d Mi

THE HARVARD ORIENTAL SERIESVOLUME TEN

HARVARD ORIENTAL SERIESEDITED

WITH THE COOPERATION OF VARIOUS SCHOLARSBY

CHARLES ROCKWELL LANMANWALES I'ROFESSOR OF SANSKRIT IN HARVARD UNIVERSITY

CAMBRIDGE, MASSACHUSETTSPutlisijcti

hQ l^arbaiti1906

ijanibcisiti)

A VEDIC CONCORDANCEBEING AN ALPHABETIC INDEX TO EVERY LINE OF EVERY STANZA OF THE PUBLISHEDVEDIC LITERATURE AND TO

THE LITURGICAL FORMULAS THEREOF, THAT AN INDEX TO THE VEDIC MANTRAS, TOGETHER WITH AN ACCOUNT OF THEIR VARIATIONS IN THE DIFFERENT VEDIC BOOKSBY

IS

MAURICE BLOOMFIELDPROFESSOR OF SANSKRIT AKD COMPARATIVE PHILOLOGYIN

THE JOHNS HOPKINS UNIVERSITY, BALTIMORE

I

CAMBRIDGE, MASSACHUSETTS

1906

Camekibge, Mass., U.S.A.

Publication Agent of

Harvard

University.

Boston and

New York

GiNN AND Company.GiHN AND Company.

Chicago and San FranciscoliONDONLeipzig

.... ....

Gihn AND Company,Otto Harrassowitz,

9

St.

Uartin's Street, Leicester Square.14.

Querstrasse,

For the

titles

and descriptions and prices of the volumes of this end of this volume, pages 1079-1080.

Series, see the List at the

PRINTED FROM TYPE AT THE

UNIVERSITY PRESS, OXFORD, ENGLAND BY HORACE HAKT, M.A.PRINTER TO THE UNIVERSITY

First Edition, 1906,

One Thousand Copies

CONTENTSPrefaceGeneral plan of the Concordance

The Concordance

is

part of a larger

Conditions of the problem which the Concordance involves

Primary nses of the Concordance1.

It is a

2.

It registers the variants of

comprehensive index of all mantras mantras not wholly identical

Secondary nses of the Concordance 1. It is a key to the liturgical employment of the mantras 2. It is \'irtually a finding-index of rites and practices3.4.

It is

a tool for future editors of Vedic texts

It is a repertory of the

5.

Miscellaneous uses (Vedic schools

Scope of the Concordance1. 2.

Certain published texts not included Unpublished texts included

General statement of the number and kind of works included

rarely reports variants registei-ed in the critical apparatus of a given edited text Orthographic details reported or neglected Occasional suggestions of emendation

The Concordance

Fntnre 'work complementary to the Concordance The working-up of material of texts yet unedited The elaboration of a reverse concordance

Acknowledgment of obligations To the University Press of Oxford To pupils and colleagues

........... ........ ........... ........ ......... .......... ............ ........... ........ ............ ............. ........ ........... ............ .......... ........... ............. ...............scheme most archaic Hindu prose grammar: mythology,:

PAGESVll vii

VJl vii

viiiviii

viiiviii viii

etc.)

IX

ixix

.

x

X

X X X

xixi

Explanations Introductory to the Use of the ConcordanceGeneral scope of these explanations Bibliography of the works cited and abbreviations Methods used in reporting the variants

General remarks on the character of the variants No hard and fast line between truly variant and vaguely similar passages True variants. tJha-padasVariants in padas other than the first Essential similarity with wide divergence of wording Cases of partial correspondence At the beginning at the end:

........ .......... ........ .....of their titles;;

Xlll

xiii

xiiixiiixiii

xiii

in the

middle

xiv

methods nsed1.

2.

in reporting the variants Variants for which the alphabetic order does not require more than one entry Variants involving more than one alphabetic entryof cross-references introduced

by ^ by See under of cross-references introduced by Compare (Cf.) Bibliography of the works cited, with abbreviations of their titles Classified bibliography of the works cited in the Concordanceof cross-references introduced... . .

Use Use Use

Sec

.... ..... '

xiv

...... .

.

.

XV XV XV XV

Abbreviations1. 2.

