the proper first side of a
leaf tVi VCT80, and the fragfinents ai'e very often placed with wrong leaves. The
illustration facing page 4 shows some of these fragments rearranged, and the table
on page 14 enumerates some 13 leaves placed with the obverse in verso.

33. The order now given does not pretendto be t he exact original order it is :

merely a compromise. It brings together those portions of the remains of the

te.xt that deal with the same topics. It is not necessarily a logical order according
to modern
34. views indeed, few medieval mathematical w^orks exhibit such a logical
:

order. Neither is the order a final one, and probably it is not the best that could
have been obtained but a working order had to be achieved even before the de-
;

tailed examination of the manuscript could be completed, and Jiow if the labour
were npt too great I should be tempted to revise the order once more. Further, if
an attempt is made, as I trust it will be, to separate those leaves that are stuck
together, then the question of order will possibly once more have to be considered.

The following tables show the various orders-

BODLEIAN ORDER.
(The laciiimileH Hbown in plait^n ii to xlvii aro in thin onlcr.)
k

Folio. Revieed order. Folio. Revised order. Folio. 1j

Revisotl order. Folio. Revised order.
1

j L . ...

1 A 11 21 E4 41 J 3 (U L 2
% A U 22 E 5-F 1 42 M 13 62 L 3
S A 13 23 F2 4.3 M 12 63 Ii 4
4 B 4 24 F3 44 i M 11 64 C6
5 c 1 25 F4 45 C 8 65 c 4+J 2
m
t

6 C 2 26 P 5+? 46 C 9 I) 1 E 1
U
!

7 B 3 4- C 8 27 A 10 47 1 67 n 6
8 B 1 28 D 7 48 M 8 68 D 4
9 B2 29 A 6-7 49 M 9 69 I) 8
10 G1 30 J 1 50 M 14 70 1 D 2
11 02 31 1) 6 51 A 6 7
12 OS 32 M 3 52 A 8
IS 04 33 M2 53 1 E 2
14 06 34 M6 54 A 4
16 06 35 A 6-6 55 M 10
M4
1

10 G7 + H 7 36 56 C 5
17 H2 37 M6 57 C 7
18 HS 38 Al-3 58 E 3
19 Blank 39 Al-3 59 K
20 M1 40 Al-3 60 L 1

c 2
 14

REVISED ORDER. &c.

•-••8

'•••*•'

SSSS5SS58“S#lf!S*
 15

CHAPTER IV.

The contents of the manusobipt.

35. The portions of the manuscript that have been preserved are wholly con-
cerned with mathematics. Dr. Hoernle described the work in 1888 in the follow-
ing words* :

“ The beffinning end end of the manuscript heinv lost, both the name of the work and its author are
unknown. The subject of the work^ howeyer, is arithmetic. It containa a great variety of problem» re-
lating to daily life. ‘The following are ezamplei: * —
In a carriage, instead of 10 horses, there are yoked
' accom-
6, the distance traversed by the former was one hundred, how much will the other horses he able t
plish?* The following is more complicated: *
A certain person travels 6 yojanai on the first day, and 3
more on each succeeding day; another who travels 7 yojanai on each day, has a start of 5 days; in what time
will they meet? * The following is still more complicated Of 3 merchants, the first possesses 7 horses,
the second 9 ponies, the third 10 camels; each of them gives away 3 animals to be equally distributed
amongst themselves. The result is that the value of their respective pri)i)orties becomes eiiual hov mucli ;

was the value of each merchant’s < riginal property, and what was the value of each animal r The method *

prescribed in the rules for the solution of these problems is extremely mechanical and reduces the labour
of thinking to a minimum.**

36. It is necessary to emphasise the fact that this early estimateof the value
of the work is inadequate and misleading. for this is, that, when the
One reason
estimate was made, only a comparatively small portion of the contents of the
manuscript had been understood, and that was by no means the most interesting
portion of the work.

37. The following a summary

is list of the contents of the work as far as its
present state allows of such analysis ;

Section.

'problems involving systeme of linear equations ..... A

A & K
Indeterminate equations of the second degree
Arithmetical progressions B & (’

Quadratic equations
Approximate evaluations of square-roots . . . ^ . . . C
Complex series . . F
Problems of the type x (l-ej
The computation of the fineness of gold H
Problems on income and expenditure, and profit and loss . . L, D & £
Miscellaneous problems M
Mensuration.

Such is a very rough outline of the work as it now stands. Perhaps the most inter-
esting sections are C, andA M
and of these C is the most complete and was evi-
;

dently treated as of considerable importance.’ Section A is also of special interest

as it contains examples which may be described as of the epanthem type.
Section M
is of interest principally on account of the methods of expressing the
numerous measures involved and also because of its literary and social references.

38. Although the work is arithmetical in form it would be misleading to de-

scribe it as a simple arithmetical text-book. No algebraical symbolism is employed,
but the solutions are often given in such a general form as to imply the complete
general solution, t.e., the solutions, though arithmetical in form, are really gene-
ralised arithmetic, or algebra. First of all a particular rule is given, which is

> tni. Amu XVIT, ISM, p. 33.

illllll
ASIATIC SOCIETY
 16

intended to apply to the particular set of examples that follows. These rules are
often expressed in language that would be impossible to intei-pret without the
light thrown upon them by the solutions. The examples are themselves sometimes
trivial, but the solutions, often expressed with what at first glance appears to be
meticulous care, often redeem the examples from their apparent triviality. Proofs
or verifications are often given with some elaboration and on occasions are multi-
plied.

39. The work may be divided roughly into algebraic, arithmetical and geo- .

metrical sections but the Vjouinlaries of these sections are not clear, and perhaps
;

it would be more correct to clas-sify the problems as (a) academic, (6) commercial,
(c) miscellaneous.

Judging by the manuscript as it now stands, the problems involving geometri-

cal notions were comparatively very few, and we can only guess at the meanings of
the remaining fragments dealing with this branch of mathematics.

One pleasing feature is the small space occupied by commercial problems.

There is only one problem on interest, and a rather unobtrusive section containing
problems on profit and loss.

Those problems classed as academic are concerned with particular mathe-

matical notions that in early media'val times had a traditional value and interest,
such as the epanthema, the rtujula virginum, certain indeterminate equations of
the second degree, and certain sets of linear equations.

The miscellaneous problems include examples where the chief interest is rather
mathematical notions involved e.g., there
in the illustrative material than in the ;

are problems concerned with the abduction of Sita by Hav.ana, the prowess of
Haihaya, the constitution of an army, the Sun's chariot, the daily journey of the
planet Saturn, gifts to Siva, etc., etc.

40.Such, or similar feature.s. are. however, common to many medieeval mathe-

matical works, and wo now turn to the consideration of the distinguishing features.
Characteristics that completely differentiate one such mathematical work from
another of the same period are generally hard to find and generally in order to so
differentiate we have to sum up a number of minor characteristics. The Bakhshall
manuscript is, however, almost unique in at least two I’espeets of some mathemati-
cal importance. The first of these is the employment of a special sign in the form
of a cross •exactly like our plus sign but placed after the quantity it affects to —
indicate a negative quantity. The possible connexion of this with the Diophantine
sign and its value as a chronological test will be dealt with in due course. The
second special characteristic ponsists of f he set of methods for indicating the change-
ratios of certain measures. This characteristic also will be dealt with in some
detail in due course and it must suffice for the present to point out that neither
;

of these peculiarities is very helpful in placing the work.

41. If the work were actually of the period to which Dr. Hoernle assigned it,
then by far the most remarkable feature would be the employment of the modem
place-value arithmetical notation but Dr. Hoernle’s estimate of the age of the
;

work was wrong and the occurrence of this notation is a common-place matter. It
draws our attention, however, to the skill of our author in the manipulation of
numbers. Large quantities are dealt with and one particular number contains
23 digfits. These large numbers do not appear to be given for mere effect for they
•occur quite naturally as the result of a rigorous logic. Indeed the author seems
generally to prefer simple numbers.
 These large quantities again lead us to note other rather special charactei is-
tics of the work, namely the apparently over-elaborated exposition of the work-
ings " of the solutions (to which further reference wall be made), to the pjcserva-
tion of the generality of the solutions throughout such workings, and to
the consequent necessity of preserving the q\iantities that occur so that the jieuul-
timate statement shall involve tlie whole formula. That is, all hough every opera-
tion is arithmetical in form, the quantities involved are not simplitied " or " can- “

celled ” without some special justillcalion until the final result is achieved. The
(unwi'itten) laile followed by the author was something like this iin* inlegrity of ;

the method of solution must be preserved, and so long as it i.s p re.se r \ ed the c.dcu-
—
lation may be simplified -otherwise not. Indeed the numerical quantities in these
probleirrs are treated almost like algebraic symbols.'

42. There is no actual algebraic notation employed but the unknown (juantity
'

appears to he indicated in certain examples by the usual syml>t>l for nought. '

which symbol, however, is never the subject of operatiou. Where it oe<-uis the
rcgiila falsi is employed — probably because of the lack of an eflieient symbolism.
Neither is there any symbol of operation. The negative sign alreadv alluded ( -

to, is never used as such. Operation is generally indicated by some (lefmite ail hor
term but sometimes by relative position. Fractions, for example, iire indicated
in the modern way but without any horizontal bar between the numbers, and some-
times division proper is so indicated.

43. If vre considered the Bakhsh.nli text to be a work of pure Indian origin,
then it would be also unique in anotber respect, namely in the raBier extensive
employment of the square root rule that may be expressed by v A'‘-+-’b + b/2A. A
For this rule, the early history of which is well known, was nevei- used in the early
Indian works to any extent, whereas the Bakhshalt text employs it for a com-
paratively large number of examples and applies the rule to second approximations
in a very thorough manner.

On the .same assumption as to an Indian origin the use of tin* sexag<‘.siiual

notation in connexion with an approximation to a square-root of a non-ast ronomical
quantity would also be uniqua, but it may be noteil that there is only oiu' nucIi
example preserved and this is given in a rather timid way. and the notion imolvcd
is not pursued. *

44. Whether of a purely Indian origin or not, the work is Indian in form. It
is written in a sort of San.skrit and generally conforms to the Imlian texf-beok
fashion, but there are certain apparent omissions. Ferhaps the most note wort bv
feature of the classical Hindu texts is their treatment of indeterridnate equations
of the first degree, while their greatest achievement is the full solution of the so-
called Pellian equation. A great part of the texts of Brahmagupta, Maha'vira and
Bh-askara are <levoted to one or both of these topics, but there is no evidmiee of
either in what remains of the Bakhshall text; and this apparent omission is the
more noticeable, inasmuch as there is evidence of considerable skill in the treatment
of systems of linear equations an<l certain indeterrninates ol the .second devve
Another omission to note is of a different character altogether. Every earlv Hindu
work of this kind has a section relating to the “shadow of a gnomoii,” Imt in our
text there is no evidence of such a section. We
must not. liowever, pa\' too mueh
attention to these apparent omissions. The possibility of entire sections of the
manuscirpt being destroyed is not great, but negative evidence and a mutilated
manuscript do not carry us very far.

* TboM rem*rkii do not apply to th#t ‘

M * nertion.

VHlatic S<»cicty, Caloitl*

tae-Ho-LiWt IX. i
 18

N on-mathematicaL elements.

A5. Many of the uiathematical examples contain references to the affairs of

godsund men and some of these references are of rather exceptional interest. The
names of certain doit ie.s and semi-historical beings occur, and there are references
to legends which may give some clue as to the origin of the work. Of interest also
are the animals, crops, metals, etc., mentioned. Before indulging in conjectures
as to tlie .significance of these references it will be convenient to exhibit them in
some sort of order.

46. Siva. The name of Siva occurs on a fragment (folio .50) that appears to
exhibit a sort of colophon, in which the gift of calculation to the human race is
attributed to the god. The actual name Siva occurs nowhere else, but another
—
name of the god— Sulin is given on folio 34, where the example refers to certain
offerings made to him. On folio 44, in a similar example, part of the expenditure
is in offerings for 611 and this, it is conjectured, stands for Sulin. The term Devi
'-“,

occurs (folio 49) and probably indic.ates Siva's consort.

The significance of these references (which all occur in section “ ”) is em- M

phasised by Dr. Hoernle’s remarkable claim that the work is Buddhist or Jain,'
This appears to have been mere conjecture, for we find nothing whatever in the text
that conflicts with the .‘^iaivisrn indicated. Neither is the authorship of the work
by a follower of Siva in any way surprising, for, it may be noted, Kashmir Saivism
began to re-flourish al)Out the tenth century A.D.,’ and Kalha^a was a Saivite.

Vasudbva is a name applied to Krishpa. The phrase that occurs (folio 44) is
vdsudevasya ch&rchanet.

Suras and Asuras. Suras are generally classed as minor deities. In our text
( they are said to dwell on Sumeru. In the same context Asuras, demons
folio 33)
or enemies of the gods, are contrasted with the Suras.

Raksuakas areclassed as evil ^irits but are not generallj very clearly defined.
The term occurs on a double leaf, and perhaps not correctly
folio 65, recto (part of
placed in section J). The problem, which is concerned with jiva-lom, is not under-
stood.

SiDDHAs AND VidtIdha^s. These two terms occur together on folio 37,
where the chariot of the Sun is said to be guided by the god Mahoraga among the
Siddhas and Vidyadharas. The Siddhas are semi-divine beings of great holiness,
who dwell in the region of the sky between the earth and the sun, while the Vidyft-
dharas are inferior deities inhabiting.the same region.

Mahoraga is Sesha or any other great serpent. The connexion between ser-
pents and the Sun is supposed to be somewhat intimate in certain mythologies but
1 have not found the source of the present reference.

47. SiTA. The name Sitft is not actually proserved in our text, but on folio 32
is an example based upon her abduction by Kavana. The tale is that SitS, the
wife of R&ma, was carried off through the air by Ravaqia, and that, in order to
attract the attention of her helpers, she tied up some jewellery in a garment and

> tnd. Ant. XVII, 1888. p. 38.

Am R. O. BHAiTDABKAm, VaUuMTfani, 8»lrUm, «te., pp. 188-181.
 19

dropped it to earth, and this circumstance forms the substance of the problem given
in our text.

RIvana is mentioned on folio 32. See Sita.

Partha, irritated in a fight, shot a quiver of arrows to slay Karna. With

exact relationship between the two terms is not certain, but tlie j)robleni in which
they occur i-elates to the destruction of an army, and tlie question is how many
arrows were used ? A very similar probletu is given twice over by Bhaskara (LU.
67 Vij. Garjt. 133) as follows
;
:

P&rtha, irritated in a fight, shot a quiver of arrows to slay Karija. With

half of his arrows he parried those of his antagonist witli f<mr tiinos .

the square-root of the quiver full he killed his liorses wdth six arrows ;

he slew .Salya, with three he demolished the umbrella, and with one
he cut off the head of the foe. How many w'ere the arrows which
Arjuna let fly ?

In Bhaskara’s example Partha Arjuna. the Paodava prince, but there is no con-
nexion between him and the Haihayas. The Haihaya Arjuna was a gr<‘at king but
an altogether different individual. There is. however, little doubt that the examples
in the two texts were connected, either directly or by some common source.'

is mentioned on folio 34 in an isolated phrase

Arjuna arjuneiut ijriddhra
Arjuna was connected with the Nagasand some contest with a form of (biruda
would not be out of place.

48, YuDHisuTiiiRA. On 37 is the isolated fragment raja yndhisthira

folio
nama varhia which implies some familiarity with the great epic of
India.

Satrudama. a similar fragment occurs on folio 47 kaschid rdja-kumdra

Aatrudama could be identified the reference to him might prove
If this prince
a valuable clue. Compare 3atrugh7ia which has the same meaning.

SuMD.\Ri. On folio 34 Sundarl, the beautiful one,” is asked to solve the prob-
lem, that is, she is addressed in exactly the same way as Lllftvatl is in Bhflskara’s
well-known work. Here is either imitation or a common tradition.

Chhajaka. The text (».«.. section “ M ”)

written by a Brahman, the son of
is
Ghhajaka (folio 50). Possibly this name the same as Sajjaka. which occurs
is
several times in the Rdjatarahgini. Sajjaka w’as superintendent of the >*^eda office
in Kalhana's time (XTTth century), but there is no real justification for connecting
this individual with the author of our text.

49.Besides these names of individuals there are references to various classes

of men —kings, princes, priests, learned men. Viankers, servants, soldiers, tax-collec-
tors. Merchants not only appear in money transactions but, in one case, with
Brahmans and others, as recipients of propitiatory gifts. P.ar.dits and learned men
are spoken of as earning wages, and two “ Rajputs ” are tle.scribed as servants of
a king. The problems are often addressed to an individual “
or individuals variously
described as “ the best of calculatcrs ” (annakotfarna). wdse ” or ” learned ” man
{hudha, prdjTm, pandita. dharmajitaya., etc.), or " triend.”

^ The mathrmattcal prototype i« common In the (Irerk AnDiology.

 20

.vO. I ho oollootiou of
allilllal^> ih liiiiite<i tu elophanls, horses, camels, cows,.
hiiM;ili>os,snakes, a worm anti a vulture. There is little of signilicance hero, but
tile occurrenee of eaineis points to tlie north-west of India. There are, however,
two clas.ses of hor.ses mentioned {astut, /rt/.//o) for e.\um pie, one problem deals with ;

certain numbers of (isra, anil it.s/ifrit leamel). Dr. lloernle translated

by " yak, which would be int eresting' if it were justifiable but on one occasion the ;

1\\t> terms setmi to be useil synonymously tlol. iS' and the " horse of uii army are ’

designated fiitj/ii and (unx/x (fol. 47).

Of food stuffs and other commodities we have wheat, barley, rice and saffron
(this eomliination lieing of some .sigiiificanee )
;
golil, iron, salt, molasses, and ( ?)
lapis-lazuli (ambkaioha).

The mention of a boat sailing against adverse winds is not necessarily

of.
' local
eolouiing ’ but we must assume that it was, at least, comprehensible.
Chariots {rntJut) are mentioned three times twice it is the cliariot of the Sun, and
on the thiril occasion thii chariots lortn one si-etion of an arinv, which is .said to
consist of chariots {rdtha), elephants foot soldiers (nttra), and "horse”
ilntifd) in the ratios I 1 ;j. Certain <li\isiuns of an army are also men-
: : ."i ;

tioned, namely rluimu, iinia/ifi, aiiilcini mid a/f{</nni/iiT,ii (I'ol. 47).

/)2. The leferenees to religious matters are lew hut are signifieant. We have
already mentioned olterings to Siva. Devi and \ asudeva llrahmaus also appear to ;

have been fed othiu' gifts are “ for the sake of reverence " (/)njtlrtha), and for hopes
;

for " the future world /xt rdtoha). etc. The " supreme spirit " {pararndtnuina)
(

and {) " ercalion (.stis/ifi) are also referred to.

(
’

/.s t/w loor/c homidjeix'Ods ?

o3. In my first e.xamination of the manuscript I noticed that the writing was
not uniform, and that, in particular, certain leaves differentiated themselves from
the rest by a bolder and, on the whole, a better style of writing and I distinguished ;

this set of leaves as the " section. M

This early ilifferenli.ntion was a most useful
one for it marked not only a diffeienee in style of wiiting hut also one of matter.
Imleed this “ M
section proved to have .so many peculiarities that the idea that
it w'as possilily a separate woik could not be ignored. But the rest of the manu-
serij)t is by no means uniform in style of w riting oi' anything else, and I am not
con\ inecd that the ” ” section is the M
work of a separate author, although I rather
buspeei that, it is. It is. however, pretty certain that it was the work of a separate
scribe l>u1, as there are slight indications that the other portions of the manu-
,

seripi were dictated, this does not affect the question of authorship conclusively.
I cannot point to any definite evidence of dictation tliat would bear exainination :

it. i^ rather an impression received.

'The peculiarities of the “M

section, althougli they may not prove hetero- ’

geneity of workmanship, call for some special mention and are here summarised.

1. Tlie script.

(a The writing is bolder and on the whole, more uniform than that of the
•
>ili.i.isnnmu oqi jo jsai

* '*
SftfTtivn liiis to thf prrsrni day niiiirtinvl n fnmoiis pn*fhirt of Kashmir.** Stfik, Tl. 428. Wf
ought also lo m»‘Titi(»ri l>ire)i-harW, although thnro is on rofrrfnr.1: !i> it in text-
 21

(6) Flouribiies or extensions of the bottom end-atrokes are common. These

flourishes occur particularly at the end of ligatures, but also in the
cases of the numeral flgures 5,” " 7 and “ 9,” and even in the case
1, the stop bars, and occasionally they even occur at the ends of the
of
frame-works of the “ cells ”(e.y., see fol. 47, etc.).

(c) The numerical symbols of section " M ” are shown in Table IV (7) line
where the looped “ 6 ’’
should be noted. 'I'his is a useful but not an
infallible criterion.

(d) The following table relating to the formation of the medial vowels a, i and
o is taken from Fart II.

Metliiil ai itlvdial 0*

hi hi O <V 'o'

‘
M '
section 0% 100% 32% 36%
Whole inunuscript 24% 76% 7 O/o 15% 10%

Here the indication appears to be very detinite indeed, but it must

be borne in mind that sut'h criteria only apply in the mass, that the
total number of ni examples is only 19, and so on.’ For a fuller dis-
cussion of tliese interesting statistics see Part M.

II. curious that all the mythological and semi-historical references (see
It is
—
Sections 46 48^ occur in the “ ” section. M
Indeed this section is peculiarly
Hindu in contrast with the remainder of the manuscript.

III. The mathematical contents of the “ M

” .section may be de.scnbcd as mis-

cellaneous juoblems, which are generally solved by simple rule of three but “ ’’
;

numerous “ measures ” and a method

a special feature is the occurrence of special
of exhibiting their change-ratios. But, of course, these points in no way indicate
a .separate work rather otherwise for if section
” were an entirely different “
M
work we might expect some duplication, and there is none here.

TV. The method of exposition is som«what different. The example is followed

by a statement and the answer is then given, generally, without any detailed work-
ing, and generally there is no proof ” or verification. There are, however, excep-
“

tions.

V. There are differences in language. Certain technical terms that are ex-
tremely common in the rest of the manuscript do not occur, e.g., pratyaya, yuta.

» The •ymboli aw? merely mnemonic. 8rf page »5.

n 2
 22

CHAPTER V.

Exposition and Method.

54. The text consist s of rules (autras) and examples. There is no explanation
whatevei’ of the processes by which the rules were obtained, that is, there is no
mathematical theory at all. In this the work follows the usual Indian fashion, as
exhibiti'd in all early texts. But there is a good deal of mathematical theory im-
plied, and the I'ules and examples are often .set forth in such a way as to convey the
principles followed quite clearly to the student.

55. The rules {s films) are written in verse and are generally numbered and ;

often in the solutions of the examples phrases from the sfitras are quoted. When
Dr. Hoernle tried to re-arrange the leaves of the manuscript, he took the numbered
siitras as the l)H8is of his order, but they were too few in number to lead to a satis-,
factory result. These sutras do not represent, as might be expected, the most
valuable part of the text. They are usually of particular application rather than
general and are often very ob.scurely expressed.

56.The examples given are generally formally stated in full without the use—
any kind and in most cases they are stated in verse.
of notation or abbreviation of ;

They are introduced by the term udd, an abbreviation for uddharaiiam “ an ex-
ample.” After the (question .sometimes comes a formal statement with numerical
symbols and abbreviations, often arranged in cells. Then comes the solution or
working {karniia), and here, sometimes, fragments of the sutras are quoted.
—
Finally come demonstrations often more of the nature of verification than proof.
Generally these demonstrations, by the .aid of the answer found to the question,
rediscover one of the original elements of the problem and sometimes several such
;

demonstrations are attached to an individual problem, but sometimes the variation

is merely a matter of the form of statement.

The
57. full scheme of exposition is therefore

Sutram or rule.
Uddharanam or “ example ” : indicated by udd°.
Sthdpanam or “ statement.”
Karauam or “ solution.”
Pratyayam or “ verification.”

The end of each sutra is marked after the last example by the device —
and the number of the sutra is also given at the end. |

The method of grouping sets of figures is of interest, and shows features

in common with mediaeval Sanskrit mathematical manuscripts, where also it is
the practice to place groups of numbers in cells. In our manuscript, however, this
fashion is rather more elaborated than in any Sanskrit manuscript I have examin-
ed. The mathematical possibilities of this scheme do not appear to have been
reali.sed and the student must always be careful to interpret any ^roup of figures
from the context, .nnd not from any similarity with other groupings. However
there is a cert.ain amount of consistency in the arrangements, as the examples ex-
hibited below will show. The real pui’pose of the arrangements appears to be to
prevent confusion by demarcating the numerical figures from the text itself. The
text is often written almost independently of the figure groups, and a word may be
arbitrarily divided 1)V’ the cell arrangement, which may also cut into several lines
of the text not necessarily connected with it. The economic necessity of utilising
 23

the whole of the writing surface of the bircii-bark available seems to have been the
determining factor. The text itself should be consulted but the following examples
may be helpful :

(a) Integral numbers occasionally occur without any marking off by lines
or cells, but often

(&) each integral number has a cell to itself, ejj.

42 39

(c) Sometimes an integer is marked off by two vertical bars thus :

|
14 |
and invariably a serie.s of integers is thus demarcated, c.</.
I
20 I
40 I 60 I
80 i evaiii 200 ||

(d) Fractions and groujjs of fractions are placed in cells or groups of cells,
e.flr.

'1

(i) 132 (ii) 1

60.5.504062,5 (iii) 1
1 1 1
1

33 _3227.520000 ^ 3 6 12
J I

(iv) 1 (v) 40 (vi) 21 (vii) /

1 1 20 12
3 21 12
+
!

1 3
4 1
1
+
5
•J

(e) Complete sets of operations are sometimes marked off in a similar man>
ner. For e.xample, (f/1 (iii) means J -*
i * I ^ i {d) (iv) means (1

(1 + i) (1^ ); id) (V) 40 (1-i) (1-i) (1 >).

(/) Series of operations may l)e Connected together by cell arrangements,

for example

2 dl° 1 di° 100000 di” phalarii di° 60000

1 1 947 947
2 2
3 cll'^ 1 di^ 157500 di® phalam di® 60000
1 1 947 947
2 3
4 “T~di° 216000 dl® phalam di° 60000
1 1 947 947
2 4

which means
100.000 . 60,000
dlnftras : li days
. .

•47
dInSras •
Mr days
] 57,1500 60,000
3i : U .
.
.

. ^47 •1
.

•
Sir"
60,000
99

4* : H .
*
.
•
t16.000
99 '
Sir 66
 21

(jy) The data of the problem and the solution may be indicated in one com-
bined statement, e.g.

16
11
I
4 a” chhe° 2187*
1
I

I
ijhalaih ^ara 2624400

4
1
3

This is a statement of pi jportion where the second term means 16 (1 + i) (1 t'-i) and
the number marked with art asterisk is a clianj:je-ratio.

(/O 0 2 1 3 3 12 4 di-i® 300

I
1 I
I

11 11 11’ 1

may be roughly expressed by means of x + 2x + 3 x 3x + 12 x 4x = 300

58. in our text of the sexagesimal notation in the form in which it

The use
occurs is of rather special interest, for there is, as far as 1 know, no other example
of the kind in any of the classical Sanskrit works. The Hindus, from Aryabhata
onwards, were well aware of the advantages of the sexagesimal notation for astrono-
mical purposes, but they never used it for arithmetical purposes.

Apparently there is only one purely arithmetical example of the use in the
text ana this example occurs, in connexion with a problem in arithmetical progres-
sion, on folio 6, verso, and 7, recto, where the fraction 178/29 is expressed
as 6 8'
( 16“+'
33"' + 6" This sexagesimal fraction is actually written thus
.

6
8
OC
16 cha^^
60
33 li‘^

60
6 vi^^
60
se ’
6
29

The upper tliree figures are missing in the manuscript buhl he restoration is certain
Of the abl)reviations stands for li/ila (Ok. lepte) which in Sanskrit works or-
dinarily means a minute of arc, or the sixtieth part of a degree; vi° stands for
ordinarily a second of arc' while .sV stands for hsham or “remainder.”
:

The purely arithmetical (and perfectly legitimate) use of the notation here
points to extra-India influence, for. although such a use is unknown in Sanskrit
works, it was extremely common in media-val Muslim works.

' On folio 37, vrruo ; lipiA and nlijM are uawi *

astronomically.*
 25

It will be noticed that the term Uptd here applies to “ tl^d parts ” instead of
" first parts,"^ and vilipid to “ fourth parts." The abbreviation cAa° has not yet
been traced to its origin.

59. The modern place- value arithmetical notation is employed throughout the
text and there is not the slightest indication that, to the author, it was a new or
strange invention. There is nothing in the slightest degree remajkable in the
employment of this notation in the text, neither is there in the text the slightest
indication of any other evidence that makes the employment of this notation in any
way particularly noteworthy ; but its occurrence in the text has been given an arti-
ficial importance by the arspiments of Dr. Hoernle and Dr. Btihler about the age
of the work. This topic will be discussed in some detail in the chapter dealing with
the age of the manuscript and the age of the work.

60. As already indicated algebraic symbols are not generally employed, but a
symbol for the unknown quantity is. This symbol is the arithmetical symbol • for
“ nought" or " zero," ana on several occasions it is referred to by the term iHnyct
"
empty,” " a cypher," and In some places by iunya, sthdna or “ empty place.",*

The symbol occurs some twenty times but only in sections B, C, F, G, H, K.

Its emplojrment is illustrated in the following examples :

a® u° 1 pa® labdharii 10
(i)
_i 1 I

j 1
I

I
•
1
1

I 1

This is a " an arithmetical progression where the first term is 1, the

statement ” of
common difference is 1, the number of terms is unknown, and the " (jjuotient " is 10
(where lOx-the sum of the series). Here the symbol • simply indicates that the
number of terms (pada) is unknown, t.e., that the place in the statement is empty.
The symbol does not enter into any operation here or elsewhere. The riving it a
denominator of unity is curious and really indicates that it is an integral number.

(ii) a® 5 u® 6 pa® • dha® .

1 1 1 1

ft® 10 u® 3 pa® • dha® •

1 1_ ^ 1 !
1

Here are two equivalent arithmetical progressions in which the numbers of terms
are equal and the sums also are the same in both cases, but both are unknown.

(iii) 1 bha® 6e° 16

1 1

which means x = 16/(1 -^) (1-|) (1-i) (1 -i) and since this is a deduction from a
problem it is a distinct step towards a proper algebraic s3nnbolism.

(iv) I • 2 3 4 drishya 200

1 1 1 1 1

which may be represented by x4-2x + 3x + 4x'»200. But the method of .solution is

by the regula falsi. Any number is put in the place of the • and the sum of the

^ In Hindu worki proper the dot ! the lymbol for negetWe quantity, and the zero lymbol to a malt circle.
 26

series is uVjtuiucd. Here the given sum multiplied by the asaurned value and
divided by the (false) sum is the correct solution. The connexion between this
nicthotl and that of operating witli unknown quantities is discussed later on.

(v) • •> vu niii • sa • 7 mu • |

1 1 I 1 1 LJ
•
which means 5 .v t x- 7 -t^ Here the symbol 1 stands for three different un-
known quantities. It sinq)ly indicates in ea<*h case an unknown number.

Ncujulive ifitjn.

(il. The only distinct mathematical symbol employed is the sign for a negative

quantity, which takes the form of a cross placed after the number affected. This
t

is peculiar and has given rise to discussion. In Sanskrit manuscripts a dot before
the quantity affected is the usual method of indicating a negative quantity. In
the transliteration of the te.xt the original negative sign -^) is, perhaps illogically, (

rreserved but it is a rather special feature of the manuscript, easily printed, and
;

fending to no ambiguity.

On this negative sign Dr. lloernle writes' ;

**
H«re, ther«*forp, there appeurH to he a mark of preat iinti<|uity. As to its 1 am unable to
linyHutisfarttirv explanation. I have been intorined by Dr. Thibaui of Benares that Diupliuutos in his
Greek arithmetic use.s (he letter th (short for «>/o’thic) reversed (thus 'f) to indicate the nejfiitive quantity.
There is undoubtedly a slight reHembhince between the twti signs; but considering that the Hindus did not
get their elements of the arithmetical science from the Greeks.^ a native Indian origin of the negative
Mgn HDems more probabbv It is not nncoiniiiou in Indian arithmetic to indicate a particular factum
bv the initial .syllable of a word of that import .subjoined to the terms w'bich compose it The only
plausible suggestion T can make is, that it is the abbreviiitifin (ka) of the word kanita, diminished.’* (Ife
also points out that tlie letter k in its unt'ient sbajH* as used in the Asoka iuscriptions is a cross and goes
on to say) “ Another suggestion is, that the sign rejire.seiits the syllable wi7j an abbreviation of nytiiia,
‘
diminished.' Thi* (ikxlntra f<»r nu (or nu) in the Asoku cluiructrrs would very closely resemble a ctorh (
However for the jiresent, the question must be left an open one.**

With this Inst rcmaik 1 agree and I would at the same time point out the
<langer of attempting to trace an isolated symbolic’ form back through the ages.
The suggested connexion w'ith A.4okii times is only part of the special pleading for
the extreme antiquity of our manuscript. The striking resemblance with the Dio-
phantine .symbol, so lightly discarded oy Dr. Hoernle, is of interest, but lacks sup-
port. From the time of Diophantus to the time of the Bakhshiill manuscript is
too great' to allow' us to ignore the lack of other examples.

Abbreviations.

Al)breviations are employed to such

62. an extent as to become, at
times, embarrassing, and they emVu’aee almost every type of term. None
of these abbreviations is u.sed in an algebraic sense, although, at first sight, when
,
we fijul ti^\ pa^ used consistently for the elements of an arithmetical progression,

an algebraic .symbolism does not seem very far off. But ha^, u°, r/o” are the
abbreviations of names of animals, ////°. 70 ® and 0 ° of plants, etc., etc. The names
< •<

* tmiian Aniifuartf, *vii (IBM), p. 34.

* Th<w A goo«1 dr«1 from them, and practically all their later aatronomy came from the Oreeke.
* Rut T>r. Hoemlc aantimoe that the Bakhi^hlllT waa praotieally cictitemporaneoiiR with Diophantna 1
 27

of ineasures, which are numerouB, are nearly always abbreviated. Certain

common terms of operation are often abbreviated, and of these the following occurs
most often

hh&^ for hhdga, placed after a term to indicate that it is a divisor.

ie° for ^esharfi, a remainder.
mu° for mUlarh, a root, a quantity that has a root, capital.
pha° for yhala'rtt, an answer,
etc., etc.

Indeed the writers seem to have used abbreviations whenever there was no
ambiguity incurred in their own minds. We are not so fortunately placed in this
matter and occasionally it has been impossible to rediscover the term implied.

Fundamental operations.

63. In the classical Sanskrit works there is generally very little formal infor-
mation about mathematical principles and method. Axioms or postulates seldom
or never occur and strict definition is seldom attempted. Rules are of paiiiciilar
application rather than general order appears to have been a matter of conveni-
;

ence rather than logic; and the fundamental operations receive scanty attention.
In a work that is essentially Indian in form we iriust not, therefore, expect mucli
formal attention to such matters, and, indeed, in the Bakhshalt text thei’e is little
We cannot even tell how the detailed processes of such operations as multiplication
and division were actually performed. All that remains are the formal statements
of terms and results, and all I can now present are examples of such statements.

64. Examples of addition

(a) 960 64 yutam jatarii 1024

(fe) 10 sa rupam 13
3 3

(c) 2 rupa samyiitaih *3

{d) 840 49 datva jsltarii 889

(c) 924 1
836 1
798 1
esharii yutiih kfiyate jata 2558

(/) 120 1
90 1 80 1
75 1 72 1
eshaiii yoga kj-ite jiita

ig) 120 evam 377

90
80
75

—
12

(A) 10 30 !
90 1
ekatram 130
1 1

(•) 960 M<1 64 added : 1024 ie prodooed. (M P*"* V V*- a) * onay eddcl together 49 hevinf; bern niven
. j- nroduoed. (c) 624, 836, 798 : tbe eum of these mey be determined they prodi’- e 2-.0.^>H. (7) 120. 90, HO. 75. 72 the
: ;

theM b mede : they produce 437. (») 120, 90. 80, 76. 12 : thus 377. <h) 10. 30. 6<) eltogethcr im
 88

(•) 29 7 anena 3rutam 36

48 48
58 58

(*) 5 de 1 1 1
10 de 15
1

1
1 16

(1) 45 sardha tray a yutam 52

2 2

65. Examples of subtraction

(a) 1 I
d I
vi4esham | 4 |

(6) 3 2 videsham 1
!
2 1 2

(c)
1^1 l_
^
J
jStaih 2

(d) 6 I
3 6uddhi 3 j

(e) 42 tryunam i 39
L
I

(/) 3 7 I
vidoddhya |
4
I

(9) 77 294 !
p&tya desham 217
11 -Ji ]
11 ..J

66. Examples of multiplication,

(a) :
2 I
dvigunam |
4

(6) 30 ashta gunaiii 240

j j

(0 2 40 gupita jSt&ih 16
5 1

id) 4 gupita j&tfi 27

1 8
2
+

(«) by tilts (7) inornMMl»a6||. (1^) f thu 16. {t)

V with three end e helf eddedn
(a) 5, 9 the difference is i. (6) } . 2 : the difference is }. (f) 5, 2 : snbtrected, 2 is produced, (d) 6, S't the dUfersoee Is 8.

W
;

(•) 42 less three is 39. (/) 3, 7 : hering mibtreoted, 4. (g) * hsTing subtreoied the difference is 2^4 •
(a) 2 multiplied by tiro»-4. (8) 80 roultipHed by eighth 240; (e) f end 40 multipried : 16 is produced, (d) and 4) anill*
piled V produced.
 29

(«) 6 1 aneiia gupitam jstam 4

1 1 J

(/) 1
2 1
a 1
4 gu 9ita j&tam 2
1
3 j 4 1
5 5

40 phalaih 16
1
1

(A) 3 4 abhyasaih 12
I

L_ j

(») 8 I
&tma guaaih I 64

(;*) 880 964 gu^ita j&taih 848320

168 14112

(*) 737 J 178 anena gu^itam jfitam 65593

58 }-{ | 29 841

(0 108625 pada-gans 6455949625

65600 32275^00

'Km) 405280 44004 saongupya jatam 179945781120

38724 77448 2999096352

67. Division. Fractional quantities are expressed in the usual Indian way
with the numerator written above the denominator without any dividing line, and
they are usually placed in cells. This indication of the operation of division often
does away with the necessity for an explanatory word or phrase.

