k*V
l<

\

'
/////

'/a//,

///////y//</'

VEDIC METRE

A. F. AVE MARIA LANE. WELLINGTON STREET. fUtpjig: F. E.] . [All Bights reserved.K CAMBRIDGE UNIVERSITY PRESS WAREHOUSE. C. Hon&Ott: lasgofo: CLAY. 50. Calcutta: MACMILLAN AND CO. Manager. BROCKHAUS. #efo gork: Bomba-g anti THE MACMILLAN COMPANY.C. Ltd..

CAMBRIDGE at the : University Press 1905 .. 10 VEDIC METRE IN ITS HISTORICAL DEVELOPMENT BY E? VERNON ARNOLD. Xrr- It . CAMBRIDGE..D. PROFESSOR OF LATIN IN THE UNIVERSITY COLLEGE OF NORTH WALES FORMERLY FELLOW OF TRINITY COLLEGE. Litt.

A 7533*l4 63023G 2.A. AT THE UNIVERSITY PRESS.ambrtoge : PRINTED BY JOHN CLAY. ^S<o . M.

.DEM ANDENKEN RUDOLFS VON ROTH DES GEISTVOLLEN LEHRERS DES ERFAHRENEN MANNES DES HILFREICHEN FREUNDES GEWIDMET VON SEINEM EHEMALIGEN SCHULER.

.

. The various forms . The less usual metres III. ix xiv 1 General Introduction II. Doubtful hymns and fragments . XI. III. VII. 327 . General conclusions Appendix IV. Rearrangement of the Rigveda proper Sandhi Syllabic restoration . Ch. Ch. . X.CONTENTS PAGE Preface Ch. I. VIII. V. Ch. .. IV. Table of hymns Ch. Ch. 22 . Metrical Commentary Corrigenda and Addenda 326 Indices : Index of Subjects Sanskrit Index . Quantitative restoration 108 149 175 Dimeter verse Trimeter verse Ch. II. IX. Stanzas and strophes of the stanza . . Ch. The linguistic evidence of date 28 41 Appendix Ch. . . 47 70 81 Ch. . . . VI. 228 244 250 269 289 Appendix Ch. 331 . The popular Rigveda Appendix I.

.

cannot determine with any exactness THE Rigveda its We by external evidence the authors for our itself : date. hymns. and was thus led to observe that even those parts of the Rigveda which were not in his view later additions. and to be associated instead with These poems have generally been the poems of the Atharvaveda. but a Before library and a literature. accordingly to delimit the popular Rigveda. it is necessary that they should be methodically arranged. and the majority of first My task in this work is ' Mandala x. described as 'later additions to the Rigveda. Almost all western students of the Rigveda have perceived that the Rigveda contains a number of poems which by their special characteristics in language.' and in my opinion this Since however Professor Maurice Bloomfield description is correct. These differences but they can always are sometimes broadly marked.' which has generally been vaguely described as consisting of certain hymns appended to the first nine Mandalas. subject. differed very materially in their metrical structure. sometimes subtle : be clearly recognized when expressed by numerical calculations. 1888). The first systematic attempt to delimit the popular Rigveda was made many years ago by Professor Hermann Oldenberg in his Prolegomena to the Rigveda (Berlin. It is . 1899) questions ' ' this view. Oldenberg also adopted the method of statistical calculations. in his I different work on the Atharvaveda (Strassburg. but have adopted from him the title popular Rigveda for these poems. have not thought it right to take it for granted. contents can be appreciated. the locality. or the personality of the we are therefore thrown back upon the poems themselves knowledge on these points and the internal evidence lends most readily to an enquiry into the relative date of the . For this purpose an examination of the language has proved the most effective method but the evidence is greatly the fact that the differences found in the language strengthened by those of : are throughout accompanied by differences of metre.PREFACE is not a book. and metre are to be distinguished from the main body of the hymns.

pp. and to give instead with all possible fulness the materials from which I have drawn my own conclusions. and should be a real assistance to the study of its I do not for a moment. Rigveda and Atharvaveda (JAOS. xxxviii. whole literature. Rigveda vii The second Mandala 33 (KZ. pp. xxxiv. by of the is to one of four successive Rigveda proper is assigned a true adumbration of the historical development of the periods. the collections themselves to be of service to future students. 429-486). of the Rigveda (KZ. I been shewn long ago) that the it is conclusively shewn (if it has not are in a true sense later additions. Rigveda vii 18 (KZ. At the same time. pp. and further that these too The conclusions I have reached have been in their main outlines hymn ' ' published by me in the following articles : L in the Rigveda (Festgruss an Rudolf von Roth. In my view this question is to be answered in the affirmative. Rigveda. assist us to obtain an arrangement of the whole see whether they can of the I Rigveda upon historical principles. 297-344). hymns of the popular Rigveda are spread over a long period of time. which each believe that the formal scheme reached in this book. pp. xxxvii. 207-219). think materials permit of any accuracy in details. I have received encouragement from more than one quarter to proceed to the publication In so doing I have endeavoured to of my views in a complete form. Grassmann in : . 1 group of these questions has to do with the proposed his Worterbuch zum Rigveda (1873) has attempted to restore completeness to numerous verses in the Rigveda in the number of syllables and more particularly with the first The resolutions ' by which H. reduce controversy to a minimum. articles also contain my replies to the objections and which my propositions have called forth from several Vedic Although my position has not yet won acceptance. On the other hand. but by which others in turn may be led to form more or less divergent I have not dropped the thread of historical theory which to views. Literary Epochs in the Rigveda (KZ. 309-320).suggest that the meaning. 491-496). xxxvii.x Preface and to the principal object of this book to examine these differences. mind binds together in proper order a large collection of facts but my : even if this thread can hardly fail In elaborating this work I found that solutions were incidentally suggested to several important questions which affect the text of the be not altogether sound. 145-148). Historical Vedic Grammar (JAOS. pp. pp. These criticisms scholars. 1893. xxii. xviii 2).

' and has not Horace said that In this case however a metrical at times bonus dormitat Bomerusl Virgil occasional apology of license.' the syllables being thereby long This general theory is frequently asserted 'metrically lengthened.' that is the final vowels which are found with short quantity in the Pada-patha text and in the native grammars. I trust that the treatment of these points in Ch. namely Th. In certain positions in the verse metrical follows. I take this opportunity of giving some account of it here. Only habit can explain the general acceptance of so remarkable still Latin scholars will faintly recollect that the a metrical theory : so long ago to require the metrical license. there have always been scholars who have questioned Grassmann's procedure. Zubatv.' which the quantity upon the authority of the two scholars who have dealt most fully But in fact with the question. the credit should rather be given to these writers of having shewn that the theory is untenable. This seems the more desirable. Although principle of resolution. and in particular with regard to the so-called Sariihita lengthenings. xi and of -aam for -dm in the genitive the native commentators justify the plural. Benfey and J. : but the poets find themselves unable to laws require long quantity words of the requisite measurement.e. and to be guided solely by the general usage of the Vedic poets. and give to those syllables an value by 'poetical license. which in fact assumes a fluidity of pronunciation not known in any settled language. They therefore introduce provide syllables of artificial is short. Ill will greatly reduce the uncertainty that has hitherto prevailed on these points. The second group Sariihita text of questions has to do with the authority of the with regard to vowel quantities. of the final vowels of what I have termed the variant final which the quantity varies in the Samhita ' text. is derived and is somewhat as from the unhistorical views of the native commentators.Preface restoration of indara for indra. The current explanation vowels. seems are we to suppose to destroy the nature of verse altogether. as I hope this illustration will justify the cardinal principle which I have to disregard in points of this character the authority followed. namely not only of the native commentators. As the method of this book prohibits me from discussing the history of this question in the text. but also of the Sariihita as well as the Pada-patha text. should always find the metre embarrassing just as they that the poets How .' i. of verse even of was thought not ' which the instances can be counted by the thousand. but (more or less regularly) with long quantity in the Samhita text.

But my study of the subject of resolution shewed me that the Vedic poets were in the habit of following very as a ' ' definite standards of pronunciation. since by means of them any piece of plain prose can be shewn to correspond to some metrical scheme. In this way he has no difficulty in shewing that current theories as to the Vedic metres were often founded on a very slender basis of fact.' he argues from the quantity ' actually found in the Samhita text to the metrical preferences of the poets. but as his shortened. Benfey dealt with the question in a series of articles on the Quantitdtsverschiedenheiten. with only a small margin of choice in the case of words which were actually changing their pronunciation . Still. Zubaty 's work is remarkable for the completeness and orderly arrangement of his material. in all previous writers that the poets preferred short quantity in the supposing ' This imagined rule had been imported third syllable of each verse.' in fact that the quantity was doubtful. from habit. Notably he corrected an error into which ' had fallen. which extend over the seven years 1874 1880 and were never completed. from Greek iambic verse into the supposed iambic verse of the Rigveda without any enquiry as to the facts.xii Preface final if reached a the metre will a 1 vowel % and why should they be embarrassed at all by they possessed in themselves the ' power to transmute at ' vowel quantity ? In truth the two principles of resolution ' and metrical lengthening together. For many years I had avoided this subject as too difficult for treatment. treatment of this subject on pages 108-140 will be but perhaps this will be more thought complicated and difficult excused by those who are familiar with the voluminous writings readily I fear that my : and inconclusive results of my two predecessors. work proceeds is used with increasing confidence. course the idea of ' metrical license ' is * ' Zubaty took up the matter about ten years later (1888-1891) from a rather different standpoint. He finds repeatedly that the long vowels are used in positions in which short vowels would be equally or more acceptable from the metrical standpoint here of : out of the question. Benfey continues to use the explanation metrical wherever the long quantity suits the metre well. if unduly pressed. Assuming that in all these forms the Vedic poets were free to choose either the long or the short syllable. namely that the vowels in question were originally long. and had intended to assume (with Zubaty) the correctness of the Samhita text working hypothesis. but unfortunately its publication was never completed. which is usually the case. but in the course of time This explanation is at first quite tentative. His studies in comparative language at once suggested to him an historical explanation. reduce metre to an absurdity.

short that is to say. carried out. All the difficulties disappear when Benfey's suggestion it is new. This discussion so far as it an unsatisfactory result in that must be felt as to the general increases the doubts may seem to lead to correctness of the Samhita text. it must be remembered. to a hearing. of later date than the forms used by Virgil but from the special standpoint of Latin literature the forms peculiar : If therefore it is urged that the bhara (2 sing. to Plautus are the archaic forms. the Samhita text has partly. that this theory is to some extent founded upon the incorrect postulate of the Sanskrit grammarians that bhara. but the hindrance of the Samhita text by the prevailing usage. and the theories of comparative grammar must follow. The metrical position is not the cause which has produced long quantity in certain cases. . not in a hesitating or tentative fashion. which the forms used by Plautus There are numerous instances in are. instances in which any real doubt remains become in this way reduced to a very small I number. and thereby seems to shake the solid But the errors of the text in structure unde omnia cognita pendent. first.Preface in xiii common speech. failed to preserve the original quantity. and secondly. It required some time to examine and classify the many thousands of occurrences in the Rigveda. I felt therefore that I could not accept Zubaty's theory of optional quantity. verb-forms in the view of comparative grammarians. older than the forms bhara. must have established itself in the interval between the primitive period referred to and the times of classical Sanskrit. from the broad standpoint of comparative grammar. and indeed as the facts shew. bharata are the true Vedic forms . but to its logical conclusion. and that a systematic study of Vedic usage must necessarily reveal in every case either a fixed or at any rate a preferential quantity.) are not. that even least if the short vowel is really more primitive (and Benfey at doubted this). : the Pada-patha text entirely. and in the Padapatha text always. But these objections seem in no way entitled In treating of the metre of Plautus or Virgil we deduce our facts directly from the usage of these writers. imperative) and bharata (2 plur. the long vowel may. objections may also be raised by the writers on comparative grammar. and it is not easy to put forward the detailed results in a form in which tliey can readily But the general conclusion is only strange in the sense be followed. am aware of these that in postulating original long quantity for many vowels. though in the Samhita text they are often. bharata. which has prevented the corruption In other endings the The shortening can be traced in the later parts of the Rigveda itself. great number of final vowels are always long in Vedic that is A usage.

A. imperfectly represent these obligations. and to whose principles of Vedic interpretation I mostly A. Ludwig. assisted by a substantial grant from the Cambridge Philological Society. Macdonell. Karl adhere). Upon this question I have of the book. To the liberality of the Cambridge University Press. the Rigveda would not make better progress if the task of emendation were more vigorously taken in hand. A. E. VERNON ARNOLD. Pischel. Kuhnau. Weber. but also for suggestions and collections of material which extend over a large part of my work. Hopkins. A. Oldenberg and and J. Geldner. IX this method of dealing with some of the hymns in the way) But where difficulties of : 1 uneven ' lyric metres. Bangor. interpretation and of metre concur in the same I have little doubt that the solution is to be found in that verses. To the writers already named I am indebted. It remains for me to acknowledge the direct assistance I have received in the preparation of this work. Rapson have kindly read my proofs throughout.xiv Preface small details are so numerous that no commentator has ever left them The question rather arises whether the understanding of unnoticed. and therefore I have suggested (though only in a tentative direction in Ch. Meillet. . E. Bendall and E. for a better knowhardly ventured to enter in the body me an indispensable preliminary to textual ledge of the metre seems to in the Metrical Commentary I have indicated correction although a fair number of simple alterations in the text which (if admissible on : other grounds) have the advantage of removing metrical difficulties. not only in regard to the matters specifically referred to. I owe it that the book has been produced with all the aids of modern typography without any cost to myself. which are here for the first time recognised as entitled to rank as a distinct class of established metrical forms. R. June. as Professor Oldenberg has also done for a large part of the book and each of these has in turn saved me from many errors which otherwise : by placing at my disposal the unon the Quantitdtswechsel. Professors would have disfigured the work. W. There are many others to whom I am under obligations in some one more particulars for their contributions to this subject either in Such are Rudolf their publications or in private communications. Zubaty have assisted latter me throughout by numerous also criticisms suggestions. and Emil I fear the references given to these writers in the text very Sieg. R. J. 1905. Professors H. and the published part of his articles C. Roth (to whose memory I venture to dedicate this work as one of his or old pupils. A.