Alphabetic list of abbreviations of the titles of the cited texts List of a few other abbreviations used in the Concordance

Vedic ConcordanceAlphabetic list of Vedic mantras with citations and variants Additions and corrections

......... .......

xxixxii

1-1076

1077-1078

PREFACEGENERAL PLAN OF THE CONCORDANCEThe Concordance is part of a larger scheme. The plan of this work dates back to the year 1892,lished

are simple differences in the order of the words differences due to the substitution of a more familiar,

;

handy, or modern word or grammatical form

for

an

when two separate announcements of it were pubthe one in the Proceedings of the American Oriental Society (for April, 1892, Journal, volume sv, page clxxiii), and the other in the Johns Hoi^hins

archaic, inconvenient, or obsolescent one of equivalent To this must be added the meaning or function.

very important point that there are alsoin

many

cases

number

University Circulars (for June, 1892, volume xi, At that time, as the reader of either 99). of these announcements will see, I sketched the plan

which a given mantra passage, composed under certain definite circumstances, was later on adapted and changed to serve a new purpose.Furthermore, Vedic literary production is often in a high degree imitative and mechanical. The poets orpriests,

and was a universal word-index to the Vedas another was an and the third, which index of subjects and ideas I promised to undertake myself, was a Vedic Conof a three-fold apparatus designed to facilitate deepen the study of the Vedas one part of it:

more or

less consciously, fell

into habits of

expression such that entire lines of difierent stanzasor

;

hymns, and considerable sequences of words of

;

cordance.

As commonly happens

in such cases, the

This similarity. to a likeness which ranges from complete identity is sometimes so vague or fleeting as hardly to bedifferent prose passages,

show much

fulfilment of the last-named part of the plan cost much more time and labor than was expected. With

recognizable, save to the practised eye of the expert Vedic student.

correspondingly greater satisfaction I result to those of the Hindu people

now present the who look upon

the Vedas as their sacred books, and to all scholars in this field of Indian antiquities. That result is, an

PRIMARY USES OP THE CONCORDANCE1. It is

alphabetic index to every line

(or

pdda)

of every

Thanks

to the editors of

a comprehensive index of all mantras. a considerable number of

of the published Vedic literature and to every liturgical formula thereof (yajus, prdisa, and so on), that is, an Index to the Vedic Mantras.

stanza (or

re)

Vedic texts, we have, for each of the various Samhitas, and for some Brahmanas, Siitras, etc., an indexThese indexes are of and they do course scattered over divers volumes not take cognizance of lines other than the fh'st. Moreover, these indexes do not as a rule register suchoffirst

lines of

each stanza.

involves.

Conditions of the problem which the Concordance The Vedic mantras represent parts of a

;

mass of traditional material which was more extensive even than that which has come down to us,material

prose-formulas as the texts

may happen

to contain:

current in

the various

schools

of Vedic

learning, preserved from generation to generation memory, and handed down from teacher to pupil

hy by

they simply register the pratikas of the metrical The advantage of having, as in the present stanzas.

work, one comprehensive index, which shall includeevery line of every stanza, as well as every prose formula, in one single alphabetic arrangement and in one single volume, will, I am certain, be prized by every student of the Veda. 2. It registers the variants of mantras not wholly-

word

of mouth.

I

have, for

my

part,

little

doubt

that this oral tradition was supported at a comparawhen we cannot say by written tively early time

tradition (see AV. xix. 72). As a natural consequence of the fallibility of both oral and written tradition,

what was originally and essentially one and the same stanza or formula was handed down in the texts of the various schools in more or less varying forms. The variants are often of the same general character as those which appear in the various formsof ballads, or in recensions of church

hymns

:

there

which occur only a single time, two or more after they have been texts, require no comment properly arranged in an alphabetical index. Again, mantras which are not whoUy identical, but are alike in their beginnings, will also fall into the sameidentical.