Examples of division,

(a) 168 168 168 labdhaih 42 | 28 |

24
4 6 7

(A) I 4 I
vibhaktaih \
35 guijitam 36
4

(c) 10 vibhaktvaih
L '? i

(«) Saadi —(i b7thia(|)iBiilUpUed,4iia prodaoed. {f) h t mnltlpIlMt. { ii prodticM. (g) 40 (I — |) (I —(t — (1 |)
• IS. (S) S, 4t Chaprodnet la mnltipliad by
U. (•) 8 itaeU— M. (») W>4ff mulWpliad ; ia produead. (8) igr
by tUa (VV) mnltlpUad. la prodnoed. (0 moltIpHed by tb« oumber of twma (}||U) - ("•) i

moHIpHad thaae togather prodnoad.

VWtf >
(a) —
t^hdkcmrnrnqvtMmt, (&) This peripbrMtio expoaition iioommon the divisor is tiimMl into a multiplier by being iurertod.
The ezemple here : 4 dividedas which multiplied into 35 (c) 10, 3 hering divided** ^ by . :

B 2
 so

{d) 447 dalita 447

29 38

(e) ! 132 I
vartyam j&tam
1
33

(/) 798 projjhya 798

1463

( 9) 2558 chchheda projjhyaih 1095

(h) 60'
j
anena drishyaih bhajitaih 1 300 jat&
j

60 1

Square-root.

The method of extracting ."square-roots exhibited in the text is of special

68.
interest. The rule is much mutilated but as it is given on three separate occasions
the fragments pieced together enable us to give the complete sutra :

akrite slishtha kvityiina sesha chchhedo dvisamgunab tad vargaU dala

saihsUshtha hfiti suddhi kriti kshayab

This means somewhat as follows ;

The mixed surd lessened by the square portion and the difference divided
is
by twice The difference is divided by the quantity and half
that.
that squared is the loss.

Then follows a note to the effect that “ by means of this rule an approximation
,”
{anaya) to the proper root of a mixed quantity is found . . .

The rule as i1 stands is cryptic and hardly translatable, but fortunately there
are examples given in some detail, and the.se show that the rule was extended to
“ second approximations.”

The rule moans that the first approximation ^/Q— ^/A’ + b is A-i b/2A or ;
but q» — Q=(b/2A)*

No rule for 'second approximations* is preserrod but there are several examples; and, of course,

no fresh rule is really required,* for v/Q*" “* * second error' may be

indicated by o, « («i /2qi)* = •

(<t) %V halved (<) 4Y reduced givcB 4. (/) lUvinff diecarded the denominator 708. (g) 2558— 1483 >
1005 where 1483 is «he donominalnr of the fraotion. (*) By thb (80) the known quantity (300) is divided and ^ of 300««>5.
1 Note that if then r— The text uaea r'-^
iae h mm A 4-

btit not
X
 31

All the examples preserved belong to section C and appear to be merely subsi-
diuy to the solution of certain quadratic equations arising out of problems in
arithmetical progressions. These problems will be found fully worked out in § 86
and I give here merely a summary of the square-root evaluations.

(i) +® while and qj = 6^1848

(ii) v^l55'a=V^l00-f 5 and qi — loi; ei=^; qi=loi— lo

2xlol

(2^)
(ill) v/*81=n/21’ + 4«+J ‘1‘ = %2= =14112 =

- 21“ /'^"V /2 21=*‘*-01®<'20 o,_ 1 r*0\4 If a ^ 40\2 ^im,ooo •

n it
^Vl =210581; 8-=8(42;
,
= 2,009.0^,302

(ir) ^W9 = ^WTU: qi = <|;‘i=AQ='=8‘^..

/(v) x/33«0U9 ® v/6T9*+ 708 ; q,^57»5^|; q.-fiT?;777,307,000*

e, _ 21,743,271.036
la ~ 7.200, 483,304, 675,000, 000*

69. There is not much doubt about the exegesis of the rule. It was neither
connected with continued fractions nor with the so-called Pellian equation.*
Brahmagupta gave the converse of the rule, namely (A + x)*t-»A* + 2Ax from which
the square-root rule given in our text is immediately deducible. But, as already
pointed out, the square-root rule itself wa.s not used by the Hindus and was not
even noticed by them ubtil the sixteenth century." Indeed the Hindus had a very
good practical rule of their own, which was given by ftrldhara (Trisatika, 46) and
Bhaskara {Lildvati, 138), namely “Multiply the quantity whose square-root—
cannot be found by any largo square number, take the square-root of the product
— —
leaving out of account the remainder and ilivide by the square-root of the multi-
plier.” For example \/41 -- v/41 x 1000000 1 000 -- 6-403 + :

This rule was used in conjunction with the usual modern scholastic square-
root rule, which dates at least from the time of Heron.

The examples in the text may be further conveniently summarised thus :

a qi qa
q (approximately
corrot^t)

41 6-41667 6 10313 6 403124

481 21 9524 21 9307. . 21 9317122
889 29-828 .
29-8161030
336009 579664 579 6615 579-65

^ In the text we find end not t by iteell, for reoNoiui exploincd in §8ff.

t The form iUelf ; hut the negative Bign cloen not appear to have boon ueed
in firat
approsimationa.
The Appendix to Chapter vi (page 53) may be helpful to those who are interested in the detaUs of the caloulationn.
^ For a fuO diaonssion of the topic fee P. Takvxbt Mimoires teieniifiquM, I, pp. 160ff. and II, pp. 157 ;

* A short historical note is given on page 45.

 32

" Rule of three”

"
70. trairaSika relating to three quantities ” occurs about a dozen
The term
—
times always employed in the sense of arithmetical proportion. Most often It
occurs in the phrase pratyaya trai-rdiikena or proof by the rule of three ” and the
proportion is generally set put in the following manner

1 1 4 phalam 18
3 1 1
2

which means i : : : 4 : 18.

The term trairddika is orthodox but the term phalam is used throughout our
text quite appropriately as equivalent to “ answer ", and is not applied to the second
term of a proportion as in the Lildvati.

There are very of the “ rule of three ” used either as the direct
many examples
method of solution or as a proof. There are also some very much damaged ex-
amples of what appear to be problems in so-caUed compound proportion. There
is nothing of the nature of a theory of proportion discussed but several examples
of the following principle occur

If
A
n,>
*'*"
^
^
rb*B’

Regula falsi.

71. rule of “ falbe position ” or “supposition ” is used in

The two ways in our
text' in solving linear equations.

(i) f solved by assuming a value e for

(x) =p is x. This gives f(e)=p' and the
correct value of x is ep/p'

An example is

The amount received by the first is not known. The second receives twice
as much as the first, the third thrice as much as the first two and the
fourth four times as much as all the others. Altogether they receive
300. How much did the first receive ?
Suppose the first receives one, then the second receives 2, the third 9, and the
fourth 48 ; or altogether they receive 60. Actually, therefore, the first received
300/60 - 5.

Iff(x)^p reduces to bx + c-p and f(e)“p' reduces to be tc-p'

-
(ii)
then X (p ^p') /b + —
e, which appears to have been considered useful when it was de-
sired to keep b and c unchanged.

» Xo ml* \b prowirred In the te*t hut Bbiaknra jrfTcie the following ( 50).
**
Aut number awnimed »t pleMrore U treotod wM epeolfled in the pertionler qoMtion* being mnltinlied end diTided, mieed end
diminUhed by freetione ; then the given qnentity, being multiplied by the earamed number end divided by the reapli yields the
nnmber sought.”
 For example if x^ + Xa^^au x,+x,»ai, x« + x,»a,
we have 2xt+ (aa-aj^a, and ; if in place of x we put e then 2e+ (at-a,) =>‘a'. and
the correct value of Xt is (aa-a'«)/2+e

This rule of “ false position " is interesting as being, in a way, a precursor

72.
of algebraic symbolism. It connotes the idea of an unknown quantity and even of
a symbol for that quantity (e.p., as in our text) but it does not embrace the notion
of such a quantity being subject to operation or being isolated. As soon as we in-
troduce algebraic symbols the rule, as it were, disappears. Algebraically the rule
solves bx»p by x-^^ep/p' where p'^^be, which shows that the transformation from
bx — p tox=°p/b was not conceived.

There is no algebraic symbolism in our text but it may be noted that Bbaskara
gives both the regula falsi and also an early form of algebraic symbolism. Neither
Aryabhata, Brahmagupta nor Sridhara gives this rule or makes use of the prin-
ciple. Its occurrence in the Lildvati therefore seems to indicate that it was intro-
duced into northern India after the time of »^rTdrara (Xlth cent.).. Mahavlra
(IXth cent.), however, uses the method in rather a special way in connexion with
a geometrical problem.*

The rupona method.

73. There are several references to the riXpona method, and the phrases rupond
karanena or pratyaya rUpoiin karanen/a occur. In all the cases the application is
to the summation of a series of terms in arithmetical progressioil according to t he
rule
8“[ (t— 1) d/2 + a^ t

The reourrence of the phrase rupond karaneua seems to imply that the rule in
question began with the term rdpoJhd which corresponds to the (t-1) of the for-
mula.* The rule is not preserved in our text but we find the following in the
Oanita Sdra-sangraha of Mah&vira (ii, 63)

rUpenono gachchho dall kritah pra chayat&di to miiirah |

prabhavena padabhyas tas sahkalitaiA bhavati sarvesham )|

" Tne number of termi it diminiahed by one, hnlved and mnlti plied bj the increment. Thia when
combined with the flrat term of the aeriea and multiplied by the number of terma beoomea the aum of all.”

The rule is exemplified in two ways in our text of which the following are parti-
cular cases :

5° 1 u° 1 pa° 19 rupopa karanepa phalaih 190

1 1
1 !
1 1

which means ;
“ a = 1, d = 1, t* 19 by the rupona method the answer
; is s = 190.’'

Similar bare summaries of the " rupona ” method are found in section " B ’’

mainly.

^ ^niia Mra saAffmAa vii, 112.

* li this faint aoho of ^^ha^orean atylo t Anatoliat writai : ** The Pythagoroana atate that their
Ik maeter. in oonnexJon
with numbfire that form a rtght>anffl«H triangle, showed how euoh oonld be oompoeed by means of unity." [| (a*— 1)]*4-
I (a* <>>1)*. dee P. TAmnmr Jlf^fnotrs# iin 14.
 84

In section *' C ” the application of the rupono method is always worked

(ii)
out stop by stop. For example when a=5, d = 3, t=^ 178/29 then 8= ((t— 1)1 +5) t ;

and the rupona method is applied as follows :

X *^?-**’ — ''37 178 85693_.»83«

t
~ 1
28 ' ~ 20 29 '
t*2x
29
4
* “ 68 ' lUt"*'
r
~ 68 ' 68 29 * 841
~" 841 *

Proofs.

74. many solutions are attached proofs, which are generally introduced by
'I'o

1 he term pratyayaih “ proof ” or " verification." This is sometimes amplified into

“
pra fyaya-trai-rdHkeiia proof by the rule of three,” or pratyaya-rupond-karanena,
"
proof by the rupona method.”

Sometimes the proof ” seems to be merely a matter of rewriting a statement

“
in another forin, but generally the answer is utilised and one of the original terras
of the problem is rediscovered. Occasionally the “ proof ” consists of a solution of
the original problem in another way, e.g., by steps but sometimes it deals with a ;

subsidiary aspect of the original problem. In certan problems approximations

are employed and then the ” proof ” may be described as a process of reconciliation.
These reconciliations ” entail a comparatively high degree of mathematical skill.
”

Occasionally several different proofs are attached to a problem. The following are
specimens.

(I) I'roblem. ^t-7 = i|tf7

•'iolufioH.

Proof. 3: 6 30; 00, 6:6 :: 30: 36 and 60-7 = 36 + 7

Problem. A 1* days, B gives 3^ in

gives days, C gives 4J in IJ days. In wkat time
(2) 2J in
will they nave given 600 dlniras altogether P

. ,

f 60.000
Solution. t =24 .

4i= -gvi. days;

**!{ 1
*'
4
and A gives 100,000/947, B gives 157,000/947, C gives 218,000/947 dlnfiras.

Proof. 2^:11:: lOO,0(K)/947 60,(KK)/947 :

3J:1J: : 167,000/947: 60,000/947

4J:1J: :216, 000/947:60, 000/947

(3) Problem. One earns « and spends f daily. How long will a capital of C laetP

Solution, t = C/ (f - e)

Proof. 1 : f : : t : F (the total expenditure)

1 :e: :t;£ (the total earnings) and F-E — C.

 35

(4) I*rohlem, Tf 7 are bought for 2 and 6 gold for 3, and tha capital is 24, what will be the profit y

Solution, p-^ C (c/s 1 )ls 24 ( g

1)^18.

Proof. 2 : 7 : : 24 : 84 (the mini her of articles)

and 6 : 3 : : 84 : 42 (the total proceeds), and 42 - 24aBl8«

or — 1 : c : : C : u , h : 1 ; : n :
(* 4-
p and C+p - C — p.

(5) /^rohletn. x (1 J) (1 - J) (1 - -x -280.

Solution. X = 280/ (1 - v\,

) -= 400.

Proof. 400/2- 200, 400 - 200 - 200,

200/4- 60, 200-60-160,

160/6- 30, 160-30-120, and 400-120 -280.

(6) Problem, l-fi (2 f i (3 4- ^(4 +} (6 4- *)))).

Solution. —

Proof. {((((U 1) 2-2) 2-3)-4)J 5- J)-0.

(
7) Problem. Solve A46— s*, x-7 = t*.

Solution. -a))> + 7-ll.

Froo/. 11 + 6-4* U-7-2*

(9) ProhUm. Dt - ((t 1) d/2 + a) t.

Solution. 1—2 (D - a) /d 4- 1

Proof by the rij»ona method : s —( (t --

1) 4- a)t and Dt —
 S6

(10) Problem. ((t - l)cl/2 + u) i.

Solutton. t - {
Y/(2a - (l)^ -f Hh - (2a - (1) j
- -
7 2 d.

Generally the solution is an approxi mation t^ or t, depending on the method of

evaluation of the surd quantity and the rupona method gives Sj or Si, neither of- ;

which is the same as the oi'iginal s. A process of reconciliation is therefore called

for, and this is given by s — s = e / 8d where ^ is the approximation to s g;iven by the
proof and e may be termed the square-root error.

Problem, a I, d 1, k GO.

First solution. 1,= (v/(2 1 - l)> + 480 - 1} -:-2-l-(2lS - l)-r2-

06«_8in.33O
Firfit proof. ».=t, (t, + i)- 1H8~ .411^

"“‘‘i
-gL = B48,sa0— 1600 _aa
fi,
= *®
r4Tf2~

/424,642 1 V « _405,280
Second * solution. 19360"
V ^ 38724

40&.2S0 444,004 170 ,045 .941.120

Second proof. :»H724 ^ 77448 “
'
2 009. 096.852 *
.
Now (’'1+2)? “d S J

179.045.781,120
and H = 83 - ©, / 8d =
2.9^9,006,862

(11) The probleinti in soctioii G arc characterised by numerous proofs as many as fiva separata — —
proofs being aitarhed to one example. The following is fairly typical:

Problem, x ^r)()0(l - J)(l - {){l - ^)(1 - J).

13
Solution. x = 158 64

Proc»f (i). x'(l i) (1 - i) (I I) (1 - i) = whence x'mOOO. This is written horisontally.

Proof (ii). x'(l - J)(l - 1)(1 j)(l - 1) - whence z'«500. This is written vertically.

Proof (iii)- x' ^ —500.

11 t) U-tl O

Proof (iv). By steps.

500 : 4 = 125 , 500 - 125 - 376

375 4^ 4 = 932 .
-
93J - 281^

281^^4-70^^^, 251J-70A = 210H;

no}5-^4«:52Jf. 21(‘H-62H = 158^1.

Solutions.

75. The solutions are sometimes very detailed, proceeding most careiully Aten
by step, BO that they become expositions in general terms of the processes involved.
Also they often give actual quotations from the rules to such an extent sometimes —
that the original wording of the rule can be reconstructed. Unfortunately, how-
ever, such helpful suggestions are more often missing than not in many cases the i
 37

particular portions of the manuscript are lacking or damaged beyond repair. In

some sections, e.g., section " M
”, nothing but the bare answer is generally given,

and in others the outline only of the working is given.

The following examples illustrate these remarks 1. The problem is DT + Dt ;

**((t— l) + a) t where a and d are respectively the first term and common differ-
ence of an arithmetical progression, and t, the number of terms, to be found. The
rule is
t= { 2 (D - a) + d + v/ (2 (1) - a) + d)" + 8dDT i
- 2d

and the particular example gives D = 5, T“6, a -3, d-4. The actual working of
the solution, translated as literally as possible, is as follows :

*The daily rate diminished by the first term ;* the daily rate is five yojanas,
5 the first term is 3 their difference is 2 this doubled is 4 *thi8
; ; ;
:

increased by the common difference* is 8 and squared is (>4, which ;

is *known as the kshejxi quantity.* *Multiplied by eight* the fixed ;

term (30) multiplied by eight is 240 and multiplied by the common

—
;

difference multiplied 960. *The quantity known as kshepa is

added.* Now the quantity known as kshepa is 64, which added gives
1024. The root of this is 32 the quantity set aside is 8 and this added
;

gives 40. *Divided by twice the conunon difference :* twice the com-
mon difference is 8 —divided 5.

The phrases placed between asterisks are quotations from the rule. A perfectly
literalrendering would not be very intelligible to the ordinary reader and one or
two gaps have been filled in but the translation is a perfectly fair representation
;

of the original.

ii. The problem may be represented by x ( 1 - i) ( 1 - i) ( 1 - i ) = x - 24

and the solution is given as follows ;

Having calculated the loss on unity* the terms become j’, J, f and these
multiplied together give This subtracted from unity gives f,
which, inverted and multiplied f by the given amount 24, is |of 24=^40.

iii. Something travels 3 yavas a day. How long will it take to go yojanas ?

Here the solution is indicated by the following proportion only :

3 ya° :
^ years : : 5 x 4,608,000 ya^ : 21,333 years 4 months.

^ The phrsM marked oil bj Mioritkt are qnotationii froin the rJIre or rule.

t Uteralljr ** divided and multiplied **

into. Thie ie generally ueed to Indicate dlTition by a fraction.

F 2
 38

CHAPTER VI,

An analysis or tub contents.

76. The present chapter contains an analysis of the mathematical contents of

the text. To those who are chiefly concerned with the history of mathematics this
chapter will possibly be of rather special interest. Indeed, taken with chapter I
of Part 111 (which will appear in the second volume) it forms the I'eal corpus
of the present study, for the latter gives the re-arranged text, while the present
chapter is an attempt at a fairly complete explanation of the intelligible parts of
that text. The analyses of the different sections are not exhibited on a uniform
plan, some are given in great detail, while others are mere suinmanes some are ;

arithmetical in form and some algebraical but these differences correspond to some ;

extent to the spirit of the text, and it is hoped that the analysis as a whole gives a
proper representation of the original.

77. In order to avoid innumerable cross references the following table is given
to show the connexion between the paragraphs of the present chapter and the ori-
ginal and re-arranged texts.

C*hapt«r VI. llluatimtioni (Bodleian order). New order.

§78 Folio(«) 29, 27 l2.

§79 „ 1,2,.3 Ai:). I.V

§«<* „ 38". .39<-. ••

58

§81 ,.59 K.

(82 „ 27 nwt<i ....... ^It

883 (i) ,, 4 rtM^to ....... B,

(ii) „ 4 vrnw) B,.

(iii) „ 8 rec;tx), 9 rwto, 7 verso .... Bf.

5«4 (i) „ 8 verso ....... B,.

». 4.

(ii) „ 9 verso ....... B^

§8B Sootion *
C ' generallv ......
88« (i) Fo]io(8) 5 c,.

(ii) „ 6, 7 recto a.

(iii) <>5 verso, 56 verso. 56 recto, 64 recto C4

,, .
. 4. C
(iv) „ 64 verso, 67 Co. 7 .

(V) „ 45 recto C.

(>) „ 45 recto, et verso, 46 recto .... C.,.,

S87 25, 24. 25. 26

». 4, h

$88 51 . . . . ... Ag

$89 10. 11. 12, IS, 14, 15, 16 recto

 39

Chapter VI. flliieiratiotiii (DodleMn order). New order.

§90 FoIio(8) 10 ver»o, 17, 18 . H,. .. r

§91 fi) (iO vejft.i L,

{«) GO rurto, G1 I'.

.

(Hi) *% 21 verno, 22 recto K4 F,

§92 (i) 61 . . .
W
(H) i» 62 .

(iii) 62 Verso w
(iv) - C3 . . .
w
§93 11 •M' . M.

§94 Folio(K) 17 recto Mr

§95 >} 14 reef-o, 43 vei-s > M
§9<> 41 verso M„
§97 If 36 veiHo M.

§98 •• 20 verwi M,

§99 * * 32 . M,
1

§l(K» (i) • i
37 recto Mr.

(ii) - 37 reeto M.

(iii) »* 37 verso M,

§101 (i) 1 •
53 recto ^10

(ii) »» 55 verso ^ JO.

X
(Hi) «f 49 recto • • •

(iv) f * 49 . • • • M,.

Sf/steftifi of linear eqnalions.

78. Examples of the following type occur

x, + x,-a,, X, + x,, ="a„ , x, ix,—a, where n is always odd.

We have a. = (a,-a,) + ^ {a»-i — a..*) ^ 2x, ; and if we assume x,==p

then a',= (a,-a,) + + (a„., — a..,)+2p.

Subtracting we get x,»p-(- (a. — aM/2, which is the solution employed in
the text.
 40

Tho following examples occur

(i) X, + x,-13, x, + x,“14, X, + x, = 15.

The value of x, is assumed to be 5, then by subtraction in order' x,=8, ‘

X, = 6 and x^ + x» = 1 1. The correct values are therefore x, = 5 + (15 — 11) -i-2

= 7, X, = 6 and x, — 8.
(ii) X, +x, - 16, X, 1^ X, ^ 17, X, > X4 = 18, x« + x, = 19, x, + x, =20.

Here the value of x, is assumed to be 7 and xi +x'j^ 16, therefore the cor-
rect values sire x, = 9, Xj = 7, x, = 10, X4 = 8, x, = ll.'

The following are implied

(iii) x, + x, 9, X, I X, -5, x, + x, = 8.

(iv) x, fx, 70, Xj f x,-=-52, x, x, = 66.

(v) X, I X, = 1860. + X, = 1634, x,4x,-1722.

(vi) ° X, + X, + X 4 + Xj = 317

X, I- X, + X 4 + x, - 347

X, + X, + X4 + Xj = 357

X, + x, 4-x, 4X4 = 365

and the following, which, however, is too mutilated to be sure of

(vii) x,+x, = 36, x, + x, = 42, x, + X4 = 48, X4 + x,=54, x, + x,=60.

The directions seem to indicate that we should cancel by six. We then get
a', =6, ai=B, ai=9, ai = 10 of which the solution is xj = 4, xi *= 2,
etc. ;
whence x, = 24, x, = 12, x, = 30, etc.

79. The next set of examples can be represented in the form Jx-x,— c-d,x„
Sx'-x,-c-d,x, Sx-x,-c-d.x,.

If we set a - 1 - d these become

Xx “ aiX, ™ 2 X “ a,x, = ....*.= 2x “ a^x, c, and aiX, ^ * a^x* * ^x

-k.

* Sino» bi tliMe two purtiruUr examplco the valuce of ftp He., ftre in ftrlthmftteftl progreasioB ft almpler aohitioD would bo
X,.. Y where m la the meftn of the aerlea ft,, ft,, etc.

• Thli oftn be ftirftnged in the form y, f yt-»'Ti=347. y,-fy 4 = 3 « 2 y4 +yi= 3 l 7 yi+yi— 367. 8ohring this wo got Xj-hVe
, .

—310, *,-f Xt**"^®* x,+X 4 * 166 X , 147, X4-l-x,=*192, which eoKed givca x,-^120, x«— 00, x^— 80, X4 — 76, x,— 73. (floefollo K)
,

Alao noto thftt moot of theoe aix exompleo con bo expreeood in the fonn
Sx— X,— ftp Sx-=-Xa— isp Zx— X,— ft,.
The exftmpleo given in the text ftre undoubtedly ftkin to the *
Kpbftnthemft \ uauftlly ftttributed to Tbjrmftridma, wbleb moy bo
exproaoed by X4 ^ , where x^+Xj^ftp Xo^-fti— x*, Xo+x,— ft,. In Aiyftbhft(ft*o io ft oimilor rale Tx—^^
where £x — x,— dp Zx - x^^d, £x— x,— d», which Cantor ( VaHtMUngen Uher der Ot§ekiehU dtr MaihemaUk l» 6Sd) ooBoidem Io
be ft modiflcfttion of the rule given by Th 3rmftridftft See ftlao P. Tanhext Af/moire# ^eieniifiguu. Tome il* pp, 103-106.
 41

Therefore 2x
= + . .
.
)k = ^ k.
A solution is x,= x, =’ x.= ^ and c-p=q^

thus
There are four examples illustrating this process which

di <*. d. d. d. p^q
—— c *1
may be

*4
tabulated

1 1 1 1 1 437
(i) 377 120 90 80 75 72
y 3 4 5 6 so

7 s 11 S5&S
(ii) '4
1095 924 836 798
IS 0 lias

(ui) 5 7 8 i
262 42 28 24

(iv) 2 3 4 17 6 8 2

Of these only fragments remain. For example, of the first we have :*

1111
1 3 4 6
1

«
120
80
90 W 76. 7i_4.37
BO
ISO
90 + 80 + 76-i-72.317, 317 + T =377.
90
120 + 80 + 76 + 72 - 347. 347 + 8
-377.
80
120 + 90 + 76 + 72 - 367, 367 + 4 -377.
76
120 + 90 + 80 + 72 -362, 362+ 6
-377.
73
120 + 90 + 80 + 76 - 366 366 S
.377.

The second of these examples may be expressed in the form

Xi + Xi-(i+i) x,=x,+x, -(i + i) x,==x, + x,- (i + i) x,.
From this [|x, —+x,==y x„ and 2x/(Sx — c}=||+y+ii=J^-

The solution given is 2x=2558 and c- 2556 -1463 = 1095, whence x, = 924,.
x,-836, X.-798.

Of the third there is sufficient of the formal question preserved to enable it to

be restored.

One poftBSMei Beven horseB, another nine mules (P), and a third ten camels. Each gives one of his
animals to each of the others and then their possess ions become of equal value.

Compaie with this the treatment of the epanihem by Jambliohua (Heath, Gr^k Maihemniir^,
i, 03-94),
»

have come across in an Indian work is given by Mahftvira in his Oanita Sdra- Sangraha
• The nearest approach to this that I

(rU tS9.S40).
another, by obtaining |, 1, { or of the oontcnts of the purse
Viva jBsrohanU saw a purse of money. They said one after
I beoome three times as rieh as all of you.
Lrt ba the value of the puiae and E„
p
eto., the original capitals; then 3(2 e“E,) f Ei. 3(£e— t) > -i-Xg, eta; —
wImbm llI*-(J+*+*+i+*)P— «t—a®i-
Bat al'Karkhl (xlth o«»t) g«v« (ia, 8 ) the following A oerUin earn ie divided emong three people, one-helf being given to
the flnt, one-fhitd to the eeooad end one-iizth to the
third. But then one-bnU the ehnre of the flret, one-third the share of the
equally by the three.
seoood and ooa-slxth that of the third is pooled and dinred
problAdi became famous as it was one of thoee propounded to I^onsrdo of Pisa and solved
by him.
In A.D. 1225

U Ef Eg, Eg be i\o shares and So be the amount pooled, then E4 — 4-c-^*,

XE-^SEg^de, whence £e— ake c—7 and than Si^SS, e, - IS, 1 ,— 1.

£e— S|-» 2o, £s-* 2Eg** 9o,
 42

The fourth example is of similar form.'

80. Tlie following occurs on some very mutilated scraps that have been pieced
together. It is now impossible to say how it originated or what its context was.

x(x + y t z) -60
y(x + y t z) = 7o
z(x + y i z) -90

We have (x t
y hz)’ 225 whence x i
y i-z = 1.5, and x-4, y"=5, z = 6.

80. (a). There is one nearly complete statement and solution of the following
pair of equations :

X I

y + z = 20
3x+!|y + ^z — 20
of which tire only solution in positive integers is x = 2, y --5, z-^13.

This type of problem was a favourite in Kurope and Asia in early mediaeval
times. Later it was known as the Regula Virginum,’ llegula Potatorum,’ etc. ‘ ‘

It was given by Chang-ch’iu-chien (sixth century A.D.),“ by Alcuin the Englishman

(eighth century) and others. About 900 A.D. it was pretty fully treated by Abu
K&mil al-Misri who gives some six problems, varying from three to five terms and
attempts to find all the positive integral solutions.^

In the earlier Indian works problems of this type do not occur, but exactly
the same example is given by both Mahavira* ( ( ninth century) and Bhftskara*
(twelfth century).

Quadratic indeterminate equations,

81. There are two types of quadratic indeterminates preserved. •

(i)x'a-a*, x-b-t*
The solution given may be represen te<l by

l b

which makes both x + a and x b perfect squares. In the actual example preserv-
ed a- 5 and b = 7 and the solution is x = 11, which is the only possible int>egral (posi-
tive) solution obtainable from the formula. No general rum is preserved but the
solution itself indicates the rule. It proceieds by steps thus 5 + 7 = 12, 12 + 2=6, 6- ;

2 4, 4 2 = 2, 2’ 4, 4 7 11. The value of this type of detailed exposition is here

: f

^ Thin ant of <tx»mplrN U inirtNfiiocd by a rule which iiiniply moann change tlie fraction
; into Tbia ilhistratee the fact
that the or rules often make no pretence of indicating in any way the general theory they are merely inten«1od to be helpful.
:

* Yoehio Mikami.
H. Suraa. Am BmcA fUt StUenkeit, etc. Bib. Math, xi (1910-11), pp. 100-120. See alao ScTaa’.H Die MttSkematik«r und AHro*
nnmen Her Amhrr nnd ihrer HVrkr, p. 43.
# *.^. of
1 jL + y-f f w
». 100, 2x-f--j -h-j -f w — 100 there are iiaid to be 304 aolutiona and twenty are given. See aUo L. K,
DiciLHO^'n /fieforv ihe Theory of Nun^ere. vol. ii, pp. 77, cf

* fhiniia Hdm Siahgrahn vi, 152.

* Vija OanUa. §158.

 43

self-evident. It seein:i to have been almost necessitated by the absence of a suit-

able algebraic symbolism.*

82. The second type is

(ii) xy-ax-by-c-=o

The solutions which appear to be followed in the text are x — (ab + c) — m+ b,

y = a + m; ory = (ab + c) -^m + h, X“b + m

where m is an)^ assumed number. The only example that is preserved’ is xy — 3x—
4y± =0 1 of which the solutions given are 15 and 4, and 16 and 5. t.e..

(34-1)^1 + 4-15, 3+ 1-4; (3 4- 1) - 1 + 3 = 16, 4 + 1=5.

Sit7iple ‘
motion ’ problems.

83. From the mathematical point of view this section is not of particular inter-
est but it was possibly intended as a sort of introduction to the section (C) that
follows directly after. The examples can be classified into three types :

(i) A travels at a certain rate r, for a number of days T, and then B starts
at a daily rate r,. When will A and B have travelled equal distances ?

Since TiT + r,t = r,t, t = r,T / (r, — r,) which is expressed by a aUtra
and illustrated by two exaniples.

(ii) ^ travels a distance a, the first day. a.-d, the second day and soon;
B travels a, the first day, a., - d, the second day and so on. When will
they have travelled equal distances i

Since ((t-l)d, ;•
2+ a.)t = ( (
t ^ l)d, : 2r a,)t, t--2(a. -a,)/ (d, d.) + 1

(iii) One goes at a fixed rate A and another goes a the first day, a ^ d the
second day and so on. In what time wdll they have travelled equal
distances ?

Since tA = t((t-l)d : 2 i a), we have A = (t - l)d +2 + a, and t- (2A -2a) + d +

*Thi» type of equation ooourn in many medieval worku from

the tiror of DiopbaiUuN ontvarda (e.q,, see Diopmaxti)^ Ii, Iff
Dhah- I ;
MAOUPTA. xviii, 84 ; al-KabkhT.
and has here no very suc^ciat interest beyond the indifiation it gives that the
p. 63 ; rt-c.)
J^khahAII
teat followed the fashion. Dr. Hoernle, however, thought that it indicated a fieculiar o^mnexion between the Balchsh&ll MS. and * ’

Brahmagupta's work ; and from this defluon) that our text may have been one of the sourctw from whence the later asfronomnra
*

took their arithmetical information.' (Ind, Ant., xvii. 1888. p. 37.)

• Unfortunately the text (folio 27, recto) is so mutilated that the oorrectness of the
interpretation here given cannot t>e
guaranteed. But the ceptation was well knovm to mediaeval mathematiciarui and baa historical int^west. Brahmagupta gave a
general solution (xviii, 61) which, however, he appears to have thought unnecessiirv. HhAskara also gives a genera) solution (Viia
Ouniia, 212-2U) together with deinonsiraiions in both alfebraicsl and gK>metrical forms but be also on another occasion fib ;

2t)8«2(y9) gives an arbitrary solution. The general solution is also given by al-Karkhf. MabAvira gives (vi. 284) the equation in a
different form, namely (a-h^) (bf-)=ab^-A which reducee to x y—} ^ o of which the solution x^^ abi A
^ is ,
^
^
).

y — -J-
+mj or j- v+ 4 ( **r)> » - T'+m- w" make +1 thrn y- *••'’**
which is the solution pivim by
HahAvtra.
Hia examplea are : The product of 3 and 6 la lA. and the mpiired product is IH or 14. What are ihr quantiticH to »>e added
or aubtraoted t (i) (3+»- )
(8-1- * )- 15 + 3. (ii) f3— J-)
(8 — ") =18— 1. His answers are (i) 2 and 7. (ii) 6 and 17.
 84. There are two examples in this section of different types

(i) A chariot is drawn by A horses a at a time. How many stages should

tiiere be in u distance f. an<l how many stages should each horse do ?

The argument seems to be the mileage done is la therefore each horse does ;
‘ ’

Ja/A miles the number of stages js A con.sisting of 1 / A miles each and each horse
;

does a stages.^

(ii) A and H start on a journey together. A goes at the rate of r, and

meets B. whose rate is r„ on the return journey. If the distance be*
tween the two places is /, when will they meet ?

Let X be the distance gone by the slower traveller then the other goes 21 -x
and tr, tr,, whence t = 2/;' (r, A proof is given in the form f : r, : :

t : tr, and / ; i\. ; : t

Quadratic eqaatiovs and square-root approximations.^

85. Ostensibly this section deals wdth certain problems in arithmetical pro-
gression that were common property in comparatively early times; but it is of
special interest because it elaborates a square-root rule and gives a number of ex-
amples, some of which involve large numbers. It also employs, although to a very
limited extent, the sexagesimal notation. Of particular interest are the methods
of reconciliation necessitated by employing approximations. There are two types
of arithmetical progression exhibited- -

(i) I)T +• !) .. ((t l)<l-r2 + n)t, where T_is u yiveii period and t is to be determined. The solution
is t- 1 2(1) a) f d.tN/(2(I)-a) + <l)> + snt] -r2«l.

The second typo is the ordinary one of finding the number of terms of an arithmetical progression when
the other elements nro given, tbo itolution being

(ii) t- {
- (2a - a) ± -f-2a.

Ihese two typHs may ronveniently bo reduced to one by the formula t « (p ^ s/q) -r 2d or (pH-q)-r2d,
where (| V U and p^ 2D - (2a - d) or - (2a - d). The negative sign for the root quantity is not used
so t-t (p -Hq)-“2d always. Unless Q i® » sejuare number the solutions given depend upon approximations,
and the most interesting results exhibited in the text are those connected with the proofs or veriHcations
of the approximate answers obtained.

In typo (i) let s' « 1)T + I)t and 8''-.((t I )fJ -i-2 a)t. It approximate values of t are used such
that l)t,, and s/' - ((t, - l)d 2 f a)t, , it is evident that s/ does not equal s^, and the problem of
reconciliation arises.

Now s
'
..l)T + Dt,*DT fl)pd-Dq.)-‘-^J-^(HJl>T + 4Dp + 4Dqj4^8d, s/ ((t. - 1) d/:> + a;t. = (4Dp
-f 4Dq| -p^ + qi*) :-8d, and s/'-s/= q/ (p*-f 8dDT)-«-8d hut p’4*81)dT ; = Q = (^^ therefore —
(q^* q*)-r8d - e'8d (see below).

^ This rxsmpla is evidently tmditional and we find it in Moh&vlra (vi, lS7-lft8) in the following form :

**
It is well known that the homes belonging to the Sun's chariot ore seven. Four horses being harnessed to the voke to draw
They have to do a iotirney of seventy yojanos. How many times are they yoked and unyoked m four T
**
it.

in the same as that given above.

His rule
Approximations are here indicated by aubacripta
> : thus a, represents a first approximation to the value of s, and 9^ represents
so approximation of the second order, etc.
 15

This form of roconcilintion orcurg onW onco, but that about to be explained occurs a number of times.
In typo (ii) l)d/2-»-a)tait*d/2+ (2a - d)t/2* (t^d - pt)/2 and since 2dt-p — q, we have 4d^t^ ;

4dtp— q*-p*, and 8ds«>4d^t*- 4dtp — q*-p*. Therefore s, - s — (qj* - p*)/ 8d - (q* - p^/M— (q|* - q*)/ 8d
>«e/ 8d as before.