Our acquaintance with the poetry of many other countries leads us quickly to the conviction that its ten MaTidalas or cycles have gathered up the work many periods. GENERAL INTRODUCTION. 1 . a. or may be of : distantly inferred from our knowledge of the general history of mankind. but a library and a literature. The object of the present work is the study of the metre of the Rigveda in connexion with its history. 1. In this study the funda- mental questions which concern the metre of the Atharvaveda and other early Indian poetry are implicitly contained. To their diligence and sincerity we of the West are deeply indebted for in this way a unique : of the history of mankind has been preserved. must naturally look for unity and tradition. except such as is contained in the pages of the Rigveda itself. 3. Men agreement in its parts. During the period in which this poetry was produced. The historical standpoint is strange to Indian literary of learning who regard the Rigveda with the due to a book of sacred authority. 2. great political and social changes undoubtedly occurred but of these there exists no record. they seek to interpret its form and its meaning by absolute standards. Scrupulously faithful in the preservation of every iota of its text. and of the history of the Rigveda by the aid of the metre. It becomes therefore our task to study the Rigveda from within.CHAPTER I. monument But to students of the European tradition the Rigveda is not a book. and that the original composition of the hymns was probably the work of several centuries. who find in it the respect inspired source of the convictions and laws which hold together the social life of their country. and no small contribution made to its understanding.

the collections of : material at least will be available for other investigators. Western fall critics of Vedic metre. in dealing with isolated they are directly guided by textual or metrical principles. derived from too hasty comparisons with classical or modern verse. and it is hoped that it will appear that the general theory presents a consistent and reasonable interpretation of the facts as a whole. As such it is necessarily has made some tentative in character. 1869). 5. If this is so. and details. feel questions. Some. because the AnuJcramams: of metre. and the nature of the evidence by which In subsequent chapters the different parts the}' are supported. the native authorities appear to be no better have the advantage of them in that informed than ourselves. and because we have a wider knowledge of the general start free On the other hand. it will have a claim to be historical students as a accepted by sound working hypothesis but if otherwise. 'Internal' evidence of metre For this whole abundant. Almost every European and American critic of the Rigveda : contribution towards the theory of its metre but the present is the first attempt to comprise in one view probably the whole of the material available. worked out study of the text of the Rigveda enables us to ascertain for ourselves the principles which guided the Vedic bards 1 and in : regard to details. we stand in danger of history of language. of the subject will be considered in detail. . 1 into two schools. according as F. We we from the prejudices naturally engendered in India the linguistic and metrical practice of the commentators' own by times. importing into our subject other prejudices. undertaking the material at our disposal is on the We have besides at our service native theories by the authors of the Pratisakhyas and the but we have no longer need of them. 6. Hymns to the Maruts (London. lxxvi. or from the instinctive desire to find short and simple solutions of the problems that arise in a great field of literature in which the workers are so few and the desire of completed knowledge is so keen. may be found to need correction in In the present chapter it is proposed to state important the general conclusions which have been accepted or independently reached by the author. having regard to the great antiquity and authority of the Samhita text. p. Max Miiller.2 4.

Now it is admitted that the text must be corrected in some points. Native as well as European commentators alike adopt without hesitation that hypothesis which accounts for the largest number will of facts in the simplest way. we make the textual correction admission of one metrical variation or irregularity will if the accord The with the text of ten verses. because without such correction the Rigveda is merely prose and it is not conceivable that the arbitrary correction of two letters only of the alphabet could convert prose into verse on so large a scale. bards. a very general agreement many important points and this fact at once : shews that we can neither regard the text as final. 7. and criticism will occupy with suggesting alterations of the text wherever it does not conform to them. is to be preferred. we admit the metrical variation. is The clearly quantitative: principle underlying these admissions on either side the multiplicity of instances constitutes the proof. If we can discover the logical ground on which the one or the other further progress will be assured. 9. recognizing the general skilfulness of the Vedic If the propose in the same cases textual restoration. Others. with the possible exception of such modifications as native authorities regard as permissible if on the : other hand a free field strict itself is left for the ingenuity of the second school. nor the metrical standards as holding good without exception. because it is a proportion of verses in one inconceivable that textual incorrectness could cause so large hymn to be defective in precisely the same way.The 'quantitative' : test 3 alarmed at any proposal to tamper with it and incline in all cases of doubt to maintain the text. and to admit a metrical irregularity. has been reached on In spite of the logical dilemma. On the other hand. inclination of the one school were carried 1 theory of Vedic metre ' to an extreme. it is admitted that we have deviation from the usual metrical rules in such a hymn as ii 11. 8. canons of metre will be laid down. the would be merely a summary of the facts presented by the existing text. again. full application of this principle can only be learnt by experience but its short statement will sufficiently indicate the : importance of a full collection of facts as a basis for any theory. as for instance in giving syllabic value to y and : v. If one textual correction : rectify ten verses. 1-2 .

(iii) the rearrangement of the Rigveda in parts . : the text of the Rigveda the metres . partly of metre. Still there is a very general agreement upon two points (i) that the division of the Rigveda into mandalas is. one of these divisions can be adequately treated apart In following the order laid down. and constitute later additions to the original Rigveda collection. and others constituting a large part of the tenth mandala. and should be assigned to some other author or period in accordance with their inner content. it is held that that homogeneous. No from the others. in order to obtain a full survey of the facts. each depend this book : being held to confirm the other. being the work either of one poet or of poets connected in time and in family. or of appearing . 12. homogeneous and (iv) the historical development of the art of versification. of the Rigveda certain features of language or metre recur with much greater frequency than in other parts. Historical division of the Rigveda Western critics are far from unanimous in their views as to the historical relation of the different parts of the Rigveda. based upon historical grounds. the different mandalas (or their : and (ii) that certain hymns parts) being relatively homogeneous at the end of each of the first nine mandalas. or to the more detailed discussion with which subsequent chapters are occupied. or " Rigveda proper. it is held that these are collection is historically misplaced in the text. it will be necessary to refer from time to time either to later sections of this chapter. are homogeneous in character with the Atharvaveda. In consequence of their great importance they will be fully examined in the earlier chapters of sitions just historical for the moment it will be sufficient to notice that they on arguments partly of language. sections of these." 11. (i) (ii) The subject falls into the following divisions . . in part at least. As before. On the other hand. if in that collection we find one or more hymns in which these common features are absent. It may be said without hesitation that the two propo- stated constitute the only possible basis for the study of the Rigveda.4 10. the arguments are If in a particular collection forming part quantitative in character.

Though not in accordance with Indian practice. they have the advantage that they cannot be mistaken for Samhita readings. fsrdyisthah. and final i % u u generally. from an " text is now acknowledged to be an early The " Pada-patha commentary upon the Sarhhita.Textual restoration (i) The Text of the Bjgveda. : . to be given to y v of 1 : \ndriyah is to be read for ndryah. in numerous words and endings the value of a separate (iii) syllable is. combines the final vowel of one word with the initial vowel of the next. ^tuvdm for tvdm in a few words long vowels or diphthongs are optionally (iv) thus sresthah as to be read as equivalent to two syllables the text e. final a a must occasionally. Our is single " authority for the text of the recension. which has been preserved Rigveda unknown antiquity without any variants of importance. 14. the initial a must usually be restored as a separate syllable. because the the metre is open to doubt. (ii) where the text omits initial a after final -ah or -e. and it may be inferred with some confidence that the other texts hymns were historically derived from it. in accordance with the classical rules of Sandhi. that of the metre being the most important. first-hand " 13. for mrlaya always mflaya. Other Vedic texts repeat the matter of many of the Rigveda in slightly differing forms. The following are most important 1 The technical spellings ndriah. svandh. be read as separate syllables. . and cannot be set in opposition the Sarhhita to its authority. but in such cases the Sarhhita text generally appears to have the advantage. The Sarhhita text can therefore only be revised upon internal evidence. 15. for thus for misrepresented suvandh almost always : In other cases restoration itself : is less certain. a few words are regularly pdvakd we must always read pavakd. gopd (v) as gopda.g. tudm indicate the occasions on which such restorations are required. either necessarily or optionally. The following : restorations are amongst those generally agreed upon (i) where the text.

is One or both of these consonants may belong to a word The sounds following. h are for this purpose recognized as full conrepresented by m . or an quantity given represents artificial pronunciation by which a slight irregularity of metre is the metrical rhythm it is and In these cases the Pada-patha usually gives the glozed over. e. the verses in which the word occurs contain we the normal number of syllables. be. without of the accent. If syllables. sometimes as given by The distinction to a large extent follows the preferences of the final vowel of is : not easy to determine whether the the true value of the ending. as is shewn by a correct text ddma d (for ddme d). but if we read instead -aam the regular number of eight is obtained.6 Undetermined problems in numerous verses some part of the noun indra follows (i) the caesura. regard to the position syllable containing a long vowel or diphthong is necessarily long. But and these verses may (iii) verses of seven syllables are in many such hymns fairly common. and the next word in the same verse begins with a vowel. In such cases the quantity of a final in diphthong is that of its prior element. tdsmd adat (for tdsmai adat): but the is quantity of a long final vowel when not combined by Sandhi not always easy to determine. many flexional endings and of several the text sometimes as short. But on the other hand indra is commonly a disyllable.g. and verses of ten syllables are sufficiently common difficulty. short value. and the number of syllables in the verse appears to be ten. but this is only evidence of the pronunciation of the word at the time when this commentary was composed. 16. the only doubt arising when succession of the short and long A the long vowel or diphthong is final. and are by some critics. in the Rigveda to suggest a metrical solution of the in numerous verses ending with some genitive plural (ii) form in -dm there appear to be only seven syllables. adverbs long. whereas most verses in the same hymn have eleven syllables. metrical value of a word depends solely upon the number and syllables which it contains. provided that it belongs to the same verse. A syllable the vowel of which is followed by two consonants long. so interpreted. interpret indra metrically as a word of three indara. In the determination of the text of the Rigveda that of the The metrical value of words and syllables is implicitly contained.

The principal metres sonants is : 7 and the Vedic Ih (also variously represented by Ih or dh) equivalent to a double consonant. which consists of four dimeter verses. each of twelve syllables. form a Jagatl . 167 io. Four trimeter stanza. we term catalectic otherwise.' the ' stanza. prescribed. 22. of metrical restoration in the Rigveda : (i) Samhita text (following Aufrecht's edition): \ vaydm adyendrasya prestha vaydm 4v6 vocemahi samarye purd mdhi ca no dnu dyun tan na rbhuksd nardm dnu syat || vaydm (ii) Text metrically restored : vaydm adyd indarasya prdyistJidh vaydm 4u6 vocemahi samarye vaydm purti mdhi ca na dnu dyun tdn na rbhuksd nardam dnu syat | || i . ' 18. Similarly. distinguish as a dimeter verse or of eleven or twelve both of which varieties are included under the name trimeter verse. verses. The following stanza will illustrate the general character 17. (i) (ii) The most typical forms of the stanza are the Anustubh.' 19. 20. The units of Vedic metre are the verse. when we syllables. which consists of four trimeter verses. such verse of eleven syllables with the loss of a syllable before or after the caesura. These shorter verses may be termed decasyllabic. trimeter verses which contain only ten syllables verses being usually equivalent to the 21. A verse consists it most commonly either of eight : syllables. The number of syllables Thus many dimeter if in a verse is not quite rigidly verses contain seven syllables only : such verses. the more general name heptasyllabic may be used.' and the ' hymn. are not uncommon. (ii) The Metres. each of eleven syllables. and the Tristubh. if dimeter verse with dimeter verses : loss of they correspond in rhythm to an ordinary the last syllable.

and describe their more . 25. and even of sixteen. The number large : of such composite hymns in the Rigveda is very but the hymns thus combined are usually homogeneous in character. that becomes probable that we have two or more hymns combined in the Samhita text. 12 or 8. Brhatl and Atyasti (12. Stanzas may also consist of combinations of dimeter and : trimeter verses. but these are comparatively rare. of stanzas in a it Where the number is. 8. A few : structures 27. 12. 12. It is also not uncommon for the last stanza 1 of a hymn 29. syllables in one stanza gives uneven lyric metre. it becomes at once expedient to treat them as separate hymns. 8). however. hymns thus combined differs in date or character from another to it is which joined. are Usnih (8. 8). but usually than three or more than fifteen stanzas. 8). 12. and is comparatively rare in the combination of a Kakubh or Brhati stanza with a Satobrhati stanza. the Pankti of whilst three Tristubh verses form the five. 12. if so. the latter being then usually of twelve syllables all metres of this type we group as lyric metres. or the metre suddenly changes. 8). 28. except in the case of strophic and mixed lyric hymns. to contain one or even two additional verses. Thus the Odyatrl consists of three dimeter verses. 8. a composite hymn. ' External may ' metrical form Stanzas contain more or fewer verses than four. 8. . The external form definitions just given may be said to determine the of the different metres. Lyric metres may also include verses of four syllables. Satobrhati (12. (8. 8. it is not necessary for the present purpose to If. the Mahdpankti of six metre known as Virdj. The most important lyric metres 4). 8. The strophe (pragdtha) consists of the Rigveda. the suspicion arises that one of the separate them. 30. rare various The combination of verses of eight and eleven contain any number of stanzas. it A hymn may ' consists of not less generally uniform in character. and two decasyllabic verses that known as : Dvipadd Virdj. 8. hymn is very large. 12.8 23. and. 26. 8. hymns are composed of lyric stanzas of very these we call hymns in mixed lyric metres. 8. 24. Kakubh (8.