Mantras

or appear in a wholly identical form in

Vlll

SECONDARY USES OF THE CONCOEDANCEThe future

[Preface

or nearly the same place in a direct alphabetic arof occurrence rangement. It is obvious that the places

editor of a Vedic text will find in a

com-

readingsthis

may be advantageously with a statement of the various grouped together, The method used for of the different texts.of a given mantra of this kind

plete assemblage of all the mantras an auxiliary of In the work of constithe very first importance.

tuting a Vedic text, the mantras are the most intractable part of the material concerned, because they are

Once xiv''. purpose is explained below, at page if the forms of the mantra in question ditfer more, at the beginning, then obviously again they will

written in a dialect which

differing, as

it

arrangement more or occupy and it will be necessary to connect less widely apart, them by some system of cross-references. This alsoplaces in the alphabeticis

was imsiderably from the later forms of Sanskrit understood by the scribes. Since much of perfectly the material of this kind with which the future editor will have to deal is quite certain, as experience shows,it is

does, con-

explained below, ataffords,

Concordance

To sum up, the an easy and ready primarily,page xv.;

to be contained in the literature previously published, obvious that the Concordance will greatly facilitate

the establishment of the

new

texts and the revision of

means of ascertaining the following things: First, where a given mantra occurs, if it occur but oncesecond, whetherit

some that have already beenas

occurs elsewhere, either with or;

edited. Moreover, since, has been already said, parallel prayers are to a large extent imbedded in parallel ceremonies, the

without variants, and in what places and third, it occur with variants, what those variants are.

if

SECONDARY USES Or THE CONCORDANCE1. It is a

Concordance will not be without value in establishing the text of the liturgical books themselves. 4. It is a repertory of the most archaic Hindu prose. The Concordance presents, for the first time and in a form ready and convenient for systematic

mantras.direct or

The above-mentioned usesprimary ones

key to the liturgical employment of theare plainly the

study, the prose mantras as distinguished from the metrical mantras of the Vedic hymns on the one hand,

expected to serve.literature

for which a work like this is The nature of the subsidiary Vedic (Brahmanas, Siitras, etc.), however, and its

and from theIt

rest of the early prose

on the

other.

that these prose formulas are in a dialect or in a style that differs not a little from thenarrative or descriptive prose of the Brahmanas and Sutras. The formulas abound, at any rate, in poetic

seems to

me

intimate relation to the fundamental texts, are such that the Concordance may also be readily put to

which are scarcely Vedic important First, since the Concordance gives not only study. the places of actual occurrence of a given mantra in the Samhitas, but also the places where it is cited in the subsidiary works on ritual and household custom and the like, it furnishes the key to thecertain indirect or secondary uses,less

or other archaisms that deserve to be collectedtreated

and

for the systematic progress of

by themselves. There is also good reason to believe that the prose of the formulas is the oldest

Hindu prose and so the oldest Indo-European prose. The study here suggested seems to me likely to proveto be a not unfruitful one.5. IXEiscellaneous uses.

It

can hardly be doubted

liturgical or ritual employment of every mantra as I hope that the prescribed by the ceremonial books.

that the Concordance will be of service in the

work

Concordance will prove to be a most

effective

means

of determining the relations of the different Vedic schools or cdkhds to one another. I am not sure

of advancing our knowledge of the hymns and the ceremonies in their relations to one another. The

but that the present time is just as opportune for this interesting and fundamental research as any that islikely to present itself within the next fifty years or And there are various other interesting questions so.

hymn

or prayer, and the ceremony that accompanied often serve it, mutually each as a commentary on the other. The subtle blend of song and rite makes

that will

suggest

themselves to

different

scholars,

a

full

knowledge of both necessary

for the under-

standing of either. 2. It is virtually a finding-index of rites and As a corollary to the use just mentioned, practices. I may add that, since a given prayer is liable to be rubricated in similar or identical rites and practices described in the large mass of Hindu ritual the

mind and habits of invesfor the solution of which the Concordance tigation, can not fail to be a useful tool. For example, it may be noted that this Concoi'dance assembles an enormous number of passages beginning with theaccording to theii" bent ofprohibitive

texts,

Concordance

will

able extent, as or identical rites and practices. 3. It is a tool for future editors of Vedic texts.

iucidentally serve, to no inconsidera helpful finding-index of similar

adverb 'md, and that even a cursory examination of them reveals the interesting fact that only a very few contain verb-forms other than in-

junctives or augmentless preterits. Or, again, the extreme frequency of mantras beginning with the

name

of a divinity has as

its

consequence that man-

^A

Preface]

SCOPE OF THE CONCORDANCEa very adopted a selective method.advisabletoIt

IX

tras concerning a certain deity are here, to

would

For instance, the mantras beginning with the name of Agni fill twentyeight pages, and those beginning with the name of Indra fill twenty-three. Consequently, in this book will very often be found, most conveniently assembled,large extent, groupedtogether.

include

in

this

Concordance

clearly be inall the

stanzas that occur in the works last mentioned.