The square-root rule coraes in incidentally

a surd quantity occurs. In this way it is actually when
given three time;* (see 1( it was not treated >eparatply and was considered
68 ) end this se»'ms to iiulicafe tbut
as subsidiary^ to the arithmetical proeression problem. To the student of the history of mathematics,
however, it will be probably considered the most interesting part of the work. The rule may be expressed
by^/ A^-f b~ A
-t-b/2A approx imate ly.^ Hut since (A -|-b/2A)* A* 4-b -j- (b/2A)’, the error may be shown
by e.»(b/2A)*, Again e./ 2 ri approximately, where r,^A + b/2A and the
second error may be represented by e,— (u, /2r,)®=* b^/64A*{A + b/2A)*. Thus for a first approximation
s, -s=re,/ 8 d, and for a second approximation t, -SMe^/Sd.

86 The examples of this section are fragmentary but in must cases they are quite unambiguous and
.

can be interpreted with certainty. 1 now give an accurate representation of these examples; but it should
he noted that those portions that I have enclosed in angnlai* erotehetB do not occur in the text as it now
stands.

Rule., t= I 2(li ^ a) +d v (2(D - a) + d)* + 8l)td ! 4- 2d

fi) Eatamfle, D — 6, T— 6, a=*3, d — 4

DT = 30. P n«6-3-2. 2(P a)-.4.

2(D - a ) +d» 8. f2(D - a) + d)* « 64. DT^ 30.

SDT? - 240, 8DTd - 960. ! ‘2(D - n ) ^d !

= t SDTd 960 64

-1024, ^/1024-32. 2(n-a)4-d*8, 8 + 32^40.

< 2 d- 8 and 40.r 8 ^ o 1 > .

Proof by t))e rule of three; 1:5: :6:25 and 25 + 30-^55.

(ii) Example, D — 7, T*5. a =*5, d— 3, ^

DT-.35, D=7, a-.5, < P-a 2, {2(D - a) + d)* =*49 >

8DTd - 840, 840 4 49 - 889. < v^SSO - x/2n» + 48 > » 29^.

_ __
2 (P - a) 4- d
x ji iv
7, 29^^ 4- t — ^ *

5^
— 2130
^'*
1
^
K
rtj
^*1
0 J
9 and
2 IJC
6
2 ^i
4
. V
^ "aiT •®4"84'16
Ql
4*3
mil ®
4-6 2^
^

ii - 1 - 1^'- 1
-W, (t,-l)d/2«^^ (t,-l)d/2^a-^V r>

* The history of this rule is of conNulrrsble interest. It wan f^ivm by Heron (Hkath. f^reek Math, ii, .324) who indioatee ths second

approximation : and it waa known to Planudes (xiiith erntury). Barlaam and Nirholas Hhahdas (xivth rent.) and to al-QalasSdl (xvtii
caii. See WosPt'XK, Jour, Asialiqur, 1864, p. .384). It does not appear in any early Indian work but it is given by Sflry Mlte (xvith
cent.), who aetributea it to his father, Jny&na RAja, Colcbrooke’s translation {Algebra, «c., from (h^ SaMcrii, p. 156) of this Indian
version is as follows The mot of a near si^uarc, with the quotient of the propf>sc<l scpiarc dividiMl by that approximate root, being
halvwl, the moiety is an approximate root ; and repeating the operation sa often as necessary, the nearly exact rentt is found. Example
6. This divided by two which is first put for the root gives 4
for the quotient, which added to the assumed root 2. makes I ; and this

SOryadAsa's approxinia6on may thus be represented by

A 4
.

divided by 2, yields { for the approximate root” loyAna

2
RAJs *<4
work on astronomy is entitled SiddhAnia Sundara. See khe referencje to Si'wdahi on folk) 34 of nor text

n 2
 46

((t,-l)d/2 + a)», '

W* BH
178
to 841
•

8,'..I)T I)t,=7 (f>+ = orttieo

and
24 _ . 48* _ h» .
^
(.
8il 841 ^ 4 « 841 X 84'"7ATBd'

Proof by the ‘
rule of three ^
(iii) Riiltr t.- )- (2a- d)-fv^(2a 8dij 4^2d.

Ejcainplr. u-«l, dn»l, H<tt60.

Solution, 8d8= 480, 2a - d =1 , 480 + 1

481,v 481 ^ j?l42= 455
= 4^ ; <andt,8 880
**
34
'

O 9rtl 8^,320
< Now s
A iA
'^(|+l)/2 V.
^108 14112

^ IGOO 848,320 lom

ie(X) 846.720 _
< But
_ r> A
- 8 = e^ c#
-8dr=
j 1

^
/ Ml
>— 4 j72*
j n.»n,ii-z«»

14,112 ^
t

r4Jl2”‘'^l4112
== do.

^
< a
Ajj'ain
.
. iviiv
sfi ^ .11
20
oj
( 20 21; V 425 042-400
““
424,042
,

’ ^ ^
,
424,842 ,v -
•
2 k2»*V”^ 10,382 19.302 ^ 19.882

*
l(i.VJH0 _ ,
(H,
.

I (1 ; ^ 4- O)^.^
w -
^
--
ixio
1
_ 406,280 ^ 444,004 “ 179,946,941,120 •
38,724 '
) :;8,724 77,448 2,999,690,362

179, 946,941 , 120 - 100,000 _ _

^ - 2,096,61^0,362
179,946 ,781,120
< nut S,
^ lOA
8— e, aa — / ^
lOO.OoO .

2 ,999,096,362
"" ^ 2 ,999,096 ,862
' “ ^
^
^ (21 |V)*

(iv) Example. <ar.7l, d-:^!, 8-5; therefore t«( -l+v/41)/2; >x/41

i» 05 1 > G5
—
___
4 1 66 _ /41 . » 66 ^ 89 ^ 66 6785 » 25 - ^
^
^TTm ^
, 1 \ / I

M
. I j ,
Hilt H, --
,

.74— 1.
I
1 .V| (* (48'*" ““ *^ '
( '
-J 4 2l‘“ ^^fl 52 84

Again a second approximation to is given by — { (TV)V^A==yi^*^“* —

' 1848
"" ^ -i '
1H48' 309C

(t) Emample, <a— IJ, d-*H, s*»2; therefore { — J-f* ^

- 3 + ^^lOo)-*- 6.

Now ^/i05n=:10| appioximaiely, and a second approximation is given by 10^ -

AVith this value t^-C 3 -f 1 / d = ^ll^on, and a,- ((t, - 1) !j- -h a)t, = 1) i + Dtf >
10225 -Mi /K**-2r) ... 10H626
iuiKi',£o .’>9425
.ni^-D vi,9uo,i^,v20
6.465,040,625 IX 625
3280l.>
X I

i
.

f iJls- •})<»- 05(9X1 ^ 4li2(K)~ ii:22'7:520.^ B, -«-6,/8d -8^27.690^01)

. _ r,, I.'> 5 0 10 62:1

, . - i; 2 rf _ i;. 455 0 10, , < 00 _
and S — :i,i.'27,5.0,0(X»
*
" 3,227,520.(XX)

(vi) Example, a -l.^, <1=1^. s -7000; therefore <t- ( 3 + v^33609)/6, and ainc* 679* « 836.241 •

.'1M,G09 - 768, we have as a first approximation to the root quantity 879 while the second approxi-

niatmn
.

is
.*^384 , 384 j 0/^-,i3^S V.
>— ^
r*^\> *
^
^
*.r,7o708 294,912
*^168“* 777.307,600
_ ^ 6,620^)00- 294,9 12
’‘<“(^79 (579* ‘
f,7») *379 579' 671.260 7fV,30T,5OO

“ ^448, 244, 346,088

< m,
51 5,225,^*88 150,576,267,588 150,676,267,588
^ ^ .w.-v

•’'•’777.307.510'= 777.!«»7.r.Oo
•
- /
iherefore , , (
>, » t-^,Y:aiW. 50O
777,307,500
,»\
•*) “ 4 .«e 8 8«.000 .
•

• Thf Ai>]Mnflix lf> IhiH ('liapirr (iiam' O.'l) may hr uxrfnl tti those who are ininexteHl in the details of the oaloulaUcme.

Here iiiHtead of il 1 J ue mnxl uxr 4.«1— 6, for wr havr renlly pven « ami 1/ new values by caneelling, namely a •• S, d > 3
and V ix four tiiiiea f^reater than it should have been (Biiur crTr^q,’— q*) ; otherwwr ainre q*-=26f,
I , r I M‘ «»
ic.Hi* .*4i.*(4.2a)*“ i.s-'r.cjw.iRio*
“^Uvsi* *
i
 47

, _
44«,W0.500j08H , 221,790^S0X)** 10,60S,l>b,OiS
Now 1)1 + 1)*. and ‘,-1- 4/W8.84S^ » *)n-.

l,6IW,815,000 ’ «.-» 2
1

1184*7 AiV.MS U>,75S,3K),782,74<t,743Ji714a8

(t, l)d/2 + « i,i3pi»,ooo' > and finally '
7,256.4HS4»4,«7(>,000,7iOO

7C^_ 31.743471,930 S0,7634H3,703.7X6,000400400

But 8, - e,/8d. . ,
*’'*>‘•1' gl'Pll « 7,000.
7,!W.4«8>»4/i7t>W0.000’ 7,360,483,^076,0(103100

Serif’s.

87. There ia a aet of aeriea that may he ropreaented aa followa, where T,, T„ T,,
etc., represent the aucceaaive terms of any of the aeries irrespective of their form.

T, f T, + T, -+ s

X + 2T, + ;»T, — 4T, -- 200

71
X d + i') 1
2T, t-.]x -
f ‘IT 1 { j
X ,
f rr.t-jx 2

2!»
X (1 + ?,) 1 2T,-Jx • :'T,- ?x •-
4T,-*x 2

X 1 •iT, + :vr. f 4T, 1:12

X (l+j) 1 2T,f,5x 1 :IT,+ |X 4- •»T, 1 Jx Uli

41 ^>
MisBing <X (l+«) 1 2T, ;.x 1 :vr,--.jx + IT, !Jx 401
2

X -h •iT, 1 3(T, fTj) f- 4 (T,+Tf i T,) -- aoo

X (1 t ?)
4- •iT, ^ 'x -t 3(T, t T,)+?x + 4 (T,4-T, |T,)f Jx - 222

X (1+2) •iT,-?x + a (T, 1 T,)-jx } i (T,+T.fT,) -?x = 78

These are obviously built up on a definite plan. There arc three fundamental
series, namely
^
a, ^ 2a, f 3a, + . . . . i na,

a, + 2a, + 3a, + .... + na„ ,

a, + 2a, +3 (a, +a,) i

•
n (a, +a, + a„.,)

and a subsidiary series which may be represented by

0 + e + e(l + d) + e(l +2d) + etc.

is no uitempt at general suiniuution, wbirh, indecMl, is not generally possible, and 1 ain afraid
There
we nan credit the author with little of nmthotnniieal value here. It may he noted that the Biinm of the
series may be giren as lOT. ^3(©-f d), 33Tj -t; (22© + 7d), 60T, i: (2tie + 7d), and that in all raseg e^-T. i,
a m mi. 41 e 21 71 29 1?»6 124 2m il »00 . 144
and« The answers are therefore 26+ or
2 -
.

2 * 2 2’
,

2 ^ 3 ^
Of these
Y miRBing.

88. There are other series of which the most notable is also a double series. It
may be represented by
' -
(
is 2.
O^M:: + (a, . a. ^ a!) ^ (a, . a,. ... a.) }
*'>•' »'
 48

There are other simple series on certnin fragments, that, as they now stand, are not of particular interest.

89. There is a set of some 17 examples that may be represented by

C (l-a,)(l-a,) (T~a.)-R-C^L

where R is the remainder and L the loss ’ after successiye deductions from 0* The yalues of a are
* ’ *

generally simple fractions with unity as numerator, and if (1 - a|) (1 - a.)«B»p/q then Cp/q — . . . .

whence C=»^and Ct*L/(l- **). The following are specimens

60( 1 - 1 )( 1
-

The following questions are somewhat restored.

(i) A trayeller takes on a journey a bottle containing four prasthoi of wine. At the end of each
stage he drinks one pra$tKa and then fills up his bottle with water. How much wine and
how much water will there be in the bottle after four stages

The solution is first giyen by

17 17
4(1 - J)(l - J)(l - J)(l - p^afihas, the amount of wine, and 2^ prasthoM as the amount
of water. Secondly it is worked out by steps.

(ii) A tax i.s paid at the rate of one-third, one-fourth and one-fifth on three separate occasions and
the total amount paid in taxes is twenty-four. What was the original amount.^

The solution given is

X (1 - I) (1 --
i) (1 - 24 and x»40.

This section is interesting’ on account of the numerous proofs or yeriflcations of the calculations,
which haye been referred to in the preyious Chapter ($ 74).

Computation of gold.

90. This topic is treated in most mediseVal mathematical works and our text
follows the usual treatment but with some slight irregularity due, possibly, to a
slight misconception of the author, who apparently looked upon the quality ' or *

*
touch factor as a negative quantity.
’

1 8ee BaiSKiax’s IMAvan^ 73.

• It Is also intermting broause of the oooiiSTenoe of a similar section in a papyrus of Akhmfm (HlATS, QfMk MtUh, Ii, 044 }

hut Heath's rsferaiioe is misleading end so Is Cantor's : the papyrus of Akhmfm was pnbliihed by J. Baillbt, JIf/mpirsf du Cai^
IX. 1892, pp. 1-88).
 49

The tundanientai theorem ia, of course, the usual one of averages, namely ;

f,w.
(i) x= "•1+ »« + •••«.

where w, is the mass of a piece of gold, and f, is its quality or touch.’ That is f is ’

the amount of pure gold per unit in some unknown measures. There is some indi-
cation in the text that f is expressed in mdshakas and w in suvartMts, where 12
mdshakas — X sti'oar7),a but this is not supported by Bh&skara or Mahavlra, and in-
deed the money measure.s in our text are nowhere well defined.

Modifications of (i) that are indicated are

• • • • *•»._ I..

(»)

(iii)
W| + W. 4- X

( f|W ,+ + +•
(iv)
"l+»J+ •
X which is not understood.

Earning and spending.

91. (i) A earns e, in d, days ; B earns e, in d, days. If A gives g to B when will

they have equal amounts ?

t = 2g-(J-^)
(ii) A earns e in d, days and spends / in d, days. How long will his capital C
last?

t = c/{^ <^)

(iii) One gives a, dindras in d, days a second gives a, dindras in d, days ; ; and
so on. In what time will they have given h dindras altogether ?

In one day the total gift is + ^ herefore 6 dindras are

^ q t

given in days and the respective gifts are Ac.

The following occurs

In li days one gives 2i dindras, in days a second gives 3 J dindras, and in

days a third gives 4^ dindras. In what time will they have given .'>00 dindras alto-
gether ?

dindras are given in day, 500 dindras will be given in

Since
60,000
+ y|- + ^ I

Knt\t
500/ i|5
- 63^’ days.

Profit and Loss.

92. C
be the capital, p the profit,
Let = C + p, n = Cc the number of articles, c M
the cost rate and 8 the sale rate [c= a /b where a is the number purchased for b
 50

drammas, Hay, but the money measure is not generally stated ; and 8*= J where c is
the number sold for d.\

The following ruUs are given with illustrative examples.

{})
C- *'
(h) p V

M _ ( +|.
(iii) C-. f » r/-»
(iv) C i-f/i

The ‘
iV *
Section

U3. The exampleh that belong to the so-called section ’ may be desoribed as miscellaneouSi but '
M
most of them are solved by the rule of three terms.* Their chief interest is hardly mathematical although
*

certain of them exhibit in very interesting forms the o ho nge- ratios of certain measures; it lies rather in
what may be termed the social nature of the formal questions (see St 46>52). The translations of the
4|ueBiionR here offered are informal and sometimes very much restored *.
' * *

94. An army cousiMts of chariots, elephants, horse uiid foot in the ratios 1 ; 1 : 5 : 3 ; If a complete
army contains 10 >; of these how luuny ot each kind are there?

The answers are

10 3^-21.870 chariots.

lO'r.- 21,870 elephants.

10*3^*5-. 109,350 foot.

10*3^'3 — 65,610’ horse.

(10* 'r) = 21 8,700 ( 1 akshauhinl)

This the traditional subdivision of a Hindu army. There was probably

is
more information given in the text for the terms akshauhini, anikini, chamu and
pritand occur. Accoixling to the dictionaries 1 akshauhirnii consists of 10 anikinia
or 218,700 in all a chamu consists of 129 chariots, 129 elephants, 2,187 horse, 3,645
;

foot, or 6,090 {i.e., 3 x 43 + 3 x 43 h 3^ 3® x 5 ^ 10 x 3 x 7 x 29) while a pfitanA is said ;

to consist of 243 243 729 1215 =-2430 (or 3' 4- 3^ 3* 4- 5 x 3*- 10 x 3*). See also
• i -i-

Alblrunl s India, chap, xlviii.

95. One nroduces ton and a half in two and one-third days. For the sake of religion he gives thirteen
and one-third in three and one-eighth duvs. He offers to VAsuoeva one quarter less than thirteen in eight
and a half days. Desiring reward in a tuture world he gives to Brfthmans for food one and one-third in
three and nne-fiffli days 2J in 5 days
(
.
) and
. also twelve and a half in thirty-three and one-third
. . . ,

days for the wine for the consuraption of the merchants. In the treasure house is stored twelve
hundred- Say, t) I’andit, how long can this expenditure continue?

”
The daily income is and the daily expenditure is ^4- if ^ The
daily loss
years.
is therefore ~ ^ and J lo
• 1 • • 1200 :x gives the period, and x- x
^ years
^
A proof is given in the following way 2 \ : lOJ :
:
^ x 360 days: x, and Xal782^ which is the
total amount earned. Also 1 day:
^ x 360 : x*,and x* — 2982^. which is the total amount
spent; and 1782 + 1200 2082^.
 51

95. A certain person euros one and a half in one und one-third days and gives away eight in five and
one-third days for Bha(viini), one in thirty-two days for pu(i'Hl(>ka), and as an offering U) Solin one
quarter of two and a half in thirty-six days. If he atrearly possesseH seven hundred when wiil it all be
consumed

The daily income ia and the gross daily expenditure is *1 t" “

J , r* i j V, i vi*

expenditure if. therefore = f,,

and 1 day : : 700 :x, whence = *

JMOB 7 months 2^ days.

A proof is ^iTen thus: — 1 dayjjjj '^^^j*'***

: and this* is the gross expenditure. Again; 1

day :
* : '•
which is the total income; and 700+1869^* =2550j?j. Furtlicr ** yeara :

* *
'
N4 » the daily expenditure.

97. A
boat* goes one-half of a third ol a yojana id us one-thnd less one quarter in oiie-bali of one-tliird
of a day, but then it is driven back by the wind one-iwLlf of one-fifUi of u t^ojann in one-eighth of three
days. In what time will it travel one hundred and eight i/ojtjnoit

The details of the question and the solution aia not clear.

98. A snake eighteen hasUm long enters its hole at the rate of one-half plus one- ninth of that anpulas
lees one twenty-first part daily. In what time will it have completely entered the hole?

Since 24 aAffulas hasfn we have | +4 ^f 1 ““^1 • :: 18x24 :x, und ^ -^=2 years 4 months
10^ days.

99. A snake gets out of its skin as ({uickly an poKsible at the niU* of Imlf an aiigula in one day. Its
length is 100 yajana, 8 kroia, 3 hmfa plus 6 nhgul<m.^ Tn what time will it be free?

\ angula 1 day 100 i/o. +6 kro, ’^3

: : 6: * or :
i nugula \ 1 day : : 77.370,077 any. : x,
and x--r 154,752,164 nays -r 420f?67 years, 1 month and 4 days.

100. On folio 37 are three very simple astronomical problems :

(i) Bhanu (The SunJ travels 600,000,000 yojanas^ in one day. State with i.ertuinty the amount
of progress in one yka(ika ’

Since 60 gk(i\ihi — 1 day (of 24 hours) we have

60 gha, ; 600,000,000 yo, :

1 yha, : x and x = 8,333, 333^ yojann.

(ii) 'Phe Sun’s chariot is guided by the god Maitobaha among the SiddhaA and Vidyadharas. The
clever scientist says that, according to the general rule, it travels half a hundred koti* in a
day and night. Tell me. 0 be.st of cjilonlators. what it v. ill go in one wuhurta?

1 The wonllng of the example is restoved on the banis of (^ortain indications given in tin? Kniiition pri‘Mi»rrn<l, ^. 7 .Vm‘^ mav aland
‘ *
.,
*
for flOLlTf. The numerical portions arc certain.
* MahAvIra (V, 23-31) gives a small section dealing with problem* of foruarfl anj l>ackwar(I niovcmcrit *, which he ilhistrataa •

by a boat problem and a snake problem amongst others. TTis boat proldcm i«

In the course of f of a day a boat traverses i of a hroia of the ocean. Init owing to a ivr'rMc winds it loses i of a krota. .Slat# 4

in what time it will have advanced 90^ ynjana, thou who hast powerful arms in ciossifiv easily the ocean of numbers,
(ifnatesr 3 years 117 daya) :

The snake problem is :

A powerful, un vanquished, exoellent, black snake, .32 htufiaa long ent^irs a hole at the rale of 7) if^fjuha in of a day and In :

the course of } of a day its tail grows by 2J of an angula. O i.'rnamcnt of arithmidiciann, fell m«; by what time it enters
fully into the hole. (Answer : 76$ days.)
* The snake is here nearly 1,000 miles long This may have been a moileat eatimat.<‘ of fVic Hi 7 of the ?v*iT>nnt. Sonha. who
! .«*

supporta the world, causes earthquakes whenever he yawns, and occasionally deairovs the universe. (See the VuiKnv Purd^$n, ii. v.
eta) But Mr. Hargreaves suggcsits that theoxaropis refers to the KAga KUpattra. whose head was at Bcjiare* and whoM^ tail rtstched
—
to Tnsila a distance of just about 100 yojanaa,
* An orbit of A. 10* yajannB or a mean radius of nearly SO million yvjanoM or, say, 727 million raiica What is tbs origin of this T

» A
H
 62

Sin(!A .'jO muhurta ^ 1 day (of 24 hnur«) we kaTe

30 mu : 600.000,000 ; : I mu. ; x and x « 16.666.6661 yojama.

(iii) If Hhanuja (Saturn) traverBoa one sign in two and a half yeara, tell me, thou who art thorough*
Ij learned, how far he will go in one aolar^ day.

Since 1 sign— 30^ i- 108000^ we have

2| years : 108,000^ : : : x and or 2 minutes of arc.*

101. The following examples^ seem to form a set hut the text is mutilated and only the hare skele-
tons of the problems remain

21681
fi) 1 tola : 5^ yc^ars : : 1 to -k- I dkii-\^ I /idr + 1 #/// 4-1 ## 4- J 1 mu : x, or 1 to :
5^ yearw : : to : X
and X — 6 years 8 days.

819|1 81
(ii) 1 to : 35 drammn* : :
\\ to4 1| md »1| an»4U 3^ : x,or 1 ; 35: 192 Ki and X— 68 128
drammoM,

(iii) 1 to : 400 dimna : I t///a4-l adr 4 I //<#4-J /»/4-l /x/d*! mu : x, or 12 dkd : 400 dl : ;

dhd : X, and X-^60^-j;;j udnvi-hi) </? 4 10 ^//d4- 1 oni,

(iv) 1 day : (3 to + 2 mid- 3 anid-S yad-l kad-l pad-l mu) - (4 ra4-4 ei) : ; 26 7^^ 6 months
20 d»y« x,or 1 day : : *01“ - 19^*®^ ; : 9170 day* : x.and x.- toIa=l bki
-f034 pii 4- f> to 4 0 ma
1 f 0 <in/ -f 3 4- 4 ka + I /w + 2 mi.

* S»^ my tfinttv Afiirmuttny, p. ft?.

* It Mhnulil hr About 122^'

’ The imporUnceof th^At* exam pies im I in tlir uss of numeroiw msasurrs and in the methods of expressin|( ohanre-mtios. The
ubjrut of mi^AAiirni ii dinrusard in chapter Ml.
 63

APPENDIX TO CHAPTER VI.

(did* to tA« calchlation/.)

PACTOBS. !>qCABX8.

384 = 2*3 107 5S4 = 2*.41* 21*;r:-441

441 =3*. 7* 108 62.6 = 5*. 11. 79 24*=B76

81 = 13.37 112 003=31.3613 29*= 8 41
92 = 2*3.41 322 752 = 3.2*.41* +1»=]681
576 = 2*3* 335 241 =3*. 193* 77*=.6929
579 = 3.193 .3.36 009 = 3.31.3613 179*=.32041
621=3».23 125 042 = 2.461* 193*= 37249
656=:2*.41 444 004 = 2*11.10091 61* -212.621
737= 11.67 64.6 r>04-2*.3.41* .679*=.3.3f)241

768 = 2*.S 1 654 6)6 = .3*.r>.179.19.3 768*= 589824

841 =29* 4 663 84.6 = 3».5.179.193 .3361 *=11290321

889 = 7.127 7 773 ((75 = 79.193

.3*.r>*.l

1024 = 2** 510 225 088 = 2»..3*.7.15973

1158 = 2.3.193 2 999 096 .3.62 = 2‘.3*.7*. 461*

1681=41* 0 455 042 62.6 = 5*.l 1.79.2377

3696 = 2*.3.7.11 17 994 504 112=2«.ll 17.149.10001

5786 = 5.13.89 21 74.3 271 9.36 = 2*»..3‘=768‘/i*

9681 = 3.7.461 113 227 047 522=2.-3* (6 290 391 .629)=?
9986 = 5.1997 221 790 2.60 ()44 = 2*.3*.11.560,076,.389
10225 = 6*409 347 892 .350 976 = 2**.3*= 768*
11858 = 2.7*.ll* 7 250 483 .394 676 = .3*.5*.l 79*.l93^
14112=2‘.3*.7* 60 753 383 762 72 r) = .3*.5*.7.1 79".! 9.3*

19362=2.3.7.461 448 244 .345 088 = 2»..3* (194 5.60 497)

38724=2*.3.7.461 50 758 883 762 746 743 271 930 = 113 227 047 .622x448 244 345 088,
40528 = 2M 7.149 i =2».8*. (0, 290, 391, 529) (194, 550, 497)

59426 = 6*.2377 = 2».8« (2'»+3..6»'.7.l79*.193*)

66593=11.67.89
67125 = .3.5*.179
84672 -=2*.8*.7* UnramUad 6 290 391 529 luxl 194 550 497

84832 = 2*11.241

Powers of 2, 3 and 5

2*=10 2*=82 2«=64 2»=128 2*=256 £*=512 2'«=1024 2*» = 268 435 466 2‘*=107S 741 824 2**
4294 967 296

3*=81 .3*=248 8*=729 S*=2187; 6*=625 5*=S125 5* = 1562B

Prime niimbera greater than 100

127 149 179 193 24l 409 461 1997 2877 8881 8618 10091 16978

H2
 54i

CHAPTER VII.

Mgasuiuss.

102. The measures exhibited in the manuscript are of rather special interest.
As a whole they are Indian and the terminology is Sanskrit but there are some ;

Sauskntised western terms such as Uptd, dramma, dindra, aatem employed. Most
of the term.H arc well defined but the values of some are doubtful. Money measures^
however, are, as in mo.st early Indian works, very ill defined and hardly show any
differentiation from measures of weight.

103. Change rat ins. The change ratios are often given with considerable care
and elaboration, and are expressed in several different ways. The change ratio
apnears to be considered as a divisor for it is most frequently marked by the term
rJiliedam] which indicates the operation of division.

Examples are

chhe'^ 80 rakti^-su''' i.e. 80 raktika - suvaana^

chhe^ 24 am°-ha‘^ i.e. 24 nhgula=^\ hastn

chhe''^ 2 <jha^-mn^ i.e. 2 yhafika 1 muhurta

chhe'^ 4608000 yai^-yo'^ i.o. 4,608,000 yava I yojana

UrdhaX chchheP 768000 a°-yo°, i.e. 768,000 anguUi=\ yojana, and in this particular
example the operation of multiplication is to be performed.

Urdha chchhedam 108000 viliptdndm rnH, i.e. 108000 vilipta ==1 r&H and multiplica'
tion is indicated.

adha chchhedam 2000 pa^-hhCi°, i.e. 2000 pala I bhdra and division is indicated.

Another form is illustrated in the following examples

tolendtti dhdne 12, i.e. 1 tola — \2 dhdna.

dh&nendsti aifi° 4, i.e. 1 dhann^andikyi,

din&randati dhdne 12, i.e. 1 dindra — dhdna

* By * change roHa T mmxk ft namber by which one rlcnotninfttion it multiplied or divided to ehftnge it to ftnother.
Thus, since SO
•bOlinipi-* 1 pound. SO Is the eMange mHo between pounds and sbillinffs. It is ft mnftiptier or a dtWsor toeording to the dimtlon o# change.
t Often ftbbreviftted ehke. T Or ‘ division by 80 rhnnges rahUkatt to sirnuryias.*

t Diet, UrdhfHi.
 66

i04. In tabulated examples the methods of expressing the change ratios are
modified. In two cases sets of cumulative ratios are given thus

I
to° 1 dhfi® 1 1* am° 1 1* ra° 1 I* ya® 1 1* si^ I 1*

1 1 12 1 48 1 60 1 192 1 480

ka® 1 1* pft® 1 V
which means primarily
1 1200 1 4800

1 to° + 1 dha° + 1 am° + 1 ra° + 1 ya° 1 si° + 1 ha° + 1 j>a°

while the other numbers are simply explanatory and mean that 1 dhii° = j| 1

am®- i to° 1 pd® « —)

A similar table is given without naming the terms in the form

which means

li to° + dhd° + Ij am® + 1^ ya® and that 1 dhCi^- to ‘,

1 am'- ^ to^, 1 ya® -
to°.

Both of the sets given above are also tabulated in a different manner, e.g.

to 1

This again primarily means 1 + 1 dhu i 1 am + 1 /a i-

ya + 1 s* + 1 ka + \ pd and the other numbers are simply

change ratios and indicate that 12 dh§==l to®, 4 aih -

dhft, H ra = l am, 3-j ya = l ra, 2i si =l ya, 2^ ka-1 si.

Note that where the change ratios are fractional they

are written at the side, obviously to prevent confusion.

Examples of the type just given occur on folios 7 recto, 48 recto 48 verso, 49
verso.
 66

105. A variation of the above scheme in which the charge ratios are halved
occurs thus
to

2 which means ma°+ a«ii® + liya®.

1
1*
1 mft The notion of halving the change ratios seems
6
2

1 * t-o be like this : J

to° multiplied by to®-
1
1 am
2
2
3 ma and so on. Another example of this prac-
1
1
2 tice occurs on folio 20 recto and gave consider-

able trouble It is written thus and means ^ a + ^ i b + .

i c. Until the change

ratios between a, b and c

could be found the problem was insoluble. The real change ratios are given by
a=b, b’^c.

106. Before proceeding to discuss the several kinds of measures that occur a
summary table of them is given for reference. The names of the measures are
nearly always abbreviated in the manner indicated the full
terms are :

Adhaka Guih Gufija Pra - Praatha

AihAa Ha Haata Pra var Praafiti

Aih

Aih ApdiM KA « KAkini Ra - RaktikA

Alii — AAfCula Ka » KalA RA - RAAi

Bh& — BhAra KbA ^ KbAri Sa - Satera

DhAnakA Kro Krosa Si - SiddbAriba

Dh& rrr

Dha MS Dhanufi Kn -- Kiidava Su - Sovar^

Di Dina U lipta To - Tola

r» DfnAra MA MAaa Va - Varaha

Dram I>ramma MA MAahaka Vi “ ViNtfiU

Dra rw DraAkiibana Ma rr. MOdriWA Ya - Yava

Dro Drona Mu « MuhOrta Yo - Yojana

Oav « QavyQti PA PAda

Qba Ohatika Pa «= Pala

 '

The
Time..

gba mu cU mA va

gha 1 »24 min.

mu
i
2 1

h
di 60 30 1 •24 hr.
aip
mA 30 1

rA
a 360 12 1

mfi pA ka li

Length.

ya aA ha dha kro ga 70

4.0(H) 2.048 1.024 208 04 16 I

 68

Measures of Time.

107. The ineaMurea of time used in the text are the udual practical measures
employed in India and there is nothing in any way remarkable about them. The
measures found in eaily Hindu texts are given below.* All the authorities agree
oil one point, namely in dividing the day (24 hours) into 30 muhurtas. The half-
rnuhurfu was possibly introduced for astronomical purposes, as was the vinadi of
the Surya SiddhO nta. ALso it is noteworthy that the tables give neither hours nor
’
weeks, tlthough he hour was used for astrological purposes and the week
t
‘
'
'

came into general use in the early centuries of the present era. Neither of these
measures occurs in our manu.script but in this matter the text is quite orthodox.

The measures of time employed therefore call for little comment, but Dr.
Hoernle suggested^ that the year of 360 days might be an indication of the age of
the work He could hardly have been aware that it was the common practice to
!

give this value in Hindu arithmetical text-books.® ftrldhara, indeed, remarks ;

**
Time is calculated according to this rule in all mathematical works.”

Measures of Arc.

The measures of arc. which with the early Hindus were lengths rather than
angles,* are such as occur in all mediaival Hindu astronomies. The term liptd is a
Sanskritised form of Xttrrrf.

^
In tbc Hoctiona that follow f h«Te uaed with ooriBiderable freedom the chapter on 'Weight* and Meaaurec * given inL-Ih
M a'k .4n/ii7Mt(tV4 of India, which le the moet ueeful oontribution to the eubjeot (of thoee at my command) ainoe that given in
I Ki'Nl’a f lulia,

"
IriiJiaA Antiquary* xvli, 1888* p. 37.

See Mahhvira and Aridhara.

* my Hindu Aetronomy, p. 54.
 59

Hindu Measure* of Time.

300 la I
 00

Measures of Length.

108. (a) Although individual measuree of leng^ occur pretty oftw there are
few examples in which the change ratios are exhibited, '^e following (fol. 32,
verso) is the most complete of such examples :

100 urdha chchhe 768000* a®-yo°

6 1
8*
3
4000*
5
24*

which means 100 yojanus + ^ {kroias)^'^ hastas + S ahgulas, the numbers marked
with asterisks being change ratios. These numbers inform us that 24 afigulas = 1
hasta, 4000 hastas^i kro.^a, 8 kronas = 1 yojana, and that to reduce yo;ana« to
ahgulas the former must be multiplied by 768,000. The term kroia does not ac-
tually occur, but it, or an equivalent, must have been mentioned in the question to
which the table was attached.

In another statement (fol. 32, recto) the equation 8,000 dhanus^l yojana is
given, hence 1,000 dhanus-\ kroia. This equation occurs in Hindu works but
most of the texts give 2,000 dhanus = 1 kroia.

In another example (fol. 36) the statement

chhe 4,608,000 - yo^

indicates that 4,608,000 yava8 = l yojana, whereas Mahfivlra and BhSskara give
6,144,000 yavas~\ yojana. The difference in the change ratios is noteworthy.

Tlie inea.sureB gadyuti and yojana are connected (fol. 12), but unfortunately
the actual change ratio is missing. The term gavyuti is not common in Hindu
works but it occurs in the Markavdeya Purdna, the Mahdhhdrata, etc. Its value
is variously given as 4000, or 2000, or 200 dhanus.

Of the Indian tables given below the last is what the SUrya Siddhdnta
designates a-mUrta or unreal.’ It was never used seriously but it is interesting
'

as exhibiting the constant change ratio of ' eight.’

108. Possibly a measure of area is referred to on folio 39* verso where a

(b)
change ratio of 64 is mentioned and possibly hhu on fol. 26 also indicates an area.
;
 61

Indi«n Metsuret of Length.

M^htn^ya PurA^. Kau^itltfa Artfui SAMra.

Ya. AIl Pa. Vi. Ha. Oa. N4. Ga. Yo.

Ya. Ati. Vi. Ara. Da. Uo. You

. 1
AtiguU *
. 8 1

Pad* 8 1 Y ava* . 1

VttMii 2 1 Angula . H 1

HmIa 2 1 V'itaHti . 00 12 1

4 1 Aratiii 24 2 1

Ni4i . 2 1 Daikda 00 8 4 1

GATyfltl 4.000 2,000 1 Goruta 4,000 1,000 1

Yojaiia 10,000 N.OOO 4 1 Yojatia 10,000 4,1X10 4 1

Sridtiara. AfffAdHrn.

Aiu Ha. Da. Kra Yo. •Sr. Yu. Ah. Pa. Vi. Ha. Da. Kro. Yo.

Ahgula 1 .Seaamum* . )

Yava H 1

Haata 24 1 A^ICiila K 1

Pada . 48 6 1

DaxKla 06 4 1 Vitaati 06 2 1

Haiita 192 24 4 2 1

Kro4a . 102,000 8,000 2,000 1 Daiirla 768 96 16 8 4 1

Kro4a 1,.536,000 192,000 32,000 KMNK) 8.000 2.000 1

Yojana . 708,00032.000 8,000 4 1 Yojana 6.144,000 768,000 128,000 64,000 32.000 8.000 4 1

BkAakara. Atniria (yn-rral) Tahir

Ya. Ail. Ha. D,. Kro. Yo. Paraminua 1

ParaaQkhamas 8 1

Yava . 1 Trasarei^iu 64 8 1

Ahgula . 8 1 Renu .512 64 8 1

HaaU 102 24 1 VAligra . 4,096 612 64 8 1

DapdA . 768 06 4 1 LdkahA . 32.768 4.096 612 64 8 1

Krola 1,636,000 102.000 8.000 2.000 1 Vak* . 8» 32,768 4.096 612 64 8 1

4 Ya¥a 8» 8* 32,768 4.096 612 64 8
Y<daaa . 6,144,000 768.000 32,000 8.000 1 1

BakJk^hdli.