its study is more difficult. but incorrectly. (c) the Tristubh-Jagati/a?m7ies respectively. There are few parts of the verse in which the poets do not consider themselves free at times to depart from the usual rhythms. On the other hand. presumed Rigveda. 32. (6) the Anustubh. These preferences modify considerably the general iambic rhythm prevailing in both parts. in the latter part of the verse for short syllables. may be obvious characteristics. in the sense that the even syllables. of the internal form. ' use of the terms ' dimeter' and ' trimeter' verse in this book is not intended to imply acceptance of any such historical theory. and so on. Hence it has fourth. . been supposed that the earlier part of the verse is non-rhythmical. and is treated by native commentators as of comparatively small importance. for In the early part of the verse there is a general preference long syllables. the different kinds of verse.' such as is found in so many Greek This supposition meets with many difficulties. much metres the rhythm of the latter part of the verse more rigidly defined than that of the earlier part. all In 33. Hence In all is it has often. The lyric metres being practically unknown in later literature. and others avoided everywhere In consequence of the greater there exist metrical preferences. metres in the Rigveda the quantities of the first and last syllables of each verse are indifferent. to belong to the more distinctively early parts of the We proceed to consider the internal form. of 31. are more often long than short. namely the second. 35.' Internal ' metrical form 9 In accordance with them the metres may be grouped as belonging respectively to (a) the lyric. been supposed that Vedic metre has arisen historically from some combination of iambic feet. In almost all metres a general iambic rhythm may be noticed. The internal form is even less rigidly denned than the external. 34. so that it may perhaps be said that there are no 'rules' of rhythm in the Rigveda. there is no considerable part of the verse in which certain rhythms are not steadily favoured. or rhythm. and the metres. but at elasticity : the same time yields results of greater historical importance than the study of external form. and (with some exceptions) each verse is independent in structure.

In this variation the verses are grouped in pairs. Hymns in which such preference consistent are said to be in the Trochaic Gayatrl metre.10 36. so that the stanza consists of two distichs. 38. which may be imply described by the term syncopation. An important variation of the Anustubh stanza found in those parts of the Rigveda which have been already referred to 10) as similar in character to the Atharvaveda. somewhat similar way In the cadence the fifth and seventh is syllables are almost 33 35). used in a in modern music. some principle of transfer of metrical weight. trochaic is fairly In Gayatrl stanzas preference is sometimes shewn for a rhythm in the cadence. 35 appear each to neutralise the other. the principles laid down in 33. 41. the third is almost invariably long. We In the opening the first syllable is indifferent. is therefore fairly typical Vedic dimeter verse : or of vald\sya gomatah dpdvar ad\rivo bilam tudm devd dbibhyusas j tudm \ | | tujydmdnd\sa dvisuh || i 11 5- 40. The dimeter verse falls naturally into two parts. and the cadence. The second and fourth syllables are preferably long ( 33). it ( not unusual to find a short 35). and each distich of two dimeter . This fact seems to is short. There is no caesura. or each consisting of four syllables. Dimeter verse mem- bers. all the verses have a strict trochaic In the following stanza cadence : ukthdm cand dgor arir d nd gdyatrdm | | sasydmdnam \ \ ciketa | giydmdnam || viii 2 is 14. so that the But if the second syllable quantity of this syllable is indifferent. distinguish the two parts as the opening 37. in which the division into parts members is marked by upright lines. according 32. The following stanza. to but often short ( 34). and is regular in ( the Atharvaveda itself. As to the third syllable. but always short ( syllable in the sixth place also 39.

as the caesura is early (that is. In order to make clear the nature of this change. first of the sixth. four syllables . and the name epic Anustubh is justified by the resemblance between this metre and that of the epic sloka. consisting of the fifth. beginning with the eighth. of which the quantity in the opening ' semi-cadence is indifferent. The following stanza shews a fully developed epic rhythm but the Rigveda shews every gradation from the normal Anustubh * ' : rhythm to this. follows : The variation may then be is defined as The seventh becomes in ' syllable. which includes the remaining syllables. the two parts consist either of four syllables. and to the opening of the second and fourth verses the title re-opening. which consists (b) the break. namely (a) the opening. The frequently long. we have . into two parts. To this becomes almost invariably long in the re-opening. only found in the later additions to the Without questioning the truth of this view. or of five and six syllables each. which regu(i) or larly follows either the fourth or the fifth syllable : (ii) into three members.cadence. 11 first In each distich the cadence of the verse and the opening of the second are modified in the direction of lengthening the third syllable. or late (that is. according If the caesura is early. as separated by the caesura. after the fifth syllable).'Epic' Anustubh verses. third syllable. vdyur asmd | tipdmanthat \ kunannamd pindsti smd kesi visd\sya pdtrena \ \ \ ydd rudrend\pibal sahd 42. In the case of Tristubh verse. and seven we have a a long short first part and a long second part and a short second part.' metre the name of later Anustubh has been given. \\ x 136 Trimeter verse may be analyzed in two ways : 7. . better to choose a more colourless epithet. after the fourth syllable). it may be it is on the ground that Rigveda. part first : if late. 43. we give to the last part of the first and third verses of the stanza the special title semi. and seventh syllables and (c) the cadence. which in the cadence ' regularly short.

a still stronger contrast to the iambic rhythm is is not uncommon (. Hymns in the lyric metres incline to the to employment of iambic variations. as is the case in particular in the seventh Mandala. In all cases the second part regularly begins with two short syllables. which therefore appear be characteristic of the more distinctively early parts of the Rigveda. ' if the The middle member ' of trimeter verse is termed the because the general iambic rhythm is broken at this point by the preference for two short syllables after the caesura. But if the caesura is late. but they are only to a small extent combined in the same hymns. and we ' ' 44. Sometimes the general iambic rhythm is not broken in the middle part of the verse.^ . Hymns in which such rhythms are forms 46. The cadence ' ' of trimeter verse shews the rhythm as that of dimeter verse. are regularly short short. Hymns in which this rhythm favoured The said to be in Cretic Tristubh (or Jagati) metre. is identical in rhythm with that of dimeter verse. Thus it is only in a small part of the Rigveda that We . the second part is in each case longer by one syllable.12 ( Iambic' and 1 'cretic rhythm In the case of Jagati verse. where the symbol M marks the position of the caesura. || (l prevalent.and - n ^ ^. which therefore takes one of the w w or v . and the eighth and the tenth are regularly long. 45. Thus the ninth : in Jagati verse the eleventh. have a farther example of syncopation in the principle that fourth syllable is short. if the caesura is early. however. 47. the iambic rhythm is much more strictly observed. iambic and cretic variations are occasionally found in all may be parts of the Rigveda. after takes the form of a cretic . The remaining syllable is usually long so that the most regular break : forms are 1( y v .v. able to trace a progressive tendency to eliminate the employment of short syllables in the eighth and tenth places. On the other hand. The opening of trimeter verse. when the break -). the fifth must be long. an early caesura. may be said to be in Iambic Tristubh (or Jagati) metre. same general syllable. but either or both are occasionally are.

Normal JagatI stanza kaksivate : ddadd dr\bham maha\te vacasydve vrcay&m indra sunvate n mendbhavo vrsa?ias\vdsya sukrato fl \ | \ | \\ | | visvet td te \\ \ sdvane\su pravdcia \\ i 51 Tristubh stanza with iambic variations : agnim vo de\vdm ^ agni\bhih sajosah \ ydjistham du\tdm yo mdrtie\su u adhva\re krnudhvam nidhru\vir rtdva n | | | tapurmurdhd \\ ghrtd-an\nah pavdkdh \\ vii 3 Jagati stanza with cretic variations : imam stomam j u drhate | \ jdtdvedase \ rdtham iva n sdm mahe\md manlsdyd bhadrd hi nah prdmatir asya samsddi \\ \ | dgne sakhye \\ md \ risd\md vaydm tdva \\ 94 49. and has evidently some connexion with the varying quantities which. The following stanzas illustrate the most important forms of Trimeter verse. the division of the 'members' being again shewn by perpendicular lines : Normal Tristubh stanza bfhaspdtih \\ \ : prathamdm jdyamdnah \ \ maho jyoti\sah \\ | saptdsias para\me vioman tuvija\to rdvena \\ \ \ vi saptdras\mir u adha\mat tdmamsi \\ iv 50 4 . 48.). 'Decasyllabic' verses correspond generally to Tristubh verses with the omission of a syllable immediately before or after the caesura this omission we term a rest Two types are used : with sufficient regularity to have given special names to the . (15 the Samhita text assigns in many cases to final vowels. The employment of these final syllables in a position from which other short syllables are excluded requires careful investigation. as has been already noticed iii.Typical trimeter stanzas : 13 a short tenth syllable is at all common whilst the short eighth syllable is chiefly found in the shape of a syllable containing a short final vowel.

It seems evident that the 53. upon the JagatI verse. the rest occupying the position of the To this syllable first following the caesura. as rdsi ksdyam I rdsi \\ | rdsi sdrdha sajosaso n \\\ indra \ ye ca \\ | mitrdm asme mdrutam nah mandasdndh \ \ j \ \ prd vdydvah pdnti \ dgraiiitim || ii 11 14. The Dvipada Viraj verse has a rest at the sixth place.which contain them but these. 50. the quantity of the seventh syllable being indifferent: the two parts are then so similar in form that the stanza of two decaof syllabic verses may also be described as containing four verses or pentads. the most common forms of the break in Bhargava verse point we may . If the rest be denoted by a give the name Bhargava. ( 1 .14 'Decasyllabic' metres : metres of the hymns . shewing division into is Here the members In two hymns of the Rigveda a similar metre is based 52. whether early or late. decasyllabic metres (amongst which we include the Bhargava) presuppose and are derived from the more common trimeter metres. fol- lowed by a trochee. | ^ | and v . following the traditional name of the author of the hymns. only of historic interest. are . The Viratsthdna : verse has a rest at the fifth place. ' ' metre. are frequently found in combination with Tristubh verse. n | \ j - \\ x 77 7. Their precise relation . 51. Example : rayir nd cit\rd \\ s4\ro \\ dyur nd prd\n6 ni\tyo nd samdfk nd sunuh \ \\ i 66 1. five syllables. the relation of this metre to the Tristubh. as well as others. The following stauza is entirely composed of Bhargava verses. the two types alternating : sriye mdryd\sa \\ an\jih\r akrnvata | sumdrutam nd pur\vir dti ksdpah divds putrd\sa n e\td nd yetire td ak\rd nd vdvrdhuh dditydsas \\\ .

The ' ' hybrid verse is again a connecting link with the Atharvaveda and later literature. Internal contamination is the combination of two parts of a Tristubh verse which correspond to different positions of the The combination of two parts consisting of four and caesura. 55. at in hymns of any length. and of Tristubh and Jagati stanzas in the same hymn. An important feature in trimeter verse or the combination in one verse. though very common in the Atharvaveda and later verse. stanza. or not correspond. six syllables respectively gives the equivalent of a decasyllabic verses of this type. In hymns of which the general character is Jagati it is quite usual for one or two Tristubh stanzas or verses to appear towards the end of the hymn. whatever verse with a rest at the caesura . consequently it is least rather the absence of such variations that calls for notice.1 Contamination ' 15 them is an important subject of investigation. down to and including the epic period. their origin may be. \\ ara\tih sdmiddhah J x 3 and seven 1 a. they belong to the distinctively early parts of the Rigveda. Contamination may is contamination. As none of these metres are found in the Atharvaveda or later literature. of parts that do hymn be external or internal. 56. to is a presumption that. the same hymn with external contamination the two features are : exemplified in the first and third verses respectively of the follow- ing stanza : uruvydcd no mahisdh sdrma yamsat asmin have puruhu\tdh puruksuh sd nah prajd\yai hari\asva mflaya indra md no u \ririso md para dah | \\ \ \ \\ j | ( | \ | || x 128 8. The combination of two parts consisting of five syllables It is often found in respectively may be termed the hybrid verse. are fairly as for example ino rajann \ common in all parts of the Rigveda. It forms therefore a connecting link between the Rigveda and this later literature. . is External contamination the combination of Tristubh and Jagati verses in the same stanza. But the contamination of a Tristubh hymn by Jagati is only exceptional in the Rigveda. like the lyric metres. there 54.