From

them, accordingly, I have culled whatever appeared to

much

of the material for the study of questions relating to Vedic mythology. Nor must I omit to say that the

have Vedic form or Vedic flavor. Unerring judgment in such choice no one will expect I do not believe that I have erred on the side of including too little. It is:

words of the mantras form by themselves a very considerable part of a word-index to the mantras.initial

perhaps more likely that I have included some things that might just as well or better have been left out. 2. Unpublished texts included. The Concordance,

scoFi:1.

or the coircoitDANCE

Certain published texts not inclnded. Although the title claims that the entire published Vedic literatureis is

on the other hand, gives more than is promised by the title-page, in that it includes a very considerableof material not j'et published. Of the four books of the Kathaka-Samhita only the first has so far been actually issued by the editor. Professor

amount

incorporated in this Concordance, yet the

claim

made with

certain reservations.

The Paip-

Leopold ton Schroeder of the Universityit is

of Vienna

:

palada-Qakha or Kashmirian text of the Atharva-Veda, to begin with, is in a sense published, being accessibleto scholars in the chromophotographic reproduction edited by Professor Garbe and myself; but it is too

a source of peculiar satisfaction that I am able to give in this Concordance the mantra materialto

me

from

corrupt to be incorporated here and compared with the rest of the material, and the birch-bark original still remains unfortunately the only one known to us for this text. We may hope that the Concordancewill prove of great service in restoring this text so far as is feasible under these singularly distressing circumstances. Again, the edition of the Drahyayana(^rauta- Sutra, promised, and in part, I believe, issued, by Dr. J. N. Reuter of Helsingfors, I have not as yet

I made a highly important text entire. to Vienna, in 1902, for the express special journey purpose of copying the material from the three unthis

published books, and the editor generouslj' met my wishes by the loan of his manuscripts. From them, by the close and arduous labour of a month, I excerpted the needed parts, and embodied them later in the Concordance. To Professor von Schroeder

So, too, the Qanti-Kalpa, edited by Professor G. M. BoLLiNG, in the Transactions of the American Philological Association, volume xxxv, appeared too late. I believe I have read all the later Upanisads

received.

owe an especial debt of gratitude. Again, by the kind cooperation of Professor Friedrich Knauer of the University of Kiew, I have been enabled to present the mantras of the entire Manava-QrautaSutra, an important text in eleven books of whichI

and Smrti or Dharmait

texts

of interest in this connection

which seemed hkely to be but I have not thought;

Professor Hanns only five as yet are published. Oektel, of Yale University, has enriched the Concordance by the not too numerous, and unfortunately

necessary to continue,

among

all

kinds of late para-

lipomena or paricidas, a pedantic search which might be indefinitely prolonged without commensurateresults.

very corrupt, majitras contained in the Jaiminiya- or Talavakara-Brahmana. Finally I should note that I have incorporated all the material from the so-called'

dedications

'

of the ritual of the

horse-sacrifice or

Anything that may come to light within the next twenty years or so may well await the day when

the accumulation of

new texts

or of

new

editions of old

ones shall render a supplement to the Concordance

a profitable undertaking. Once more, the claim of the title-page calls for a word of explanation as to certain doubtful elementsof late or less important published texts. What I have endeavoured to embody in the Concordance

acvamedha, and human sacrifice or jniriimmedha. Not all of these are mantras in the stricter sense of the word but they have been included because they figure in the Saiiihitas and because they sometimes interchange with real mantras of the same or similar import.;

included.

General statement of the number and kind of works It appears that one hundred and nineteen

texts in all have beento the Concordance.

drawn uponlist

for contributions

A

of these, wi+hall

with absolute completeness is the following all the stanzas and all the prose passages of formulaic:

of the text-editions used

and with

an account needed biblio-

character contained in the Saiiihitas, Brahmanas, Ara-

graphical and other relevant notes, is given in the proper place below, at page xv*". The works so drawn

nyakas, older Upanisads, (^rauta-Sutras, and GrhyaSutras. On the other hand, in the case of the later

upon areI.

listed

under2.

the

following3.;

ten

classes

:

Saiiihitas;:

Brahmanas;

Upanisads or of the metrical Dharma-Qastras and Smrtis, or of such a text as the Rig-Vidhana, I have

4.