Ya Ah Ha D'a Kro (la Yo

Ya 1

Ah 6 1

Ha 144 24 1

Dha 676 06 4 1

Kro 676,000 06,000 4,000 1,000 1

Qa 2,304,000 384.000 10,000 4,000 8 1

Yo 4,608.000 768,000 32,000 H.OOO 2 2 1

* Thcae tebiM are Msoompaiiioil by th* AmMrUt t*bla.

I 2
 62

Measures of Capacity.

109. The following is the most complete table given (fol. 13).

4* dro®
pra° 0
4* a® pra°
ku® 2
4* prasthi

which means 2 &dhakas^ 0 prasthas ^^2 kutlavas. The figures here marked with
asterisks are change-ratios and indicate that 4 a<j[hakaB = l droua

4 prasthas = 1 adhaka

4 kudavas ^l prastha.

The terms ghataka and pala are mentioned together (fol. 15), where possibly
ghat^aka may be equivalent to drona an<l in another place (fol. 53) ghataka is
;

obviously used as a capacity ternr

A comparison with the Hindu measures shows that the Bakhshftll measures of
capacity are strictly orthodox and it rna y be noted that the Hindu measures of
;

capacity are generally more consistent in the matter of change-ratios than any
other Hindu measures.
 63

Hindu Measures of Capacity.

Aihann Vtda. Kau$iliifa Artha ^Mra,

Ku. Pra. Atilt. Dra Kli;!l. Klim. Va.

Kn. Mi. Pa. i*ra. A^h. Dro.

Ki^jluinba . 1

Kmh^iaU . 1 Prattba 4 1

Mtehftka . A 1 A^baka 16 4 i

P*la . ft4 1 Dro^ 64 16 4 1

PrMtb* 32 1 Kbkrt l,C)24 256 04 10 1

1-28 4 1 Kumbba 2,048 5.120 1.280 .320 20 1

Dro^a 612 16 4 1 Vaha 20,480 51.200 12.800 3,200 21K> 10 1

Vnrdha Mihira- Dry Meaauro. ynrdfia Mihira Lirjuttl Moaaiirc.

Pa. Ku Pra. Adb. Pii Kii. Pra. Aiilh.

Pala . 1 Pala . . 1

Kudava 4 1 Kudava . K I

PraatUa 16. 4 1 Praatha . 32 4 1

Adbaka 04 16 4 1 Adbaka 128 10 4 1

MahArAra, HnkK^hdU,

Sho. Ku. Pra. Adh. Dro. M6. KhO. Pra. Ku. Pa I'ra Ku. Pra. Adh. Dro. Klia.

Pa. 1

Shodairfka . 1
Pra. o I

Kudaha 4 1

4 Kt!. 4 V
Praatlia 16 1 1

Adbaka 64 16 4 1
Pra. 32 8 4 1

Dro^ 256 64 16 4 1

Mini . 256 64 16 4 1 Adh. 128 .32 16 4 1

Kbkri 1,024 256 64 16 4 1

l>ro. 2.'»6 128 64 16 4 1

Pravaitikt 5 1

Kumbha • 5 1 Khft. 4,006 2.048 1.024 206 64 16 1

'
 64

Money mecuuree.
1 10. Although the terms dramma and dindra occur pretty often thep rela-
tionship with each other is nowhere indicated and, indeed, there is very little de- ;

finite information in our manuscript about money beyond the mere money names.
In early times in India there were no special measures for money beyond the weight
measures for diffeient metals, and sometimes the difference in these is more appa-
rent than real. Thus Mahavira’s gold and silver tables given below have much in
common.
In our manuscript we have only the following scraps of information
(«)' chht 80 rakti° -su°
chhr
{b)
(c) I
1
1
I

2
1
9
11 1 1 1 ( 108 pha° di° 1 dha^ 8 am° 1.

\
.

_ 2 2 _ 2_J
Of these examples (a) is orthodox, for according to Manu 80 raktikas or gunjas of
copper or gold are equal to one Huvariia. Example (6) is isolated and cannot be
interpreted with any certainty. Possibly gu^ stands for guhja, and possibly
for valla Kxample (c) means 1 (i)‘ ;108 I d*“ + 8 dhd 1 am®, where
.
:
= : : t

dimira, dhd -dhanaka, and am -arh^a and it is implied that 4 omias — 1 dhdnakd, ;

and 12 dhanakds 1 dindra (fol. 33) and the .same relationship is given on folio 49
recto. The term kdkini (cowne) occurs once satera also occurs (fol. 34) but with ;

doubtful import.

The Guptas adopted the term dindra* from the Kushhus together with the
coin of weight from 118 to 122 grains and in a number of Gupta inscriptions certain
;

gold coins arc termed dmaras* But the dindra was not invariably a gold coin*
and in the Bakhshall text it probably is a copper coin, for a day’s wages is stated to
be from 11 to 3 dindras (fol. 60) and also according to Mahftvlra (vi, 231) ;

wages work out about 18 dlndras a day per coolie.’ The term occasionally occurs
in Sanskrit works of a more literary character and has provoked some discussion.*

The term dramma (SpaX^v) also occurs in various Indian inscriptions' and
in Indian mathematical works. Bhaskara makes it one-sixteenth of a niahka* and
in our text wages are about 1 dramma a day.

Such are the facts but Dr. Hoernle writes*

The way in which the two terms are used in the Bakhshall arithmetic seems
to indicate that the gold dindra and the silver dramma formed the
ordinary currency of the day. This circumstance again points to
some tiine within the first three centuries of the Christian era as the
date of its composition.!’

The subject monetary measures is thus treated in a vague and unsatisfac-

of
tory nranner in what remains of the manuscript but this is by no means a pecu- ;

liarity almost a characteristic of Sanskrit texts of mediaeval times. The

: it is
connexion between measures of money and weight was then fairly intimate and
they can be hanlly considered apart.

• (a) W) rnktikAii^y Muvarna. (6) 9 -1 m‘ (c) 108/64

.
(iknarn -f 8 dAdiwiAxStf -f I atfiF,
I

• Ok. 8t)v<ipiov. Lai. dtnatinr,

• J. Allan, VnUdogMf. of the Coins of the Gupta Dynasty, etc., p. oxxxiv J, F. Flxbt, CI/, iii, not. 5, 7, 8, 9, 62, 64.
;

• Comparo our penny and read Sir A. STxm note on the Kashmir dindra in hia Kfdhafui*s Rdjatara^f^i ii, 308ff. See also seo.
133 below.
• We not, however, place very much reliance on these text-book wages.
must
• JR AS,1907, pp. 408 and 681. etc.
^ RpigrajJiia IndUa, I, 167 ; P. VoonL, ChamJhA Inscriptions. 204, etc.
• IMvaR, sec. 2.
Ind, Ant. xvii, 1888. p. 37,
 66

Hindu Measures of Money.

 66

Measures of Weight.

111. Besides the examples already given (sec. 104), the following are note-
worthy, :

(a) dha^ 1

am*^ 1
4*
ra^ li* bha®
ya*^ 3 * bha®
ka*^ 6i* bha® 1

(h) 3
to®
ma° 2
12*
am® 3
4*
ya® 3
4*
ka® 1 #
6i
pa® 1
4*

<c) ra'^ 1
ya° 1 3! * bhft°
ka° 1 6i* bha°
pa® I
4*
6e®
K
<d) chhe° 8* (e) 216 bhft®
dh?l° 2 270 pa°
chhe'^ 12* 2000* chhe®
guTh° 3 6 to®
chhe® 5* ,
8* chhe®
ya° 2 3** bhn" 8 dhft®
12* chhe®

These measures are combined in the following table.

Md'’. Pk. Ka^ Ya®. Rn®. An®. Dhl®. To®.

Pida . 4 1

KaU . Id 4 1

sr . 40 10 2* 1
Tav« 100
320
25
80 20
Ol n
8
1

Raktikt 1
400 100 25 10 4 U 1

T)lilnaka 1.000 400 100 40 15 S 4 1

Tola \ 19.200 4,800 1,200 480 109 60 4H IS I
Pala . 153,000 38,400 9,500 3.840 1.585 460 384 06 6

(•)
r«*4-l
In sn the m marknd with nstaridi* (not, ol oonxM, in tho orisfauJ). Th* tnUe 1 dMP + 1 «h°.f I

{*) 3 faP 4 2 <rta*-*.8 ya^+l Iw® M pS^.

(e) 1 f»* +1 jrt“+l hi®+l* ja“

(<() 2 MiP-ii gtimP+S pa*’.

(«) Sia 6*a*+27« |irt® 4-5 fcr>+8 <(»4®

 67

To this we should add 12 m&shas - 1 tola, 5 tolas = 1 suvarvu and 2000 pala •= 1
bhdra.

This aigrees with the Hindu tables generally but more particularly with that
of Varfthamihira as given by Albiruni, which, however, contains neither si° nor
raktikd — hoih. interpolations here, as their fractional change ratios show. Obvi-
ously Albiruni's mdri and our mudrika are identical the term occurs in no other
:

Hindu work known ito me. The remainder of the table dhdnakaa to 6Adro5- does —
not seem so orthodox. Compare with Mahavlra. As stated before there is some
uncertainty as to monetary mea.sures in our text and it is possible that some of the
measures included in the above table should be treated as such.

The combination that occur in the text are as follows :

mft, su, to fol. 11

ya, ra/ am, dh& ... 49

mh, pa, ka, ya, ra, am, ma, to 49

dhft, to, pa, bhS . . . .

9f 48

dha, gum, ya ... 1 * 48

mQ, pfi, ka, si, ya, ra, am, dha, to 55

ya, am, ma, to ... . 55

Possiblv the term dra° should be included in the above table. Tt occurs only
once (fol. 2b) in the phrase chhedarh 6 dka'^ dra’^ which means 6dhd — ldTa°,
and this must stand for 6 dhdnakds = 1 drank-oharfa for Albiruni gives 6 vulsha - 1
,

draiikshana; And Mahavlra gives 6 hhaffn^A drankshvna. and dhnvaka, mdsha

and bkdga are synonymous terms as applied to measures of weight.

The term occurs thrice and the values given are 40

si^ -- 1 tnd°,
2^ si^ = 1 ya°
and 480 st°=^l — all of which are in agreement. The abbreviation .<a'° probably
stands for siddhdrtha. Connected with it there is another measure kti^ also un-
identified. The statement in which both these occur is as follows :

I ku° ebbe° 128 ma'^-ku° ^ mg,° cbhe^ 40 8i^- ma | sa^ 55 |

Here we have 128 md° = \ kv^ which so far has puzzled me. Sn° is an abbreviation
for satera; but its connexion with other measures is not clear. Mahavtra makes 2
dindra \= satera. The equation 8 suvarya -1 kii° can be deduced, but is not help-
ful. ^dna (Diet. -4 inRshakas) occurs.

The measures are generally expressed by the abbreviations mv. pd, ka, si, ya,
ra or gum, arh., dhd or md, to, pa., hhd; but occasionally the full terms are employed
and in one place (fol. 49) a fairly complete .set of terms is given, namely

dhdnakd aifidika raktikd yavd kaln pdda mudrika

 68

Hindu Meaturet of Weight

7. J7a»« amd i'dj^valkya, 2nd — 4lk Ctntury A.D. 2. Nintda, Ukr-Mt OMtt^ A.D.

Ya. fU. Mi. K*. Pa.DlM. KA. MA. Apk Dhik Dl.

Trasa^rapa 1 Kikai^ 1
Misha 4 1
LikabA 8 1 A^^ikt 80 80 1
Dhinaha 320 80 14 1
RAja-sarahapa « 24 3 1 DinAra 3,840 080 48 12 1

Qaura-iarahapa 72 0 3 1

Yava 6 1

RaktikA . 18 :t 1

KauHUyon
Miaha 16 6 1-
MA. Quo. MA, Sn. Pa.
Karaha 240 80 16 1

MAaha 1
Pala eoo 320 64 4 1 Gunja 2 1
MAahaka 10 6 1
Dbararia . 9,000 3,200 640 40 lO* 1 Suvarpa . 180 80 18 1
PaU . . 040 320 84 4 1

Tbo BfibaapAii Smpti

(a) gives 4 karahai***! <]lh&naka and 12
dhinakai^l dinAra.

3. VafAha-mihira , 6th Cfntury, 4, MahAvira, 9th Cenhtry,

{Mint Albiriint.) 1

PA. Ka. Ya, Aiha RhAi Dra, DL 8a,

Md. PA. Ka. Ya. Ap. MA. Su.
I’Ada . 1
Mdri . 1 KalA . 4 1
PA<jfi . 4 1 Yava 28 oi 1
KalA . 10 4 1 100 28 4 1
Yava 1(X) 26 6J 1 Bkftga 400 100 16 4 1
Aii<^f . . 400 100 28 4 1 Drakaliupa 2,400 000 00 24 0 1
MAaha , 1,000 400 100 10 4 1 DinAra 4,800 1,200 192 48 12 2 1
Suvarna . 28,600 0,400 1,600 280 04 16 1 Batera 9.G00 2.40(> 384 90 24 4 2 1

BakhshdH,
 69

CHAPTER VIII.

The sources.
112. In considering the question of exotic influence the period of conroosition
of the work is of importance. We
must forget for the time being Dr. Hoernle's
that the work was written in the early centuries of our era and bear in
mind that it possibly belongs to a much later period,* by which time western mathe-
matics, although almost all traceable to Greek sources, had assumed new forms and
had included some new notions. On the whole western medieval mathematics
tended to become less rigorous and more mixed than the mathematics of classical
Greece. Practical calculation (logistic) altogether supplanted the earlier pure
arithmetic, mensuration took the place of pure geometry, and algebra slightly
developed. Such changes are, indeed, indicated in the later Alexandrian works,
and what we sometimes term degeneration had already set in there. The intro-
duction of a place- value arithmetical notation made calculation easier and more
popular, and more intricate arithmetical problems than those, for example, exhibited
in the Greek Anthology, appeared in the later mediajval text-books. The general
body of popular mathematical knowledge became more diffused. (Identical pi'o-
blems occur in Chinese, Indian, Arabic and European text-books of a comparatively
early period). Indeed the mcdiicval mat hematical works of Asia and Europe had so
much in common that at first it seems almost impossible to pick out that which is
definitely western or eastern in origin. In this connexion it should be remembered
that the early Hindu astronomers (Aryabhata and Varaha Mihira) were among
the first to exploit Greek mathematical learning that later the Arabs, after sam-
;

pling Indian works, turned to those of Greece and that it was from the Arabs that
;

Europe received once more the learning it had previously rejected.

113. Such facts indicate to some extent the difficulties of making a direct
comparison of, say, a twelfth century work with those of the classical Greek period.
A knowledge of the development or degradation of mathematics during the inter-
vening period is obviously demanded, and without such knowledge sound judg-
ment IS impossible. That knowledge must be sought elsewhere, but I ^ve here a
summary cnronological table of the period, that may serve to recall the chief mathe-
matical writers and their works.
A.D.

6th century Hypatia *d. 416

Prod UR 410 485

Boethius b. 470

Aryabka\a h. 47G

6th century

Eutocius

DamasciuR “i

visited PeiRio 629

Simplicius )

Dominus
Chang ch’ iu-chien 650

VardAamhira d. 687

Isidore of fievillo 570-006

^ Bpeoifin reasons for arriving at the later date are given in the next eliafiter.
 70

7th ceaturj Brahmagupta b.

Fail ul Alexandria 640/1

i*uj>yru5 of Akhtnin

Bede b. 786
8th CAyninry Muhammad b. M&sa
Aicuin d. 804
9th century IB aka vita

Tabit b. Qorra 886.401

lOth century al- BattOni

Avicenna
Pope Sylvester ii (Gerbert) d. 1003
11th century Albiraiil visited India lOlT-lOSO*

Srldhara b. 091
ubKtrkhI
Psellus

Omar Khayyam b. 1046

l2th century A deland at Cordova 1120
Bkaskara b. 1114
Leonardo b. 1176

Although the manuscript was found at Bakhshftll it cannot be «.a« nTnf>d

114.
that it originated there. The evidence of the script, however, limits the area of
origin to the neighbourhood of Gandhara. Roughly the d&rada script area is limit-
ed by longitudes 72 and 78 east of Greenwich and north latitudes 32 end 36. This
area includes, besides Gandhara, Kashmir, Kahgra and territories near by.
Vogel enumerates six inscriptions from Gandhftra, some of which belong to the Swat
valley. The evidence therefore does not militate against Bakhshftll itself as the
actual place of origin.

ll.'j. The position is peculiarly interesting with reference to the routes of trans-

mission of knowledge and the consequent liability to outside influence. Between

the fifth century B.C. and the fifth century A.D, the country round BakhsJOtSlI was
actually under the dominion of seven different nations the Persians, the Macedoni- :

ans, the Mauryas, the Bactrian Greeks, the Scythians, the Parthians, the Kush-^
ans “and" writes Sir John Marshall,* " it may be taken for granted that, with the
:

exception of the Macedonians whose conquest was merely transitory, each of these
nations in turn left .some impress on the arts and culture of the country." From
the sixth to the tenth centuries of our era Gandhara was more or less subject to
the Guptas and their successors and then came the Muhammadan invaders. The
following are some notable dates relating to the country with which we are con-
cerned.

B.r. Alwxnnder receives the suhTniseion of Ambhi, kihg of Tazila.

\^i) Demetrius of Baoiria conquers tbe Punjab.
8.'S-r»0 Matios, the Rrvthian, conquers Tnzila.

^ A OuiiMio TmHlat fw fS,

 71

A.D. 20 circa Gondopkarnet tha Parthian ruling over Kabul, Tanila and Arachoiiia.

60 Harmnua and Kujula Kadphiaea annex GandhAra.

160 Kaniahka tha KuahAn, king of (jk^ndhilru.

S19 Tha Gupta ara begina.

400 Fa-hian antarad India bj way of GandhAin.

400-500 InTaaion of Indio bj the Epthalitiaa or Whita Hum.

500 Sung Tun in Qandh&ra.

630 circa Hiuen Thaang in GandhAra.

720 Kaahmir aubjact to China.

763 Ou-K*ong viaitad Qandh&ra.

1001 Mahmud of OhaanI defeaU Jaipal near Peahawar.

1175 Muhammad Ghorl attackH Multan.

116 The achievements of the Greeks in mathematics and art form the most
.

wonderful chapters in the history of civilisation, and these achievements are the
admiration of western scholai's.* It is therefore natural that the westoi-n investi-
gator into the history of knowledge should seek for traces of Greek influence. in
later manifestations of art and mathe natics in particular. The position of Bakh-
shall in the heart of Gaiidhfira, and the political history of that country are such,
that not only warrant the search for traces of Greek influence, but make it practi-
cally imperative. Indeed the neglect of such an enquiry would stamp any investi-
gation of this kind as incomplete.

Evidence of Greek influence in the realm of art has been discovered in profu-
sion in Gandhara and the surrounding country. Sir John Marshall wntes* “The
monuments and antiquities that have recently been recbvered from the soil at
Taxila and other places, all consistently bear witness to the strong hold which
Hellenistic art took upon this part of India. This hold was so strong, that long
after the Greek kingdoms of the Punjab had passed away, even after the Scythians
and Parthians, who overthrew the Greeks, hhd themselves been supplanted by the
Kushans, Greek art still remained paramount in the North West, and continued
to exercise considerable influence until the fifth century of our era, although it was
growing more and more decadent year by year.”

It would not therefore be unreasonable to imagine that in mathematics also

there was atleast a possibility of Greek influence in the same country and at the
same period and Sir John himself supplies a sort of connecting link between art
;

and mathematics. He continues" “ This persistence and this slow decadence of

Greek ideas is best illustrated by the coins, the stylistic history of which is singu-
larly lucid and coherent. In the earliest examples ev€«ry feature is Hellenistic.
The standard weight of the coins is the standard established by Athens the legends :

are in Greek Later on, when the Greek power in India became consolidated,
the old Attic standard gave place to one, possibly based on Persian coins^e .

bilingual legends were substituted for the Greek ; and little by little the other
Greek qualities gradually faded .

^ The modem Tndian doee not oeem to be ettreoted In tide wnr.

* A fe Taxila, p. 5M.
• Ih, p. 27.
 72

The caee for mathematics is almost exactly parallel, allowance, of course, being
anade for the more abstract nature of the subject and thte history of mathematics ;

gives abundant illustration of the same types of change as Sir John traces in the
Indo-Greek coinage.

Knough has been

said to show that it would be stupid to exclude the possibility
of Greek influence
in the realm of mathematics on general grounds ; and if such
influence is to be negatived finally it must be for special reasons to be discovered
by subjecting our manuscript to a detailed examination.

117. It is unfortunate that the question of origin of the Bakhshftll Manuscript

was discussed and judged upon before the work was thoroughly examined. The
announcement made by Weber, on the authority of Biihler, at the fifth Interna-
tional Oriental Congress was based upon no real knowledge at all and Dr. Hoernle ;

himself only examined in detail about one-third of the manuscript. We do not,

therefore, feel bound to treat the pronouncements of these eminent scholars on this
particular subject with the reverence that is usually due to them. Also we know
that Weber was misled by Btihler and Dr. Uoemle told me that he was prepared
to modify his earlier views to some extent. With reference to the last point the
quotations given below must therefore not be taken as expressing Dr. Hoernle’s
final views. However, they are the basis of all that has been subsequently written
and it is on them that he position of the BakhshalT Manuscript has been determined
t

by the historians of mathematics. Dr. Hoernle summarised his views in the follow-
ing words :

“I believe that, it w jjyenerrilly admitted that Indian arithmetic* and algebra, at least, are of entirely
native origin. While Siddhfenta writera, like Brahmagupta and hia predecessor AryabhaU» might have
borrowed tneir astronomical elements from the Greeks or from books founded themselves on Greek science,
they took their arithmetic from native Indian sotirces. Of the Jains it is well known that they possess
astronomical liooka of a very ancient type, showing no truces of western or Greek influence. * In India
arithmetic and algebra are usually treated as portions of works on astronomy.* In any case it is impossible
that the Jains should not have possessed their own treatises on arithmetic, when they possessed such on
astronomy. The early Buddhists too, are known to have been proficients in mathematics.* The preva-
lence of Buddhism in North-Western India, in the early centuries of tmr era, is a well known fact. T^t
in early times there were also large Jain communities in those regions, is testified by the remnants of Jain
sculpture found near Mathura and elsewhere. From the fact of the general use of the North-Western
Prakrit (or the Ofitha dialect ’) for literary purposes among the early Buddhists it may reasonably be
*

concluded that its use prevailed also among the Jains, between whom and the Buddhists there was so much
similarity of manners and customs. There is also a diifusednessiu the mode of composition of the Bakh-
sh&li work^ which reminds one of the similar characteristics observed in Buddhist and Jain literature. All
these circumstances put together boom to render it probable that in the Bakhshftli manuscript there has
been preserved to us a fragment of an early Buddhist or Jain work on arithmetic (perhaps a portion of a
lar^r work on astronomy) which raav have been one of the sources from which the later Indian astronomers
took their arithmetical information.’’

118. There is not the slightest evidence in the manuscript itself of its being
connect.ed either with the Jains or Buddhists. It is Hindu (ftaivite). The author
was a Brahman (fol. 50) to Siva is attributed the gift of calculation to the human
;

race (fol. 50) offerings to iSiva are mentioned on more than one occasion (fols. 34,
;

44) references are made to certain incidents recorded and persons named in the
;

Hindu epics (fol. 32, etc.) ;* and there is not a single reference that could be con-
strued as indicating any connexion with Buddhism or Jainism.

^ The reference {a probably to the tSQrjfaprajnajdi, See my Hindu pp. 19-21.

* Not before tbe adwent of weatern aatronomy.
* 1ft Dr. Hoernle here tbinkinf^ of the incident related in the LaUlaviMara and populariaed by Arnold
T The legend, according
to Weber, ** carries no weight whatever.'*
* 1 abould not inyaelf have noted thia aa a characteriatic of the BakhahSlT work.
* And theae referencea occur in the {lortiona of the MS that were not critically examined by Dr. Hoernle.
 73

119.Further there are indicationa of a connexion with Mualim mathemati-

cians ( § and there is other internal evidence that points to a much later date
120)
than Dr. Hoemle’a thesis aUows. It is rather curious that, the connexion betw'een
arithmetical and astronomical works that Dr. Uoernle points out only holds for
those later works which distinctly show western influence. For the major assump-
tion “ that Indian arithmetic and algebra are of entirely native origin " Dr. Hoernle
was not himself responsible. It was a common opinion, that still obtains to some
extent. One point appears at first sight to be in favour of Dr. Hoernle's argument
(but of which he could not have been aware), namely that in some matters of
detail' the Bakhshill work more closely resembles the Oanita-Sdra-saiigraha of
Mahfivira than any other Indian work on mathematics.
120. Although the mediaeval works of East and West have so much in common
yet there are differences even in the later treatment of topics and notions that had
origfinally a common origin. For example the Indians, although they adopted the
western sexagesimal notation for astronomical purposes did not utilise this nota-
tion for purposes of arithmetical calculation. On the other hand the Muslim
mathematicians commonly employed this notation to express ordinary fractional
quantities. Now in the Bakhshall Manuscript is an example of the transforma-
tion of a simple fraction expressed in the ordinary way to the sexagesimal notation.
This transformation may be represented by
^/“-6 + 8' + 16" + 33'"6’'|i

No such example occurs in any early Hindu work and there is not the slightest
doubt that it indicates direct western influence. Indeed our author could have
hardly provided us with a more conclusive piece of evidence.

Again our manuscript exhibits a method for finding approximate roots of surd
quantities that is not Indian. The method may be represented by s/A* + b = A4-
b/2A approximately, and closer approximations may be achieved by continuing the
process. The sutm embodying this method is given three times and a number of
examples of first and second approximate evaluations is given. Indeed this
square-root method is one of the most prominent topics of the work. Its history
is quite well known (See §69). It occurs in many western works from the time
of Heron onwards but it ocours in no Indian work earlier than the twelfth century :

indeed the earliest record of this method fn an Indian work (other than the Bakh-
shall Manuscript) known to me is of the 16th century !

There is an interesting similarity between part of our text and the arithmeti-
calpapyrus of Akhmin (See §89). There are problems of the type of the Epanthem
which can also be traced to a definite Greek source (See p. 40) but in this and ,

other cases it is possible that the problems reached the Bakhshall Manuscript by
way of other Indian works.
121. But, of course, this evidence of western influence does not mean that the
work was not Indian. It is, indeed, almost as Indian as any other mathematic.al
work of the period. It contains references to Hindu mythology and to Hindu
deities and the language is Indian of a sort the script is an off-shoot of the classi-
: :

cal script of northern India the form of presentation is Indian and the material
; ;

of most of the examples is Indian.

The general conclusion is that the work is mainly Indian, but that, as was to
be expected, it shows signs of outside influence, and it gives rather special promi-
nence to the non-Indian material used.

theM hare not tbo tlightesi JaiiiA slgnifioanoo.

 CHAPTER IX.

The Age ue the Manuscript and the Age of the Work.

122. Dr. Hoernle held that the mathematical treatise which is written out in
the so-called Bakhshall Manuscript was considerably older than the manuscript
itself. Indeed he thought that the work was composed about six centuries earlier
than the copy we are considering. He excluded the possibility of our manuscript
being a tran.slaiiou for he largely based his estimate of the age of the work on the
anti«piity of the language employed in the manuscript. If Dr. Hoernle were right
in his differentiation between the age of the work and of the Bakhshall copy then
we should have to consider the possibility of other copies being preserved and the
probability of t he work being known to other mathematical writers of the interven-
ing centuries. Dr. Hoernle, indeeil, did suggest that the work was one of the
sources from which the early Hindu mathematicians drew inspiration, but without
any justification nfortunately Dr. Hoernle's reasons for his views as to the ages
1

of the manuscript and the work are not satisfactory and we are compelled to reject
his conclusions altogether. Of course it will be impossible to say definitely that
the manuscript is the original and only copy of the work but we shall be able to
show that there is no good reason for estimat ing the age of the work as different
from the age of the maiiu.seript to any considerable degree.'

There are certain general causes

for Dr. Hoernle’s mistaken conclusions. He
examined in detail only a (comparatively small portion of the manuscript; the his-
tory of the .'^arada script was not well known thirty years ago there has since that ;

time been light thrown ii[>on the type of language used in the manuscript, and the
knowledge of rnedijcval mathematics has been extended.

It is proposed therefore to re-examine in detail the whole qjuestion of age and

in the course of the re-examination Dr. Hoernle’s arguments will receive due con-
sideration.

The age oe the manuscript.

(a) The circumstances of the find.

123. The eireuinstanees of the find have already been described in detail (63).
They led Biihler and Weber to suggest that the manuscript might prove to be of
the age of Kanishka, i.e., of the second century of our era. need not labour We
this point. The suggestion was based upon a misunderstandi^ and there is not
the slighte.st evidence to support it. It was discarded by Dr. Hoernle, who. how-
ever. follows much the same line of thought in arguing for a slightly later date.
He writes " The country in which BakhshUll lies and which formed part of the
Hindu kingdom of Kabul, was early lost to Hindu civilisation through the con-
quests of the Mvihammadan rulers of Ghazni, and especially through the celebrated
expeditions of Mahmud towards t he end of the tenth and the beginning of the ele-
venth centuries A.D. Tn those troublous times it was a common practice of the
learned Hindus to bury their manuscript treasures. Possibly the Bakhshfill
mannseript tnay be one of these. Tn any case it cannot well be placed much later
than the tenth century A.D. It is quite possible that it may be Somewhat older.”

Dr. Hoernle assumes that the manuscript could not well have been written
after the time of Mahmud. Regarding the alleged burial custom I can say nothing.

1 Thtve 1m that the MS not a ropv at all. Tf 1*0 not t)ia work of a tlngla arrlb# ! tlier^ ara oroM ralaranoaa to loaaao
<rf tha mannanHpt ; tbore ia a raaa of wrongly ntimWin(( a tStra an^ tb# mlataka la noiod in another hand*wrHlng.
 76

It may have been prevalent among learned Hindus

* ’
; but there is not the slight-
est evidence to show that the Bakhshfill manuscript was deliberately buried.

124. (6) The material.

The material on which the w’ork is written is birch-bark, which was the
common writing material for a considerable pciiod of time in Kasluiiir and its
neighbourhood. Unfortunately we have very few birch-bark manuscripts earlier
than the fifteenth century preserved, so it would be rash to fi.\ definitely the earlier
limit of its use. However the earliest known birch-bark n^anuscript belongs to
about the second century of our era and we know t hat this material was in common
use in Kashmir until about the seventeenth century. There will be little danger
in placing our manuscript wdthin the limits here indicated.

But there was fashion in birch-bark manuscripts, and the process of prepara-
tion of the material developed. The strips of bark from which tlic leaves for writ-
ing upon are obtained can generally be split up into a number of laminfc. Some
scribes were content with comparatively coar.se material, obtained by dividing the
original atrip into two similar strips but sometimes the sub-division was curried
;

on much further and very thin strips were obtained, two of wdiich cut to the re-
quired size were pasted together to form a writing leaf. Portions of the Bower
manuscript and all of the Bakh.shrdi manuscript are of the cruder form, while the
Kashmenan Artharvu Veda consists of the more elaborately pieparcd and finer
writing material. There might appear to be little doubt tliat the u.se of the cruder
125.
material denotes an earlier period, but there is a good deal of doubt really and the ;

scarcity of known specimens reduces the value of any criterion based upon this
fashion.

(c) Format.

The shape and size of the bii-ch-bark leaf might also be expected to give
indication of age. The format of the Kharoshthl Dhammapada from Khotan
some126.
was probably due to early western influepce while the Bower manuscript format
was probably due to the Indian palm-leaf potht. The Bakhshall format differs
considerably from both of those and is certainly of a later date. Dr. Hoernle
thought the Bakhshi&ll manuscript was the prototype of the early Indian paper
book, but it might have been the other way round (See §16).

The Script.

script wdll be described in some detail later on and the chronological

The
evidence it gives will be indicated. The earlier orientalists had rather inaccurate
notions about its chronology. Buhler stated that the oldest ftarada inscription
was that of Baijnath, which he dated A. D. 804 instead of A. D. 1204. while Dr.
Hoemle took the ftarada script back to A. D. 500. The main facts relating to the
chronology of the ftarada script are as follows The earliest known examples are of
:

the ninth century and are found on certain coins of the Varma dynasty of Kashmir.
There are at least two inscriptions of the tenth century, namely the Sarahan in-
scription and an inscription of the reign of Oueen Didda. Many dated inscriptions
of the eleventh and twelfth centuries have been preserved. At the beginning of
the thirteenth century is placed the Baijnath inscription, wdth wdiich. according to
Vogel, the history of the Sarada proper comes to an end.
 76

Of the script of the BakhshS.il manuscript Dr. Hoernle first wrote' “ Some
—
of the forms which very frequently occur in the manuscript, especially of vowels,
very closely resemble the forms used in the Adoka and early Gi^ta inscriptionB.’'
The implication here made he slightly modified later* by writing as follows :

“ The SSradS characters used in it exhibit in several respects a rather archaic t3rpe,
and afford some ground for thinking that the manuscript may go back to the 8th
or 9th century.” “ But,” he wisely continues, “ in the present state of our epigra-
phical knowledge,' arguments of this kind are always somewhat hazardous.'*

Every letter of the manuscript has now been examined and the script has been com^
pared with other available examples and the following age criteria have been
applied :

(») Vogel thinks that the form of the letter n is a fairly reliable test of age.
Table I, Part II exhibits three distinct types. The Sar&hap example
has a horizontal connecting stroke in the middle of the letter, the
Baijnath example is without this horizontal stroke but has a tail
turning inwards fi'om the left. The Bakhsh&ll manuscript shows no
examples of either the horizontal middle stroke or of the left hand
tail, and thus seems to place itself between the SarShap inscription
(tenth centuryl and the Baijnfith pra.^astis (early thirt-eenth century).

(it) The common method of forming medial d is to add a knob to the top
right hand of the matiikS. and of this method there are hundreds of
examples in our manuscript ; but in the case of ja modification of the
matfilcS was the rule. There is, however, an isolated example (folio
16, recto, the sixth akshara of the third line from the bottom) where
the commoner method is applied for According to Vogel this
method of writing jd came into fashion about A. D. 1200. This test,
if reliable, is rather intriguing, as is also the next one.

(Hi) Medial i and i are generally formed as in Devanagarl, but there was an
older practice of forming them by sickle-shaped curves above the
matfikSrS and of this older type there are two or three in our
manuscript (folios 1 verso, 52, recto et verso, 60 recto). According to
Vogel these superscribed short and long fs dropped out of use about
A. D. 1200.

(tv) The slanting superscribed medial e tends to become horizontal in later

ftaradu, e.g:., in the Baijn&th inscription. In our manuscript the
stroke nearly always slanting but there are a few examples where
is
it is Also another method of expressing medial e is by a
horizontal.
stroke behind the m3.trik& and this method is exemplified 269 times
in our manuscript. The anecdote related by Vogel (p. 96) shows
that this back stroke method was not in general use in the fifteenth
century.

(v) Medial ai is formed in two ways and according to Vogel the change
took place about A. D. ItOO. In the Bakhshall Manuscript there are
19 examples of the older method as against 61 of the more modern
method.

* Indian Antiquary xii, 18S3. y, 8P.

* Ib. xrii. ISBU. p. .^6-

* Onr knowWdfif^ of the tcHpi han bfcn largely extended by the reeearcbee of T)r. Vogel, whose Tolume on
Tka AnHquiiUa of (Mamhd hee been of the greeteet help In the present enquiry.
 77

(vi) Medial o is expressed in three ways of which the most modem method
largely predominates in the text.

The following statistics are interesting. (See pages 95 — 98.)

MEDIAL B.
Older [-ej Newer [fe]

70% 30% Sar&hap inscp. xth cent.

57% 43% ChambS, No. 25. xith „

44% 56% Bakhshall manuscript.

0% 100% Ant. of ChambS^P- 63. xiiith „

MEDIAL 0.

Old [ o ] Middle [6] New [5]

58% 12% 30% Saruhap inscp. xth

10% 20% 70% Chamba, No. 25. xith

13% 10% 77% Bakhsh&ll manuscript.

0% 10% 90% Ant. of ChambS, pp. 64, 65. xiiith

The second table may be combined thus

MEDIAL O.
Older Newer

70% 30% Sarahap inscp. xth cent.

30% 70% Chamba No. 25. xith „

23% 77% Bakhsha.ll manuscript.

10% 90% Ant. of Chamba, pp. 64, 65. xiiith „

This scriptual evidence is rather remarkable and is as convincing as such evi-

dence can well be but I must add a word of caution. Fashion in writing is not
;

altogether a matter of chronology it is also largely a matter of locality. Indeed

:

no tangle criterion based upon script is infallible, but it is significant that all the
evidence that the Bakhshfill script gives points to some time about the twelfth
century, and there is nht a single item of evidence of any type against this
conelurion.
I. 2
 78

Symbols and notation.

127. Not unconnected with the scriptual questions are the forms of symbols,
and particularly those connected with the arithmetical notation.

The negative symbol, or minus sign, is a cross + ,the use of which for this parti>
cular pui*po8e is unique. The nearest approach to this use is the inverted ^ (i^)
employed by Diophantus to indicate a negative quantity. Dr. Hoemle rmected
the Diophantine origin of the Bakhshall symbol on the ground that the Hindus
did not get their elements of the arithmetical science from the Greeks.”* That
such a sweeping assumption as this is altogether unjustifiable has been already
.shown, and the implication that the work is wholly Hindu in oiigin has never been
proved. Dr. Hoernle attempts to trace the sign back to the A4oka ka but, as he
confesses, not with much success. However he suggests that the BakhshS.ll symbol
is " a mark of great antiquity.” See § 61 for some further discussion. All that
we can now say is that the use of this symbol cannot be traced to an Indian source.