. then those to These are usually termed the family Indra. of bards. It is fortunate for historical investigation that this attempt was carried no further. and extending from x 1 to x 84. being based almost Within each collection entirely upon mechanical considerations. It has been thought. and the Agni collections are on the in the descending order of the number of hymns in each. collection There remains the possibility that the hymns in each may nevertheless be on the whole homogeneous: and this supposition is very generally confirmed the contents of the separate collections. 59. of from i 51 to the end of vii. the use of a special refrain verse . The ninth mandala contains only hymns addressed to So?na Pavamana. that this book marks the beginning of an attempted new recension of the Rigveda. in which the arrangement hymns do not necessarily stand first. is by an examination of The most homogeneous (i) striking features which suggest that a collection are the following : the use of the family name of the authors in the separate hymns (ii) (iii) . in which the whole body of hymns was intended to be arranged according to the deities addressed. hymns of the Rigveda are 57. From i and in viii we have another is less series of family collections. not without reason. From x 85 to 191 we find single hymns arranged in the descending order of the contained in each. and consists of fourteen collections of hymns arranged in the ascending order of the number of hymns contained in each the hymns to Agni coming first. each being ascribed by tradition to a particular family 1 to 50. and the use of metres or rhythms peculiar to the collection. These average much shorter than the series first mentioned. also the arrangement is generally mechanical. in each collection. 58. : collections.16 Chronological order of the (iii) hymns The Rearrangement of the Rigveda. In the tenth mandala we find a series of collections arranged regular. number of stanzas It is clear that is the arrangement of the collections as wholes of little or no importance to us. and therefore without value. In the Samhita text the which the largest extends arranged in three principal groups.

There seems a presumption that all these collections belong to a period of originality and rivalry between the bardic families. find reason to ascribe the collections 3643. Grtsamada. iii. and that in these there is great regularity both of external and internal form. and x 20 26 to Vimada. this period we should also be inclined on metrical grounds to assign the collection i 127 139. i 44 191 50. In other collections we find that two metres are almost exclusively employed. Tristubh and GaVyatii. tions are 1922 to Sobhari. and ix 197. Such collections are i 111. 2 . and to conceive it as one in than originality of form was the aim. 32 59 to Kanva. ii.The 'bardic' and 'normal' periods ' 17 Some or all of these indications combine to confirm the ascrip- tion of the hymns of Mandala it) may shortly express the Vasistha family. 94 115. and v each contain a nucleus of hymns of similar metrical distinctiveness. which may be ascribed to Dir- ghatamas. or (as we to Vasistha and that of the hymns of vii to : Mandala and 49 vi to Bharadvaja. It is natural to these hymns to the same general period. and we name provisionally the normal period. 1223. of this group may easily be supposed to have extended over a long period of time. which we may provisionally name the bardic period. viii 1 6 i 11. and those which extend from x 33 to x 84 employ almost exclusively Tristubh and Jagati a. i In the same way we 5. the name of Kusika being on the whole the most prominent and : in the metre there is little to differentiate refer all one collection from another. 2430. All these collec- marked by the fact that their metrical forms are not only different from those which prevail in later literature. Refrain verses and authors' names are very rare in these collections. but also almost peculiar to each family. 58 64 34 and 7493 to Gotama. each of which corresponds to some one of the collections in this class. which perfection rather Such a period would seem it 61. The composition of the hymns 60. The groups i 31 35. to be necessarily later than that last described. iv. namely. 2326 to VyasVa. viii 165 to Mana. 3538 to Syavasva. and in each family over several generations. The collections i 140 164. and the hymns ix 98 111. perhaps to be connected with To the name of Divodasa. and Atri respectively: but very many of the hymns in these collections are of a more normal character.

or are appended to single hymns with which they are not homogeneous whilst a very large proportion are included in the tenth Mandala. These hymns. and some hymns. . and of other metres Rigveda as a whole. and include a large proportion of hymns in which cretic rhythm is favoured. shew the characteristic of the normal forms of the Anustubh and Tristubh. these variations are only occasional. as we have . partly on account of their subject-matter. of the Rigveda. already noticed ( 10). begin to appear with some freAlthough these groups form a comparatively small part quency. with reference to the distinctive often found in these hymns. though seldom such as contain more than four or five stanzas. are usually ascribed to a later period than the rest of the Rigveda. Its metrical characteristics are the 'epic' Anustubh and 'contaminated' Tristubh. and characterize the Atharvaveda to belong to a period of transition which may be provisionally named the cretic period. or of those sections which comprise group of hymns of the Agni or Indra hymns. In many of the hymns. however. but of well-recognized importance. though The linguistic forms which refrain verses sometimes occur. x 14 Other hymns appear at the 19. Authors' names are again wanting. but mention that in this part of the Rigveda are some abstruse philosophical and cosmogonical speculations. and charms.1 8 The ' ' cretic period metre. must not also forget to contained We may use this convenient title. colloquies. rhythm Last in order. The only considerable this type recognized by the Sarhhita text is that of the funeral hymns. It does not. in contrast with the hieratic diction : ' ' ' ' of the remaining hymns. although there is good reason to think them earlier in date than the existing recension of the Atharvaveda. ' seem well to call them here by the by which we should be pledging ourselves to conclusions that are merely provisional. The name popular Rigveda has recently been given to these hymns. 62.' name myths. which at the time must certainly have had a popular character. they seem to have a character of their own. end of the respective Mam] alas. however. and may therefore be supposed to appeal and partly because of the relatively chiefly to the common people modern form of language. is that the part of the Rigveda which resembles in form and character Atharvaveda. which includes dramatic late Rigveda.

that the the quantity of initial and final syllables is always indifferent in 22 . They frequently compare their craft to the highest kind of workmanship known them. without any We have already noticed (32) that regard to their quantity. and to form some estimate of the skill The sketch already in handling their material. the bards 19 We have thus outlined a rearrangement of the contents of the Rigveda into four principal groups.Workmanship of 63. (iv) The Historical Developement of the Art of Versification. except the 'normal' period. can be made of the literary and aesthetic powers of the Vedic bards: but it is important to examine the conditions of the mechanical problems presented to them. as well as in producing verse following established models. the ' titles easily be found misleading if they are supposed to any adequate way the general history of Vedic poetry from the standpoint of the developement of social institutions ' popular may indicate in : such a study lies beyond the purpose of this book. and given to each group are intended merely to facilitate reference and to aid the memory. each containing subdivisions corresponding to the arrangement of the text of the This rearrangement is necessarily tentative. of the Rigveda has shewn that each given of the metrical contents The bards were period. to be a special gift of the race. either ancient or modern of inventing and appreciating new and delicate variations faculty : of rhythm seems 65. that displayed in the construction of a war-chariot and they shew confidence that a new song will be more pleasing to to : ' ' the gods than one which is old-fashioned. was inventive. The time has perhaps hardly come when a fair estimate 64. But no such statement can justly be made with regard to the rather the poetic literature of India. It is a common-place of Western criticism that in many of the mechanical arts the Hindu workman follows too submissively ancient rules and models. shewn by them occupied in constructing fresh metrical schemes. The comparison first of the earlier parts of the Avesta indicates Vedic poets were not far from the period when verse was measured solely by the number of syllables. Such titles as 'bardic' and Rigveda itself.

which determine with varying regularity the quantity required in each position. : 68. the motive of the change seems evidently to be the same. The origin of the decasyllabic verse will is a matter as to which a conjecture be hazarded later (Ch. some of which have been described above ( 34.20 Metrical motives is the Rigveda. had they so desired. vili). which is probably generally recognized as the most beautiful metrical form of the Rigveda. but it is only . simple. 67. and we have every reason to suppose that this feature inherited from the earlier period of purely syllabic measurement. The historical relation of Tristubh and JagatI verse has often been discussed if : and is historically the earlier. to this cause. because it is quite possible that the caesura and the two short syllables that follow it may have been earlier in date than any such rhythm. and may in some cases have been due also a ready But it is instrument of the inventive temper. 35). ' But the ' ' cretic rhythm : certainly and if it be suggests a deliberate aiming at variety of form granted that the epic Anustubh is based upon the Anustubh of ' the Rigveda proper. the Tristubh has Either then Jagati certainly the better claim to this position. are not perhaps entitled to ascribe variety to their metre. or a dimeter verse added to a similar half-verse has been assimilated to the rhythm of the Tristubh. 66. one of the two has been derived from Tristubh by contamination with dimeter cadence. and thus give to mechanical perfection. carried The principle of and we may well it alternation of short and long syllables is believe that the Vedic bards could have But in addition they were guided by other metrical motives. The contamination at a later period of Tristubh with Jagati did not lead to the construction of any harmonious metre but perhaps in principle it is not to be distinguished from that mingling of the lyric metres from which grew the BrhatlSatobrhatl strophe. Above all. 'Contamination' readily suggests careless workmanship. But in all other parts of the verse we find rhythmical tendencies. the 'break' of trimeter verse to a deliberate design of interrupting We the iambic rhythm. they were actuated by the desire to combine these motives. By one method or the other a metre not without a beauty of its own has been constructed.

' ' . they will be glad to abandon the easy but untenable theory that of form employed by them is due to chance. to think that professional reciters and their instructors could by mere accident have a perpetual burden left to the stanzas in a shape which must make them memory.Merits of Veclic poetry 21 through this stage that we can account at all for the existence of the Dvipada Viraj or Pentad metre. which was actually brought to as polished a perfection as 69. In these 'irregularities' there may be meanings not easily recognized. and holding up an ideal which has led each in turn to seek rather to enrich his successors than to grasp at his own immediate enjoyment. 70. This observation cannot fail to suggest reserve and care before bringing forward any suggestions of incompetence against the versifiers or of carelessness against the editors of our text. or the the variety it that purely personal bias of individuals and to recognize instead we find all the signs of a genuine historical developement. any metre of the Rigveda. that is : of united efforts in which a whole society of men have taken part. To whatever conclusions we may further be led in detail. And in proportion as modern students come to appreciate the skill displayed by the Vedic poets. and for this reason they deserve to be carefully studied. They seem indeed to bear the same relation to them as the rich harmonies of classical music to the simple melodies of the peasant. If this was so. then the Vedic bards also are to such as sought be counted amongst great men. must be plain that as works of mechanical art the metres of the Rigveda stand high above those of modern Europe in variety of motive and in flexibility of form. It is difficult to think that a professional left his bard should without motive have verse with an irregular rhythm. creating an inheritance which has passed through the generations from father to son. sometimes apparently ending in failure. not hard to recognize that apparent In all these cases it is irregularity is the result of the inventive spirit. can put it into order for him with It is also difficult hardly a perceptible alteration in the meaning. sometimes leading to the construction of new and harmonious forms. and out musical tunes. when any European scholar. without serious practice of the art of versification. and set forth verses in writing.

16 i-ioa. 129. according Atharvaveda being distinguished by the affixed ii letter a : i23i6-i8a. 97 a. 158. 10a. 181-183. 16 a. 184 a. 34. The hymns which. 59 5-10. 42. 88. 86. 58. often with slight differences in the arrangement of the stanzas and in phraseology and in their general character these hymns can hardly be distinguished from the remaining verse : portions of the Atharvaveda. out from the remainder by differences of language. 174 a. 7 -8a. 55 5-8 a. 14 i-i2a. . 164a. 103. 8 a. 135. 22-23 vii 50. to the popular Rigveda are the following. x 9 6-7 a. v 40 32 6-8 a. 155. 136. a. 166. 146. 60 7-12 a. either because the characteristic features of the two groups of hymns are mixed. 163a. 19. 145a. 109 a. 71. 43 . 57 a. 98. 173 a. 159. 50 10-13 a. 177. . i3~i6a. iv 1 8 5 ? 4-8 a iii 53 17-21. 121a. . 81. 191 2-4 a. 162. 114. 84 a. subject. 15a. 67 31-32. 29-31 a. 114. 47 26-28 a. 82. 72. metre. 8-9 a. 151. 161a. 103 a. 117. vi 28 1-7 a. In srme hymns and fragments a doubt arises as to 73. those of plainly belong which either the whole or more than one stanza is repeated in the 72. 154a. 87 i-2ia. to the definition just given. 162. 244-6. 190. 165 a. 163 186. ' Sufficient evidence in each case is indicated in the Table of ' Hymns at the end of the book. Of these hymns a large proportion are repeated in the first nineteen books of the Atharvaveda. 19-21 a. 85 a. or because they are almost or entirely wanting. 104 a. THE POPULAR RIGVEDA. 59 12 a. 137 a. 107 a. and are marked 71. 112. 191a. and and by their position in the collection. 42 9-na. 17a. 130. 152a. 128 a. whether they should be referred to the popular Rigveda or not. n-i4a. 56 4-7. 189 a. 164 a. 169. 113. 125 a. 142 i > . viii 58. 90 i-i 5 a.CHAPTEE The ' ' II. 74 a. 22-24 a. 83 a. 18a. 91 a. 75 a 5-9. 59 6-7. popular Rigveda consists of hymns which contain about one-sixth part of the matter of the Rigveda. 78 5-9 ix 5 8-11.

hymns and collections of hymns characteristic features must necessarily occur with fair frequency. If however we can group together in any way these shorter hymns and fragments. strictly In the first place the two groups of understood. in which the can only be relatively few. But in the case of short hymns. Here the only real doubt is whether ' the transition period lies nearer to the one or the other of the two main groups: and this question is clearly one which should be reserved to a late stage of our enquiry.' list of the hymns of the ' popular are going further than our general theory. In drawing up a definitive Rigveda. For example. Such generalizations are often of : use in determining roughly the position of the single stanzas. as for instance the collections i 31 35. but overlap. the single Tristubh stanzas is appended to many hymns. and those which extend generally from x 35 to x 84. and do yet relatively they not affect the broad contrast between the two groups. taken together. which we have already assigned to the cretic period ( 61). it may be possible that a general character will come to light in the group as a whole. characteristic features altogether wanting. In order to deal with the doubtful hymns it will be desirable to review the evidence available.Doubtful hymns Although hymns of 23 this kind are on the whole fairly numerous. resemble in character the Rigveda proper whereas the corresponding Anustubh stanzas resemble the popular Rigveda. The difficulties that occur differ in kind according as they affect large or small amounts of matter. this concurrence is not existent in Against this theoretical objection must be set the great errors practical convenience of a definitive list. although the inference may not be finally justified in certain cases. consists of probabilities the evidence available. certainty. . only: and though the concurrence of many probabilities con- hymns stitutes historical all details. permits us. and difficulty In the case of long only arises when the features of the opposite groups are found in the same hymn or collection. In such cases the general inference must be that the hymn or collection belongs to a transition period. i 94 115. we In the second place are not sharply divided. form only a small part of the Rigveda. when analyzed. and still more in that of detached verses. or so scanty as to give room for the play of chance.