Sutras,

5. Qrauta-Sutras Upanisads Mantra -Pathas, and related texts;

Aranyakas 6. Grhya7.

;

H)

b

FUTURE WORK COMPLEMENTARY TO THE CONCORDANCE8. Dharma-Sutras, Dharma-Qastras, and Smrtis texts of the Veda; Vidhana-texts 9. Ancillary texts. I have not thought it TO. Four miscellaneous have read, worth while to catalogue other texts which I late Upanisads and Law Texts, but for the most part which yielded nothing that appertains to this work. The Concordance rarely reports variants registered text. in the critical apparatus of a given edited in critical apparatus of Vedic texts contains The due to all kinds of the main worthless;

[preface

;

askant at such a thing in such a place, I trust that they will take my long and intimate acquaintance with the materials involved in such questions as in

some measure justifying

my

procedure.

rUTURE WORK COMPLEMENTARY TO THE CONCORDANCEThe working-up of material of texts yet unedited. The heading of this section implies faith in the futureof Vedicstudiesto:

incorporated these defects into the Concordance would have swelled readings the its bulk inordinately, and have been a task upon

readings, To have of tradition.^

I

am

confident

that

the Vedasis

will continueterest in

be studied as long as there

in-

the history of

human thought and human

institutions.

Because of the half-mythical nature of

whole no

less futile

than

difficult.

it has in some rare cases seemed from the critical apparatus of a advisable to report not adopted by the given text a manuscript reading editor of that text, which manuscript reading is then added note var. led.' distinguished as such by the Hence the user of done. but this has not often been

On

the other hand,

Vedic tradition, we do not know certainly how many different books were originally elaborated by theancient

Hindu

poets

and theologians of;

different

schools and

'

;

but India is a very large and the seemingly fabulous statistics of the country, Carana-Vyuhas have proved to contain an estimablelocalities

measure of

truth.

Still less certainstill

may weextant,

be as toto

this Concordance may at times profit by looking into the critical apparatus of those texts which are cited forms of by the Concordance as containing the various

how many how manycome

of those works areof those

and as

which are extant may ultimately

the same pada or formula, especially when there question as to the original form of such pada or formula, or as to the interrelation of the Vedic schools or mkhds.is

Orthographic details reported or neglected.

Cer-

tain peculiarities in the orthography of the MfiitrayaniSaiiihita have been reported for the most part, if not

with absolute completeness.

They are duly explainedin the Introduction to

by

Professor yon

Schroeder

Nevertheless a hands in manuscript. considerable number of Vedic texts their names need not encumber these pages are even now known to and although their publication exist in manuscript be deferred, it is in the end sure to come. It is may plain that, all in due time, a supplement to the Concordance will become a neces.sity. For that supplement it will perhaps be worth while to wait until all, or nearly all, the remaining texts shall have seento our

;

his edition, and they are included here because they On involve differences of alphabetic arrangement.

the light of day. Meantime I take the liberty of the attention of future editors of these texts callinj,'

the other hand, although the KS. and ApMB. write final s before the three initial sibilants, respectively,s and s, I have not thought it advisable to burden the Concordance with the details of this variation which has no bearing upon the alphabetic arrangement, and is not carried out with strict consistency even

as c and

importance of giving with each text-edition an index of metrical lines and of prose formulas, after the manner of this Concordance, and with constantto the

to it. When I reflect upon the large of texts elaborated in the present work, or consider, for example, that the forty thousand padas

references

numberof theI

by the manuscripts of thoseto

I beg the reader texts. understand the general statement just now made as sufficing, for the sequel, to cover all such details. Occasional suggestions of emendation. The Vedic

Rig -Veda will not have to be indexed again,glad to think that the final supplement, con-

am

siderable though it may be, cannot possibly involve anything like the labor of the foundation-work.