The arithmetical notation employed will be referred to again (see § 130). Here
we are concerned with the forms of the symbols only. In my paper published by
the Asiatic Society of Bengal in 1912, 1 attempted to place these symbols epigraphi-
cally. Unfortunately we have very few examples of numerical symbols written in
early AaradS, but on the whole the Bakhshall symbols resemble most closely those
of the Devl-rl-kothl Fountain insci-iption’ of A. D. 1159. The AaradS numerical
symbols are fairly consistent in form but there are rather peculiar variations of the
" four ” and “ six.” The Bakhshall " four ” most closely resembles the example in
No. 15 Chamba which probably belongs to the eleventh century.* The evidence is
really too scanty for us to form any definite conclusion.

Language.

128. Dr. Hoernle writes’: —

" The BakhehiUI arithmetic i^ written in that peculiar language which used to be called the GAthil
dialect, but which is rather the literary form of the ancient North-Western PrAkrit (or PAli). It exhibits
a strange mixture of what we should now coll Sanskrit and PrAkfit forms. As shown by the inscriptions
U.g., of the Indo-Scythian kings in MnthurA) of that period, it appears to have been in general use, in
North-Western India, for literary purposes till about the end of the third century A.D., when the proper
Sanskrit, hitherto the language of the BtAhmapic schools ceme into general use also for secular comMsitions
, Its use, therefore, in the BakhshAli arithmetic points to a date not later than the 3rd or 4tb
. .

century A.D. for the composition of that work.”

It would be pie,sumptuou8 for me to contradict Dr. Hoernle on linguistic

matters, and his opinion here carries great weight. But from the evidence before
me I am compelled to disagree here, as elsewhere, when he is speaking of the age of
the work. lie gives a number of examples of the " peculiar characteristics ” of the
language of the Bakhshill work, on which he appears to have based his views. Now
every single one of these " peculiar characteristics ” is common in AEradi inscrip-
tions of the eleventh and twelfth centuries. Indeed their very occuirence helps to

• Ind. Ant. xvii, 1888. p. 34.

• .4n#. o/ Chnmhik^ p. 1512.
• The reaemblancr of wimi* of the numerical lymbols to lctt«-i i§ noted in part ii Bhhiar (/ad. Pah p. 88) itataa that
ymbole for 4 and 0 are *
ancient letter nuoieraU *; but the * letter numeral theory '
bat MMoumbhd !
^ Ind. Ani. xvii, 1888, pp. .38 37.
 79

confirm the evidence supplied by the script (§ 126). As it is proposed to devote

the second volume of this work particularly to linguistic matters 1 give here only
a list of Dr. Hoemle’s “ peculiar characteristics ” with references to the Chamba
inscriptions [in square brackets] where the same peculiar characteristics occur.

(a) “ Insertion of euphonic consonants.” [No. 32.] (6) ” Inseiliou of s.”

[Inscp. No. 32, p. 210.] (c) “ Doubling of consonants." [Nos. 15 & 32.] (d) " Pecu-
fiar spellings." [Everywhere.] (<>) Confusion of ri and ri. [Nos. 15 (p. 165), 24
(p. 184), 25 (p. 188), 26 (p. 198). J “ The jihvaniuliya and the upadhmdnlya are
(/)
always used before guttarals and palatals respectively.” [“ It is one of the most
notable characteristics of the Sftrada,’’ writes Dr. Vogel, " that we find the
jihvCimfdiya and vpadhmdniya used with great regulaiity.” (p. 58) See Nos. 13,
12; 15, 128; 32 (p. 210), etc., etc.] {g) “Irregular sandhi.” [Nos. 14 (p. 161),
24 (p. 184), 26 (p. 198), etc., etc.] (A) " Confusion of the sibilants.” [Nos. 14 (p.
161), 15 (p. 165), 24 (p. 184), 25 (p. 188), 26 (p. 198).] (il “ Confusion of n and n."
[Nos. 14 (p. 161), 15 (p. 165), 26 (p. 198).] (;') “Elision of a final consonant.”
"
[No. 14 (p. 161) where the same actual examples occur.] {k) Interpolation of
r.” ? (1) •' Etymolo^cal and syntactical peculiarities.” [Passim.] (m)
[ .]
“ Peculiar forms.” [Passim.] (n) “ Peculiar meanings.” Dr. lloernle’s inter-
pretation of the two examples he gives cannot be accepted.

My references to eleventh and twelfth century inscriptions here given make

no pretence of being exhaustive. They are merely given to justify my non-accept-
ance of Dr. Hoernla s views. On this matter I shall await the verdict of those more
competent to judge than I but my tentative conclusion is that the language of the
;

manuscript is not appreciably earlier than the script itself.

Metre.

129. In 1888 Dr. Hoernle wrote': —

''It appears that the enrliest niatheinatical works were written in the iloka measure; but from about
the end of the 6th century A.D. it became the fashiod to use the arya measure. Aryabhata o. 500 A.D.^
Varftha Mihira c. 550, Brahmagupta c. 030 all wrote in the latter measure. Not only were new works
written in it, but Hoka works wore revisied and recast in it. Now the BakhshSli orithmetic is written in the
Aoha measure; and this circumstance carries its computation back to a time anterior to that change
of literary fashion in the 5th century A. I).”

This statement is altogether misleading. Mahftvira's (Jat.iita-sdra-aangraha

(9th century) is largely in Sloka.t, and the Sdrya Siddhanta (c. 1100 A.D.) was
written in that measure and a number of other works dealing with astronomy and
;

mathematics written in the Moka measure and rather later than the Sdrya Sid-
dhdnta are known. Also we can point to a number of l^ftradfi inscriptions of the
eleventh and twelfth centuries in which the Hoka measure is employed. It is un-
fortunate that Dr. Hoernle's obsession regarding the age of the work led him to
employ this rather disingenuous argument, for it was repeated and emphasised by
M. Cantor, the historian of mathematics, in his ^eat work.* The subject of the
metres employed in the Bakhshftll manuscript will be more fully dealt with in the
second volume of this work.

* tM. Aiu., stm. ism, p. s«.

* FotImii^ iihet QuMchU dtr Maihemaiii (3rd «ditioii), toI. i, p. 698.
 80

Arithmetical notation.

The arithmetical notation employed throughout the manuscript is the

130.
modem place-value notation. If we knew the period when this notation was first
employed in the north-west of India we should have some criterion for the earlier
limit of the date of the work. Note that we are not here so much concerned with
the date and place of the invention of the modern notation as with its introduction
and early use in the north-west of India. Obviously we should search for evidence
of this in other manuscripts, on coins and inscriptions generally. Before we deal
with such evidence, however, it may be as well briefly to refer to the broader aspect
of the question.

According to the Hindus the modem place-value system of arithmetical nota-

tion is of divine origin. This led the early orientalists to believe that, at any rate,
the system had been in use in India from time immemorial but an examination of ;

the real facts shows that the early notations in use were not place-value ones and
that the modem place- value system was not introduced until comparatively modem
times. The early systems employed may be conveniently termed (a) the Kha-
roshtl (6) the Brahml (c) Aryabhata's alphabetic notation (d) the word-symbol
notation.

(a)Kharoshthi " script, written from right to left, was in use in the
The “
north-west of India, Afghanistan and Central Asia at the beginning of the Chris-
tian era. The notationis shown in the accompanying table. The smaller elements
are written on the left.

The “ Brfi.hml notation ” is the most important of the old notations of India.
(b)
It might appropriately be termed the Indian notation for it occurs in early inscrip-
tions and was in fairlpr common use throughout India for many centuries. The
symbols employed varied according to time and place but on the whole their form
was fairly consistent. They were written from left to ri^ht with the smaller ele-
ments on the right. Several false theories as to the origin of these symbols have

/ t i Si
4—-T " ‘- T - r f 9 10 ip ^^—sv —6e 70 fe Co /CO »»o ioomouoo
7 D :: Q a
T 'r-'' ^ -y i , r-"

Z Z]
t I I , -i

4. . fK XX ? // Iff

w»ja BBg DBB B D B Q B aa m e 1 7 oc 7- •J

i Bg
B BB DBBB n D B B aa B B J ?
p< 71 BB
B BB R B _L H BBD n d m K BB
SjSjQSi aBB S B s 9 A,S- QE BB
B0B P m aQaa9 s B
n 7 9
NUMERIOAL NOTATIONS.

been published, some of which still continue to be recorded. The earliest orienta-
lists gave them place- value, but this error soon disproved itself it was then sug- ;

gested that they were initial letters of numerical words etc., etc. :
 81

(c)Aryabhata’s alphabetic notation also had no place- value. It was written

and read from left to right but differed from the Br&nml notation in having the
smaller elements on the left. It may be exhibited thus :

Letters. k kh g gh ii ch chh j jh n.

Values. 1 2 3 4 5 6 7 8 9 10.

Letters. t th d dh n t th d dh n.

Values. 11 12 13 14 15 16 17 18 19 20.

Letters. P ph h bh m y r 1 V i sh s h
Values. 21 22 23 24 25 30 40 50 60 70 80 90 100.

The vowels indicate multiplication by powers of one hundred. The first vowel
a may be considered as equivalent to 1000, the second vowel t = 100' and so on. The
values of the vowels may therefore be shown thus :

Vowels, a i u ft It e at o au.

Values. 1 10* 10* 10* 10* 10“ 10‘* 10“ 10'*.

The following examples taken from Aryabhata’s Gitikd illustrate the applica-
tion of the system :

khyukhfi^ (2 + 30).10* i-4.10*«4,320,000.

-1-30 + 3.10* + 30.10* + 5.10* + 70.10*+ (50 + 7).10*- 57,753,336.

The notation could thus be used for expressing large numbers in a sort of
mnemonic form. Aryabhata's table of sines was expressed in this notation, which,
by the way, was used only for astronomical purposes. It did not come into ordin-
ary use in India, but some centuries later it appears occasionally in a form modified
by the place-value idea with the following values ;

1 2 3 4 5 *6 7 8 9 10
•

k kh 9 gh A ch chh ;
jh fl

th d dh n t th d dh n
P ph b bh m
y r 1 V sh s h 1

Initial vowels are sometimes used as ciphers earliest example of also. The
this modified system is of the twelfth century A. D. Slight variations occur.

(d) The word-symbol

notation . —
A notation that became extraordinarily po-
pular in India was introduced about the seventh century A. D., possibly from the
East. In this notation any word that connotes the idea of a number may be used
to denote that number e.g. Two may be expressed by nayana, an eye, or karna,
:

an ear, etc. ; seven by aSva, a horse (of the sun) fifteen by tithi, a lunar day ;

twenty by nakha, a nail (of the hands and feet) twenty -seven by nahshatra, a lunar ;

mansion thirt-v-two by danta, a tooth ; etc. This notation, it is said, was used
;

by Brahmagupta.
 82

The modern place-value notation. - The orthodox view is that the modern
(e)
place-value notation that is now universal was invented in India and until re- ;

cently it was thought to have been in use in India at a very early date. Hindu
tradition a.scribos the invention to God According to Masudi a congress of sages, !

gathered together by order of king Brahma (who reigned 366 years), invented the
nine figures Fatanjali and other early writeis are supposed to make references
!

to the place-value system. An inscription of A. D. 595 is supj)osed to contain €*1

genuine example of the systeiti according to M. Nau' the “ Indian ligures ” were
known in Syria in 662 A. 1). and certain other mediieval works refer to Indian
;

numbers ” and .so on. ;

On the other hand it is held that there is no sound evidence of the employment
in India of a place-value system earlier than about the ninth century y\.D. The
suggestion of “ divine origin ” indicates nothing but historical ignorance Masudi ;

is obviously wildly erratic the in.scription of A. 1). 595 is not above suspicion*
;

and the next inscription with an example of the place-value sy.stem is nearly three
centuries later, while there are hundreds intervening with examples of the old non-
place-value system. The references to India in mediseval works do not necessarily
indicate India proper but often simply refer to “ the Hast ” and the use of the term
with regard to numbers has been further confused by the misreading by Woepeke
and others of the Arabic term hindasi (geometrical, having to do with numeration,
etc.), which has nothing to do with India.’ Again, it has been assirmetl that the
use of the abacus “has been universal in India from time immemorial l)ut this ’

assumption is not based upon fact, there being actually no evidence of ils use in
India until quite modern times. Further, there is evidence that indicates tiiat the
notation was introduced into India, as it was into Europe, fjoni a right -to-left
script.*

131. As this subject is a matter of general controversy it behoves us to be

circumspect abotit drawing definite conclusions from the occurrence of the modern
place-value notation in our manuscript. As indicated above the proper procediii'e
18 to examine the evidence relating to the earliest appearance of this place-value
notation in the north-west of India. Of manuscripts there are very few earlier
than the twelfth century and none of these gives any example of the new notation ;

of coins there are plenty of examples but none earlier than eleventh cenf»iry gives
any example of the new notation and the eaiiiest inscription of this part of the
;

world with any evidence of the new' notation is not before the tenth century A.D.
To suggest that the Bakhshali manuscript wsis independent of all the circumstances
that governed the epigrnphical appearance of the new notation generally would
be unscientific.

Dr. Hoernle was very dogmatic. He wrote, in 1888, “ It is certain that the
principle* was known in India as early as A. D. 5(X). There is no good reason why

^ Thr wero adtled at a lat'Cr dat^. A eiiriM.)rv rxatnination of tbo plate {Epigrajthia Indira^ vol. ii. p. 2U) makoA
thh nlivioufl.

* Journal Aaiatique, IPIO, p. 2H9. M- Nan's authority ' do<>9 not riHtd
'
wHK He sfioaks of the Hindus' noble diseovftrtes in
—
aatronoray more ini^enious than those of il»e Greeks and Babylonians.
• About nine hundrotl years atfo AlblrQni wrote (India, ii 211) ” They
— .... relate all sorts of thini;s as hrinji^ of Indian origin,
of which we have not found a single trace with the Hindus themselves.*'
For a more detailed examination of tliis question see my papers on (1) Indian Arithmetical Kotationn, JASB 1907, 47A>008,
*
(2) Ahaevn in ancirni India, JASB 19()H, 203-287 ; References to Indian Mathematics in certain Meditrvdl teorksn JASB
The. wsp of the
801 -H0H. In 1017 in Hoivktia (1017, pp. 273-2R2) Baron Tarra de Vaux pursued a line of inTMtigation completely different
|rt»m that I had myself followed and came to the same oonolusion, namely that the original home of the modem place- value nota-
tion ooiild not well have been in India. For a popular exposition of the other side of the controversy see StnTH and K ARpnraKT The
Hindu Arabic Numerals.
• That is the place- value principle. The quotation is from the Indian Antiquoiy of 1888 (xvii), p. 38,
 88

it should not have been discovered considerably earlier. In fact, if tlie antiquity
of the Bnkhshali arithmetic be admitted on other grounds, it affords evidence of
an earlier date of the discovery of that principle." Btihler is bolder still. " If
Hoernle’s very probable estimate of the antiquity of the arithmetical treatise, con
taiued in the liakhshali manuscript, is correct,” lie writes, " its (i.c., the place-vaiue
notation’s) invention dates from the beginning of our era or even earlier.”'

132. There are two other criteria suggested by Dr. Hoernle, namely (a) the
length of the year mentioned in the text, and (6) the occurrence of the term dindra.
Regarding the former he came to no definite conclusion but seems to have thought
that there was some correspondence between the year of 360 days that occurs in
the text, and the accepted estimate of the length of the year at the period of com-
position of the work. Had he gone into the matter further he might have come to
the conclusion that the Hakhshall work had been composed in pre-Vedic
days It seems hardly possible for any one to be misled by such “ internal evi-
!

dence ” but it ma}' be as well to point out that the practice of reckoning t he year as
consisting of 360 days for the purpose of arithmetical examples wa.s quite common
in mediseval Indian works. Mahavira and Srldhara actually give 1 year - 360 days
in their tables of measures, and the latter adds the remark "
'I'ime is calculated
—
according to this rule in all eiithmet’cal works.”

133. Dr. Iloernlc's argument regarding the use of tlie diiitlra is as follows :
-

In the early centuries of our era the dindra in <ise. in India was the gold one only,
and “ the Bakhshali arithmetic seems to indicate that the gold di/mra and the
silver dramma formed l.he ordinary currency of the day. 'I'his circn instance again
poittts to .some time within the first three centuries of the (.’hrislian era as the date
of its composition.'” The only reply that this statement calls for is to state that
all the evidence of our text points to the use of a copper dlnOra and there is not the
remotest •ud'cation of a golden dindra:' (See § 1.10.)

134. There are other indications rather than evidence of the age of the work in
the material of the text. The occurrence of the square-root rule already referred
to would not be an anachronism if it were found in any Indian text from the time
—
of Aryabhata onwards but it occurs in no known Indian text until very late indeed,
and its appearance in the Bakhshali manuscript is probablv due to direct western
influence, possibly to Muslim influence. Such evidence is valueless to the inex-
pert it can only carry weight with those who have a very fail- knowledge of the
:

mathematical field of the period. Likewise the employment of the sexagesimal

notation points in the same direction, but the general reader must bear in mind
that this notation had been used by Hindu astronomers from the time of Aryu.
bhata, and that the arithmetical use of it in ovir text is rather a matter of western
fashion than the introduction of a now idea.

The emoJoyment of the regvla falsi is evidence of a slightly diffpi euf character
It occurs in no Indian work until the time of Mahavira,* and it was prol)ablv, evnn

' JnS. Pol. p. K!>.

* A>«r. rif., JIT. Tt. ih not^wo^tby tbai thr mimlim arithmetiu^I works off.«n s s«v<f.ion fo tlii» fHrham itnd iHnArt. Sw
F. WflCTiFMANN *. d. Xatur riv, np. 50 »nH

• S**! A. Stittn, in 1000. wrn^j^i “This word, iindotildofllv dorfvod from t.hf» tie.nnrivB of ihr M'rst., is woll known to Sanskrit
leiricovranby as the dosiimstion of a gold noil) nsnsJIv sTwIt dfnAra. Bwt thr manifest lmpoM*biIitv of emeptlng ibis moaning for the
paaesrns of the Chronlnln wbinb mortion sums in D^nnAran. has almadv stmnk T>r. Wiltvm. Votioin^ in two passages H^ires arc
given wbinb. If natcmlsfnd in gold woitld be large beyond all rreilenm, he sngfrested fbst the IMrpirs * meant might have been of
‘

ooT)T>er. rnriofsiv enough, however, none of the subsequent int«»rprntors soems to have foll'jMod up the suggestion tlHown nnt by
Wilwm. or have oiberwisn paid sitnnllon to the niiblnct.” Knihann*^ /fdifi/nrrnfKnf, vol. 11, p. 308. He then goes on to
prove tli'it the ttlnrita used weu a nopner one. The whole of the dissertrtion should he rerd, pp. .T0H-32R.
« Soo his vii, 112. His employment rf the method is rather spnr-al and limite<L In Nurtlu^rii India
the firr.tIrnowr ubo of Iho method ocours ’n the tv rlfth c»entwry.
 84

in the west, a fairly close predecessor to an algebraic symbolism. There is also

other evidence that our work was probably produced not long before an algebraic
symbolism came into use.

135. Of the evidence as to age discussed some is of doubtful value but there
;

remains a good deal that must be considered as giving no uncertain indication, and
the period indicated is in all cases about the twelfth century. The script, the lan-
guage, the contents of the work as far as they can give any chronological evidence,
all point to about this period, and there is no evidence whatever incompatible with
it. Bhaskara was born in A. D. 1114, Omar Khayyam was flourishing in the early
part of the century, Adelard of Bath visitod Cordova in 1120, Leonardo was born
in 1175 and it was during this period that the Bakhsh&ll work was probably com-
;

posed. ft is possible that in the future more light will be thrown upon the script
and language of the locality and period of the Bakhsh&ll manuscript and that,
;

if the inan»jscript be required as suggested, further internal evidence will be forth-

coming. Such further discoveries may modify the conclusions now drawn regard-
ing the age of the manuscript but they cannot very well put the date of composi-
;

tion of the work back to any great extent.

 The BakhshAU Manuscript
PART 11.

i, — The Script ....... Pace

87

ii. — Transliteration of the Text .... 105

iii. —Facsimiles of the Manuscript . Plates l to XLVII

M 2
 i. - The Script.

The Bakhshall manuscript is written iu the /^arada character, which is a des-

cendant of the Brahml, in a line distinct from but parallel with the Nagarl line of
descent. With the last mentioned fciarada has many features in common, but,
generally speaking, it is much more archaic in appearance. The ^arada script,
although not nearly so well known as the Nagarl, has been examined and written
about to some extent, f^eech’s (i ram mar of the Cash meeree Lniuf iiatje (.JASIi.
xii, 1894, 399 sqq.) gives the alphabet. Biihlers tn<Iisrhe f'ataeorqraphie, gives
useful, but sometimes mi.sleading information Vogel's Antiquities of Charnbd ;

State (1911) contains the most valuable contribution and to thi.s I am chiefly in-
debted ; Ojha’s lihfiritiya Prdrhina Lipimuld gives some useful tables, but, .so far
as the J^arada .script is concerned, is largely b/ised on Vogel’s work. Sir George
Grierson’s paper in the Journal of the Royal Asiatic Soicety (1916, xvii, 677 .vyy.)
contains tables of ligatures, etc., of modern Silrad.H and in his note in the lAngiiis- ;

tir Survey of India (V’ol. viii, Part ii, page 254) he states that the J^arada charac-
ter is the ancient indigenous character of Kashmir, and that it is still generally
used by Hindus and is taught in their schools in that country. 1 have not yet been

able to obtain a copy of Burkhard’s tables.' which, according to the editors of the
Kashmirian Afharraveda, contain transliterations of 31() characters and ligatures.

The principal examples of the .'^arada script that are kriown come from Kash-
mir' and Jammu Gandhara’' (in which the village of Bakhshali is situated)
; ;

Ladakh Kiingra, Kulu and Manjl the Shahpur district of the Punjab, and Delhi.
; ;

The distribution roughly correspond.s to the area between longitudes 72'’’ and 78®
east of Greenwich and north latitudes 32° and 36°.

The earlier epigraphists had rather inaccurate notions about the J^arada
script. Cunningham described it as the Gupta character, which ho thought had
persisted in use in certain localities;! Bvihler .stated that the oldest >^arada in-
.scriptions were the two Baijnath pra^askis from Kangra and that their date was
A.D. 804.1 whereas it is ,4.1). 1204; Dr. Hoernle dated the development of the
.script from about ,4.D. 500 and laid down the maxim that ftarada characters were
no guide as to age, and Kielhorn seems to have entertained a somewhat similar
view.

The conservative characterof the script puzzled many investigators and it

seems to have retarded the formulation of any chronology based upon palaeogra-
phic considerations. Vogel, however, in spite of the weight of the opinions of
Kielhorn, Biihler and Hoernle, holds that the historical development of the ftfirada
script can be traced. “I believe,” he writes, § “that a clo.se examination of the
characters will also enable ua to fix the approximate date of any undated SiLrada.
record of the pre>Mnhammadan period, pro\dded it is extensive enough to lend
itself to a detailed study.” Whether or not Vogel has been perfectly successful in

ProorodlnitH of tho Academy of TlrnoA. Vol. p. 640.

^ KMbtAir {«• indnod, known
*4 iikmdH AtiAcim, or land of the goddiwN Samovotf.
tha Paihawar diatrict and tho rnirronndinff hill iraeto,
f A. R. R. aiv. 121.
t Tnd. Pat p 67.
§ A Champs, p. 4$,
 88

formulating his chronological tests remains to be seen but, at any rate, be has ;

provided material and detailed notes that are invaluable.

Ur. Hoernle first gave 500 A.D. as the approximate period the birth of the M
S&radS, script, but later he modified this view and stated that it orinnated directly
from the Gupta script in the course of the seventh century. The earliest known ex-
amples are on the coins of the Varman dynasty of Kashmir which start from the
middle of the ninth century; while previous to this the acute-angled script was,
it is said, in general use in northern India. It is, of course, possible, as Vogel
points out, that the J^arada was employed as a literary alphabet considerably ear-
lier than the ninth century but there is no evidence.
;
There is a Kashmir inscrip-
tion of the reign of queen Didda that belongs to the tenth century, as also possibly
does the Sartlhan inscription and there are many dated inscriptions belonging to
;

the eleventh and twelfth centuries. The following table gives the dates of some of
the more important J^firada inscription :

A. D.

Varman dynasty Kaahmirf 855-939

Coitifl of the

Inacription of the reign of

Devi-ri-ko^ fountain inscription

of

Queen Didda ..... 992

1159

Angom inacription • . • 1197

Baijnath inacription (Kahgra) 1204

With the Baijnath pra^astis, according to Vogel, the history of the Bftradft
proper comes practically to an end and the script of the later records differs so
;

considerably rom the pre-Muhammadan inscriptions that he proposes to give it

f

a special name. “ The thirteenth century,” he says, " forms a blank which sepa-
rate.s the two palaeogruphic periods.”

The relationship of the Bakhshall script to other Indian scripts is illus-

trated in its main outlines in table i, which shows the BrRhml, Western Gupta
(Bower), Acute-angled, darada (Nos. 4, 5, 6, 7) and the Nagarl alphabets. It is at
once seen that the Bakhshfill alphabet is more closely connected with numbers 4
and 6 than with 7 and that numbers 4, 5, and 6 are differentiated on the one hand
:

from number 3, and on the other from number 7. Indeed numbers 4, 5, and 6 are
examples of the J^aradii group proper, which is supposed to be a direct descendant
from the acute-angled script (No. 3) and number 7 is a direct descendant from
;

the earlier ftaradfi. .script.

The most notable differences between the i^&rada and the acute-angled are
with respect to the na, ta and .ia; while minor differences are seen in the ga, na,
pha, la and ha, but in these latter cases slight variations occur in one or both of
the scripts which, more or less, blot out the differences.

The differences in the cases of the ga, ta, ia are possibly due to the emphasis
given in the later sciipta to the overhead horizontal lines. The most import.ant
—
difference lies between the two example of na ^the middle horizontal connecting
line being consistently employed in the acute-angled script, and consistently
omitted in all but the very early examples of the dSradS.

Between the ftaradfl proper (Nos. 4, 5 and 6) and number 7, which may be
termed modern Sarada, the differences are much more marked. The i, e, ka, ja.

f Cl’NNlNOH AM Cot.. »/ MtHeMot Mia, pl.to hr.

 y® (Table I) are particularly noticeable while the na in both the Baijn&th
;

example and in the modern d&rsidfi shows an inturned tail on the left.

The alphabet.

Further details of the alphabet are given in Tables ii to iv. Table ii shows the
consonants and their more important combinations Table iii gives the various
;

methods of writing the vowels and Table iv gives most of the ligatures, the nume-
;

rals, etc. It is proposed to examine the whole alphabet more or less in the order
of these tables.

The Consonants, k-, kh-, g-, gh- ,

ii-

The normal k- of the manuscript differs very little from other early d&rada
examples. Indeed the variation in the manuscnpt is almost as much as in the
inscriptions. The right-hand curve ex Mbits, on the whole, a tendency to close
inwards towards the vertical like the modern dfixada, particularly in the “ M ” sec-
tion of our manuscript but there are some half doxen examples where it is almost
;

straight. The Sarahaij and BaijnSth examples have a much shorter and more
open right-hand curve, while in modern darada the curve is quite closed. Vogel
says that the left-hand loop is generally more rounded in the older inscriptions
and in the Bakhshall manuscript, but this is, at least, doubtful.

As the first element in a ligature k harks back to the Brahml type, e.g., in
ku, kfi, kt- kr-, k.<th-, etc. (Table ii, 1), as it does in all the J^arada examples but
;

as the final consonant in a ligature it generally retains its normal shape.

In the Sftrada, g and gh have fairly constant forms, which show very little
deviation from the Bower manuscript examples. The left-hand shoulder of the
g is sometimes slightly rounded in the Bower manuscript and the acute-angled
script and in the Bakhshall manuscript similar examples occur once or twice.
; In
ligatures each of these letters retains its normal shape.

The letter h- is only found in combination with other consojiants, as is gene-

ral in the daradS. The form of fi- is the same as in the Gupta period but it has
developed a swelling at the right-hand end of the top horizontal stroke, and this is
sometimes mistaken for the sign of <3. In the BakhshSll manuscript such right-
hand protuberances can. however, generally be distinguished from the d sign. See
and t-

ch-, chh-, /-, jh-, h-

”
Buhler speaks of the “ quadrangular cha as characteristic of the Sarada,
but it can hardly be deemed to be such, for in the Sarahan and Bakhsh&ll examples
it is often somewhat rounded, or in the latter has the right side vertical, and in the
Devl-rl-kothi inscription it is almost triangular. Biihler possibly had the Baij-
nfith and modem SftradS examples in mind.

The akshara chha is also fairly constant in form, although the BakhshMl ex-
amples are more cursive than in the inscriptions. The digraph chc.hh- frequently
occurs. These letters are illustrated in Table ii, 2.

The a conservative character in the daradfi

j- is In the BakhshElf manus-
cript it differs very little from the Bower manuscript and acute-angled examples.
 yo

except that tlie top horizontal stroke has developed u much more pronounced knob
oil the riglit-hainl. Accordiug to Vogel this knob or wedge disappears in the
Muhainiiiadan period, but Grierson gives it for modern daradfi. The akshara ;d
differs i onsidci ably from ja. The top V»ar disappears and a nearly vertical stroke
is added to the end of the tongue but instead of this modification of the mdtfikd,
;

in late inscriptions occasionally wo meet with a jd formed by the usual medial d

sign added to the indt nku, and in our manuscript there is one such example
(folio 16 r.) The later script seems to revert to the Bower manuscript type.

The letter ;7i- only occurs in jjhi and jjhi/a. It is something like the ;& with
the left-hand stroke turned backwards. It shows little alteration since the Gupta
period.

The h also only occurs in combinations (jfia, jna, hrlia). The superscribed h
differs considerably from the subscribed form, but both forms are almost identical
witli e.xainples that occur in the Bovver manuscript.

A/i-, d-,

Of the linguals (Ji does not occur at all independently, and it is doubtful
whether it is intended as a subscript. The generally, but not always, has a top
bar, and always it has a well developed right-hand knob. As the first element of
a ligature A occurs twice, but as a subscript to s/i it is common. The scribe evi-
dently made no distinction between sht- and shth-, for the open and closed letters
are used indiscriminately for the same wonis. 'I’he normal A or th as a subscript
to s/i is generally accompanied by another curve attached to the right of the sh
near the bottom, lisualiy this line curves slightly inward.s towards the subscnpt
but in several cases it is a long .stroke slanting outwards. This additional stroke is
ab.sent in the S;irrihai.i inscription (and in the late Kashmir SiM’adA) but it is fairly
common in other early Sarada inscriptions.

The d~almost exactly like the Baijnath and modern i^arada examples. It
is
from those in the Bower manuscript, and its development is
differs considerably
somewhat uncertain. Many examples of rjid- occur in our manuscript.

The akshara dha is very like the ph/i but has a kink in the downward stroke.
It.s form has not altered essentially since the Brahnil period.

'I'alile 1 exhibits three distinct types of the >^arada cerebral nasal n- The
Sarahai) example has a horizontal connectingstrokeinthemiddleof the letter, like
the Bower and acute-angled examples; the Baijnath example is without this hori-
zontal stroke but has a tail turning inwards from the left, and the modern d&radft
has the same tail somewhat shortened the BakhshalT example has no horizontal
;

middle line and no left hand tail, and is exactly of the same type as that exhibited
in the L)e\M-ri-kothl fountain inscription and other inscriptions of the twelfth cen-
tury V^ogel seems to think that we have here a reliable criterion for age, and if so
the Hakhshall manuscript places itself between the SarS.hap and Baijnftth pradas-
tis But like all other particular tests for the Saradfl this must not be reliM upon
with too great confidence, for in the lat«ir dAradS. we find the Bakhsh&ll and Baij-
nflitli types nse<l promiscnously.
 yi

In the Bakhsh&ll manuscript the form of n is very constant but different sec-
tions of the manuscript show slight variations. The tendency is to make the left-
hand stroke shorter and slightly thicker than the others, paidicularly in the “ M ”

section the third example in the table is peculiar in having a long thin tail but it
;

only occurs once or twice in some hundreds of examples. The principal conjuncts
are tiy-, rn-.

t-, th-, d-, dh-, n-.

The dental group is quantitively very important. It is, on tlie whole a

conservative group, the chief changes being, perhaps, in the genf?ral shapes of the
th and dh, which in the early inscriptions were, more or less, crescent shaped,
in the SarShan pra^asti. The employment of tlie top bar modified the crescent
shaped letters into lozenge shaped. In our manuscript the crescent shape is
partially retained.

As a subscriptt is something like itiitial v bnt without the thin up-strokc

(See Table 4) th in
ii,ligatures has dilTercnt forms as a second and third element,
;

<ig., as in tlhy-, sth-, rth-; dh as a subscript usually has a triangular form (ddh-,
hdh-, rdfi-, ddhv-), but it is sometimes ditlicult to distinguish ddh- from dr-.

In the Devi-rl-kothT fountain inscription the rth- is rafher rcmarkablv differ-

ent from the Bakhshali examples and so if is in modern .Saradil. 'I’he e.vampics
in fig. 1, from the Sungal grant, are interesting as Au Afii tK>* nthys
showing the transition stage in the development of
th as a subscript and medial.* I” ^
Fig, I. .Suhgal.

Hand mnde uopiea. /nfttan Anfiguarg,

• Biihlnr (Table VI, column viil, 50) given an inrormri form fi»r rtha. I may an W'dl iu»int out liero that numbers 17 (j7 ti), 24
am
and
and 38 (hha) Of BuhloFf tiabla are hnnllv rcprencntativo cxamplca.
thoac, at Dr. Hucrnic wat v^ell aware, are not perfertly roliabla. For example,
He had only hr HMomle'n hand-innilM ciioiea to fidloer.
tho triosoript of Ki t icclo given in the Inclmn AntiijUfiry
(xvii, 1B88« p. 43) minn^pmaenta iya and iu in the penuHiruafn line ; ^patk., ddhtfa, etc in lh» list line ; and that of FI 4 veisn gives I. 1
.

; 1. 3 uhpUt lAm,9tha ; 1. 6 A/tl ; 1. 0 tpra ; Liat lino #r.*ama, inaccurately.

N
 92

p-, yh-, h-, bh-, m-.

The labialgroim ia also very important quantitively, and^ this rather ac-
centuates certain difficulties that present themselves. The principal of these are
due to similarities between the v and dh, the pA and (fA and the m, d and s under
certain conditions. Generally in the farads, no distinction is made between the
6 and v, and in the BakhshalT manuscript this lack of differentiation is almost, if
not quite, complete. The resemblance between the h ot v and dh in ligatures has
alretudy been noted upon the ; m
is distinguished from the d by having a complete
vertical stroke on the right, but the ligatures nd-, nm-, rd-, rm- are sometimes
rather difficult to differentiate and badly written examples of
;
m
and a are also
confusing. Final rn (with the virama) is very different from its mati'ika and occa-
sionally is only to be diffei’entiated from the numerical sign for “ six ” by the length
of the very long virama stroke.

y-, r-, 1-.

In table i are shown two types of daradS y, of which the Bakhshall ex»
am pies may be described as representing the transitional stage between the Bower
manuscript example and the Sar&hap example. Table ii,* however, shows that
the open (Sarahap) type also occurs in the Bakhshall manuscript, but it occurs
with comparative rarity.* Most of the examples are of the type shown in table i
and the first example of ii, 6 while the t)q5e illustrated as the third example in
;

table ii, 6, occurs generally in the A, section of the manuscript. In most fiftradS
scripts the latter (open) type prevails but in the DevI-rl-kothI fountain inscrip-
tion there is exhibited a tendency towards the prolongation of the central vertical,
although the left-hand loop is never actually closed, as it is in most of the Bakh-
shnit examples. As an element of a ligature the form of y is masked. Many ex-
ample.s are given in table iv. Occasionally the upward right-hand portion of the
curve is lengthened as in yyli (iv, 1) and ddhya (iv, 5) otherwise there is little to
:

note. As the middle element of a ligature y seldom occurs, unless we count U as

one of the elements.

The one of the most consistent in form in this family of scripts. The
lettor r- is
bottom serifhad developed in pre-SAradA days. This serif is generally rather
largei- in our manuscript than in other dlrada texts and occasionally develops into
a loop (F 24). As a final element the r seems to be formed by a thinning and
lengthening of this serif (table iv). For the peculiar shape of rU see page 95 and
for other modifications see below.

Two types of I are shown in the Sarftha^ and BakhshAll examp.Vs exhibited
in table i. The former has the left curve attached to the vertical by a horizontal
stroke, and this seems to be a characteristic of the earlier S&radfi scripts the :

latter has the left curve attached by another curve ^practically always in our
manuscript (table ii, 5, etc.). According to Vogel the former type was still pre^
valent at the beginning of the twelfth century..

i-, ah-, a-.

Table i illustrates fairly well the development of the ^ftradS i, which, in its
later form, is exactly the Nftgarl a. In our manuscript the S is always square in

8m FT 4, e. 38, 41. M. 5S.

 03

shape and genei*ally has the triangular wedge on the left emphasized See table il,
6). This wedge is never open but always blocked-in solid. The ah is consistent in
form (See table i), but in the ligature hshn the portion of the vertical above the
cross bar is suppressed* (tables ii and iv) See also shsha in table iv, 4.
:

The a is like the i but open at the top. In the Bakhshall manuscript it is
<)uite common for i and a and ah to be used indiscriminately.

The letter h is fairly consistent but occasionally tends towards angularity,

like its Acute-angled prototype.

Visa rgo

The visarga is represented by the two usual dots like a semi-colon. Vogel
states that in several Chauiba inscriptions “ composed in corrupt Sanskrit ” the
real meaning of the visarga is misunderstood and that it is regularly used to
separate words and sentences. The same remark applies to our manuscript.

The jihvamullya( x-) and upadhmamya ( are often employed. In one

case 33) the former seems to be used in the middle of a word, dvxkha ( ?) for
(fol.
duhhha. It (x) is similar in shape to the h or v, while the latter ( 0) is very
much the same as in the Ohamba copper-plate grant of Soraavarman (1. 2). Vogel
states 170) that the upadmaniya dropped out of use in Chainba after about 1200
A.D. Examples are shown in tables ii. 1 and ii, 5. See also page 79.