The unthematic flexional forms of nouns. 75. and infinitive systems in verbs are particularly conspicuous. though subject in detail to the and in the Table of Hymns will occasional play of chance ' ' : be found stated the number of instances of each kind to be treat found in the hymns and stanzas. obtained by the steady progress of certain The different conclusion. yet it is of great value in giving in almost cases a first presumption as to the class to which a hymn 84 87 will be found lists of the distinctive In belongs. though enriched certain sounds hardly known to the Rigveda proper. aorist. and the extended use of the perfect.24 74. support the inference of earlier date already drawn from the grammatical forms: nor is it to draw possible . ' ' ' ' Other words. features of each class of hymns. having to do with the familiar objects and activities of Hence we are led to the contrast between the ordinary life. developement of the Sanskrit language for shew that multiplicity of form in the stem-formation and they flexion both of nouns and verbs which marks the earlier history of all Indo-European languages. rivals. drawn up upon linguistic principles which are mainly mechanical. tendency to uniformity. and to the possibility that both may have been simultaneously in use for different purposes. On by the other hand the language of the popular hymns. number of Amongst the popular words are a great Indo-European origin which must certainly have been ' ' familiar to the poets of the Rigveda proper/ though seldom used by them. The linguistic evidence First in practical importance stand divergences of language. however. subjunctive. heroic and the popular language. numerous that there are few stanzas in the Rigveda which do not contain one or two and though the evidence of a small number : of instances all is not final. These are so in which we include both grammar and vocabulary. so far as it is necessary to them separately. An a whole shews us that the earlier stage in the examination of the distinctive grammatical forms as hymns of the Rigveda proper record an . It remains to consider how much im- portance can be attached to this kind of evidence in individual cases. differences of vocabulary lead directly to a somewhat Here divergence of style and subject-matter ' seem more prominent. and therefore free from the influence of personal bias. shews a favoured types in destroying or assimilating their 76.

whilst our lexicographers find many In so-called 'American' neoterisms in early English records. In Anustubh verse a single occurrence of a long 7th syllable in ' ' 1 semi -cadence ' is a decided indication of popular character. The evidence : of metre that of language for : did not appreciate the rhythms used by his predecessors. ' ' in much particular passages a minute criticism of the separate words and forms can lead to no practical result: and therefore in doubtful cases linguistic evidence seems in to be of comparatively small importance. the 'epic' Anustubh in all its stages ( 41). is. yet this would seem to be for the younger poet either impossible as regards the rhythm 78. so the popular words at least may find a place in the older hymns under the influence of ' special subject-matter. in the Atharvaveda is a strong of any part 72 that It appears from the list in indication of the later date. whilst the absence of a short 3rd syllable in several successive 'reopenings' On the other hand too much stress must is almost as significant. which are in practice associated in all the hymns in which they occur. 79.The metrical evidence 25 a line between the two classes. and the contamination of Tristubh metre by JagatI ( 54). or desired to improve them. and even later words and forms anticipation of linguistic developements not yet established some by in Amongst ourselves Bible phraseology is still found current literature. 77. The . on the whole. not be laid on single instances of contamination in trimeter verse. As older words and forms may appear ' in later hymns either by direct imitation of particular models or by the unconscious effect of literary tradition. and consequently of later date. or fragment. There seems every reason to suppose that the metres of the from those of the Rigveda popular Rigveda are directly derived proper. viz. more cogent than the external form of older hymns though may be occasionally imitated later. literature. The characteristic metres of the popular Rigveda have been described. or even repetition of a complete hymn of it beyond one stanza. for these are not uncommon in the hymns assigned to the 'cretic' period. But judging of large quantities of matter its considerable and this appears to make it is really of later date. cumulative weight is certain that as a whole the popular Rigveda .

namely at the end of complete collections. viii 24. 154. as is conclusively shewn by the language and metre. belong poems to the Rigveda proper but this presumption fails when the deity : is not known to or not is homogeneous with the Pantheon of the ' a presumption that charms. vii is still of some weight. Oldenberg (Prolegomena. which plainly belong to the popular Rigveda are that the hymns almost all found in certain positions. 50 the quotation of single stanzas. i 22 16 32. 81. 32. to that Hymns which disturb the mechanical arrangement of each collection also deserve careful examination. belong to the popular Rigveda but even Rigveda. x 40. vi 68 10 n . It is only discussion by Professor as a starting point but when other ' indications fail that a consideration of the subject-matter of the poems becomes necessary. To the position of hymns and fragments in the Samhita 80. viz. : So there amongst the early hymns we of conscience or disease tions find a few in which relief from pains is the main theme. 19 11. especially if they appear in the Rigveda to be detached from their context. the Atharvaveda from all parts of the Single stanzas are drawn by but the repetition of such stanzas in the Atharvaveda Rioveda. it With regard full arrangement : will be sufficient to refer to the H. As we observe text considerable importance is to be attached. except by 18. Mythological narraand dramatic dialogues are not entirely wanting in the Rigveda proper. Chapter 11) it must be kept in mind that many of the oldest hymns in the Rigveda are found either out of the order or in the tenth Mandala. or in the tenth Mandala : to find other hymns of the same character similarly placed. 19. or of the Agni or Indra sections of those it follows that we are likely collections.26 The evidence of subject-matter of the more than one half hymns which clearly belong to the with more or less completeness in popular Rigveda are repeated whilst of those hymns which clearly belong to the Atharvaveda: the Rigveda proper only about a dozen are repeated. iv 2. ' There is clearly a presumption that hymns in the strict sense. though they are much more common later whilst the converse appears to be true of poems with liturgical and : .' that is poems in which the profit of the suppliant or the harming of his enemy is the primary object. that is in which the praise of the gods is the primary object. v 46 7 8.

Philosophical and cosmogonical poems. hymns would exceed the scope of this book. that * some period either to the normal or the ' ere tic period. but there will be found in 88 a list of all those hymns of the Rigveda with regard which reasonable doubt seems to exist: and an asterisk is prefixed to those which appear to be more properly assigned to the popular Rigveda. therefore that we may have reason ' ' other than that of the earliest literature. and this to ' table should also be consulted in each case. and to refer it later to is. The evidence of language and metre is given in the Table of Hymns 'at the end of this book. . To discuss at any length the evidence affecting particular 82. and those which contain proverbs or riddles. are generally late. In those cases in which the evidence justify us in assigning a generally be probable that is not quite sufficient to hymn to the papular Rigveda.The evidence of subject-matter 27 ceremonial subjects. it will it is not very much earlier in date.

ghrnti . A 84. F. Lanman's Noun. 63 5d. 1877). Forms characteristic of the Rigveda proper. as the subject is treated in the later chapters of this book. three occurrences are reckoned. the lists on pp. and Avery's Verb. found in the popular Rigveda only. Consequently the eighth Mandala.Inflection in the Veda. In lists of compounds the Pada-patha forms are often given for convenience of tabulation. dagger f marks occurrences which depend upon correction of the text. This reckoning agrees with Aufrecht's second edition of the Rigveda (Bonn. in -a. Notes. 440 References to the occurrences in the Rigveda are only given when they are not readily accessible in such works as Grassmann's Worterbuch zum Rigveda. so that the refrain stanzas and verses which are found in many hymns are as a rule only counted once. that is.APPENDIX I. ix 97 16c. 437. and with Max Miiller's editions and Index. An asterisk * is prefixed to all references to passages in the popular An asterisk prefixed to a word or form denotes that it is Rigveda. In reckoning the number of occurrences Aufrecht's printed text is followed. A. and give in a corrected and amplified form 452 in the same article. 36 16a. THE LINGUISTIC EVIDENCE OF DATE. xvii 4. The Valakhilya hymns are counted as hymns belonging to the 59 inclusive. Words and forms are accented when any of the occurrences are accented.Inflection in Sanskrit. ghana i the instrumental singular masculine or 8 3b. From nouns -a. 436. neuter in Masc. Where one or more verses are repeated many times in a single hymn. pp. 1. The lists given in the following sections 84 87 are drawn up upon the mechanical principles described by the author in Kuhrvs Zeitschrift N. as viii 49 remaining hymns of the Mandala are counted as viii 60 103. The forms connected with metrical restoration are omitted here. not as viii 49 92. 83.

indrapd-. and abl. sarvarathd v 35 5c. iipasruti. from stems in -mant in tuvis-. rjunlti. For details see Lanman's Nomi. ndvyasl. sumati. 20 . in . yusmd-isita. somapd . The forms arydh. -vrdha. -datta. in -dsu. i 52 6a. vajri-. vira-. -yiij . 4. cdru. vartani. text often -i. tavisi-. 12. tuota. in bhuriddbhyah. vrsatvd i 54 2d. The form dive-dive. dhuti. in krstiprdh. pi. -data. sdnil. varivodhd-. s'rustl. sdkfl. in rtupd. saptati. hiranydyi 139 iii. rsi-. niyut-. vdsvah. vivid i 80 15b. avi-. sadhri. havis-mah. The ablative or genitive singular in -vah from -u stems. pasusdh the voc. sukraputapd. gnd-. svtapd. citti. supaptani. in which tud. pp. sucipa. sarvadhd-. prdyutl. . From stems in -vant in adri-. in dprabhuti. sugopd-. vi 48 la bis himti x 37 10b. -ah. haviskrtl. 8. -mant. sakti-. indicating early date . vdsatkrtl. ?iitiktl. sdnu. and at the end of the verse. in -dbhih. : The neuter 10. bhdnu-. mahitThe form mahitvd appears to remain vand. *sumiti. 159). in -a. purojiti. : : 3. sucipe. 334 34 lb. mantu^. sahasrasd-. dyu-. kavitvd x 124 7a. The instrumental tud (only in the phrase tud yujd): compounds in tud-. iv 43 6b. namely tud-isita. -uta. in -d. navati. see 158. sakti. *bahu. dhiti. -nlta. havirde the gen. susasti. sus'dml. dhiyamdhe. The instrumental singular in -i. the loc. pasu-se. updbhrti. 337 some of the instances are uncertain. saras-. marut-. jdspdti. asdnl.Forms . isdnl. see tridhdtu. before consonants. in rtupabhih. 2. -hata. smdddisti. prdniti. prdbhutl. madhupd-. : dat. prdmati. 86 Id. prdyuktl. rju. in asvadd-. bhuridd. sing. purU (puru). x 160 lb. *mddhu. dvyah from stems in art-. 1 1. So too yuvd-datta. prdyatj. ojodd-. sakhid iv 16 x 23 7a. uti. in uru. -atama. nisiti. -datta. in klldlape. for the occurrences see : . viti. 53 9d. ix 107 20c candrd fiv 48 lc camasd x 25 4e yajnd i 168 la bis. pi.Inflection in the Veda. -nlta. sudlti. vrsatvand viii 15 2c. sahas-. v 73 5d. -a viz. saddsd-. rdthid viii : i 20c. suriitl. Neuter: ukthd iv 33 10a. suvrkti. suprdnifo. 141 4d. tapani. -uta. suhdntu and from a radical stem mayobhu. subhamye. mati. 9. 7. vasudd-. ratnadhd-. hdstacyuti. The accusative singular mahdm. longer in use. jdspatyd. vdjasd-. sahasradd-. gopdbhih -abhyah. didhiti. diistuti. svadhd-.has instrumental meaning : -vasu. : Old declensional forms from radical stems in -d. 68 10a. somapd-. sing. . in mddhvah. sdhuti. sucis-. visti. vilu. the instrum. Comparatives and superlatives in -atara. trivisti. plural in -u (-u is rarely correct. The locative sing. sing. in dtdsu the dat. suddtu. asvasd-. 5. vrsti. taranitvd i 110 6c. in -ah. -nid. suMvd *x 141 4b in -tvand kavitvand viii 40 3c. in agrepdbhih. rayi-. dcitti. ratnadheyd iv . sadhdstutl. pi. *avyathi. . sustuti. saci-. sumati. heti. pd/rivisti. dustuti. 6. The vocative in -vah. and indraalso tud-diita. In some cases similar forms can be restored by the aid of the metre viz. the -e. de. -mah from stems in hari-vah : -vant. vdsu. *just%.