In this matter I have been very abstemious, but I have not refrained entirely from making an occasional conjecture as to what seems to me to have been thecorrect

texts often invite to independent emendation.

The elaboration

of a reverse concordance.

Withthan

the present work serving as a basis, a reverse concordance, that is, one which arranges alphabetically theverse-lines

and formulas by

their endings rather

form of some nowIf

phrase.1

any of

my

unintelligible colleagues are disposed to look

word

or

by their beginnings, should not prove too difficult an undertaking. I am sure that it will prove a valuableone.

We may perhaps except parts of tlie AV. as edited by Roth and Whitney, e.specially book xix. and the Kuntapa-hymns of book xx. Here the editors have practically rewritten the text at many points, and it might have been well to report the of themanuscripts rather than those of the editionreadings are easily accessible.;

A very considerable numberThis matteris

of Vedic verse-lines

readings but those manuscript

are nearly or wholly alike in their endings, such as sukrtdm u paravie vyoman (or vyoviani), or. . . . .

.

lokam.

discussed once more in another

connection

below, at

page xiv".

A

reverse con-

Preface]

ACKNOWLEDGMENT OF OBLIGATIONSreveal

XI

cordance wouldagain may yet bedirection;

these relations,

and would

in its

own way present many new uses. It my own fate to continue the work in thisI claim

no special right to the task, and any competent scholar is welcome to undertake In that case, however, I hope that he will first it.but

Greek in the University of Cincinnati; Dr. D. H. Holmes, Instructor in the Classics, High School, Brooklyn Dr. J. T. Lees, Professor of Greek in the Dr. H. W. Magoun, someUniversity of Nebraska; ;

time Professor of Latin in Redfield College Moore, Professor of Latin in Vassar College

;

Dr. J. L.;

Dr. J. A.;

communicate with me,uselessly duplicate the

in

order thatlabor.

we may not

same

Ness, Instructor in Latin in Wittenberg College Dr. R. B. Steele, Professor of Latin in VanderbiltUniversity; and, finally, the late Dr. A.W. Str.VTTON, who was, at the time of his deeply lamented death.Registrar of the University of the Punjab and Principal of the Oriental College at Lahore, India, and Professor of Sanskrit.Professor

ACKNOWLEDGMENT OF OBLIGATIONSThe external University Press of Oxford of this work is such that any specific appearance

To

tlie

comment upon the technical skill, the good taste, and the patient fidelity of Mr. Horace H.\.rt, M.A.,Controller of the Press, is quite uncalled for. I am certain that I owe much also to the valuable help of the Oriental Reader at the Press, Mr. J. C. Pembrey,

Arthlr

A. Macdonell, of the University

of Oxford, kindly furnished me with a printer's proof of the index of Vedic pratlkas since then published

Hon. M.A.

especially care of details concerning the alphabetic order of the index. The services of both these gentlemen I grate-

of Oxford,

for

his

watchful

on pages 105-114 of the first part of his edition of the Brhad-Devata and a similar kindness was done;

me by Professor Moritz Winteenitz of the University of Prague, who made accessible to me the materialof the Mantra-Patha of the

acknowledge. To Pupils and Colleagues. A good many years have gone by since the first preliminaries of the work on the Concordance were begun. It is pleasant to think that in those early days I was aided by the members of a modest institution within the Johns Hopkins University, the so-called Vedicfully

Apastambins before

it

was

Hanns Oertel published. Feiedeich Knauee of Yale University, and of ProfessorTheservices of Professorfessor

of the University of Kiew, and in particular of ProLeopold von Scheoeder of the University of

Seminary.textsfor

The work

of reading the several Vedic

the purpose of marking off properly the and formulas I did of course myself; but, padas this important preliminary done, I would often turn