Virdma.

The final consonants k, t, f, n and m have each a very long virftma symbol
^rawn through the top of the letters on the right. See table iv, 6.

The V^wela.

a and d.

Initial a and d are fairly consistent in the Sarada, and the Bakhshall ex-
amples differ very little from the acute angled and early Gupta types. The a has
at the bottom of the right vertical stroke a small triangular wedge, while the d has
in place of this wedge a small curve something like the bottom part of the t. Both
•a frod d have open tops, while in modern Afirada the tops are closed, as in the
Devanftgarl.

Medial d.

The symbol for medial dis a wedge or serif attached to the right of the
m&trik& at the top. Examples are seen in table iii, 1. In the Bakhsnftli manu-
script this symbol is generally rounded and is often* written with a very small
flourish at the top.

• Powlbl; to pmvont annlution with Ik ; but ia ligatuR* Ik U coaaidenlil; modifiMl.

N 2
 94

There are three methods of forming medial & to consider (i) The archaic :

method of joining the symbol by a horizontal bar attached to the left top of the
m&tfikg. Vogel thinks this method dropped out of use about A.D. 1000, and that
it may have been merely a local fashion. When there is any choice this method is
not employed by our scribes, (ii) The more common method is to join the symbol
to the top of the right of the letter, and this is generally done in our manuscript
with each of those letters that has a dexter upright or a top bar. (iii) The third
method entails a modification of the mfttrikfi., but this only occurs in the case of ;.
Examples of jd are shown in table ii, 1 but there is an isolated example* where ;

ja is formed by method ii, and, according to Vogel, method ii, as applied to the j
came into fashion abo\it A.D. 1200.

It will be noted t hat each of the matrikas of n, /, and I already possesses a mark
similar to the medical d symbol. In the cases of n and 1 the usual d symbol is
added according to method ii, but, as already stated, to form the jd, method iii is
generally employed. In the nd the symbol is often slightly curled inwards, and
this hook-shaped symbol is said te be characteristic of early t^aradS..

Note that the -d symbol is joined to the left upright of the d-, but this is be-
cause the dexter upright is not a complete one and thus the dd is differentiated —
from the md.

The vowels i and i.

The initial vowel i does not occur. The short vowel i is of the form that is,
with fair consistence, employed in all ftarada scripts it consists of two dots placed :

above a, more or less, semicircular loop. (See table iii, 2.)

Medial i and
almost without exception, of forms that are essentially the
i are,
same as in the Devanugarl, and consist of left (-i) and right (-i) vertical strokes,
sometimes considerably longer than the matfika and joined to its top by a bend
inwards.' But our manuscript contains three or four examples in which the -J is
formed by a sickle-shaped curve above the matyika and with its convex side up-
wards and turned slightly towards the left. (See table iii, 2.) Vogel says, " We may
assume that about A.D. 1200 the superscribed medial i and i dropped out of use.”

u and u

Initial u is the same character as in the Bower manuscript. The left up-
stroke, which is generally continued to the level of the top of the letter, differen-
tiates u from ta. The long u, as an initial, is the symbol for the short vowel with
a streamer hanging down from near the right top of the letter. It is distin^ished
from ru, which has a similar “ steamer,” by the long upward curve on the left.

Medial u is expressed in three ways (i) By a triangular wedgef attached to :

the bottom left of the right vertical, or, where this vertical is absent, to a vertical
attached to the bottom of the inS,trik§ (ii) By the addition of the initial u as an ;

ordinary ligature and (iii) by attaching a downward steamer to the right of the
;

• F Ifl fMtO).
* Tht r aiffolEa it uftan 1oii|«ir{thjiii tb«t of •«. Compare the akaharM and It in table ii. 2 : iheae are contiguoua akabaipa
from F 33 t.
t Oftan doaod lolid, Imt occaMionnlly open.
 96

letter near the top. Each of these methods generally applies to a definite set of
letters only, as follows :

Thtt wedge in altik^hed tu The curFod u Hymbul in attaohed to The streamer ejmhol to

d-, dh~, n-, p-, b-, y-, i-, nh-, h- k-, g-, hh-, s-

This practice is quite in accordance with that of the early Sarada but in our ;

manuscript the hi has either form, and this is not surprising, since the h and <•
are occasionally employed indiscriminately. According to Vogel the wedge syro.-
bol is earlier than the curve.
**
Medial ~u formed most generally by a, more or less, horizontal " streamer
is
attached to the left bottom of the vertical or to a vertical providetl for the purpose
{hu, dyu). The ru and the hru* are exceptions to this method of formation the :

former is very like the initial n but without the left up-stroke, and the hru is some-
times formed by an angular attachment to the middle of the right vertical, and
this is possibly a modification of the initial ii. In one case hru is formed by the
addition of two hanging streamers on the right and the r- stroke is shortened.

fi

Initial ri occurs once (63 r.). Medial ri is invariably the initial sub-joined
as a ligature. Examples are shown in table iii, 5. This symbol differs from other
^Arada examples by being less rounded. Biihler terms it angular, and Vogel states
that this angularity is only found in the later inscriptions.

The vowel c

Initial e does not generally differ essentially from the Bower manuscript-
examples. There are, however, examples in our manuscript similar to the 12th
century examples given in Biihler's tabte VI (II. x-xi) and these examples sug-
gest one of the medial forms.

Medial e is formed in two distinct ways (i) By a slanting stroke over the :

letter, touching or nearly touching the letter at the right top corner [^]. In late
.^AradA this stroke tends to become horizontal, (ii) By a short thick stroke or

knob attached to the left top of the consonant by a horizontal line ['e].f Vogel de-
scribes this as a wedge but in oxir manuscript it is never wedge-shaped.
; It is
very like a reversed medial d, bvxt with two alight differences the connecting hori- :

zontal stroke is comparatively long, and the terminating knob or stroke has no
little flourish but is perfectly smooth.

These two symbols, so different in appearance, seem to be used indiscrimi-

nately, at least for certain letters and they seem to be very unevenly distributed.
;

For example, F 11 exhibits the second method [ e] 11 times and the first or top-
stroke method only twice while F 60 does not show the second.
;

* ll is ihs r In ths licstnir thst ocoahIqdii the so-enlled exception.

t The symbols shown In square himekete in this and the following parsp'aphs art mnemonJo only. They do not aoourately fn-
present the symbols employed in the manusoript. The tobies and text should be consulted in all cssm.
 06

The table ou page 97 shows that the two methods are used almost equally in
the M
" section and that in the remainder the ratio is roughly 3 to 2.* The
foUowing table gives more details for sections A to L, and M :

I
ko go ohhe to no mo y® ro lo •0
1

C h Pero*nUgM . 38 100 100 68 82 100 20 60 88 100

At6 L]
( >• 62 0 0 32 18 0 71 50 67 0
" X . .

-I

100 60 66 13 33 0 67 22 50 0.
•

*• 0 60 44 87 67 100 83 78 50 100
t

No. of eumplei 21 21 S6 46 2S 16 24 21 62 7

The values of these ratios depend largely upon the number of cases. As a
whole they seem to indicate an inversion of fashion. The M section, which is pos-
sibly the more ancient, generally shows a larger proportion of the second method
[•e] but notable exceptions are ke and ye.
; In one or two cases complete inversion
is exhibited for example, me is formed by the first method [^] only in sections A-L,
:

while in “ M ” it is formed by the second method [ e] only and the same is the case ;

with se. The best test aksharas of those given in the table appear to be ke, chhe,
te. ne, and 7ne; although all may be used with fair safety. It should also be notetl
that ne is formed by the first method only in all sections by reason of the shape
of the matfika. Similarly the first method predominates with le.

Vogel relates an interesting anecdotef of the 15tn century which tells of a for-
ger changing me into dam. The original narrator w'rites " In order to express :

e following a consonant the clerks used formerly to write a stroke behind the con-
sonants. But«as. in the course of time, the senpt became changed the writers of
to-day write the stroke e.\ pressing e over the consonant.”

Medial ni
does not occur in ftarada. Medial at is formed in two ways
Initial at :

(i) By two top strokes [ai] (ii) By a coihbination of the two e symbols [Ai].
; In
both cases the ai symbol is thus a .symbol for ee.

Ofmethods the latter [M] is the older. Possibly the earliest example
the.se
of the former method [ai] occurs in Chamba No. 2.'>, and Vovel assumes that the
change took place about A. D. 1100. In the Baijnath pra^astis the double top
stroke [ai] is the more common. Later it is invariably used and the top strokes
are horizontal.

In the Bakhshall manuscript both methods are employed, but the second [-^i]
predominates indeed out of 80 examples 61 or approximately three-quarters are
:

formed by the second method [ ai], and 19 by the first or double stroke method [ai] ;

but in the M section there is not a single example of the first method [ai].

Medial o
Initial o does not occur. Medial o is formed in three ways (i) By a large :

circumflex shaped symbol placed above the consonant fft] (ii) By a combination of ;

• A l lA 2 •«.

t of ChnmhH p. 6.^.
 97

the e sign [ ej with the symbol for & r o] (iii) By the top stroke for e combined with
;
the & symbol [6]. See table iii, 8.

Out of 252 examples the distribution is as follows :

First method 6 189 or 75 per cent.

Second method o* 58 orl5
Third method . .5 25 or 10 ,,

But in the M
section the percentages are remarkably different, namely 36 per cent.,
32 per cent, and 32 per cent, as against 85 per cent., 14 per cent., and I per. cent, in
the remainder.
Medial ou
Medial au is, in our manuscript, always expressed by the circumflex sign com-
bined with the medial a sign. See table iii, 9. The more ancient method, which
combines the second o method [ o ] with the top e stroke [^] (and of which three
examples occur in the Sar&hap inscription) is not employed.

The following tables give statistics of the methods employed in the formation
of the medials e, di and o, etc.

JjhvAmQlIya Style
Section. Modi el f. Medial oi. Medial o. and of
tipadhmlnlya. writing

*© h ‘ill Vi •O’ O* 5 X i

A . . 39 46 13 9 6 0 42 1 7 a, «,

B . . • 22 28 6 1 0 0 35 3 1
«f

C . . 28 29 1 1 3 2 26 0 0

D . 22 28 2 0 0 0 10 1 0

B . 8 10 2 0 0 0 9 0 0 •j

A
F . 13 28 6 0 0 0 0 0

a .
52 36 1 1 2 0 20 2 3

H . 10 24 8 3 1 0 11 1 «f

j . . . 5 7 0 4 0 0 3 0 0

K .
1 4 0 0 0 0 3 0 0 ®1

L .
11 41 4 0 0 0 4 1
3 0 «*

0 26 23 23 6 4 a
M .
67 77 18

Total • 278 358 61 19 38 25 191 17 16

Perfteniagea,

Section M e 47% 63% ^

1002i 0% 36% 32% 32%
1

Remninder • 48% 67% m'o 30% 14% 85®;

Whole ICS e 44% 86% 76% 24% 15% 10% 7e®i

 m
A coraparisou with certain dated inscriptions.

Period. Medial r. Medial o. lUteenoee.

e ^ *o* 6- 6

xth century 70 30 58 12 30 SaWthan Inscription.

xith century 57 43 10 20 70 Chamba Inacriptions, No. 25.

? . 44 5G 13 10 77 Bakhnhili Manuacript.

xiiith century 0 100 0 10 90 Ant. Ohainb& pp. 63, 64.

Ligatures

The ligatures can usually be easily analysed into their constituent ele-
ments and only occasionally give difficulty. There are about 90 different digraphs
and some 30 different trigraphs.* Some of these occur very frequently, while of
others (< (/ . nch-, tp-. fph-, shn-) there are only isolated examples.

The consonants, as elements of ligatures, may be classed as (i) those in which

the inatrika is considerably modified and (ii) those which retain their original
shape. Of the former are k, fi, th, y, and in certain cases r.

As the first, or tniddle element of a ligature the letter k reverts to the form
pertaining to the Brahmt script ('I'able iv, I and a), but as the last element it re-
tains its normal shape, e.f/.. Ik<i (iv 3). 'I’he letter h occurs only in ligatures and,
except as a subscript to j, only in isolate*! examples {itch- iv. 1, shn- iv. 4). As the
final con.sonant its fortn has little resemblance to the matvika.

as the middle or bottom element is also very different in shape

The lettej- f

from the mat]ika. In our manusenpt it always takes the form of a spiral and
there is no example of the S shaped -th (See page 91).

In H/n/a (table iv. .')) (he spiral shape is lost but see the 8uhgal example given
on page 9 1

As a final consonant y is also disgui.sed in form. See kya, khya, gya, chya,
jyu, etc., in table iv,‘. Generally the final y is of the t3q)e shown in khya but occa-
sionally tlie liottom curve is more prolonged upwards on the right as in the gya
shown in the table.

As a lirsl element r is often shortened and sometimes loses its bottom serif,
e.g., nia, rth<t, rya (table iv. 3) and in rva it is marked only by a small excrescence
;

on the left curve of the v matiika. As the final element it is changed altogether
and is something like the virnma but placed at the bottom of the ligature. Nearly
all the examples in our text have the stroke continued left and right (table iv) ;
but. as a middle element the r con.sists of a stroke slanting upwards on the left of
the vert ical only, e.g., tryu and irya in table iv. 5. In hru it is treated as a middle
element and also loses the right half portion.

* « ^ . c'.iunting ciiuMdiiiints unly. hut ttihcn thn vowel aignn form p«rt of the .Toriical cheme they aro eMential elemeata alao.
Sec for « xampic. truu and tlhrU in table iv. n. Siniilarlv (hr cirniimflcx o, althoulfh detached, la an elemfmt of the yertical acheme.
tlifniun'M are rfvilly i'xamplofi of vortical writing white modial land the loft-handod v art example of r<ght-lo*left writing.
 99

In 5^^- and the original forms of the ^ and ih are, more or less, prQgerved,
“ streamer ” is added to the right of sh (iv. 4). "
but a The word ashta Eight
occurs some thirty or forty times but the form of the conjunct is not consistent.
There are, at least, three types (1) that shown in table ii, 3 as shi-, (2) that shown
:

as shth- in the same table, and (3) that shown in table iii, 9. The last type occurs
some half dozen times in the " M
” section only.

The top bar of the second element (or third) of a ligature is often oiniltcd.
3oe tta, nta, pta, ptta; sometimes it is given in a shortened form iv. 4) and ;

sometimes in its full extent {mhha iv. 3).

Numerals.

The numeral figures areshown in table iv. 7.* There is some resemblance
between certain of these symbols and certain letters, e.g., the 2 is not unlike the -n
in jna, the 4 resembles the k but instead of the top bar has a loop to the left, the .I
may be said to resemble the p, the 6 the final m, and the 8 is not unlike the h- and
is also rather remotely like the h. But this resemblance to letters has probably
—
only a fictitious value the letter-numeral theory having 8uccund>cd

There are two types of symbol for unity the semi-ciirulai enjAe and the
almost horizontal line V.)ut generallythe.se have separate functions
;
The curved
symbol is ordinarily used while the straighter symbol is used in fractions. 'I’he
other differences shown in table iv. 7 are a matter of style in writing. The ex-
amples in the first line are all taken from the M .section, while tlu)se in the .second
“
*’

line are taken from other sections. The example of 2 in the second line only occurs
occasionally. The chief differences in form are shown in the cases of the 6, a and
9. The “ SIX ” indeed is a fairly safe criterion in differentiating between the work
of the scribes. The symbol for zero besides being used as such is jilso occasionally
employed as a sort of symbol for an unknown quantity.

The symbol for “ minus ” is

a small cross placed after the numeral. Tloernle
attempted to connect this sign with the Brahml ka and to show that its use was
an inmeation of considerable antiquity. But he was hardly successful and the
details of his arguments in this matter do qot now call for any remark. See § 61 ()

and 127, Part I. No other signs of operation are employed in connexion with tiic
numerical notation.

Punctuation.

Besides the visarga the single bar |

and double bur ||
are used, and the

end of a sutra is marked by — The viramaappearatobeu8edtoinarkapau.se

or the end of a verse. On folio 5 recto t wo punctuation signs. wTiich are pos.sibly
cancellation marks, are employed.

Cancellation.

On several occasions letters and words are cancelled 1»y marks like accents
placed above the letters (See folios 8 and 9).

The flfnm ! Ikt Indian AMiqnary. XVII, 1888, p. 36, »itd in BQhl«r*t UbU IX at# hnrdly n pr^ut^tivo.
O
a .s »f» •5^

-< fr V X CP

a 40 at JET to to* TT ID-

qr « cr BT fcPr

i -o o •cr *r fcr t3 m kr

•s •p •? T rr
t a-v r
% Jar ?T 53“ CO* far

t >0 > a- ;7 tr

-fc •r *o •C' 10 X
a O o fcr rv IP

4. -• *0 O
A sr > > y :j tr

« H •r Hr H- •r tr- ir •r

o <=» IC7 ie^

4 1^ •7 a o
a e irr ezr ID py
49 -< 1C O' •o tr^ •r

i. H r t fr

4. •• >9 h9
"V K •r» ur^ hP

4. o O o o o K>

4. KJ O o t4 Kj

4 M K If |y

4, X. K
i. Ui tfr «r lo IP

=1 -6 •€ »« « »« Kfr

4 TJ M *IQ tr 1^ IF

4 U 1^ C( U 1/

4.^ B ET 3* r 2r X
<r s cr «r (T f;r 17 IT

4 r~ ?? e & e e
4 •t' le

XT r E r>

-4 •:> *P •o 99

•
* • t4 ;o ;r 2^ :n
r
ac Icr vf» pr> (?

ar SB* JR CP

MS

s
ad 6LE0
BAKHSHAli MS BAUNATK
’S E 'S
22
ACUTE-AN sarahan

BOWER

'd .1
ea 95
 TABIC N Con»o'*)Ant«

Aa, Aa Ca itn iA.». ttiik Ol».

2. Palaiali

cX». €,\», «A4* cA<tAA«

j^T‘'

3, CtA^I/ndt

^ f { ^« ^Ju. *k{x *ilfa J.4

n« nA na ^S
 105

TEXT.
bhyftsa i
” i ru [Ir.]

turguijiaihpamcha^^aihhastagatamdhaQa^ I

palhchagu^am 36 ||
naTamasatraih 0 ||
sfitruih ||
gtipaupri

thagrupayutauyachanjlyuktisamguoaulii ||
gupane^agu^e . ai

rupahinenabhajitau |
viparltayachanakshiptaugupaaSsyorayaThvidhil) ||
. .

evaihautraih ||
dvitlyapatrevivaritasti ||
daiSamaaatraih 10 || "^jr^ tl

II
sOtraih ||
aih4amvi4oddhyachchhedebhyakury&tatparivartanam ||
. .

sasyarhtatapi'ojjhyadhananvi^avinirdiiiet ||
udS |
paiiichanaravapijamadAye

wTlrnyatekilati tatroktamaoivikrltfimapimulyamkiyadbhavet

daih I
ardhatribhSgapadan^aihpanichadAa

I I I I [ 1

syam^atoprejjhyab^adriiarhkriyate |

. , . .
jatii 180 90 80 I 76 78 tatraprojjhyajataih 180 .
0
| |
I

.
.
I
esh&myogakii 60 60 60 ! 60 W tejatS 687 atosa . .

. AeshaihSTT e^amapimulyaiii |
chaturparasa^asaryasyam ||
prathamasyaaaiika ardAam

«0 1
76 I
78 chaturpamyoga 817 prathamardhepalashtibhiryutaih 877 pratham . .

.
pmthamadhanaih |
tjitlyachaturthapaihchamasyadhanaxiisarvasyam 847 dyitlya tribbagaih 8

.
.
yntaih 877 edadyitlyasyadhanambhayati ||
punaprathamadyitlya chaturtbapaihcha

. . syarh 867 tTitlyasyapadaih 80 eshayutaih 877 e8hatn%a«ya . .

, . . ti punarapiprathamadyitlyatritlyapaihchasva seg
II

. . . MAaoh .... bKavaU ||

 106

2r. J
syadhanawbhavati ||
athapratha . .

tyamsashti^esharii 8’^^ alhadvitlya 76 sya 180 QV&fh

||
80
ya ayakxiyate 1$0
. . . . mdvitlyasyabhavati ||
athatfitl 78 80
7« 180 9
377 tritlyaayadhanambhavati ||
chaturthasyaknya I te .

! 78
90 2 .

evarii 377 chaturthasyadhanabhavaiiiti ||

pamchamasyakriyate |
stha
80 7.
panaih 180 evaihpaihchaiuasya 377 ||
eahamaoimulyariipra
90 15 7.
80
il II II
anyonyaviditavibhavaih 78
75
18 vanik . yaiii tvi dalarii^atha
| II

2v. ]* n
aTTi^aih 7+ 8+ 1

ioddhyavisodhayet rioarhsthitarh e 18 18 6 Isha

|

18 18 1 6
yate 19 7 11 kuryatatparivartanaih 18 4 6 rAchhede

18 4 6 19 7 11

ihjataih |
asya |
984 836 798 projjhyajata |
984
1
1

M68 146^jah&ihyutiihkfiyato
1

. . . . e 1463
|

krit jata I
2668 |
chchhedaprojjhyam 1096 etanmaijimulyam

The bottom portion of thin pajctf i* bUiik.

 107

tv&SBamadhanajataprastamulyarntaduchyatam f
4 . 6 go |
6 43 |
e

praatamulyam 2 | 8 | 6 |
dattaissamadhana jatfi 17 |
17 |
17

iamaafitram 18 || sutraiii ekayutanasaiiikhyadvi

||

. . . Aindcha ||
evamtavatkaryaThyavatpurushaisarnabhavati

. ptam patrebhilikhitasthitachaturdasainaautrarh 14 ||
->««^

. tram |
gratisyaivavi^eshaihchavibhaktariipurvagariitvinah tenaivakalaihbhavatis^Aa

. . . . kenatu ||
uda ||
addhyardhayojanagate4atahaintasa . . .

j^asyahayannavab ushtrada^atyitlya . *

pradattaihchaparasparaih | piithagdhanarhtuvaijijainmulyaiiivapi'ayinarhprithakyad

vaktuihtat-omechchhindhisaihSayab ha |u 10 vapijjaka Sd.ya.

I®

. Qikpi 9 d<i'batam |
pioda ^ f 8 |
10 |
deyaiii 3 4uddha4esharii 4 6 7 lata
| |

paraaparakritaihgu^itajatam 168 |
168 | 168 | sva^esheoafuvifcAo;^

labdhaih 42 (
2B | 24 eshapratyaikamalyaihekaikasya
|

gt/Qitajatani a4vaihayai ashtrebhyab 294 |

262 |
240 | ekaiA'a

ihjata 282 I
262 I
262 I
 108

[ 4r. ] . t .
. 7 . . . oh

yojanapamcbakarii |
saptadin&nitasyaivagatasya |
paratadvitlyanavayojanatfta gatake

"l . 4 -
.
tSm 1
1 di 6 yo 1 diiia 7 gatasya gatayojana 36 dvi 1 di 9 yo gatbyaava viteshaihchaka
|

L» ^ .
' ....
^

yate |
gati 6 |
2 |
vi4esbam vibbaktam 1 purvagata86
|

bhirdiDaisamagatlbhavantinavayojanam ||
pratyayatrair&4ikena 1 di 6 yo 85
1 1 4
ud& II
ashtada^ayojana ekenadiney&ti |
tasydshtadi
1 di b yo 86 pha 81
gatasya |
dvitlyapamchaviih^eyojanadineyati |
kenakile 1

. aSsyatSrh ||
evamekada^amapattrebbilikhitapurrepi ||
paihchadaiama Biltraih 16

.... yorvi4eshakartavyaih uttarasyavi^eshatab vibhajctamsuttare

l^as&Byaihg . . i . .

14t.]

taraiii 2 vibbaktam 1 adi4esha 2 jatfi 1 dvigupaib 2 I lupasamyutaih I 8 I esha

A 1 1

. saihkalitepratyayfipadamhlna ubbayesthSpitavyfi rapo9&karapephalaiiil

181.
'217
kimprabhutepilikhite ||
ghoda^amasutram 17
II

nillrubliitiiiiinftHti

.... ddyorvi^ghadvigu^amchaya^uddhiivibhajitaiii nip§ . i

patis&gyamtad&bhavet || ||
ud& 1 1
dvaySditri chayaacbaivadvi .

. dikottarah dvayochabhavatepaththakeoakSleDasagyatarii ||
Bth&pananikpyate |
e.

pa 0 dvi I a 8 lu 3 pa 0 karaoaih fidyorri^OBha . .

|
1 1
111 1
. i
 109

ft 5 u 6 pa 0 dha 0
j
t«r.]
1 1 1 1 karaoaih
1

|
idyorvi^eahafftfidi . .

10 1 u 8 pa 0 dha fl~|

1 1 ehaya^uddhichayam 0
VJ t^uddhi 8 Ifidiioaha f
j

1
.. | 3 |
>

dvigU9am 10 uttaraviiSesha 3 vibhaktam i

10 sarupaih 1 18 anenakd
8 3

aamad hanfibhavanti ^ pratyayaih rupoi;^karanepa phalaiii :

65 >

I
dvi 66 I

ashtAdaiamasutraih 18 ||
dinagamana madirahitaih .

pamtachchhottarepa saihyutam |
pratinihita atmagupaiiijneyam

o . i I
ashthotturagupitekshepuaaiiijbakodatvfimulaihprati . i

hatam [ 0t. ]
|
|q |

. . . dinagamanamSdirahitamdinagamanayojana^paAcha 6 idi
j |

. . .
8 rahitarii jatam 2 dvigupaih 4 tachchottarepaBamyutaih
I
I
|

|
j |
8
j

. .dtmagupam 64 e^akshepasaihjfiakorftsi |
ashtottarasaihgu . i . .

| |

. /a&dharSahi 80 ashtagupaih 240 I uttarepagupam uttaraih 4 I .

| | j j

. . .
gupitaihjataih 960 k8hepaBaihjfiakodatv& |
tatrakshepasaihjfi . .
1
1

4 I
yutaihj&tani |
1024 I asyamtllarh
|
82.... |
pratinabito

i m 6 ywtamjatarii |40|u....ifa....

0ikepratyaya 1 5 5 phalaiiiy
[«r ]
1 1 1

nenaseaha 56 etesainabdhanam ||
uda ||
adipaibcha uttavarhtr

pinaroyojanagamyate |
dvittyapratidinaibssaptagatasya .

napaibchakaih kenakalenasamatfiihkatthyatfiingapakotta . .
|

u 8 pa 0 pratigati 7
5 j 1
11

1 11
1

yojanikaifayojaiia
'
85 I karapaih |
dinagamanam&dlraTlt

tatradinagatnaiiaih 7 Rdirahitaih |' 8di I_

rahitaifa
( |
 110

*
i ^ ! anenagu9itariij&tam |
84 ;
. . .

san^jiiakodatva tatrakshepaiashi |
49 |
datvajatmii I
889
|

. . dinadadatlBamaih |
karaqiokriyate |
siitraiii
||
akrite^li . .

. . . chhedodvisaihguii^ab tadvargab dala |

aami^liBhtha |

.
. yab anenasuti'eiiaslishtbaitiulaiiiaruiyasvaniatima . . .

telahdham mulaih 29 pratinihitarh (

7 anenayutaiii ;
8tf

48
48
. 2188 dvigiu.ioltai'abhajitaih |
ta . o
.
I
58
I

[It.]
•

I-
7

9
dalita

sfiayeyu
16
60
88
cha

li
1

tarn
58 _
787 padaghna tatrapadarii
m 29
j |
80
anenagupitam 6 vi jatam 58 86698 slisA . . .

60 841
tya4eshaihkyiyate | 4e 6 66689 bhii Igehyite 7
29 841
.
.
pratyayarhtrai r& ^ikena 1 7 yo 17

phalaiiiyojana 42 28 i
niyatamtena 1 1
1 29
^n|e.o. 1
1

j
vim.4atiina sutrai

[7t.]

. . matkiidanab >1
&
11
3

1
u 4 ,

I
pa 0 1
nitvadatta 7
1
j

8 niyatam 7 vi^eddhya 4 juUarardhenabhajitam |

uttaiaihj •

I j j
j

..... jitam 4 jfttarii 2 labdhaiiisarupa | eaharupSdhikaihI

8 e4ak&la| |
I

2 -- L. .J I

. . . .
3 u 4 pa 8 rupunakarapenaphalamru 21 || dvitlyasyatrairi
, .
I I

1 I
1
I i di 7 8 di i
I'd 81 II
eshassamadhan&jata ||
?tda
1 1
li 1

ka uttaradvayariidvitlyapamchapratyaharh |
kenakalenaaama^dm .

pakottama fi 1 u 2 pa 0 nivatanityarh 6 h

^
1

1 L-1 ‘1 '
lJ
 Ill

II
sdtram dvigu^a/nprabhavarnyucldhadviguva :
daya . C8r.]
j

*f* ttare 9 abhajechchhe 8 haihlabdhamrupaiiivinii'di 8 et ||

ud& ||
va

bhvitakaxkaschitatiaikodushaiaa^akaiii |
pratyahaiiikurutetatrakarmarabha . i

minavab dvitlyamkriyatekarmamdvyaditvitayaruttaraih |
padaratatra

. kenak&lenas&syatam 1 a 2 u 3 pa 0 prati 10
1
1 1 1

t, . : . . ...

dvigupamprabhavaih^uddhS |
prabhavamj 2 jdviguQath 4 niyatupunadvi

0 . . tarii 16 (uit.arardher>abhajayet)* uttaraiii . .

*ThA portion within the brurkela ia deleted in the MS.

gantatal eka ^chagatnanajfieya yutdaaajnguijya [87.]

niyorathosvairda^abhiryujyatehayapaiiichakaiii |
gamtavyaiiiyojana
II

. . oiivakimudbhavet 1 ha 10 i
havalagnamthaava
‘
5 l gantavyoyojana
'

1 J
1

yorvibhajyagantavyaTh |
tatrahaya 10 gantavyaihyo 100 atobhiiga . . .
| ^ 1
|

bdha 10 tatrayukta^va etais8amguijyapariyogaj5tam . . .

I
j
|^6 |

. yojanfinyaikosvarudha (
pratyayalj paThchabhi44at:a8aThgu9yajataii) . .

kfiyate ||
yadida . .sAoyojana ^iparhchaa .... ...

r . shth . . ch . . . udS ||
Ote.g

. . . . m&ptamdvijanmabhi |
tatpunastesamaihbhaktvhda .

# _
ssamaptaviln saiiikhyayax katimiiiachakshiikativipraj^kuti . r . . rh

1 'U 1 pa 0 labdhara 10 karapani ||

labdhamdvigupitaihkfitvatatra
111 1
'bdham 10 dvigupam 20 jtathfidyunam 181
L
j

— j j

1 I

vibhajitaih |
atrottarani anenabhaktvajataihtade

. kaih 18 ayaihprashQfiprShdQfi ekonavim4ati |

sth&pa

pratyayarh a 1 u I pa 19 rupo^Akarap^enapbalaih I 1
1 1 1

*Th» fint m it in th* MB.

y 2
 112

[ ] yo\a,7ui serathagama

JO 4ayo Ijyo 70 gantavyaih ||adhva»an)yo^d) 7 t I bhaktam |

1 I

111 l-lJ [l\

enaguqit&j&t&nlabdha dviguQaih 20 eshftlpasyaU atha
I
1^ |
j
||

nft ayaihkftlojfieyab anenakftlenashshat yojana,nigantavyarii |

S,

. yamekayojanikasyasamSgamobhavati ||
tadyathatrairfi^i . e . . .

yadyekasyashat yojanatadaviihdanarhkirh 1 6 yo 2
| 1
saptati^oddhyatesha atraasaptati 70 agatapaiiicha^a 60~ adhva
[ j j |

m 1 di 1 yo 20 di pha yo
1 1 /

sutram 24
[lOt.]

.1 .
dhdntn snw'g«i)a pra/vritit>j7 u

yinirdiset ||
uda ||
tribhagainalada^dAasyatridhantasyaiva .

111111
a«Mott.ara^atanidatt.amkiTii^eshanivadapandita ii 108
1 1
kiitvfirupakahyaiiiparthajata j
82 Isesha ||
pra I 8+ 8+
tbamabdhantekshayaTii 88 I ^esharh 72 dvitlyabdhsuitekshayarh 24 I
ieskam
j j j

tjitlyabdhantekshayaib 18 ^esharii 82 pratyayaihkriyate 4thfipanarii

j j
|
j

0 111
1111 bha ^eshani 32
1 ;
phalaihmulft 108 ||
” atha
3+ 8+ 8+ TnajStikriya^aroharii . J . .
^
 113

. . . . . ^xe^lyayam [lOv. ]

tribhitryashtabhagasaiiiyutaih

tfl^da '(
1 shtottara4ata x kiih 27 1 108 plia i^e

3+ 8 1 1
I

yadyelcasyatrayastraya ashtha

bhagatadadvatfin^anauikimiti 3 32 phalarii 108 uda II

•lI 3
kjiddhantasyalohasyadaaamshakahiya 8 teatrayaiii saptatedviguijia
|

achakim^eshariivadapaydit'ati 3 140 kritvarnpakshayaihpasthamiti

10 1
rupaih kshayainkritva jataiii^esha I 7 ! iiiulam
^ :

1 10

40 I
anenaguijitariijataiu [_98J kshayaiit) 42 I evaiii 1 40 pru

na 7 I
1 I
98 I
phalant 140 ....
1

pratyayah 0 [
Hr. 1
. .

1
latribhagaiiikshayavrajati ashtada^n
palakrlte pa 1 |

3+ 1 18!
thotambru 10 hi ||i 3 I 1 karanaih |
arldhyardhapalartisflihe.
, .

! 3 bha
2 kptvarupaksha 2 yaiiiruparii 1 ks!iavaTiikrit,v:ijatain 7 [181
idaih ; i
j

9 1
-
I
I

gUQitaihj&taih 14 kshayath 4 pratyayatrairiisikeua ||

addhyardhapab
|
j j

krltet/ribhagariikshayagachchhati |
ashtadaaapalakri takirtik shayarh vad.

pdita 1 1 18 1
phalaiii ^
I
punatpbhSgadivard haii)tadacha.tu
1 3 1
1
1 *
2 1

kimiti 1 1 I
4 I
phalani 18
III
udS. II
ohaturbhagamala . .

3 1 i
1 I

2 1 I rn a datapath
 114

[ ny. ] ... I
•
jJ
athapratyay . .

1 4 phalaminal& 600 ||
punarevaprastfirakra
1
5
4+ mam 1
|
600 1 1 1 1 phalam
1 1 1 1 1 1
shaihsthapya 4+ 4+ 4+ tediti
4+ 64
1
4+ ^esha 168 to 1 1 anyaihchaturtbapratyayamkiriya

1111
|

64
1
4+ te 0
1 1111
4+ 4+ 4+ 4+
bhfi4esha 158
1
6
phalaih 600 ||

1
ddyarnkshayam 126 ^esham 376 dvitlye 64 kshayadi 98 to
||

a6a 9 ^e.^^haiii !
881 kshayam 70 desbaih 210 ksbayaih 69
1 1 6 16 47
sharh 168 e^asarvaL 4 trakartavya 16 16
| | It

[12r.]
tuprasthama

dhunastathab ambhasa !

kfitvarupakshayampaBtamiti. tatraksbayaih.pastainiti .tatrakshayab I'O . . .

• • •

aya^esbam 3 3 3 3 4 gadyutigadyutigatvStprastbaib . . t . . . .

4 4 4 4 1
gadyuti _ yojanaih |
cbatuprastbai fidbakaih J tadfidb&nta^orguna

. . tatab '
91 1
gvrittipraTiittirgupanamto* *tab 11 4 Hanena . .

I 1

tji . tamjritaih 81 eiamaddbubb5gfibbagebritelabdhaiu |

madbuprastu 1 .

Ae 1 1
aihbha j |
bhlgApraata 2 kudava 2 16
18, 16

. ktiprakahepake S.dkak34o(^ashakuday&bbavanti |
16 |
atoma

^esbaih IS
 lit

16 e [12v. 1

pragthaku<}av& | 4 I
8 4eshachatv&ra

. . ku<}avaU 2 2 toshEchakudavfipIta | ma 7 9 ;
punacha(«i7r
1
A 1
M 1
X 1
4 4 sham 81 176 jalabhS 4+ 4 1

j
gam I
madhuku(Jnva 6

jalaku^ava 10 4e 16 evaihkudava 16 lula ||

datvft^ulkrtw
|| ||
1 16

chaturbh&gaihashtau &Q.Itakuihkuma |
chatuaulka4alai8tukiih6eHhaiiiva<i«

^4.ta 8|l
*
l^karaDam |
kntv&rupakshayariipaatam |
pastaiii |
8 8 jgu . i
“ 1 1 4
. . . tarii
ay; 6 I
4ulke I
8 iesham 0 1 ane nagu

«+j
til . kshayaih 1 deshepa 4 1 datvagupitaj&ta 27
1 |

1 1
2 2 4+

. . to

1111
1+ II- 1-1- 1-1-
guiidtaihjatom I

'
81
82
u
[
18r. ]

4 4 4 4

phalaihjSljphalaih punSnyaih 8 phalara'. 81 jpunapratyayaiii|0 bha

||
1
phalaihkumkuma 8 ||
uda 1 + II
tribhagashadbhaga IpaibchaT.
4 j

gudapipd68ktabhfirakam
1 + kimtosharhdattabhirbhavet
1 +
. | 4 4
1-f- 1+
S II
2 I 6 1 4 I
gupitafl^jitaih 82 jetatphalani || || 4
/ II 8 I 6 il 5_|

chatu 4 paihchakala
h 9

I bhenada^adropStprayojita
14
4

dvaityibhiBtukiihl&bhaihkatthyatRmgupakottama :i
10 I
5 [ 5

A ... 196 I
 116

i3v. I
0 phalaihdro 10
( ]
S 1
4 10
III
'ids II > i“ys
2 1 4
I
I' a
1
... jjakasya
^
4 dro sbashthisvadale
^ uakshayaihgata
pra 0 1
4 I

u vfiddhyatril'ha, |
ge 4 a pra svapadena 4 tatojjhitam |
' .

ku 2 ;

ddhyatuj)anichabhagenastatha ku 4 praathiJ vpddhidvayogatam kavriddhi.

|

. syakiihva^eshariitadiichyatarh 80 1 1 1 1 i rupalft .