ksdbh. raghu-. tvdvat. vasutvand. -yd. ddn. 1. 2d. bhid. are as follows mdhya i 22 8d . viii 67 19a. sds. khd. Pronominal stems in -ydt. x 56 2b. hdvvman. stubh. . Stems of neuter nouns in -tvand: janitvand. stdrlman. are included under No. visvd: manus). 14. 90 2b. yavayu. and ndkih. vadat viii 45 5a. 32 21c. rasdyia. vii 74 5d. The following pronominal forms : yuvoh. dhdrvman. 114 5d. bhuj. didhisdyia. *nddh. van. mils. gfbh. vi 20 13c. 30b. i 68 6a. 20. the verb jmdy and pdrijman). mahitvand. bhis. tubhya. and sridh (with asridh) and nearly all the words which are rare even in the Rigveda. asmdbhya. vid. The occurrences in which these forms are suggested by the metre vi 27 8b. 129 3d. vip. ris. sU. vrd. sometimes found in the text in Sandhi combination. sometimes to be restored in accordance with the metre. tuc. usrd-. The instrumental forms. somaparibddh). 81 6b. yusmavat. vfdh (with the infinitive vrdhe). 22. pas. yaman (with dkhidra-. asmdbhya i 10 8d. ix 32 6b. mdrta (with dmarta. pis. 21. asmayu. mfc. mahdyia. nds. 79 4a. prthu-. yuvayu. tvdyd. xi. martabhojana). 30 20c. mdnus (with mdnurhita. ksdp. yd. mdvat. srit. dnusra-. 17 la. 81 6c. rip. bahupdyia cf. tubhya iv 3 16a. -vat. daksdyia. sanasi. id. arc. kfp. ndbh. Many monosyllabic stems remain in use in classical Sanskrit. uttamayia. v 52 10c : . nrpdyia. jn&s. 16. sti (with stipd). dyutdd-. 326 of historic tenses. For aldyia. x 39 6b. vdrlman. dd. vii 29 la. 116 7a 7c. sdriman. prdthista x 61 5a. 47 10a. C 2. gdh. yuvavat. pandyia. 148 4a. sdviman. 62 7c. 15. psur. istd-. J. . nid (with devanid). These include several common words. 2 dftasi. sakhitvand. and in tvaydt. bhrdj. stut. as mahitvand. C 3 and C 22. 18. tuj. rdp. iv 5 2a. rujdt ix 34 lc. sravdyia. ksu. v 6 5d. pft (with prtsutur). 1 7. 40 la. vol. Unaugmented forms Avery pp. mahaydyia. mahd. In the list given by in the Journal of the American Oriental Society. forms belonging to the following stems anarvdn. ahnavdyid. svayd. ksd. bhdrlman. 117 2d. 63 8c. x 7 2a. viii 1 32a. mar- tiatvand. rit. trd. hr&t. the following corrections are needed: add tdksat i 127 vi 12 5b. 26 12c. vds. 49 lb. tdr. bhdrat v 31 lie. 42 2a . 39 3a. viz. dhartiasi. viz. bhuvat i 61 14d. du. mih. ahamyu. 78 Id. Stems in -asi atasi. viddyia. ud. The forms mdhya. ksip.30 Forms indicating early date 13. sthd. viii 2 ix 88 la. pravdd-. kundapdyia. yusmaydt . tur. bhuman. -yH. paribddh. sdh. pro-. 29 5a. dmh. trayaydyia. pur. 361. : vfs. pre. su-yaman). bddh (with jiiubddh. vft (with vrtariisayd). : : 19. tvayxi. Stems in -dyia atasdyia. vij. ndbh. vitantasdyia. vi 53 5c = 6c = 7c. 35 7d. . mdh (with the adverb malidh). purumdyia. panaydyia. dudrdvat v 50 4b. dkdyia. ksmd (in ksmdh). 84 6b. 71 48 9c. ayd (pronoun and adverb) had. patitvand. gmd (in gmdh). Stems in -iman ddriman. iv 55 8c 6b. navanta *x 120 4e. puru. star. In the case of heteroclite nouns. but a large number are peculiar to the Rigveda. ran. mtt. 46 9c. iii 14 2b. pfks (with supfks). and sdrdha (with sdrdhaniti). said-. tmdn (with purutmdn). pdr%man. sprhaydyia. sahdsra-. mur. dds. 1 dhdsi. jmdn (with j may d.

One or two forms remain in use later. krse. piprdyah. subhdna. bhiddnt. sdna. dchedi. piprdyasva. mamandat. yujand. yasi. dyutand. mdtsvd. vavdnah. hise arcase. The active participle of the root-aorist rdhdnt (with rdhddri. susuclta. dbodhi. alddasah. dsadi. sprdhand. tatdnanta. nesi. jujuvat. rdhddvdra). ydjase. stubhdnd. saksva (sac). For the quantity of final vowels see 159. manand. cakdnanta. vrsdnt (in vrsadanji). gmdnt. vartam. dadhisvd. dadhdrsat. rardnah. bhiyand. dhdksi. 23. drddhi. budlmnd. rnjase. sthdnt. dyutdna. 158). citdnt. jabhdrat. mdtsi. susudah. gantdm. asoci. tatdnah (1 tan). and imperative forms of the perfect 25. saksva (sah). yuksvd. sasahisth&h. sdsdhah. cydvana. vesi. sacand. namely kartam. aroci. vidand for the last see also 3. mdsi. middle. dhdta. daksi. tatanah (2 tan). vividat. heta. naksi. Forms of the second person dual and plural of the root-aorigt 27. krdnt. varktam . kdrta. gdnta. 328 only be necessary to quote one form from each verb. stuse. : : The 31. tatdpate. All subjunctive. The aorist imperative forms in -si. krand (including krand). carkrse. bhaksi. 26. *dhrsand. varta. 29. citana. rucand. dasvams. The first person singular subjunctive in -se : : ohise. asarji. viddnt (in vidddvasu). whether from budh or bhu. rurucanta. jujusan. ddhayi. piyana. pisand. ydtdna. avedi. yatdnd. gdntana. dsrayi^ Cf. ydmsi. Rigveda appears to be 1043. sasahise appears to be 2 p. pdrsi. vavrdhdsva. mdmahantam. except vidvdms: namely khidvtirhs. cdkramanta. yotsi. dudhuvlta.Forms omit rocata indicating earl// date 31 The whole number of these forms in the iv 10 5d. prdsi. niddnd. namely *kdrsi. in arand. susravat. vdvrdhdnta. asavi. srjand. namely cakdnah. 24. vdvrd/idte. vain-dhati. sahand. paprcasi. srosi. mdmahasva. josi. Whitney regards it as a present. mdmrjita. to Historical For details refer All forms of the perfect subjunctive active. dhdtana. vdrhsva. -varna. dyutdnt (in dyutdd-yaman). aorist imperative form bodhi. dmyaksi. gr/ilse. yodhand. dvari. mumurat. caksamlthdh. astdri. vdksi. dadhidhve. svitand. djndyi. optative. namely andsamahai. sotana. dadabhanta. kdrtana . dstdvi. mamdhah. jujosah. No. : 32. ydntana. sing. : The augmented passive forms dceti. mdsva. dmoci. mdmahanta. ydksva. sdtsi. saksi. sasdmate. and sdhvdms. krdhvam quantity of the final vowel see dhisvd. ydksi. hosi. of which 32 are in the 'popular Rigveda': in the independent parts of the Atharvaveda there are 29. yuyojate. ciketasi. 330: here it will as a rule Vedic Grammar. agami. dbhraji. . Perfect participles without reduplication. ayami. dpayi. trddhvam . * f ddrsi. All forms of the root-aorist imperative middle krsvd (for the trdsva. vipand. urand. dhdvi. dyoji. rdtsi. : D of the root-aorist dkdri. djani. vavrdhlthdh. jesi. prathdnd. pdnt (from pa 'drink'). 30. cikitah. mumucah. dadhidhvam. dhrsdnt (including dhrsdn-martas. jugurat. paspdrsat. mumocat. having strong stems. ahetana. avdci. jagraslta. sbta. milhvdrhs (with mllhusmat). 22 above. namely jusand. 33. The middle participle of the root-aorist. rdsva. jaghdnat. vdvrdhasva. puriise. srota. vidhdnt. susuvat. 28. -vin). rdsi. dlddyah. bubodhah. ddarsi. pp. dhrayana.

Adverbs in -vdt. The infinitives in -dhyai. : : 4 : the flexional forms may be classed a number of pronouns and adverbs. 40. irdya-. The infinitives in -vane turvdne. bhrgu-. as these also belong to the grammar of the language : With 41. sacd-. iydnd). say a-. urdhva-. pratna-. akta 'ber. virupa-. vdvrdhd-. sddddyoni. huvd-dhyai. -suta. ddhd nu. ddha tmdnd. mandayd-. namely iyar-. : redse.(present). risayd-. s'obhdse. atri-. But cdksase andjlvdse remain in use later. vdja-sdtaye). vrdha-. dvitd. havydddtaye . vltdye (and devd-vltaye) . : The in -tdye and their derivatives istdye (and vasya-) . manu-. jard-. vrdhdse. pitr-. pipl. maksUjavas. maghdttaye. from the s-aorist. bhdra-. tyd. The pronoun syd. dhruvdse. dsascat). apnavana-. I- (in imahe. vrjd-. isdya-.(except the participle pdvamdna). *-budhna. vdha-. vasistha-. ddhd (for ' 85. 36. in krtddvasu. guhddavadya. prdkse. *rudhdnt. -mdtar. satra-jit.(including asascdt. ddha smd. sdnant (and sanddrayi.' atyd ddri swift. vrnjdse. 1. pdva. sisdnt. All verbal forms from certain conjugational stems. nabhdka-. rnja-. sdha-. manda-. sded (and in sacabhu). namely iyd-. angiras-. dh&yase.m. 3. namely manus-. bhojdse.aksu (and in maksumgamd. : 39. atharva-.32 Forms indicating early date 34. sisacsasc. yaydti-.the desiderative vivdsa-. rabha-. sucdnt. The middle participle in -sand. nd as.' and its compounds adri-jd. purvdeittaye . saha-sdnd. rsi-. vandd-. bharadvdja-. purana-.' m.(not including the causative). ddhd ddhd ha ca. vdjayd-. soma-pltdye) . ddha (ddhd) in combination with other particles (of which tmdnd. paka-.and its participle plpidnd the aorist pddrink' (except the augmented indicative).(from yu 'separate'). stone. dhurvane. stuse. sthurayupa-vdt. gdma-. same form in area-. vrdhdnt. yuyo. B. mddayd ydja-. 38. jraya-. -barhas. . -se. maksuyu). hdrase. aurvabhrgu-. jise. -sdh. -bhid. cardse. stavd-. dohdse. The active participle of the a-aorist. sanddvdja). .(from yu 'join'). -juta.' ' 2. dvitd are themselves old. ddvdne. nama-. tard-. -sdnu. -hdn). piba-. purva priyamedha-. ksddase. namely ddhd cid. huvdnt. vyasva-. see 161): dnusdk (and sdnusdk). sued-. and parti35. the quantity of the final a. -dugdha. spurdhdse. ird-. kanva-. rdjdse. javdse. -ha. spdrase. tu (tu). and the aorist optative asyd. manda-. risant. yuvd. syandayd-. 37. risant. yama-. Mm. dhiyd-. Words characteristic of the Rigveda ' proper. ddha dvitd. rnjdse. ksdra-. sriydse. jamadagni-. see No. isd-. cipial adjectives of the : dhiya-. tujdse. . pltdye (and purvd-. -sdha. tarhsayd-. vartayd-. stdvathe perfect plpi. ndsayd-. jigd-. oha-. -samhata. sadh. sim. 42. infinitives sava-. bhiydse. saddnt. 15 and below). trpdnt. vdsuttaye. satrdc. card-. jami-. sdtdye (and medhd-. 43. sdtra (and in satrakdrd. prnd-. duhd-. The infinitives in -ase arhdse. pusydse.

singular.' with rtayd.' and arusdstupa.' 24. dhdnu-. kiitsa (proper name) with kutsa-putrd. &a 'regard. amd-.' with rsvd-ojas and rsvdvlra. dsvia 'of horses. su-. 17.' arusd ' red. and satydgirvahas. 14. with su-uti. carsani mortal. and djuryat. sand-jitr.Words indicating early date 4. 12. Hay at. 32. -sdh. sadyd-. a. khddo-. dhiyd-. abhika 'face.' a ge. hdri-candra candrd-agra. gdbhasti 'arm. sahdsram-.' with uru-. rta-.' with canasy. 3 . 35.' with carsanx-dhft. visvd-. ws'y (proper name) with ausijd.' ' 18. jur 'aging. 7. 22. su-. visvdyu. drnas 'stream. rtay 'order. and 21. mddhu-arnas. and 29. candrd itself remains in use later. and dwelling. visvd' yuvepas. and -varna. muhurgir. dvya 'sheep. 31. rzy) stretch. -prd. gomant ' rich in cattle. ai/w asuria (text asuryd) 'lordly. ni-. girdh 'songs in the plural gir voice remains in use in the Also compounds of gir. 8. urusy widen. -nirnij. girvdhas. su-dvas. Compounds beginning or ending with candrd gold . 15. -vat.' with gdbhasti-puta.' 13. su-ksiti. ' 28. rtayin. avitdr and avitri helper.' cdnas 'delight. dnti-.' ara^S 'minister. 27. 19.' with a-. ^1 (except the dative ictdye. visvdyuposas. 9.' with dyusdk.' Cf. 10.' ' with camusdd. arnovft. 23. rbhuksdn.' urusyd. sw-. -savas. -ratha.' with akdyia (also included in 17).and visvdcarsani remains in use later. ' rtdyti. su-abhisti. * ' : 33. cdnohita. vi-carsani. svd-. dhdraydt-. -vatsd. -budhna. 6. girvanas.' with avydya and dvyaya. later). 26. ekayu. 34. su-abhisti-sumna. 25. rvd upright. help/ and its compounds abhisti-kft. -dyumna. and purnd-. camu ' cup. which remains in use an-. ' r/ (r/. rbhuksd (proper name). rjisin 'direct. dvas 'help. -pd. rtayu.' 16. and pro.' go-. which remains in use later: ' also avasy. samgir.' except the form dvase. sahdsra-. 30. satdm-. namely asva-.' with arnasd. D 1 2. itd-.' and asuryd (text asuryd) 'lordship. gir' : ' vanasyu. puru-. 5. avasyu. mc?w 'drop. 33 dndhas abhisti * sap. -mat. 20. 11. urvi-. namely giravfdh. cdnistha. and urusyu. ksiti ' A kutsid.