Vienna I have already described above, at page ix*", and acknowledge them gratefully. The editor of this Series, Professor Chaeles R. Laxmax of Harvard Univer.'^ity, teacher, colleague, and friend, almost from the day when I began mySanskrit studies in 1877, has brought to bear his great editorial talents upon this Concordance in

over to one or another of these my pupils a text or part of a text in order that they might transcribe the mantras upon slips. Portions of the first draftof a rough manuscript were also prepared by some of them from the slips after their partial arrangement. Not a little help was derived from their willing

every possible way. He has also found time to read one proof of a number of sheets (i-io and 49-50). I can only regret that I have not had the benefit ofhis trusty eye and his sound scholarship for the of the work, to the exclusion of many a blemish.It is a pleasure,

bulk

hands in these ways, and I desire here to recall the unselfish and enthusiastic spirit in which each gave what he could spare of time and labor. I hope that even those of them who aided least have not remained unremembered. The names of these scholarsare as followsin the Johns:

pupil and

finally, to remember friend, Mr. Henry C. Waeeen,

my

former

of

Cam-

Mr. Warren gained the respect bridge, Massachusetts. of students of India by his scholarly volume, entitled Buddhism, a series of illustrative translations fromthe Sacred Books of the Buddhists (volume iii. of the Harvard Oriental Series). To his love of Oriental

Dr. F. R. Blake, Instructor in Oriental Languages Hopkins University Dr. G. M. Bollixq,;

learning

it is

due that

this

Concordance

,ind the other

Professor of Greek

and;

Sanskrit in

the

Catholic

volumes of this Series are issued in a suitable andI record dignified form. At the close of this long labor, the kindly and substantial aid which he gave, while

University of America M. Beandow, A.B., sometime Assistant Librarian in the Johns Hopkins University;

Dr. H. L. Ebeling, Professor of Greek in Hamilton Dr. C. J. Goodwin, Professor of Greek in College;

alive^ to this

undertaking

;

and

I

am

glad to say that

Lehigh University;

Dr.

J.

E.

Haery,

Professor of

the enlightened provisions of his last the studies that he loved in life.

wiU now

further

Baltimore, October, 1906.

MAURICE BLOOMFIELD.b2

EXPLANATIONSINTRODUCTORY TO THE USE OF THE CONCORDANCEGeneral scope of these explanations. Since this Concordance has to do with no less than one hundred and nineteen different works, it is, in the first place, obvious that a complete list of those works should be in a systematic order and with proper biblio-

uha-padas, which consist of an otherwise identical phrase varied, for instance, by the substitution of the name of one deity for another. Such are

given,

mam mam

agne bhaginarii kuru,indra bhaginarii kuru.first.

and

and that graphical descriptions of the editions used, the bibliographic list should be followed by an which the alphabetic list of the abbreviations by This list of abbreviaseveral works are designated.tionsis

Variants in padas other than theportant, next, to draw the differences between

It is

im-

attention to the cases

where

two or more stanzas cropSince, foris

body

of the printed just before the beginning In the second place, while of the work.

mainit is

some other pada than the fii'st. the Concordance, the unit of comparisonout in

that the several more or less arbitrary devices employed in reporting the variants will be it is neverreadily understood by every Vedic scholar, theless possible that uncertainty of one kind or

probable

the pada and not the stanza, the Concordance reveals very many such cases which would of course never come For example, to light in a mere index of pratikas.

and I have another should arise in some mind accordingly thought it worth while to explain these devices systematically and in connection with a few;

stanza 3 of the RVKh. 10.127 is in part parallel to Ai-S.3.7; but the first three padas of each stanza areso unlike those of the other that no index hitherto

available could have brought to light the parallelism. Not until we come almost to the end do we find the

illustrative examples.

correspondence.

The fourth padas are vi9vasya jagato nic^am, RVKh.,

and

GENERAL BEMAKES ON THE CHAKACTEB OF THE VARIANTSbetween truly variant and vaguely similar passages. The similarities between related Vedic passages are of very varying so close that one degrees, and range from a likenessfastline

vi^.vasya jagato ratri, ArS., a parallelism which the alphabetic arrangement of the single padas in the Concordance makes apparent

No hard and

automatically. Similarly the stanza RV.1.141.13 shares its second and third padas with several Yajus texts (KS. 7.12 MQ.1.5.2.11), while in padas one and four;

be called a true variant of the other, to a likeness so slight or vague as hardly to be worth reportNevertheless, if we take a few typical examples, ing.

may

ApQ.5.9.10; it is wholly different in subject-matter and diction. The RV. stanza reads:

the differences of character as between several of the

many

possible classes are palpable enough.