1
|1 1 1 1
ta 36 pratyayaiiipu 5 I naayaivo
II

1 1 1 IbhS 86 plialani 60 ||
punanyariipratyayam 80 phalam
+
I

1 1 1
2 3 4+ 5 1
1+
ayevarnulamnajNayate + + phaj 2 lafh
||

[
i4r. ]
lat . dhan .
|
yasyataninayatachakshu .
.
y . . . .

.
palijitasulkapii.Kjaiii 24 ||
karapaih ||
kyitvarupaksha d

pasta 2 3 4 jatusariigupyajatain 2 etavadapirupasaiiii^u

3 ! 4 5 5
ajatam 3 aTienabhaktva^ulkapii.idaihgiinitamjatam I 40
j

6 I

esapindeibpratvavaiii 2 40 gunitajatam 1® ^esharii 24

' j j
'

: 5 1 I

ni 40 I I
anyaiiia '
syapfatyavarh ;
40 phalarh 18 kshayaih 2 . .

vain 40 ndh gndapindal ^ jfiatatulyo4chatvari

II II II 34.
1+
avvegudaii) |
tricbatn paiiichasbadvriddhya 4 chatvariih^avekshaya
1
• 5+ . e .
 117

u I Uv. 1

uda 11
ajfiatorambhalohasyati'i ^

chatu^paihchakaksliaye sapt .
|

viiii4atipi9<jasyatpdhaiitaie8hya(l)‘iahyate kiihsarvaiiivadatatvajnakshayamcha .
|

makatthyataih 1 1 1 I ie Ij karaijam kyitv^rupakshayaiiipilstha 2 8

1 |
j
8 4
gu^itamjStam 8 rupakshayam 8 1

Isl
1 i

[Sj
anena^eshaiiibhaktarii^esliaiii 1

—
27 1 bba

til jataiit 45 aayasaptavinsa patyaseahaiii 18 etakshayaih udfi

| || || ||

rikshlnasyalohasyatvidhantariiparhchamaf^akarii |
najftfiyatetpravrittikaihna

. . ahapradp^yate |
pravrittiieshariiyopi^dadikevalamvuh^atisthitam |
a

thpravfitUsyakirhvai^eshariivadasvaTue 1 1 1 I
kfitra
|

4 6

. . . . kahayasyacha . r ^ab
[ 18r. It
yamkalchiyadi^akyastaduchyatfirii ||etanine8aih4ayamprajiiaddhdbgntakBhaya7^

vich&ra^&b ||
2 8 I
4 ksha^e 82 |
kara^aIh II
-bdhautasaihgu^yagunt
8 4 ; 6 1
a

jfitaih rupamdadyg 8 bhagehiitelabdhambhaktve^t

5
j&ts 82 phalaih 20 e^asgpravyitti ^eshaih 12 . .
|
1
82 path chavirn Fatima sOtraih 25
II ||

X
t He bottom portiun of 16 recto ii blank.
T
i

y pravjittibhave^jaAAe ||
[ 16v. 1

1 i| 1 4e 16 |(karaoaih ||
dhanta^oghatitaihtena |
rupa
3
I^BHayaifakritvajataih 2 2 2 2 gupitam 16 bha.
j
8 8 8 81
deshepaguKiaye ^eshani jl6j i gui^itajatS
j81 | ^

|
8

pravrittintyarthab ||
athfinyavidhikalasavarpe |
chaturdhanta

lohasya ekR^ltischadattavfin kiih^eshatiivadadhannajftaya .

ijitekritartigramaTh 81 1 1 1 1 II
phalarh^e 16 || pu
"
||
A
1 A
1 1
* 1
‘-r + ,1

yaihkyiyate ni ii larii na jfigyate

III .1

1111
+ + + +
1 1 1 1
bh&4e 16
1
i
phalarhlohapala 81 ||

3 8 8 8
 118

I l«r. 3 vibhaktamjata/n I 4e 10 I dh . . .

’1
0 I

aoena guijdtamjataiii 90 2 bhagehfitelabdhaih 12 ||

|

asyapratyayatraira^lke I
10 pha 12
j
j
il
i 6

H
udu II
makshikaggbatakasyaivadvitiibhagapravardhi

. iyedvipamchamobhagot|'itlyedvi.saptakod&/mvamcAa^artAe .

. Tambhagaihevaiiijatapalatrayaih {
babhuv&saulkikaihritv&

rvamvadapapdita 2 ,
2 2 2 4e 3 dh&ntaiogAd
,

8 1
5
;
7 9 1
II

atimiti I
k|itva

[ !«. 1 sutram II
idfiniauvaroakshayaihyakshyi
. 1.
.
yasyedawisutraih ||
kshayam aaihgupyakanakastadyutirbhftjo

yettatab Bamyutairevakanakairekaikasyakshayohiaab ||
udft ||
e

. . . saamkhy&suvan^amashakairiQai |
ekadvitrichatussamkhyd

* *rahit.a8amabhflgatnrii BVapananiknyate eshArh

|| |

.+ 2+ 3+ ;
4+ li karai^am ||
kshayaihBathgnpyakan^k&di . i
!

•
L .2 . ^ Is ii 4
^ — -|kBhayenasaIhgu^yajfitaIh |l|0|9|ltf
ti I
eshayxiti 30 |
kanaknyuti 10 anenabhaktyfi la

[lit.] uda ekadvitri 1

II II
10 “80 4 phamft4e 12
»<lr^aprojjhita ime 1 1 1 1 iia&4akadyi

tntSihchaivachatu0pamchakarS.Th^akaihkiihkshayaih 1 2 8 4
1 1 1 1 1
kBhayamsamguQyakanak& e^asthApayate 2 8 ;
4 I
0
|

1 2 8 4 stadyutirbh aj ayetatatab hara8SByek]iieyutam

2 8 4 5

•68 Baihyutaixkanakairbhaktvatadakanaka 10 anenabhaAtom

•
0 |

e^a ekaikasavarpasyakshayam
j&taTh
I

^ ||
• • •

^ikenakM
I

•
•
||
— —
 110

• • • . .900 . . $rut.iuHa7ne |
kram«nadvayam [ I7r, ]

*0 188 4 pha 168 sh&di uttare ekahlnat&m suvaroamme

|

1 60 1 160 sammi^ryakatthyatamgaoakottama sth&panafl^

||

6+ 6+ 7+ 8+ 9+ 1+ 8+ kshayaihsamguvya

L 6 7 8 9 10
1

1 ^ 8 ! 8 !
4 jatam |

80|4S|66|72|g0|2|6| ei»amyuti |
880 ||
kana

. . ihyuti 46 |
anenabhaktva labdharii 880 |
pamchada4abhfigechchheda
paihchada^abhfigechchl
|

I
phalaih I 7 4e 1 I eka ekaikain64a j
kakshayam | pratyaya
I » I
* *
*
880 1 phaladi evaihsarveshfiihpra^
I I
i 1

k aly k3)i^hanaiyadbbavelabdhasakBhayajAatam&4aka ||
udft ||
.... [18r.]

shakopr^ptodTau .
praptaihchapaihcbabhi |
trya^chakatibhi ^praptfishadeva

Dikevalaih |
chaturbhimashakairhlflamkatidnsbtbvamayasakhe trayafeba
|

katibhi ^pr&ptftauvarpariiinaaakovadab ' 2 8 karaqam

1

aprftptaaaihgaA&katIditi 6 Li-. 6 0 4+
1
i ty&ra 4 aaifagupyajatarh 24 '

kaTbcban3nitatojjbitaihdvabhy&m
j

kapaiiioliabhidyayaihaaihgu^yajatain 10 tad 3ruti 12 bi . 3)?

2 |
j |

hitv3j3taih4e8haih 12 |
apr3ptaga9dikai . . i ....
R 2
 120

[ 18t. . shtaviiiiiatimM&trMii

^ sutram unaisaamguvyakauak&tatpit^^jamchavidodhayet su

var^akanak&bhyastara^isheshamvibhajayet apraptaga^<jii

. ^a^uddhenakanakenatu |
yallabUhaiiitatpramapamtuga^dikayfi vi .

. . t II
uda II
ekadvitrichatuasaiiikhya apr&ptainfi4ak&ni tu

. kadvitrichatuasaihkbya ekatravartitakilab gapdikSjnd . .

. . kfi dnaikSda^amHahakai |
apraptajftiltakanakaipra . .

yab 1 karaii^aih u . ai
1

[ 10 *]
FoUo 19 it blank.
 121

[20r. J

f. I aayaiva^prashQasyft rovi

. . 20 raktl dh& 1 su 1 chhe 80 raktisurakti 1

1 a 1 1 1 1
. * 0
||

. pd 1 .punat)'i ya 4 tiyfisyaiva 2 20 1
II j

1 |
8
chhedarii 6 dha draphadhd 4 ya 1 L * 1 1 chhe"
1 1

8 2
pft 2 mtl .
II
Buvar^asyam&oamsama
1 chhe 1
4 4
dft II
Bapaihchanavabh&g&Qidin&niirayodaBal)i

Qftihkiih II II

•4 B 1 chhe 7 chhe 2ov. ]

1 I

pa 12000 1 1
1
I

ara9am |
sarposhtada^ahastopravi^atyftrdhfimgulaiii^anava

. i ekaviih^atibhagammapaharaihti |
pratidinenab kiiiik&le

vilaihsaihprfipyate 1 1 1 1 18 chhe 24 amha

2 21 -f" 1 860 1 1
. 8 mfi 4 di 101 1
9
II
kita X kil&rdh&ihgulaihdivasedivase

parkchfiid . . . ke . .

tribhdga [ 21r. I

. . dinetatha |
trirUpaparhchabhidinai |
. sh&ihda

rQ 1 ru 8 djishya 100 1
karaoaih ||
11 11

1 1 1 _ 1_ 1
i:
1 1

di jj
8 di 1
,
krityft |
8 |
2 1
8 dri 100 1

1 _ii !
1
'
1 1
]
kJ 1 1

. bhaktath 6 |
dfishyena
!

tThii iMf eoMbto of tis ocroiio borito tte portioa nurM (a).
 vardbamtjitlyasya jlvalokat eshfirndlnSr

. . ka»yakimbbavati 9 dl 8 dl {

1 1 1

rivartanaiiik}iyute 2 0
||
1 1
86 dji 600 1
praksbe 1 di 1 di
10 1 1 2 :

[2*r.] 47 6
.

947
rtitajfitaphalaihd! 500 ||
asyapratyayatrai

dikena

22v. .] dvigupaiiidvitlyasyaprathamd

. . .
.
I pratbamachatur^upamchaivachaturthechaivadattaySn eha

. . . am^atamekaiiidvayanugaih ||| vadasvaprathamedattaihkiiiipramftQaih

sya . ijdyishya 200 4l!inyamekayutaibkiity& 1 9 9

| |
I j j|
1 1 J j

kahepayuktyapbalaih |
90 |
40 |
90 | 80 |
evarh 900 ||| eshfiih

S. 90 ;
90 pa 4 rupopSkarapenapbalam 90
1 1

‘II
sfltram || yadrichcATi'fipinyase^ll^ye tadfi ih
 123

dftchatfigu^a7nd [ 23r> ]

prafAamasyatukiihbhave/

yadTicAohh&vinyase^UQye

. . chchhi || 1 1^ tad&vargamtuk&rayet 2 2 8 6
|
^ | |

. kahipegu^itam || 1'
|
^ I ^ |
M |
I
_ J j
i

prakahiptam 88 ||
drishyamvibhajet 182 '
vartyamialaih |
4
88 I
I
1
^adattam ||
atonyftaab ||
4 |
8 |
24 I
94 I
....
eshayargakramaga^itam || ||
athayutivargaih/rrt . .

s taih II
. S . i . am^A^yeyinyaatamtadachaivakraineguoaih

kritvachaturtha E 23 v. ]

prathamasyatukiihbhayet . .

. .
8
1 I 1
^
.=«=— ndkaih 1 ||
eshaoyastariipra

ra4 tadgchaivakrameqLaguoitaih |
1 | 8 | 9 |
48 |
eahgiiiyu

. 60 anenadfiahyaifabhajitam 800 j§ta '

® ®
.

^
I
I . i

. . ayadhanaih ... anenakshepamguvaye | 5 |

10 | 46 |
240 |
e .

. .
yutivargagauitaih II II
ad& II
prathamasya .

. . . thariidattaihchavaidhanam |
^sachadv&rdhayi/todattaih . . .

4atam chatu^chatva^middAt [ 24r. ]

'dattariichaivachaturgupaih* |
Amprathamasya

4uQye4u 1
1 .

yutaihchaivagu^aihtatab yutamchaivaguqaiiikfitv&kSraycgaija ....

. . natu 6 gu^aih I
j^are uparaihadhe adhaihguoayoj U) sardhadv .
.
yutaib

. . tlyarll ^y&gU9ai(^aih |
sSrdhaissaptabhitrlitii j
45 ,
sUrdhatrayayut^m
I 2 I

. chaturthar&f^iguoayeshshadvim^atibhi |
jata j
208 I
| _J sardhachaty&riyu
1
2 I

i evarh dyi ^yarii |

sarvaihtadevajaiatn

Tbi* phcu* ia Mated ia tha M8.

 124

I 24t. ] tris&rdhayu .

urgu^a7/lcAaturthenanav&rdbayutamdattaIh |

tildv&vim^adhik&kimatraprathamasyadatt&Bi^

'

'^^'121212
d I
2 5 I

I
8 7 4 9 1 ekatramdattaih 222
*
4uQya

j
^

erik datva || ^ II
yutagupitayutakraraepajatarii ||
sthfipa .

67 867 drii^bya 222 prakahepenajatgm 229

2 12
I

,
||

^ 1 _) ntab dn.4yab 222 '|| jft .

II
udg II
pratbaniariinajanaini |
divardbayutaib

pariipamoA-arrfAat/Mtariiprathama

( 25r. ]

ru^amdiyate |
yutamjatarn dvitlyaguparh parya

tritlyn ekatregiipitam |
yutena |
ynt ani 1 1
10 j
2
I 2 S

yutarh I 8.3 gupitam 132 ripaihjatam

1
pgr^a
| | |

e^anygaa i 6 1 5 |
28 I 128 I
driabya 78 e . .

!
2 '
2 i
2 !
1 . . .

. ! . , 6 1 vibhaktavyaTh I 2 1 78
2 ,1
'
11.. I
 125

rao • • dQpyasthSne [ 26V. ]

rupamdatvS |
1 | 7ut&j&t&. 5
2
16 prathamfttritlyasyatri guiiaii)y utaiii j ata m
e
rgu^aihnav&rdhayutarhj&tarii 29 ekatranyasS 16
2 2
. 9 dri 71 ,

prakshiptaih 71 1
bha 1 ktamdriabyannatkin 1
2
• 1
dnena 2 sarvamgu^itamtadeva I
® 1 22
1

ekatraih ii
«da II
2 1
.
2

. sh&mapai'ovidhi^ ||
prathamadhanaiiidattarhnaj 6 tamkiihtudivavdhnyutaiii |
tadfidvi

tjrMijamdattaihpaihchardhahlnam |
tad&tjitIyenatpguijan»dattaihsa]f>/«/-dha

. . thfichaturthe 9 achaturgu 9 amnav&rdhahlnaihdattaniekatramta . . .

im m 2 5 3 gw 7 II
* gu
1 24 " 1 1

hara^afl^ . . .

[ 26t. J
||

rQpaihdatv&li yutaihjfitam « i P
2 I

m 6 prathamatritlyam 1
tfiguiijam
2
m& chaturtbaThchaturgaQaihnavftrdharahitani |
i^eaharii 11
2
6 6 ;
8 I

i
dn 29 prakahepayuktib 29
2 2 i
2 8 !
2 2
bhaktam
29
2
il
jiltaih
U
. gUQitaihtadeva |
evaihrivarft^lbhayanti |
triprakararh

samSptof/t ||
^fiQyasthftnerupaihdatvS |
tadanuyuktara |
guiiita
8
 126

[ aav. ] phala . .

athadvau 4 |
SO asyadalampAa

aMdshtha 8 |
S2 |
dalarii pha

bhu 86 16 I
28 dalarh pha

84 I
10 dalaih pha

athatiipi usara da

36 I
20 I 4 I
asya tfi

32 I
20 I
8 atho chchhed

28 I
20 I
12 puna

24 I 20 I
16 .

[ 27r. ] chh fasmStki

.
y

.
« kaia^aih |
prithakrupariivinikshipya |
pi’ithak rupam kshiptaihjfttaih .

||

» jl_^j !i bhyaao |
tatragupa |
8 '
4 i
abhyasarii I
W I rupahlnaiii |1 . .

abhdgdch&iu<f>pnxhcha ^.5 |
atrakshiptamjatain j
15 | 16 | e^atrigup

. . . tamula . nichatu<^paii)cha || 8 | 4 |
esha

[ 27v. ]

maayadhanaiii |
eahaiiinmikkramepapurvokt sth . . n •

9 pra 7 dvi 10 tri 8 cha 11 paiii yutariijfitaihpratyaiA;

1

. m 10 tri 8 cha 11 parii 9 pra 16 I

17 I 18 I
10

aih sarvatra kSrayet ||

17
(a)* 25 dn
[ 28r. ]

tamdvigupamdvigupaiiibh&ramas

labdhaih 14 ||
punakfiya

*Thia i» iu>l homogonrouH. Thrrc bxv two leaver. Ther» are other gmall sorapo.
 127

(a) * * • • • [28T.1
II

1 1 1
gu^aye |
16 | 8 |
guiji . jatft 9

ahiitva afjho shagU9a

11
abhagasyadivardhaxlum 96 2 phalaih

phalam 5 . o an

<a)« . . —• II
siltraih II
g\3in<iu .... t 29r. J

kadhanam ||
guQanyasorupahinariilabdhamriiparii

(b) prathamai^yasya |
tatrechchhapaiiichab |
6 |
tatprathamacha

4 I
16 I
tadadiiiodhayetkramat adinu . .
.
pathchaiadv

(c) kr ch 4

etochaturda4abhi4oddhya4esham npamcha

(d) . . 'diaet ||
ud& ||
dhanfi . .

Byadvitlyayonmi^raihdhanaihtatrattrayodashab dvitlyatptlyayonmi .

.
.
I
adyatvitIyayonnii4ramdhanaitipaihchada4aamritab ekaikaayadha

cbchhichekatthyat&Thmainab 18 1 14 1
15 1

1 1 i

<b) 1 n
19 2 karapaih ichchha [ 29v. ]
||
1
. . danllodE^eCkram&t tatrfidl 16 4ud . s . e .

(o) tlyftyauitoddhya 7 chatuitbayam4oddhya 18 pain .

<d) y . . dvitlyay . . miiraTridhanam

yatyitlyay . nmi^raihdhanamsaptadaahasmritab tritiya^chaturthay

. . chatu^paihchakaTni4ramtudhanaiiiekonaviih4ati |
prathama

titatracha ekaikasyadhanamkimssyadvechchhi . .

. W II
M II
1* II
M II

<a) tr . bh .

*B«eldt8 th« pitCM (a), (b), (c) and (d) Art two othor imAlI AcrApt In tht envelop#. It is doubtful whether (a) ie in the
proper eoeelopA.

8 2
 128

[
80f. ] !•••••
•
«
pl*li • t • « • •

suiraih |
ekayutanarasa

8arvashsha<jlbhipala . . . anenala6(2A

hltapratbam 38 42 48 1
64 ,e.
|

sh . 78 7. 1
.

SOv. kfi
[ ]

^(^dtsadfidakyi

bhflgaharaihktiyate 284
70
8 mudgadbe 1
24 1
47
kriyaXe 70 • •

8
[
31r. J* 10

traira^ikena 8 dine 168 dind

1 1 11
iyasya kriya 2 J te I
. . .

. , dram, 168 dinS 140 prathamenadattaih .

|
11
. . J ssaptal^ lI^i datiaissamadhanfi
I
]&t(l |
h

. shadri^aih 77 1204 1
217 dvitlyasyadatta . .

11 |llJ 11

. 77 e^aasamadbianajAtS || ||
punAnyaiffaarva&M . .

, . Ijhfi 4 dine |
drani 15 jIvyS ||
dvitlyasya |
bh'S 8 dine | draih
1 1
44 2

*F. 31 eoniitts of two letTcw linck tomth^r »th»r. Tli« trunftlitorntiofi girfn it that of the outside surfeoes of the two learei
only
1y blit it ia not itln'tyn (‘My to difForenliate.
:
 129

[ 81v. I

bhfiga

k&ranaih 2 1 11

chchhedasarii guQe dram 1 4 tarii dram 16 mudgil

1 1 1
•
2 2

. . r({AayutIhritaphalaih ||
asyagU 9 akaradvay 1
9 18
. . uparariigunaye |
ddautava 5 dva 2 2
\

.
.
gu^aya | 2 | dvinavabhagesu

chanmambhftna [ 32t. I

tugaganaiiinita* *rayai;^e |
r&

yarntyaktasutaya^etayg |
sakaikeaaparavartaihdhanurbhagasapa

apapatamSQ&aaudatobh&gaihnidharyate |
evaiiitutpuriinaoa

hiyamfinaihtunitya^ak kiyatastuparavartaibhuniiihprapyayateja .

dha 1 1 par& 2 I
yoja SO chhe 8000 yoja i: phalaiiipara
1 10 + 1 1
6 varta

18181 6e udS I
n&gaiivacllchhandag&niidmtacAada^a
11.1
. i . . . nam . . . / . .

eha .. (f, ... 4 •

[32y.]

yiksvaga* *ndhi'te eladviko

# • • .
1001 iirdhachchhe 768000 ayo adhachchhedaih I
0 1 J
phalaifaya *
8 !
1
mOT Tn& 1 di 4 ||
uda ||
vra

eAarliyft 4000 ktftpatitaiiibhGiiiitalepataiii ||

trii^atfirh

n^ibtusa s ptayojanaAlyate |
cha%urda[$a($twko<t^ .

yatapaih 24 chafoahticha |
kaidinaibhutaleprfipyaya

ko/ “S II
nySsasth&panaihknyate
 130

[ 33r. ] feAiaardhaiva*

1 chhe 0 guva gha ....

1
2 udS. dlnarakonamavi4d . .
II

. trldux khar janiyaihsukhabhojanecbal tasyardhamardhamchayadardhamardhamtone

.
guru prasadaiii 1 I 1 1 1 1 1 1 108 pbadi 1 dba 8 . .
^ 1

panadhanabhu 1 i
2 2 2 2 2 8 II
1 ktaiii ardbam
. II

aasatanirha 1
dvada^aatiticharmapikatiroraa . . . .

ii 12 24
1; 1 1
900 12 24 12 24 i
pb aroma
1 1 :
1 1

*
I
33v. 1
nya sth

1 Z90 *gha 12 chhe 88

1 1
1

112 uda II
suiiicrupvithivI^amkusurai>aihparim&f^rayaTh ||
iru

. . x^<>^^hi^&fA^8uramadiraiu ||
satatam8apta8ardh3.n&m8apdgadhya8e

II
sati'ibhagati'ipaiiicbaiii^anityamevaihchagachchhati |
yojanSn&mBatopttina

chiitura.4ltiruchcbb)'itam |
kenakalenasaugachchbevadameta^uniiicbi .

di 1 yo 84000 adbacbcbbedam 360 di

1

34r. . . . . . sasya . jat

1 ]

. . . . rtbinab kbage . k . d . bAuktapraaptidichaivamevacha |

. .
j|
abtauvadasakbekiihAsAagamvadasundari pra
1
|kba U khft
j

phalaihk^glT 63360 esbababupramawam ||

ka^cbitpumamsuvarpastukalapadayutaihyavaih ||
pratyabariisulinesfl .

kiladattavaih |
paThcbabdaima^ainevamtu dinaipariichada^astathab datva

ayaaarvayajnatuTnichchhamitatvata di 1 1 6 bha 0
1 1 1
. . chcbhedaih 122 yavatola 1 4
|
4
 131

y\xkta [
34V. J

chattritaihgai |
tuul^/a7a//t&-am]parayarjuaenag|‘iddhra

tayfirspri^ariinti 1 1 yo 777 8 phala

8 1
940 ma $aka ja 222 1 chhe rdhayutovya/f<
II 1
7
9
. stapanichapaih chasJasatei’enava jramaija ilabd ham
1

. trakathaya^va 5 I
ini na
1
. . gasya kii 1 chhe 128 tnaku 1 inachhe 40 sima sa 85
8 1 2 1

i
II
a m il

8) jata purusha^ 1 8 8 >

[
35r. J
( I
4 2 1 !

dhanaih 19 anenagui.iitaiiijiitam 4 eshapratbamaByadAanam

|

• dviguparii 12 dviyutarii 14 etadvitiy&sya

| |

gupaih 21 dvigupaiii 42 I ryupaiii 39 eshab nySsab pratyeAra . .

|
| |

da^amagravrindanfim cliuturdai^a ekonachatvariih^a |

ta<pMa):dhatfibhagan . m
1 pha 4 I e.variidl 21 . eshaprashna etainml
' '

4 1

(b) . . . . t

. smfitkRraofl |
tayoryogaviyogariiavaviyogas

rwRabhSjitfipuniaha 18 anenabhakt.va dhanaih j

9
IS
tadniipaddhayasahitaih ||
. . . .
 132

[8Sv.] (ft) . iisha)^ 4 5 I

sat^/tdhayutikri ajatd, 00

II
eshagavaavamahishlpratyaika^fileshu hhaga

1
ddl 180 ga [ 4. phalaiit 45 . .

180 1 ^
4a 1 a4va 6 phalaih 30 4+ S3
1
ni 4ala 180 inahi 4 phalaih 86 5+ P

1
(«>) mule^a ^ etadviguQaiii 3 1 dviyuta
1

i
2
1

11
J
yasyadhanaiii |
tadevaailidhaiii asyardhaih i

1 yutaiuny&aa
^3 |
|
||

i iJ
dvitlyas

36r. ]
ydchaturda
* •tasakhe
hyapaiiichatiiguoi

TO 21 eahadeiiapi'aniaQani aamaptaih II II
sa i

lavanasyarashekoshthatanivakritanirharai |
eBhamchaik3.mr&dipuna8a .

p/odhanltS |
saptapilmmapichaik9ra4igtulit&Qi | paihchasaptatyS

hasrariibhavet saptashtagupamkiiii r& 1 1

1076 66 adhachchhedaih
1 1 1
2000 pabha |
phabhS. 30 pa 2 ^ 00 II
esha

dil&vapaprama^am ||
kiikinldadabh&gasyadadyftdasht&dail

& I tasyamvim4atibhSga«va^atabliftgaihprayachchati |
naroraksha
 138

I . .
3av.
f

yftchajacha * • 1 . \B00g

yojapaihcAakaiii | ajhiMthoi.

1 1 yo 6 chhe 4606000 yayo phava 2 . .

1 860 1

ma 4 . .

yojanasyatjibhftgftrdhaihsatiibhfigapadonakam )
yanaudinatt^ibhage

genagachchhati | iftpuna^pamchabhagardhamyojanasyatathaahtamani

tiniyartaihteyayuyegayalfthatS, | yojan&naihBhtootaraiatam

lenagachcAAati di 1 bha 1 1 gu 1 1 18
8 8 2 6 2+8

[ 37r.

sttyam&nasyadiy&karaayaghatikai x kirhprayatasyavada

Buni^ohitaib I
1
30 muchhe 2 ghamu 1
600000000 gha 1
i

1 1 1
yo 8888 ‘
^ 1

. 78 I
rbhaporathaihsuramahoragasiddbasamhaividyadharaiVparivrita/?/

. rfttirau |
kotl^atardhaihsarathaiiiprayaayat tadbruhii^aatruku^alo

. . yartaih || muhurtamekenakimgachchhebruhimeganakottainu J|

. . . . ()
600000000
1(
gha £ ^ . yo 16666 . .

atubhfiga .

[ 37v.

bh&gebhavedr&ii |
urdhachchhedam 106000 viliptapaih . .

lipt& 6 pamchardhasamvatBarebhuktera^aik&yadibhiinujab bruhi

ka tatyajfiasamaAyay&BareQakiih 2 Tft 1 1 am 1
1 1 1 860
Ordhachchhedaifa 106000 yiliptfi^&m 2

rS4i I
adKaohohhedam 1 viliptftlipta || phalaibyiliptfi 9 || e
60
sliagraliagatiifa ||
udft || rftj&yudhishthironfiiTia ^ pSpduyamaaAi
Qiipat . m . m
T
 134

[88r.] (a)
j
U
dvitlyapankyayoga 75 vartyam 5 p&niyam ||
tfitiyapaAktya 84 M
15 1 11
Iqiyate yogam ,
90 vartyarii jatam 8 payasaih 1 . . .

15
m
I

. . zh . 1 .

(c) bhaga vimsas chadasagu^a |

saptam . . ksh .
. }fia

drishthaihchaaatani . . ai

• ••• • • • • o

[38v.] (ft)|
. . dri prakshepayukti 80 vibhaktarh 1 dhanaib
J J

Q.itajata ^4 | I
. . . e . . ao

k raulyam : . . . tall

1 . a 1 jam mulye 8210 ma^ale a im ||

pha
|| |j

pha l40

pha 160

^(b) Should go with 00 (») ;

and (a) poaatbljr with 30 (e).
 186

(a) *. iady&d&!<anrii^a4tath& |
saptap . ch . . ch [80 1.1

dhasaptapaihch&n&mtridvimeka^prakalpitaih |
tasyavahasya kiih k . m
tatramama | kshetrasyasthapanaihkriyate | |
karaQarii ksheltra

(b) dfisA^A^Aaksh . . ,

. . daia |
chaturda^atiitTyasyachaturthasya n .

sntushtha a . ahth

(c) tt . sth . n (d) shnthnsnne a 9 ksh

15 15

(e) ihsthapya 4 . 5 .

15 15 15 15 20 . 000
4 5 6
. . syakfi yatc II
chaturbhipaiiu'habhishshadbhi j/j/
15 15 15
4 5 6 |16 25 a
thamsLrd yoga 60 vartyaiii 4 inadhugha
.. . 15 15 I

15

(a)
chh u t
3»v. ]

II sahhi niapdalakatt>allik5 |
eiachchhodaihbhava^f'

yathechchh . . karyk |
^uthatalakpya udaharaijani talasyay .

. . mekam tadva ioishtiSatdndThdaM .i.d.m . irh

X

<b)
dhanatii 1 ^^11 asha ekaikabhagaguvitajk

16U I
180 I
200 1
800 I
evam

dhanaifa 1200

180 (c) . . . im
<d) pha
200

*yet foghekshepa 16 anena trabhkga 88 labdha | 8 |

<«) 18
M I

pha
pEa ev labdherbhkgam I 88 jatk 14 \abdhak8keparh. vi
12 I €%
j
I

*Th« portion to the left of yef poiiiblj belongi to another leaf.

T 2
 186

40r. ] (a)*, ch .
.
yatra . . . . i

. agaihchaivak&rayet kahetravaipuly ... t ... y

II
priahthaiiatadvayamchaiva uchareilatamekatab vaipuly<S(2vt

m . . e . . m . i

(b)* aga»*miarfcAad . a . . t (c)* syatudv .

yam^iataih |
aarveini4rapi ah . riibhagaatasyaivapaihcha .

(d)*n

bdhaih bupayaaoghatab ekami^rlkrit .ri..r..k

karapam havyatulyamdhinikshipvah
| 4*1 6 I
kuruprakshepakaiTi tata |
prakah
4 6|e'
*ThtN tbuuld poMibly go with 39r (a).

[ 40t. ]
(a) . h . ntyachaivatatphal .

gupitdjita I
6210 eahavfihaayakfipijiaprain&pam

a . kemulyaih kartavyam |
odhachcA/^edaThchat . . . i . . i

(c) *dhonam . . 00 pratyayatraira (b)* [Blank]

(d) * Ayitva petu |

eko .

kritaih |
1 ^atatrayai-pichabhi^puruahairZabdhaTTikimfidyamprathamfidhanaih

I
120 II
d . as ]1 . . pain . . . m
* These consist of portions of two, or more, leiivos stuck together. The knot on I) verse does not appear on b recto.

[41r. 1* drammd . abth . dv&chatyaliih^abhirdinai |

tataaptat

. iva 42 dine 1
draih 8 jivya ^ 70 puruaha 4f
1 1
1
I

dramtna 660 |
yadipariichadata^ashtyfidhiAra

. alim^abhi |
tadrammai a«A^bhi katidinam .

*Beaidea the portion iranaliteratcid the envelope contains siv small anraps.
 137

uparfid 8 f41v.]
1
ya 8 adhed&payedatta)^ 17 adheno 4 pariaaih
8 61
parimara^Tdvayagu 8 vaye 8
8
2
yet upariyuktakriyate ekapathcha^an&m
8
5
sthfipanam I ^ j
^ ||
^ . m i phalam a 17 u 2
6
/ i

1
1 [
42r. 3t
2 pha 64 kfiyate |
1 1
8
JJ
saiiaySaardhajruktetrayodai^as&rdhaThbhavati 40 bhfi 1
1
160 13
1 1 1
. pi eshfiiihchchhedSiiikritujata ekepa 1 2
.

ri saxdhatrayoda^abhikimiti i 1 1
4 27 pha 64 eshiiiiiapa
j

. . ekenalabdhachatvfirishshadbhi ^ 1 !
2 sampady at ok atham 1
1

II n ekol&bhati chatyfiri^ahsardhasyatukiriibhovet 1

. ra

[
4‘2v. ]

aalata 64 |
6adtri 24 ( . . 12 ||
ardha 18 ||
ekattraih 54 ||
eshamu

tmtr&^ikakaraQapratyekamulyavidhi ||
aparamvakshyaini |
vim^ano/ndi

kiihpratharaekhandhakodvayo'bhilikhita apasyaprash^avidhi 20 1 1
|
113
2

^igu^aye guijitfijatft 20 3 1 chhedam 20 1 1 ;i bhuf/f'

. |
12 3
jfitamphalaihru 10 ||
esHavim JJ L dt .

bhayati |
atra uparima^khandhakasya eahagUQ6kilrambhavati |
t^gya

khUyd

tB«8id« the portion trenilitoretod there ore flro rery mall srrapi.
 188

[ 43r. ] va .g

. . Arayet sardhadva ^ tubbojanaraadyamuttamet

. . .

aatribh&gatrayaBtriibdaidinaidvaoijyakasyatu |
bhgQ^ftredvSda

i§aiatava;V2ra9aThsthit{l8yavai | esbavyayasamutpattaukaxkAlam brd

hipaQditakaraoayidhanenadvada^a4atasyabh9,pdaresthitata

10 2 bh& 13 8 bha 18 8 bha 1 8 bhft 1 1 bh& 1 5 bhS 8

11 11 11 11 21 11 11
28 88 4-1-2 86 2 8 44
2 38 bha 1 1 bha^d& 1200 gu^ita . i
1 1 880 1
3

[43v. 2 10 800 . a . . chchhedarii 380 diva bb

1 II
1 i 1
re ,
8 2 727 1 oasaha ||
ayupipdaih
I
adhunavyaya pio<j[am di 1 1807 800 ; 488 1

Qrdhachchhedam 360 ^1 240 727 j 727

^

diva 2982
1
puna 800 2982 1 a . . .

il

<!am 486 727 486 1

1 880 diva
1

1 l'

li

phu 727 I
laiii 1 727 J pratidina

1807 evamaai'vatraira^ikena |
ud& ||

[i 44r. i] i 2869 di 1 • • • • «23 evaiii ....

* 1 144
,
.. vyayapra manam ||
udft ||
eka

i&rdhainutpatisatTibhagadin(ulvayet pujSrthaih aatpbhagariicha

t&4chayet sashtabhagadinatrloiva^udevasyachftrchane

p&donatrayoda* "ia^ariicha ashtas&rdbadinfini khet ||

brfihmaoSbhojanedadya paralokahitarthinab satribbftgaih

• • .
jjfiraihBapamchabhllgadinattrayet pa

. . . lacAendamchakaXk lakhanavalikd .
|
sardhaifisUrdharhAine
 139

[ 44y. J

pta^tSn&mkaxkSlaih arjanabhakaha^e |||

nyaaasthapanamkiiya

bhfi
11
8 di 6

8
bha
11
1 pa 32 bha 4u

2
2
11 4
86
1
bha!
'
1
1
1
860

. r&6i I
888 I
ayara^i ||
880 j
etatkilleQa &rjanabha . .

IM 61 i

. di 1 uitlhachcbhedam 86
1

8669 •
. . de 1 11
eshavyaye II*
61 1

va 6 m6 di 2 de 28 atha aya di 1 1 1
28
61 1
‘
. 1

. SS5 dalita 10555 .

55800 I
65600
^1
diyutaU 108626 paclaghnapa . .

66600
. ta I
6466040626 |
a topariichavi/iiS .

3227620000

U
• • uparah 6465040 .

3227620 .

. . telabdhaih 2 esbadhanaiii ||

. ih ll
a 1 II
u 1 I
padu 0 h dhanu 7000 1 gah^o

\ .0.
chaturriarii

7 3^6 asyavarga 147466 a,sya

, . . . y •
{
:

L 4ev. ]

. 48 ^ 271986 1
eahasarvagupitakaraij^ipafi

. . kTitvfibhajitajStab ^IIM aih^airam^Sgupaye^

[
^ |

, . , ft^ivarjyajStab L 679
. . 679 768 29* *4912 .