' with nitya-aritra. navy a and navy as 'young. -vrata. dhunlcuirmrl 57. *rta-. The participle dasvdms is already included in 26. tuvi-. 62. turayd. 49. -svdiii.' ' Compounds beginning or ending with dyumnd brilliance. -kurmin. -yd.' with tdvisi-mat. 44. 41. 47.' Cf.34 36. -griva. roaring. wondrous. -grdbhd.' with tuji. antaka-. ddasu.' with dasasyd. tuvismat . -ksd. niy&t 'car. ddasuri. 50. prtsu-. ddmsu. turasdh.' ' devdtdt and devdtdti 'divinity. vrtra-. dasmd dasrd ' ' wondrous. tuvi-magha. -kratu. -badhd.' with niyutvat and niyudratha. worship. Words indicating early date tdnaya 'offspring. -grd. -prati.' ' 55.' insight. asma-. -s'ravas. -brahman. -kiirmi. 46. -rdva. ' ' Cf. Compounds ending dhanv dhlti ' in -druh : a-. The simple noun -vdrdhana.' with *turd-gatu. dji-.' with durona-yu and -sad. duvasyu.' namely ap-. puru-.' daSasy 'do worship. dhuni-mat. dtuji. damsdnd. puru- dasmd. 56. D 25. -svan . aksnayd-. 42. puruddmsas. ddmsas. -irdvas. ddsuri.' ddmsistha. -rdvat.' with dasmdt. turd ' strong. -maghd. supra-tHr. -manyu. . tunjd. -stotra. su-dyumnd . namely tuvi-ojas. 58. vibhuta-. tuvistama. suddmsas. 60. Compounds ending in -tur 'conquering. radhra-.' with duvasy. tvesd-. 54. nftama most heroic' ' paty 'rule.' and ndriapas. -mdtrd. dustdra hard to overthrow. and the noun das' in A A 15. ratha-. dyumnd-vat. -yri.' with dhunay. 61. dyumnin. and dhtini dhuntii. anddvah 2. damsdndvat. dams 'excel. -rddhas. -vat and dnibhrstatuji.' nitya own. tutuji. -y. ddmsujuta. 38. visu-druh. dasmdvarcas. 37. -sagma. 39. dnabhi-. turvdsa (proper name) and turvdsayddu. -hotar. 45. -desna. -nrmnd. 'run. -susma. rajas-. and also the proper name *pdsadyumna. tavisi.' with ddsu-adhvara. mithas-. 52.' . dyumnd remains in use later. -huti . -dyumnd. tdvisl 'strength. -mraksd. durond dtivas ' ' house. abhi-. 40. puruddmsa.' namely abhisti-. 43. visva-. 51. tutuji. duvoyd. 59. and rtddhiti. tuvisvaiids. 48. -jatd. -vdja. Compounds beginning with tuvi-. ddsmia. ddmsupatnl with damsdna. for which see 52L D 2. The verb and adjective tuj 'grow. dM 'do worship. in D nary a heroic. -sati. duvoyti. -sdh. The noun tuj remains in use later.

vanus ' 86. mdmhistha.' with pratndthd. viti-. -jathara. purvd-.' with prdyasvat. ' 80. vdra 'hair. vi-. -pati. vidydt-. 92.' with pavdkd-varcas. gir-.' with yahvi. ghrsvi-. puruspfh * much desired. : -focis. hdri- varpas.' with a-. -vat. dsva-. purilvdsu very wealthy. yajatd worshipful. vrddhd-. -sft. 89. -sd. -vdhas. -bharman. vdja-ratna. 73.' with vdjaydt and vdjayu. A 67. bhdrad-.' with mamhdna. -yu. -dravinas. 104. vd/ias ' tail. pavdkd brilliant. puru-. stoma-vdhas.' 90. ' ' mamh 74. 70. -marwian. ' ghrtd-. sdma-.' paumsia 'heroic' ' A pratnd ancient. already included in 66. -dhd.' with rddhaspati and rdd/iodeya. yajnd-. satyd-. 82. -sdni. -patni. citrd-. 65. su-. vdrtis 'path. -vlra : vipra poet. bi'dhma-. -gandhia. Compounds ending and su-mdnman. mamhanesthd.' with ghord-. -rdjia. satyd-. 79. prdtijuti-. mandrd mdnman 'thought. yahti 'quick. vipra-. in -mahas 'might': citrd-. be mighty. 72. rayi. 35 pdyu ' 'protector. with the compounds ukthd-. arddhds remains in use later.' ' 64. 85.' with varivas-y. jyd-. most of the former are plti drink. -varna. bhardd-. vdjapra-.Words indicating early date 63. nr-. hitd-prayas. prd-. manma-sddhana. vdja-kftya. vitdvdra. and mdmhlyams. and the compounds anavabhrd-. 1.' 88.' with prasastikft. 91. bhuri-. su-vipra. -vid.' with mandrd-ajani.' with ydjatra. with vdravat. sandd-. vipra-. -jihva. mdmhistharati. ' ranvd rddhas ' pleasing. vdjasati is included in No. spdrhd-rddhas. and ydjiydms. citrd-.' 'lord. 75. -prasuta.' rather later. 81. su-mahas. prdyas 'sustenance. 1. ydjistha. varivo-kft. 78. purio-. ' 69. vdja 'prize. tuvi-. with manmasdh. with puramdhivat. a-. sd-piti 36 above.' charm. * 32 . sa/idsravdja -dd -ddvan. 87. and the compounds : tuvi-. 83. wealth. and purumandrd. vdrpas 'wonder.' yajndvahas is in use mrktd-. visvd-. -sruta. waggon. puramdhi (proper name).' except as used of Indra. -pastia. vaks 'grow. vdjinivasu 'liberal. -sravas. su-. with matsardvat and matsarin. 77.' 68. mitrd-.' 'glad. and dur-. vdjarnbhard. -bandhu. vdrivas 'free room. 93.' with vanusy. puruvdra. satydgir-. satd-. wealth. maghd. excited. -vacas.' maghdvan matsard ' 76. pankti-. t -pesas. 84.' with vipra-juta. and prdsasti plenty. 71. pratnavdt is included in 42 above. -yd. sindhu-. maihliayddmarhhayu. -pramahas.' with ranvdsamdri.

: All words containing the palatal t 1. saddvrdha. dpratidhrsta-y abhisti-. *kfplta. which are found in use earlier) namely aratvd. duroka-. vdja-.' 116. srdvas 'glory. devd-. asdmi-. itdtah. spdrhd 'desirable. dyumnd-. -yu . sdndd. sumnd 'blessing. and A -hotra. sdvistha .' with arkd-. D 6.' 108. ajird-. urdhvd-. prthu-. satyd-savas. wise. and 12. spdrhdvira. -id. srava-esd. 36. *vikata. savasi. upamd-. *dandd. savasin. mdrtdndd.' with sdnukd: sukrdtu ' also adri-. su-srdvas. kikata. *ma?iduradhdnika. 99. drna-. hiranyapindd. susravasyd . 110. dvi-. vrddhd-. brhdt-. 106.' stotdr ' singer. . sdnu 'back. su-. A tokd-.36 94.' with spdrhdrddhas. Compounds in -vrkti. suri 'lord. srdvistha dlrghd-. srestha-. 97. tuvi-. viti l Words indicating early date and obtaining. vrddhd-. vdsu-. 114. su-adhvard 'happy in the sacrifice. 95. Words containing nd\ dkhandala. *mahdvaturin. bekandta. se-vrdha . *renukakdta. haryatd 'desirable. svd-socis. *m'anduki. kdtd. devd-. ghrtasnu. Compounds ending in -socis : dirghdyu-. kuta. . gurtd-. 101. *kuta.' suviria ' heroism. prd-. and ghauri-. citrd-.' and suracaksas . suvitd 'prosperity.' havismat 'bringing offerings. 107. included in A namo-. tigmd-. 96. asurd. *dghdti. dndd. sdti 'attainment. pfddku-sdnu : snu. 113. 115. and svd-vrkti. namely ndmo-.' with vlti-rddhas vitdye is included in vlrdvat ' rich in men.' suarvat 'glorious. kevata. vrdhd 'furtherer': vrdhasnu tutivrdha is a-. *vaturi?i. sdndika. susna (proper name) and susiiahdtya. citrd-. (except vdsat. sird-. dhdna-. -yd. sdra-sdti.' stir a 'sun. 98. gdthd-. 103. marud-. ' 111.' ' Cf. 112. dgrbhita-. *kdtuka. 36.' rdtha-viti. medhd-. satyd-.' with sukratuy and sukratuyd. kavi-. *pi?ida.' with srdvas-kdma. prathamd-. 117.' with savasdvat. srausat.*punddrika. pavdkd-. *manduka. vdja-.' 86. sdtdye is included in 104. -y. Forms characteristic of the popular Rjgveda AND ATHARVAVEDA. urdhvd-. 102. 109.' ' strength. . C. sdtpati master. sdmsa sdvas 'praise. 105. kundrndcl. ndksatrcv-. 100. vrddhd-. ghrtdsnu. sukrd-. *sakatt 2. kundapdyia. sravojit . birita.

jdldsa. *va)iani. an-. *pdncapada. ati-. which are established in use (chiefly as desideratives) : Precatives in -dsam the Rigveda proper *jesyd-. bsadhi. *ndbhyah. *nirrtyai. jiva-. not including the in participles. kila. 9. with *nipadd. *bhuyasam. Nouns adapted to stems in -a (a) in -anta . *purusa. 12. : *stosani. *sakiini. which are estab: riila. *dvau. *kaijL. 151 iii. *sakunti and *sakuntikd. kevala. *pra-yajd : *adhirajd. and bhumid 11. 1 29. pfhiyah. karisydsi i 1 6b. *panta. and 576 are to be added astau. dharaPast tenses yisya-. karisyd-. yuvatydm. list palitd. -tai *yajatai. bhujyai. *anu-. *sam-. *rdjani. 6. -misla. *bhajdni. yuvatyah. 258. 342. bahuld. *davisani. sahasra-. *nirrtyah. : 7. 340. kaldSa. vowels (except u) -dv is found throughout the Rigveda but see 1 28. bhrtyai. *vasantd. namely bhumi with dydvabhuml . iv 30 23b. The accusative divam. *hemantd : (b) compounds in -aksd eye . rally) see * ' : 10. *gamani. 13. lished in use in the Rigveda proper A pp. with *purusi. tarantd. Subjunctive forms in -sai. janisya-. For the md bibhi- kriyasma. *dadhdni. 14. 259 4. Certain other noun-stems. sdmgatyam. . -raja .or treated as the beginnings of the transition] (f) *mamsd. occurrences see Avery. 15. The accusative The plural dpah. Duals in -au before consonants or at the end of an even pada. indicating late date /. *pddakd. : dlarsi. vasayisya-. : pustyai. and purusia are established in the Rigveda proper. devdhutiai. dlarti. sad-. 8. Forms from bhumi. paHruseya. 347.Forms 3. sdmanta. adhi-. 16. stavisyd-. *bhrtydm. manisyd-. Before 341. and dvicacali are to be added to it. valg. void. srutyai. 361. and add sathdh viii 79 8b. purusatrd. sanisya-. 346. dratiah. brdvani. of other words containing I is given in Historical Vedic Grammar. ulokd. *janghdndni.(vac). -isya. purusatvdta. *srjdni. purusaghnd. Future forms from stems in -sya. To the lists given in Lanman's Noun-Inflection in the Veda pp. *pacani. -yam (-idm) from nouns with -i stems *dnumatyah. hiranyaaksd : (c) compounds in -yajd. bhuri-. *gani. *randhaydni. phaligd. rajakd : (d) *pada. with *hrdayavidh and *hrdayia. vitfpdld. dpiirusaghna [the forms purusdta. xi pp. *ddnta. -ram. : : md with the aorist injunctive middle or passive. -ratdm *duhdm. vaksyd. heti are included in 10 and D 12. and should either be derived from a stem purus. -yah (-iah). *vaisantd. with *mamsabhiksd : (g) mdsa : For svarajd (conjectu(h) hfdaya. *apurusd. carani. *ndyani. *purusdd. 37 Words containing excepting the following. 52 below. nau. JAOS. turyai. of this formation must be looked upon as desideratives such too are : : karisydh i 165 9d. 5. : Case-forms in -yai (-iai). Imperative forms in -dm. *padagfhya : (e) *purusa. *bhavisya-. in -dni Forms of the 1st person singular conjunctive (or imperative) kdrani. s%oka. locative form mayi. catu?'-.