True variants. Uha-padas. Thus, if we take the two padas occurring in the Rig -Veda and the AtharvaVeda respectively, and acitti yat tava dharma jmyopima, acittya cet tava dharma yuyopima,plain that the latter is a true variant of the In this instance even the motive of the former. is plain: the Atharva-Veda transfuses the changesitis

astavy agnih cimivadbhir arkaih samraJ3'aya prataram dadhanah, ami ca ye maghavano vayarii ca

miharh na

siiro ati nis:

tatanyuh

;

while the Yajus stanza readsut samucb-an

madhuman

iirmir agat

samriljyaya prataram dadhanah, ami ca ye maghavano vayaiii ca

isam lujaih madhumat sariibharema.while to consult Accoi'dingly it is clearly worth the Concordance for every pada of a given stanza, if such parallelisms are not to escape notice. Essential similarity with wide divergence of word-

hieratic language of the

Rig-Veda

into its

own more

popular wording.

Another well-defined, though not

common,

class of true variant passages includes the

ing. Once

more,

it

sometimes happens that the

XIV

METHODS USEDis

IN EEPOETING

THE VAEIANTS

[Explanations

stanza wording of one and the sametext

so varied in

the one the different texts that not one of the lines of

lUEETHODS USED IN

A

QG.

is quite like the corresponding is afforded by AV.5.25.8 striking illustration The text of V. reads 1 9.1.

line in the other.

REPORTING THE VARIANTS

andVariants for which the alphabetic order does not require more than one entry. These cases are so simple that little need be said about them. When,1.

A

:

adhi skanda virayasva garbham a dhehi yonyam,

vrsuyavan prajayai tva nayamasiwhile the text of QG. reads:

vrsasi

;

abhi kranda vilayasva

garbham a dhehi sadhaya,vrsanam vrsann a dhehi prajayai tva havamahe.the fourteen words of the latter stanza, only five of the prior correspond in form and place with words stanza and yet the two stanzas are beyond a doubt

example, a single word of a passage differs in two more texts, the reading of that text is put first which stands first in the usual order of the texts, RV., AV., SV., Yajus texts, Brahmanas, and so on, and the variant reading or readings follow in Thus parenthesis.for

or

:

yah

praniti (AV. pranati)

ya im 9moty uktam RV.io.

135.4'';

AV.4.30.4^variants occur at more than one place,its

Of

Similarly, when

they follow in parenthesis, each in and with an indication of its source. vasunicarurvi

proper place,:

Thus

;

alike in their purpose and meaning, and they are in all probability two versions of an originally identicalstanza.

The

like

is

true, for;

and AV.6.36.3 (Ag.8.9.7reads:

example, of SV.2.1059 The SV. text gg.io.11.9).

bhrjasi cayyo) jivan AV.I9.24.6''; SMB.i.i.6'i; HG.i.4.3'';;

(SMB. carye bhajasi (SMB.

;

ApMB.

caryo;

HG.sa)

HG. bhaja

ApMB.written

2.2.8"*.

ya idam pratipapratheyajnasya svar uttiranrtiin utsrjato

This means that the four texts concerned, out in full, would read as follows:

if

AV.

:

va^i

;

while the AV. text reads

:

SMB. HG.

:

vasiini carur vi bhajasi jivan vasuni carye vi bhrjasi jivan vasiini

;

;

:

cayyo vi bhaja sa jivan

;

sa viva prati caklpe rtunr utsrjate va9i yajnasya vaya uttiran.

ApMB.

:

vasiini caryo vi bhajasi jivan.

Cases of partial correspondence At the beginKut so numerous ning at the end in the middle:

;

;

So much for variants that concern single words. The case of a variant which concerns two or more the successive words is as a rule no less simple to leave no doubt as to the device employed ought:

are the

cases

of partial

correspondence

or

minor

precise

wording of the severalis

parallelism between padas or passages, that no amount of mere attentive observation can be expected to dis-

the variant

texts. Here, again, within the parenthesis as before. placed

Thus:sa

them all. Set phrases, groups of two or three words what Bergaigne used to call formulas are, as every Vedist knows, the commonplace of Vedicclose

no

mayobhiih;

pito

avi9asva;

avi9asva

AQ. pitav avi5eha

(QG. PG. pitav MQ. pitur avive9aTB.2.4.8.7'';

;

SMB.

pitevavi9asva)

TS.5.7.3.4