. . 616620000 294912 -h 1168 777807600

. . 777807600 777307600 desham 679
616226088
darii

, . 67588 II
dvayenamule
 140

. 14668845
i
4668845000
21790250044 dalita e
5 5 4 6 1 6 0 0 0
110896126022 adisamyuta
3227047522 1
1554615 000 padaghnak
j|

il

1 5 5 4 6 1 6 0 0 0 6 0 7 5 8 8 8 8 7 6 2 7 4 6 7 4 8 2 7
rakarapipata 72604888946760000
^1743271936 patitajfita uparauyasasthapa
||

26000000000 II
bha

[
46v. ]

! i
made 8 made 6 ma(;}e
:
!
I
--
ka ;
20

aparaprashtat^ para a i e vlhujabavlhanhai

r.m
I
gorejarnacha uppaboaaniale a . . d/mpa

(.Uiale apot4ii.ie agabovlhujaija ehuvl

karaQam |
traigodevarehahipa^ehi s&

. a iJudhed . hivo

[
47r. I
n . . r . .

vichakshabab chamua^upritanfistisrastisras^hamiZ .

niklnida^aguD&mfthurakshohanlbudha^ ||
akahohi

ra 0 ^ 8 8 8 gu
ga
1

1
1

1 aha
11
^
^ ^ 111 10

1
guQitajft

na pa ratha 21870 eaha akahohi

6
|
11 |
tu 8 1 ti 1|
gaja 21870 pramSoam II II

dft II
ka4chidrftja nara 109860 kum8ra4atru<lama |

haya 85610 • i . . o .
 Ill

. . 15 . 10 ii phatom i50 ||:

udd [47v.]
. 1
1 1
. . . tavyu 1 1 1 1 haparthaihhehayaki
f

TMghnatasftyakaidchaiva^pattisvap&dadala^o^ftsai |
a

anyachatasrSvaihatatenamahansavaTh j|
^ar&n&mchapailma^aih . .

<ravi4arada &i 1 18 4 achhe 21870 phalaihdara

1 1 1 1
2824400 1 anya
II 4
1
ipramaoazh || 2 sutraih ||
ekorathogaja

[48r.]
1
1 raktipa
ma 8
1 i;iitaThjatam
12
1
41994^
di
SO 116200

9 pale to 3 ^olenE stidh^ IS

. . A 7 diiAnen&sti aih * dha ||

am 2

*
.
phalaibbha 2 . .
48t.
[ i|

«£tpala 2000 bha |

pa 970 ||
pa

8
dha S to 8 tolapal to 8 tolenastidhaoa 19 dh

. . 8 <jhhe 19 yadidinamekena eshadataihtsulvada^avarshelia

ar/i 9
- yaea chhe i 11
di 1
916 bha varshe 12 8
1
1 1
ya 2 8 bha 970 pa
pbalaih 1 1 bhara 9000 chhe 95
«
8 to
varoaprama oaih
8 chhe
^iraseii 8 dha
It
 142

6
[ 49r. ] . ya 8 yekvekndstika X da 4 kal&n&8tip&
4

p&daaastimudri 4 pamu 8 mu 8 ||
udahara9 ath ||
. .

sukhyairyajaiiitidevlpratimahnikechit dadamidevyart^a

kaihcAaliL krltvadlnara^atanichatvaritadh&naka am^^ikara/cft

kayavkkal&p&damud|ik&cha |
etatramulyaihvadametatia

syakirh 1 to 12 '
mu 400 dhd 1(1 phalaihdl 50 dindra
I,

firh 1
I stidhfipe 12 dha

ra 1 J
bha

10 dhaoe anasti am ya am II

I 4»v.] . hdsobh .........

raktikshaya . . . th . . . . yasyati pamchagupan^

sarvapaihchchassamarii |
divasavim^atikaihkiriiii/TndT/atimahaih
1
vadani^chayarii
1
to 3 kaha 4+ va 26 chhe 360 madAya
1
ma 2
ya 60 ma 6
rtlsiguoitam js
12
12
6 S 8 2 1 II ksha am 3 81 4 yara(k>dhya
1 9 2 0 0 4 8 di 20
80
ya 8
4
6 0 8 8 1 dhachchhedafA 2
a ka 1 |

.8 2 0 0 sa 6 UUs
A bha tt
I
7 toUptlaifai
pa 1 rvagupitaih ||
668 . 7«

[«0r. ]
. . . .

ntikavBti
vaHsh ^putraha
aikaayartheputra pautra upayogyaihbhavatub likhi

tamchchhajakaputragapakarajabrahmapOna |
aarveshariimevaaaatrapam

gaiiitariimurdAmtishtati |
adyaTasanesamsare utpamnna gapi

. . hit paachaaHshtitadakartuihaivepaparamatman

yadyaihcha8utpaihnnaThgapitaTh8akhyakarapath (
yacha

hmom . . yadivel h . drh, . i

 113

guQayebhftjaye eAar .
.
|
ud& ||dramme .
.
pusa^ataihfila&d/tam I 60V.J

ardhenalabhyatexkati |
ekarditBtukalanigaoitaprakiiyft

kuru]i 1 dramme |
phalam 60 aparaih |
ud& ||
s&rdhadvaye

trajtk 100 drapusS sardhadivardhelabhyataxkati 2

1
sattaA aixlhenoparisaihgupya 2
2 j|

1
. . var^Aakramepacha |
ardhepa urdhaihgupayema 2
1
paxhchasamguQe |
bhfijayelabdhapanj^aih . . e 2

(a) •
[ 61 r. It

tl I 1 I
s I 8 I
T
1
1
1
1
1
eshaiiiyuA^ff
6
1

sha I 48 iesha^purushaaa 4 ||
anenabhajitfirlabdhg
I

. . . aya bhavati (
12 j
13 |
14 15 ekatrarii 64
|
j ||

ud& II
ka^chidrajadaded&naThsaptapaihchfi^akatiibudha |
paihcha

chapm\&kshjS.mySinupurvashah dviguijadviguoamchaivaiTiparupottare

0
prathamepr6ptaiiikirhpr5ptamaparejane
1

. bhAja^ohitvA |
tatrottarfi 1 |
1 )
yutam 2 | •
|
| |
j |

. . 9 I
. esh&^p . . sh

(ft) dfi Si9 kavaiiaib |

uttard bAr/jya
"ST
ll

81 1 [ Mv. ]
1

•
i

. ..J

ll

eshadhanfidrishy&^odhanlygjgta 24c

jatS I
purugha ||
^ |
9 I
27 |81 |
yoga 121 ane .

j&tS Ashadvauprathamasyadhanam 2 6 18 54 16^

I
2 I
|| | | | |

uttararfiAlsaihyutaiiijiltaih II
2 || 16 || 48 |(
1
147 |'
1
444 j
esh5m
1 2 '
2 1

__2_J 2

II
uda II

(b)
• II . II
.
gy , . yoga 111 sesl)il ^punishabhAji . . .

I
37 I bhajita \60

tPTwo leaves.
'
The left portion or (h) is not homogeneous. There are also three other small scraps that are poasihly out of place.

U 2
 144

[«2r.] • • ft • • • •

. n& 57 ^«edaihj&tam 8 anenach&ltfli4agu9ayaiiijfttft IS

1
. . s«avam4urfi4am pratyayatrairS^ena 1 vaib 120 1
||
1 1 10

II
udft II
dhan&svamardhoBaihtoddhya 1
1
vaih m1
1
1
8
~T'.
1
chaturlyakaiii |
tatseBhapamcharaobh&gote 1 vam 120 1
1 1 4

datadvayam |
ailtySdhikamdhanamchaivakim&dyaihprathamaiudhaiiaih |

u . . £00 aayadvayanaih^at&D&riipada
II

dAt/eaxh4ataihbhayati 160 atr&pipaibchabhfiga 80 eraih

||

1 1 pbapi 9du £80 ||

pamcbamlj&tlkara^aibkri
£ 4
1

P* am4ayuti 28 bbaktam 40 dhanu 880

40 88

400 esbapbalambbavati ||

1 yo ^ di 6 .

[ Mr. B
y •

1
i}aTh 2 1 1 .

II t -T 1

syai . . vi^esbamtu tatragati 8 8 videsham 1 1 1

vibhaktam 2
|
8 1 1 « 1
1

sarvagatl yojana nenaguvaye 18 anena •

| L 8_i ft | 1

gate bAavisbyati pratyayatrairftiikena 1 di

. I ||

0 a oyojana 0 1
.
yo £7 pUgdtad&na, •

ya I 1 di t yo di .
phalaihyojana
1 1

1 pAa 0
12
 145

ri 2 [63V.1
.
y • •

18 yojana 20 dina phalamyo ^5

1 j
*
20 ghatike

85 ghadina

26 yo 20 . . phayo 36
7
20 ghatike

55 ghatikedina

kaihsya

(a)'*. . . . ksh •
y • • xk .
.
y . . ksh ri
[ 64r. ]

yasya t|
. . yadhanam |
dA^anateyiita |
tayoryogaviyo

. krit&ihrashayab 2 1 dri 82 1
1 1
I
}+
, . bhajyahi T 2 4 'rl- I
tveti I
tatra uttarar6 . i
1 1 1 I
1

uttaraihri^amj&taih

atltram jatam 76 oshaprathamasya

(b)

(a) . . dvigimaxhchatrivnachatritlvasyadhanambhavet tryuna 54V. J

[

. I
. saihyutaiii |
ekayiinshati&Aix^retodfnaratftara . .

Syatudazhsaprithagaodak ||
karapam ||
yasyapadaibnajiiayate

. . . . y5 ^ II
eta<iprathaina8yadhanam |

4
1 et chadattav&n hasta
2

rveshSrh || 0 || 2 ||
f+ ||

(»> And (b) thert Are eight «mA]l AcrepA; And (h) Appeani fn roniiit of two unrelated piecce etark top;t4Her.
 140

a . . rik .

55r 0
va to pha va 6 6e 9 asya
1
10
dhfi gupitaiii 7227 athadaddrammako
12
1 1200
jjSradvidhana am 4
kaidramam 4a

. . virh^atipS ra 1 1 hh& 1

lahataidhanaka |
asyaivaska
1
to 1 4 i
va 5 to 1 dha 1 1 aih 1
ya 1 8 bhs 1 1 1 12
1
8_
6
ra 1 1 ya SI 1 2 bhS 1 1 si 1 1 ka
1 60 1
2 1 192 1 480
ka 1 2 bhft
1
2_ _
pa 1 r

66v. i) kaayaya

paih |_JL0_1 ehatriwi^atem |

divardhatolamkasyadivardhamdTfiiak&syadi

rdhania^dikadivardhayavaayakiramiilyaiii |
nyasa II

to 1 86 1 to pbadraih 68 4e 81 punanyoTfi
1 1 1
2 128
1 1 ma
to 86 1 6 1 1 1 1 1 1 1
2 1 1 12 1 48 1 192
1
2 2 2 2
1 1 m
1 2
m 68 2 4e tatha a m
1 1 ya
1 2
2
 147

[68r.)
•
• y • 1 { • 20 . . . aU taatat
21
kfityunan^eshachchhedodvi sanigui^am tadvarga dahi.
|

shthab hritii^uddhikiitikshayaU ||
^eshachchhedodvisariiguqakn

21 21 t)h5 lieshampanya
20 400 dala 1 sariislishthaU 20
21 441 2 214- tv&bhaji
dhamupare uparamguoitavyamvargamyavarjaye

]ii| 420042| 400 ^eshaiii 4246

. 880 964 gu9itaj&tam 848320 chatvS^rin^a

[ se-- ]
84 168 14112
prithaksthMaiii va rgam \ J
•

aha uparapatya^eshaih 846720 vartyajatam I. «» I

14112
11

*Th« lower hftif of tbif psgo if blank

11888 184 57r.]

1
f
5 1848 hri 1848
8 1

kritikahya kritam :

nmulariiekaih 1 eshajadp^ep&titajata 9085

1848
sambhaktam uttaraihdviguoaih 2 aneoabha

9985 eahapaifachakasyapadaih || ||
aayapratya

5096 atttraih || ekonar54idvidh2lsthfipya iipaae . .

 148

67 V. • • • • IT • • • •

Bhtottaraghnegujgdte ^ dvighnam&dicha yo.

nikshipya I 41 1 inulam ]
® 6eshachchhedodvi saihgu^i
'

® 1 . - _
4uddbal^ taswrtt akjdte^lishtha 8 J
kyityuna^eshachchhedo dvi

dalasaiiiAlishthahyiti^uddhiki'itikHhayab akyite^lishtha

dSdvisarngu^akyita |
6 tadvarrjra 6
5 6 25 dala
12 12 144
12

[
58r. ] uda II
dadvimi^adchatyipamcha^a ekonatrim^evacha |
dva6a

4hadvixii4achatu^chatvalimda8aptati | chatushshashtinava .

. . . m^anaiiit ararii |
t|dradlti ekavim^a ashta

. . . pakam |*

.... ;?&55296226447064994 . . . SS^IS

e4a

1 86v. J . . sthapanaihkyiya^e

ahya 1 yuvi 1 stidba 1 dyishya 20 ja

1 I 1
1
1
do mam mamda ^
maihdo 3

tadattajatammada 2 |
ya 5 |
sa<fAe
 149

. . II
vidS.
II
kor&sipaiiichayuta u sa r&^issapta [69r.]»

muladakoBor&4iritiprashi}iak 0 6 yumu 0 «a 0 7+ niu 0

11 1
1

1 1
ta . _
1

laih 6 dvihrl^am 4 dalarii 2 vargaih 4 hlneyutiihchakartavya

J j j j
j |

7+ anenayuti 11 edasarasi ||
asyapratyanayane .

I | j

11 yu 6 mu 4 11 7+ mu 2 paihcha^amasutraiii 60
1 1 1
.] 11 1

II
Butraih gavarhvi^esha kartavyam dhanaiiichaivapuna

*Tbe Wft-buud page of this I«mf in blank.

. . . m |
60 r.]

ekonaviih Fatima gavi 10 rupa 8 . V .

1 1

yivaritSBti ||
'

cha^amaButraiii 61 II
sutrarii ||
ayavyayavis^eshatuvibha
A,
—••^narii yallabdhara»abhavctkalaihayaiiiprashpe
. d^tahyasarngu |

ya vidhi .
||
udS *
||
dvidine avjayepamchatridinenavabbakshayo

5 dJnara 9 dji
j

nam ftyavyayavi^eshantu ]
tatrayaih I 6 2 1
dina 8 i| go
|

e . m
X
 150

[60v.] «0 di phalaih 180

dvipaihchasamasutram 52 ||

II
sutram ahadravya hara6autatadvi4eshainvi

yallabdha |
dvigu^athku lamdattaaamadhanaprati ||
udd
. . dine arjayepaihchabhfitakomekapaofjitab dvitlyaihpaihcha

t)«8et>«8amarjayatebudhab prathamenadvitlyasyasaptadatta . i

. . tab datv5.8amadhan5j§takenak5lenakatthyataih 5 rO ||
^

rii . . m . ati

[01r.] anenaAcd/enasa^/iadhan&bhavanti ||
ipTAtyayariitrairdUk . kfi . .

• ^ 30 pha 50 prathamedvitlyasyassaptadatta | 1^1 deshaih i8 |

1

•
^ 1 I

.
6 80 30 I
4^ I
48 I etesainadhanajMa uda
|| ||

japutrodvayokecluiiripatissevyasantivaib mekasyahnedvayashsh . .

ga I
tUtJUipi99'BpBiC8«Xl'|lApBU9mBq^^W.ld
I
UIBJ^BqpMAipBXsBiCl'lUp

ttayan kenakalenasamatamguoayitvavada^ume ||

8 dattaiii 10 karapaih ||
ahadravyavi^esliaihcha ta^rd
|
1

. . ^ 1^1 pha . .

Ceir.] 1 13 30
1 6 1 pha 65
I
‘prathamemdvitiyas

56
r IT SO
pha 46 65 samadhanSjft
rdattajata | | 1 2 1
i
II

sutramtnpaihch&samab sutiarh 58 l|

i
«#•
yenakrayaihbh&jyaihrtlpahlnaifapunaibhajet la

guoayetatranlvlbhavatitatracha )|
ud5 II
dvibhi X^^Oati • .

ptaviknoStitiibhishshat aahtada^abhavedlfibhs xha

katthyatam 7 8 18 Iftbhft karaoaih yi

|| |
9 8 1
 161

. . gu^itam i&tam nlvljata |

syaprat^a . .
[ 62r.]

. na II
yadidvibhissaptalabhyate |
tadachaturvirii^atibhixkuii 2 7
1 1
laihril 84 asyayikrayaihkiiyate yadishahadbhitraya
|| |

. 9dlabhyatetadachatur&dltibhixkim 84 phalam 42
||

L?_ 1

mulam 24 I
patyatosbaih .U e^alabhab cbaupamchasamusutraiu 64

vtkrayaiubh3.jayechaivagfU9ayetkrayapiodAtaih |
rii —•••—
mulagu^ayelabdhal&bhamchaprapyate ||
uda

. 6Atkrl9atiya8Baptavikfii^g.tit)ibhishshat mulacha . .

^ *
II

2 7 24 pha 84 6 3 84 pha 4 .
[62V.1
1 1 1 1 athavikrayam 1 1 1

M I
p&tyadesham 18 |
eshalabham ||
puihchapamchaHamasutraiii

vikrayaihbh&jay echai vagui;i ay etk ray api Qd avat vi bhak tarn

sachakartavyaihguijayemi^rakaihbudhab yallabdhaiiisabhaveninii

deshaml&bhapiiidat&iii ||
uda ||
tribhi^chalabhaterashtaucha

otArayaihshshat Bamulalabhainutpaupa^atamBashthiviTTiii^ritaih |
kirn

2amka4chalftbhaifachakathayedgaiiakottaTnah 8 6 Tuidra 160 )

8 4 1
. . . krayambhSjayechaivaguQayet

[Mr. ]

pratyayait 6 4 240
1 1 1
]
phalam I mulaih I 90 I pfitya^esbam i
7'
i
I

chasamasutraih 68 ||
vikrayamchavibhaktayyaihguuitamkra

iivat Iqitvftriipakshayaihchaivavibhaktammulaina

ykt II
udft II *f* parhchabhi4chatuvargamtugnhltaihkenamanai>('

kenashshat vikrltaihshshatptaihcha^apuamkntaTh |
.krayavikrayasaih

uyanlTistasyaiyakatthyatSih 16 6 ni^arh 66 +
1 6 1 1
I

. bhAjayechaiva 6 ]
X 2
 152

] 63v. 6 1 884
84 pundsyavikra.y& |
1 1 pKalam j
6

mulaw
L 120 chatushshashti . patya^eshaih j
66 I eiari^amkri

ptaparacha^amasutram 67 vastra^ulkamyadbhavatita
II

^.n'tavastrataih |
trai rSri^ikavidh 9.nena 4 ulka
j

yatatvatab ||
uda || 'f* patasya^ulkavim^an^ariika . i

<r*iiataih |
putakanampapakritcdvaupatauh|ita 4 au i

papada^astathab kiiiimulyath

[«4r.]
. .
5 2 8 0 4 4 4 0 0 4 tayardhe tus . . . .

8 7 2 4 3 8 7 2 4
ardhamkartavyani ta .

5 2 8 0
. .
4 4 4 0 0 4 samgupyajatam |
a . .

3 8 7 2 4 7 7 4 4 8
hrarahtareshugup . . .

270945941120 aaya urdhaihka .

2999096352 shashardham . . .

2 6 0 0 0 0 H
I

• shachchhedodvisawigupaiii 6 deshaparhchakamprithak .
[ 64v. ]
5
an 4 aByam 4 am 77 tanmula varjitam |
tanmOlaih 1
12 1^
5 d vigup ottarasa mbhakt am 65 eshapadaih ||
1 24
navanam 1
6. 1 u 1 pa 65 rupopa 41
!i
1 1
“J 1

b dalita I 41 |
adiaaihy
 163

etatkd^a [®6r.l

&AtVmanu8hya a lagyantiU
. 1.
“t*r” II
ap^iraprashijalji yadyekapurushasyadrammashshat
1
bhirdinaijlvaloka |
tatkaryaiiiprastuM. ^saptatlnfim

9amp&kar&k8hak&nSindrammai8bshadbhikatidinajivalokam6Aav

karapaih |
idautavayadyekapiyrushasyadrammilahshat tripi^a

jivyft I
tat8aptatlnfimkirii 1 pudram 80 di I 70 pu I pha
| 6
|

didraminapamtrlpidataH9.rdAasa m in

• • u
1 u • [ 65v. ]
ft
chyate 1
1 karanarh

shtbottaraghnegaiiite |
aahtbaghanam |
480 |
uttaraghana

. va I
dvighnamSdi |
ftdidvigupa |
2 |
chayojjhitam )
cha

uttaram |
ato uttarampatayitva ekambhavati | 1 |
vurdAi

^ranikshipyadhanasya |
481 |
mulaih^liahthakarapya 21
40
vami^am 882 seahaihchatvan'riisapnf haksthapya I ^ J 42
40
yojyaih 42 922 tanmulavarjitam '|| tanmulain
42
Oottara am ...8 8 0.

yxxktabhdgaha
[eer.i

nyckirghariitupaijyanaraekadvitfichatuahshat. sa

mpapamsa . . pyanimSnayab sthapanainkriyate

1 1 dram 1 2 dram 1 8 ; dram 1 4 dram 1 6

1 1 1 1 1 1 i
1 1 1 1
]
1

petusa sam

1 e 4 2 II
dram
1 3
 164

[ ] . , ^ikeua ;
1 dram 1 ru 12 dram i phalam rupa 1 4r

LI J„i _i . I

draiii 2 rupa 6 dram pbalamrupa

1
12
j
1 i
1 |

dram 3 rupa 4 dram phalaihrupa 12

1
1 |

. . rH 1
S dram . . rupa 12 . . .

1 07r. il
8 ’
8 I jsta
I

. . 11 I
16 61 ,

• • sadn^am 1 8 8 |
. u . . . 8 jata

. . sadri^amekasya 161 yutam 77 jatarii|98i eshaphalambhavati I . r . . y

161 16 116'

pratyaya'o 98 1-f 8 8-f 2-f 8 3-t> 3-f 2 4-f 4-f

. . .

16 I
1 11 1111
i
I

I
1 1 1
|

! 1
.

Ij
II
hupdikaBamanayanasutram || dinabhaktavi^eshaihchavi

. . . /cTtyatechaivakalameshamvinirdi^et traira^ikavidhanena . .

. I
. . adattaxhchapatavyamaukshmedattamchatatBamam ||
udaharapaih ||dvi

m . . e '

. . o . . . dhanaih

e7v. • •
»«
[ 1

narh y ksh . . . viie

achchheBhaihtadvigunaj . id |
nirgachchha

r»pr . vi^amanechatvaridattak punadvigupam/diaJJi . .

. . aBu^yaha^tathgatashtasyakimatumdla panaaj/a • •

. . bha 2 8 S bha 8 I
3 2 bha 4 I
4 8 bha 5 ;
mH
^ ^ 1 I
1 1 ^ '

Jl A } i_
4 .
yntarh y . . ib .
.
gupaikcAa . ri . . .
 165

(a) 168 (b) . de^adv. . . . r 08t. 1*

i

4 I
• p&tyaj&ia^eaharn 81

ekatram 89

h&rika . . dram 8

diy

*Th« raTerv* u( thii bUnk.

yojana ' • • • •
• 69i
[ ]

drishja II
atosadrisa

hakaih |
uparimaih ^amtamttul&blmvanti

tv&limiia I
dun&chaura^I . . . tatamktnld |
dvichatv&rim

rh . . iti . . vanti etevrihakSsarvattrab sthapanaTnasya ||

. . .
yatraira^ikena 9 d ;
8 m 810 pha ID 84
1 1
1 1
. .
yasyakjiyate

yate ||
^ • I
^ •
I ^ I

. . . k . tarn 106 1 6»v. ]

II
tribbirdattattfiji^uoatrigupenatu vi
II |

miveA«>taduchyatfim 1 8 9 drishya 180 prakshepa

1 1 1 1
M 1 80 1
90 e katrain 1

vSSL II
tarn iataih^nbhirdatyaiparavaptr&pavaptii

phalaih .
ohh . . . y . 1
4 6 9 dri 190
1

!
1 1
1
1 !

80 90 ] ekatraib i
 156

[ 70r. ] (a) tola 5 . . akbhito (b)

86 etebh&gft 5
8
. 17 1 U7
I
70

(0 )
• e .

2 to 0 rlta 7 pala 2 tola 1

traiiipala 6 || II
udft gamft . i
II

. . pipesIkritfintcAa 1
dvochatisraA . ih

tisrasamfiddyatulitfinitrayoda^e |

vkaikaaya^drdAaaA
|
WWW d .

[70V.] (b) ri ri (a) prakshepajuktyfiphalayA

. . gu^yaphalar&^l

(c) . . . katrarnpala 0 || ||
udft || ardhatri

dAd4ariipamchaiiashtia)ipodadau |
aevakftnSihtud

1 dfiahya 66 U sadri
8 iJJJ.
90 SO

II II
 PART II

iii . — Facsimiles the manuscript*

 INDEX
TO THE

BAKHSHALl MANUSCRIPT
 INDEX
(The Durobort marked * FoL* refor to th« loUoa ol iho loit. Tbt otbet noinbetii refer to pag«e.)

A
Abbreviationa, 26, 27.

Addition, 27, 28.

Age of the work, 74 sqq.

Akhmin ; matematioal papyrus fionig 48«

Akshauhinl^ 20, 50. Fol. 47.
AlbIriSni, 58, 65, 67, 70.

Algebraic notation, 17, 25, 33.

Anatolius, 33.

Anlkini, 20, 50. Fol. 47.

Animals, 20.

Approximations, 30, 44.

Areas, 60.

Arithmetical progressions, 43.

Arithmetical notations, 79.

Arjuna, 19. Fol. 34.

Army : constitution of, 20, 50.

Aryabhata, 33, 40, 69.

Astronomical problems, 51, 52.

Asuras, 18. Fol. 33.

B
Bakhshalf, 1, 70.

Bhanu (the sun), 57

Bhanuja (Saturn), 52.

Bhaakara, 17. 19, 31, 32, 33, 42, 43, 69 , 61, 64, 65, 68, 10.
Bhavani, 51. Fol. 44.
^
Birch bark, 3 iwiq.

Boat problem, 20, 51.

Bodleian library, 2, 12.

Bower MS, 11.

Brahmagupta, 17, 31, 33, 43, 70.

Buffaloes, 20. Pol. 35.

Biihler O., 2, 83, 84.

c
Camels, 20, 41. Fol. 3.

Cantor, M., 40, 48, 79.

CAamd, 20, 50. Fol. 47.

Change-ratios, 54 ff.

Chariot of the Sun, 20, 44. Fol. 8.

Chhajaka, 19. Tol. 50.

z 2
 u
C conid.
Covmty B., 5.

Cowv. 20. FoL 35. 60#

CowiM, 64. Pol. 36,

Cunningham Gtenoral Sir A., 1.

Cypher as Symbol for unknown quantity, 26. F<^ 4*9^ 22, 23, 24, 41^ 59, 55.

D
Devi, 18.
IHnara, 64, 83. Foil. 21. 22, 33, 49, 54, 60, 61.

Diophantufl, 26, 43.

Divieion, 29, 30.

Dramma, 64. 83. Foil. 31, 4.1, 50, 55, 05, 66.

E
Baring and ependii^, 49.

Blaphanie, 20, 50. Fol. 47.

Epanthema, 40, 41.

F
Paine poBition : rule of, 32, 33, 83.

Foot soldiers, 20, 50. Fol. 47.

Format, 7 aqq., 75.

Fractions, 23, 24.

G
Gage, Col. A. 5.

Qandhara, 1. 71.

O&tha dialect. 11, 78.

Geometry, 16.

Gold : computation of, 15, 48. Foil. 16, 17, 18.

Greek influexioe, 71.

Greek terms, 54.

Grierson Sir G., 87.

H
Hargreaves H. 51.

Heath Sir T. H., 41, 45.

Heron, 31, 45, 73.

Hindu origin of the text, 73.

Hoemle. R., 1, 3, 10, 12, 16. 69. 72. 74. 78, 79. 82.

Homogeneity of the text, 20.

Horse soldiers, 20, 50. Fol. 47.

Horses, 20. Foil. 3, 8, 35.

Hui>dikas. Fol. 67.

I
ladiaa origin of th* text. 7S.
Iron 20. Foil. 10. 14.
 iii

J
Jambliohua, 41.

JftjiM Rija, 45.

K
*1-Kluirkhi. 41.

Knott in the biroh-btrk, 6, 7. PoU. 12 k 13, 32 k 36, ft 49, 61 62. 63, 66, fto. fig. 2.
;

L
Langutgt of the text» 11, 78.

Lnpit lasuli, 20. Fol. 14.

Lenticeli, 5.

Leonardo of Pita» 41, 84.

Lllavotl, the : 31, 32, 33.

Linear equationn, 39.

Liptk, 58. Foil. 7.37.

M
*
M *
section of the MS, 20, 21, 60-02, 09,61.

Mah&vlra, 17, 33, 41, 42. 43, ti3, 04, 07, 68, 70.

.Uakoffi^a, 18. 52. Fol. 37.

Maahall Sir J. H.. 70, 71, 72.

Meaturet of Arc. 58.

Capacitor, 82.

Length, 60.
Money, 64.

Time, 56.
Weight, 66.

Measures, 54 : Mperial (All the references an* to fulion of the teat).

Adhaka. 12. Kharis 34.

Aibla. 37. Krote. 32.

Audfka. 49. Kudava. 12.

Ahgula. 20, 32. Lipta. 7. 37.

Bhara. 36. 48. Maahaka. 8. 14. Hi 17. 18, 34 >5.

Ohanaka. 20, 4!) 55. 3Itidnka. 20 49, 66. Vilipta. 7. 37.

Dhanus. 32. Huhurta. 37. Yava. 20. 34, 49 56.

Uinara. 21 33. 49. 54,

. (5(), 01 . Pada. 20, 34, 48, 49, 56, Yojuua. 3, 4,6, 6, 7, 8, 9,

Pramma. 31,41. 50. 55, 65, 66. Pala. 11,15.16,36,48. 12, 32, 34. 36.

Drafikflhana. 20. Piaatha. 12.

Drona. 13. PrasntL 34.

Gavyuti. 12. Raktiki. 20, 18, 49.

Qha^ika. 37. RIM. 87.

Gnflja. 48. Satora. 34.

Hatta. 20.54. Siddhfittha. 34,06.

KftkiQi. 36. Savama. 11,20,34.

Kala. 34, 48. 49. 55. Tola. 11,34,48.40.45.

 IV

M—conitL
Mem see 8uinoru.

Metro, 79.

Motion prohleruM, 43.

Multiplication, 28-29.

Mythological roferencoB, 18-19.

N
Niiu M.,82.

Nogativo sign. 17, 78. Foil, i, 10-18, 20, 48. 49, 08, 09, C7.

Notations : arithmotical, 16, 79 sqq*

Numerical notations, 79 aqq*

Numerical symbols, 78.

o
Order of the folios, 12 aqq.

p
Palm leaf manuscripts, 9.
Paper as a writing material. 9. 10*

i’andavas, the, 19. Fol. 37.

Partha, 19. Fol. 47.

Profit and loss, 49-50.

Proofs, 34 sqq.

Punctuation, 99.

Q
al'Qalasadl, 4.^).

Quadratic equations, 44.

Quadratic indeterminates. 42

R
Rajputs, 19. Fol. 61.

Hukshakas, 16. Fol. 60.

Havana, 19. Fol. 32.

UeyuUt Jahi, 32-33, 83.

Reijnla virginUmA*!,

*
Rule of three ’, 32.

R<Apona method, 33.

s
SiifFron, 20. Fol. 12.

8andhi, 79.

Ssrada script, 10-11, 75, 88-94^

&f€ni.54,64. FoL34.
datrudama, 19. FoL 47.

Saturn, 52. FoL 37.

Script, 10-11, 76, 88-99.

Series, 47.

Seconds of arc, 52.

Sexagesimals, 24, 73.

Siddhaa,18. Fol. 37.

Sita, 18.

,18,72. Foil. 34, 44, 60.

Slokas, 79.

Snake problems, 61. Foil. 20, 32.

Square-root, 17, 30-31, 45, 73.

Srldhara, 31, 33.

Stein Sir A., 20, 64, 83.

Snbtraction, 28.

ddkhyas, Fol. 49.

Salin, 18. Foil. 34,44.

Sumeru, 18. Fol. 33.

Sun, the, 20, 61.

Sundarl, 19, 45. Fol. 34.

SQryadasa, 46.

Suras, 18. Fol. 37.

Suter H., 42.

StUras, 22.

Tannery P., 31. 40.

Thymaridas, 40.

Varahamihira, 63, 68, 69.

Vaaishta, Fol. 60.

V&sudeva, 18. Fol. 44.

Vaux, de C., 82.

Vidhy4dharas, 18. Fol. 37.

'Hja Oanita, 19, 42, 43.

Vulture, 20. Fol. 34.

Vogel P.. 76, 87 sqq.

Weber A., 2.
Woepoke M«, 45.
WritinK • of. 11, 97

Year : leagUk 68» 83,

YndUeUdra. 19. Fol. 37.
 Plate I

40 RECTO, t

40 RJSOTO.D

:i9 f£cto.e

38 RECIH^A

tl:-) l^)r;'"r\’.D

RECTO,

^3 RECTO,

f)l VERC.o B

35 PECTO A

4tuTit Minnb^M^en^
 1 necTo

1 V£RSO
 Plato III

2 VBItSO

3 necTO

8 VBMO

/("(aTIR ?.j-:i»iTt- * a'

Plate VI

RLCfO

: vtnso

S RECfn
Plate VIII
Plate XIII
*4 RLCTO
Plato XVIII

ht cln

\x»

:v) [ [hso

J: f^f CJO

.
<»!
Plato XIX
29 Rf CK)
!•
SJJIETt.
:i5 vtRso

',6 HI CTO
 Plate XXV

36 VSRSO

'y>

V fit cm

!j'^
:iy vcnsif

*X***rvf’^*^*^ ti-;

AftiATit 8jjirrt« ^Ak*»Bn

Plate XXVI

f V';. ,
J
RCCTO -
vf
iJ'
'.'
13'"’^'.'.'’

.'kV / C-*'
,
*

.,'8 ^/fff6’()

ir’V •/. :* >1*.

' '»v.«-
4
.

nTftr »
*>

30 Old 10

I
 Plate XXVII

s-^' V^X
' ‘‘
^ 39 Uh3l0

Kmr^ ,.
«v. J* - ''•

Vi.
-* •’••
'

^#J '^'
^
^
W; 3.4, ^t.v ^
^ v> ^ ... / ,
{fjJ

crs^^ca 11 ii. -a . vv/Tf-.

J!4'~. -3 -« ^ ,'.
':6Ja

I \*?®®5^»a. •vTlIc.y'

[tii5?f5»?if’
<«fr* - 5 48r
54. 4 ^**
"L*.*

*i^.i..‘*il^-'a0

v-f^s
St*‘%. '
>-^ ^

^7 '

fM ? V -n
* -r -
-^.jjt:' JO. .
S ^ «l ,. *'v*-.-.

'Y ?7
Plate XXVIII

II lit CIO

41 Vth'WI

42 HICTO
 Plate XXIX

^
-'
vf!'*' .. i

9rJ
42 VfRSO

!«'J5
TV.
••••V. ^ V. ^ J-' "" '
^...

(J.O’fT'Ep'^s Jfc; "rJf^'ifc----

J^?>..

”'
rfvx

4 'A ft F CTO

4 -i VERSO
 Plate XXX

'iff

44 HtCJO

'^"** *^

4
i; 'r.r^i^jf’.r-s'fifiR^'aWi'^
r
• . V,
' 4
'S' •

•' '
«'* •
S-- -^-s 1|4^

: t V

»>
^'L/iWO
J-/

“ «»
“ 'j!

'Tk'*!,

•'•^ '
- •Mcet'^-

"jw.',:

‘ -* «-• 'V '',^ %x-

s ,^->|f/"* St •st.R^v

4fj RECTO
 Plato XXXI

in ycHno

^ }* •sW

40 HECrO

40 VCffSO

A-uftTfK »*-* r'-TT. r» •

Plate XXXII

47 HLCTO

4/ VfHSO

4H HfCTO
 Plate XXXdl

4n HLCTO

‘".r ^o-
'
''
-; a‘f^ ‘, '? » ’''^ •
l^.X" 'i** •

•'
<?•- t • .
'.

'**
^
’
•ym y at*- v'*:.o-»t*'

•
IV »/

?• V^.,
'

4v t'/:/?*90

SJJim. ^AL^UU ^
Plato XXXIV

hO ULCTO

51 RECTO
 Plate XXX

Vf h'SO
«f=lf •ifijii;«>_f r.-^tj^
‘'^
.1 . W. , ‘M*-'
'''
_ A.*
^

V..
i..«^
' ^,. i'’ 5 ) 5t;

nrf»# w •^'- •

RECTO

AftiATIt SJJIETY- ^iii«

Plate XXXVI

52 VfRSO
67 RECTO
 Plate XXXIX

VERSO

SH Rfcro

68 VERSO

^rik«cir>
IRlt
 Plate XLI

cl flhCTd

61 VHiSO

62 HE CTO
 Plate XLII

09 VEKSO

ar, h'ccio

63 VERSO
 Platu XLIII

<4

<0
 Plate XLIV

/;6 VERSO

6f) RECTO

6() VERSO
 Plate .XLV

67

.
-'
-Jl,;; « ^
•
?
-:i P<
'
3"
-._.

' vv S-4^
-
>•. •>*
.. lii
•
i’

i;:

67 ^f6S0

68 RECTO

Atialie SJJICTY.
 4') r’-K(.n

•V)

S'? KKC*:

.*y HI'iC’

HhC"

»-

,y
'

;rj^l ••
'
if^ ^ ')r^'’

.4 -;',-r.:iff ^••ff

'’
‘
T •% ‘If.

.;'• UF/''

h.>,’

f-l \lF'r<r^

hE>’'

^^^•.yiif Iriiii*. ^ifli.'A* nd. iuLii li’i;.*

RE - ARR^NOK 1.) FRAUMKNTiv

^“bAKHSHAlT MANUSCRIPT AS PRESERVED IN THE BODLEIAN LIBRARY

illllll
&CTiTTr. RfflFn/