of the body and its parts: dnga 'limb. yuktvd. pfddku serpent in . mitvd. 20. following : dtho ( 171).' The pronoun adds. 18.' with *anasthd. evdm. ainutah.(2 dhd). : pasin. and visvdtah: and also the : antard (and antarabhard). the pronoun ena the form etdd. *kuru-. aghnyd cow'. *ksind-. bahu (bahuld is earlier) with *bahu-annd.' nald 'reed. sthdpaya-. : 87. for adhardt see No. *snatvd. pdsa 'snare. and saydm required by the metre in place of iva (see below. cdksus 'eye. hdr- including with the last the noun hdr as ' taking.' with *dghora-. grama-jit. bhittvd. ksepdya-. pums (pumdms) ' man ' ' ' . anydtra. with the exception of dtah. and *asthlvdt 'knee'. and bhuripdsa . brdvl-. and *grdmid . with its adverbs addh. *dpasya-. Forms Causatives in -p : indicating late date arpdya-. *bhictvd. D. odand 'broth'. abhitah. mukha 'mouth. (the adverb end is earlier) for uccaih see itara. tdtha. pracaih. pdrus 'limb'. aricya-. pltvd. hitvti. grama 'village. sthipa-. ydtah. Augmented forms of the imperfect passive : aniya-. aslt. *krsi 'ploughed land. All forms from the following verb-stems: 2. and f va when durdm. (1 (1 dhd). pdrdc (for pardcaih see No. whether neuter singular or adverb and the pronominal words ddhara. 21.38 1 7. udaktdh. and amuyd: 22. (hf-). *grhnd-. srutvd. cdtur-. adhamd. 23. kslrd 'milk. *vdpdya- vd). and sanakaih dpascdddaghvan) . Words characteristic of the popular Rigveda AND THE ATHARVAVEDA. *rchd-. rsabhd 'bull'. Adverbs in -aih. nudd- (active).' with dkrslvala . sahd. 23) tdvat. mrna-. ydtama: ydvat with *ydvanmdtrdm : sdrva all (but not sdrva whole '). Absolutives in -tvd (-tud): 19.' and *brahmajayd. in -at. dmucya-.' with *pdsadyumna. 22: in -tah. *darbhd 'wisp'. iXdac (Hdaktdt. pascdt (and in Sdnaih. *jihipa-.' with visvdtomukha . vai. 'ox. *dundubhi 'drum.' with brhdccharlra .' with ndU.' with *prsti-dmayin . *vrtvd. The imperfect forms aslh. sdrlra 'body. tirydc and its derivatives. *tfna 'grass'. asvatthd 'fig-tree'. tdtra.' 1. *hrade-caksus. sura-grama. : ' ' ' ) -hfd. muncd-. anddvah 'waggon'. *dhdpayasndpdya-. dpodaka . *bahudhd. kutah. *asthanvdt. *mohayitvd. *anasthd?i. *-ni. pardcaih. dhdpdya. and *cdksusmat . 2.' with *kumbhin. 129). hatvd. dntitah. kumbhd 'jug. Names of other objects of common life: *aksd 'die'.and su-dnga . namely uccaih. bahupdyia. *visvdtah-. and uccd are earlier : : : No. uttardt.' with arista-. *glapaya-. jayd 'wife. nicaih.' with *dngia. 23). namely apakdt. *prsti 'rib. caru 'pot'. asthdn bone. mahd-. *bahutrd. uru 'thigh'.' with *anudakd. *srotra 'ear. majjdn 'marrow'. kathdm. *-hut. dtidjdya-. aprcya-. with *sarvdtah *sarva-anga. *bahus4van : yatard. ndri 'woman'. *sarva.' with kslrapdkd gosthd 'stall'. *udakd 'water. *bahuprajd. bibhl-. ' Names vilu-.

' with dchidyarndna. dus 'pollute'. 'sacrificial butter.' with dvyanant. dsat. *gopdy tend crt knot. . sar. dsna. *kravydd. jyotirpisdci will jarayu . 1 ' ' ' aranydni forest-sprite water-sprite asurahdn and *dsurd. ' ' ' . mar (mr) 'die'.' with *dtmadd. *sarpd 'serpent'. pad 'go.' with dmrtyu and mrtyu-bdndhu .' Verbs of common life compounds of ad eat.' with prsaddjyd . ' ' ' ' magic.' with nrti. *khanitra. *mrtyu ydksma 'disease.' with *dndviddha and *hrdaydm.' -ydna. sis 'remain over'. bhdgadheya 'portion'. with : ' . nrtu .' with sugdrhapatyd . *rajayaksmd . and popular belief and * dnjanagandhi . with gandharvi.' with *dnasna7it. di4 and pradis region to the devas'. sam viddnd fort'. havydd . ddhi bru 'comvad 'speak'. ydtudhdna wizard. prdnana . dhimsyamdna. nah 'sew'. prdnd. *ajnatayaksmd. jards. sir 'break'. -jndud 'knowledge' in *sa7h-. dsa 'zone'. ' ' ' . khanitrima . bhutd 'past'. jard 'age. devajand troop of devas . cittd 'thought'. -dfijana 'salve. -rogd 'disease. *d-. ' ' . ukhdchid. . .' with vidhdtdr .' with jdldsa-.' with *sa.' with *khanitdr. khan 'dig. *suyavasdd.' with slca/mbhddesna . skambtid 'post. dtmanvdnt. and manands . yavasdd.' with avisd. *sus 'be dry'. stri ' ' ' : *urjdd.pdtha.' with sabhdvat. apsards dsura (hostile deity). devaydna 'directed force'. an 'breathe. *virdj (a metre) ydjus 'rite' *rathamtard (a metre ?) vdrcas 'energy visvdkarman 'first cause'. ghas (jaks) swallow.' with *dnastapasu. atmdn 'soul. satdtman . prajdpati (proper name) with brdhmaria . ' ' *visdvant.' with iksenia . a 'eat. nas be lost. ' ' ' ' . prdM. vasd 'cow.' with *anipddyamdna . dhirhsana. dhdtdr 'creator. sarp 'creep'. ' ' ' . ' . *s'vdydtu and sunesita . *somdd. prdsdvya. ' ' dchimiapatra. *visvd-bhesaja .' with svdhakrta.' with *brahmacdrin student. 3. sabhd 'meeting. and *himsrd. Iks 'see. *sapathia . *suvi-jndnd . papa 'sinful' (but pdpatvd 'feebleness' is earlier). *sabhdsdhd. sapta-rsdyah 'seven singers'.' with vicft and sarhcft . ' . with *yami and *yamdrajan . hinis 'hurt. pitdrah 'ancestors. nirrti destruction o' the wisp brdhmand by magic bhesajd medicine. pac 'cook'. sabheya.' *amdd. *havirdd. . 4. krtyd 'witchcraft'.' with ayaksmd. vitivdd.' with *pitr-yajnd. bandh 'bind' (but baddhd is earlier). and *sdsandnasand . spJidti fattening. *jagdhvdya .' with *yatudhani. *svargd 'heaven'. death. *madhu-dd. woman/ with *strai?ia . yamd (proper name). Words connected with disease. dvis hate '. sukrtdm 'good work'. ajya non-existent : ' ' . and -vittd) . .dh . tdpas 'penance. .' with tdpasvat . dsat dsu existence. sumanasy 'be kind'. *gup 'watch'. brdhmand Brahmin. suhutdd.' with jardyu. apiprdna.' with *asutfp. ' Technical terms chiefly of ritual and philosophy 5. gdrhapatya householdership. jiv 'live' (butjlvdse is earlier). and. visa 'drug. ' . chid break.Words indicating J late date 39 pfddkusdnu. visvdrilpa (proper name) and *vis'varupi .' with tvdpada. possessed of sap 'curse.' in *hrdrogd . gdyatri (the metre) *vratacdrin . *chidrd (but dchidra and its compounds are established earlier). svdhdkrti. ' ' ' ' ' . vi-dnjana. svdhd amen. muh be bewildered vyadh wound. jagdhd. tejas 'vital devdtd divinity '.' in abhi-. sraddhd 'faith'. nart (nrt) dance. 6vdn 'dog.' with vaJdnna . *yavdd.' with *ghdsi. -vedas. karambhdd. dsuniti. *gatdsu . chdndas 'hymn'. bhdvya 'future'. *purusdd. gandharvd (proper name).' with dhimsat. gandharvd.

' in ddhipatya. sam-vatsarina.' vrksd ' +ree. *kasydpa (proper name). abhi das 'plot. vdrman shield. 42. osadhi 'herb. *tlksnd 'sharp. 11.' . *svdsura 'father in law/ with *svasr4.ratrd. pratisthd rule.' bhrdtrvya clanship in abhratrvyd. and in *anisavyd. sugdndhi *gandhd and sugandhi. 15. year ' in pari-.' vasas ' robe ' : one or two compounds are with vrksdkesa.' 22. 35. isudhyd. *varsia. virudh 'herb.' with vdrmanvat and varmin.' 36. 36. 14. sds 'six. Words indicating t late date Cf.' and manigrlvd. *asann-.' But suklrti belongs to the Rigveda. 29. B 4.' ksiprd-. 40. Cf. ' ' ' 30. and su-isH . *-bala. *-hasta .' ' and aranyd.' with *tlksnd6rnga. *tarh 'destroy.' ' with *abhirastra. 33. isudhyu 12. 16. 10. -mat. prajd 'family. 28. with *dnjana-. The phrase *ndma(h) astu worship be *ndth 'need.*aho. dyana dranya *dkitti ' going. -dhi. raani 'jewel.40 6. ' The compounds are common in the Rigveda. upa-. vasin 'ruler. *vi-.' with isu-kft. ' supa-ayand. *rajania rastrd ' of royal blood. isu 'arrow.' except in the form osadhlsu. ddhipati lord. 32.' 43. 7.' with *tdrha?ia and dasyutdrhana. attention. sajatd 'related.' 17. *tigmd-. strange. rudh. B 26. 'smell': ' 18.' t rdtrl 'night. ud-. 21.' with varsd-nirnij.' 8.' in *d-. 23. 24.' 20. earlier. ' ' done. candrdmas moon. 27. standard. *dnumati 'reflection.' ' 34. klrti 'glor}'. is earlier. 37. pdyasvant * rich in milk.' and *anathd. 38. supra-. varsd 'rain.' puny a 'fragrant/ with *ptinyagandha. 25. *dhumd-gandhi. B 103. sam-vatsard . 9.' bhuti 'subsistence.' 41. *nir-. *para-. royalty. 39. 13. pari-.' paridhi 'enclosure. 19. Cf. The group isudhy. 31. rundh -vatsard ' hinder.' with *ati. *ni-.

vi 97 2.' with *vll_uharas. 28 1-6. Quoted AV. *x 88 15d. 48.' with *jdgratsvapnd. 20 are repeated AY. 46. 22 19-21. for which parallels are found in *i 163 13b. would naturally be referred but it can hardly to the popular Rigveda in consequence of its metre be separated from the other hymns of its group. 'lance. : 22 i6-j8.Words indicating 44. and stanza 15 in AV. *deva-. separate these stanzas from 7-9. xviii 1 6. and not of the character of a Stanza 16 quoted AV. both here and 13 9-12. and sendnt thunder. Out hymn. 51. II. 49. the form of question and answer compare *v 44 14. AV. 1 Aetf. 2. and 47. and like 28 1-6. ix 73 ca. : 50 1-9. 52. vii 6.' with *samsfj. xiii 2 16-24.' -4i. APPENDIX 88. send ' army. 29. much depends on 27 13. : *24 1. These stanzas seem to have the character of a charm the interpretation of the phrase pitdram ca matdram viii 69 15d. iii 9 9. late date 41 *sapdtna sarh srj * ' rival. stanayitnu ' 50. those in the popular Rigveda being the nearer. srastar. The linguistic indications receive some confirmation from slight contamination in 15c. I.' with iigmdheti. 18 are repeated Stanzas 19. *hdras 'heat. and the it is Stanzas 1-3 are quoted AV.ala The second section of the Apriya hymns. svdpua sleep. sdrva-send. hdrasvant. 32. *indra-. Mand. ii 5 5-7. 7. *citrd-. possibly later. *84 16-18. sapatnahdn.' with dyuddha-. . *sdrii- sdmiti 'meeting. 15. 45. i 139 11.' provide with. *24 6-15. of order. ^sarhsrstajit. The metre shews some trace of epic rhythm.' with *asapatnd. and sapdtnl. This hymn has to some extent the character of a charm . which are similar in their subject: and the whole hymn may therefore be assigned to the transition period. but only in ix 5 *8 n elsewhere. For 5a. contains late linguistic features do we find either of the characteristic metres of the popular Rigveda. syond (siond) 'soft': and siona-kft. Stanzas 17. List of doubtful hymns and fragments. Part of stanza 9 is repeated AV. can hardly be earlier than the transition period. if standing by itself. 35. Uncertain: cf. This hymn. mahd-> vrddhd-. vii 26 26 4. 5. vii 83 3. On the other hand it is difficult to linguistic notes indicate late date.

Doubtful hymns *90 Fragment of a cosmogony. but this is un39. and 4-8 is out of *93. and disturbs the order of the hymns. : This appears to be of the nature of a charm are quoted AY. Uncertain: cf. vii 97 1. 139 11. iv 33 1-8. 40. See on i 13 9-12. Stanza 1 quoted AY. . *32 4 5. Quoted AV. *29. xiii 2 35 but the hymn can hardly be separated from the other hymns of its group. hymn. The linguistic notes indicate late date. assigned to the late Rigveda on linguistic grounds. iii 20 1. besides shewing contamination. xviii 4 89. also iii 52. Mandala *8 9 10. This fragment has late linguistic notes. the traces . 7. whole hymn is late. 115. xix 9 6. and similar in metre to the next' hymn. Soma is so strange to the Rigveda proper that it is probable that the 9. *126 * 133 6. : it *120 10-12. 188 8-11. Cf. 11. quoted AY. this fragment is very uncertain : Appended Anustubh stanzas. vii 48 1. A typical mythological hymn. Appended Anustubh stanza. *179. *28. vii 6 i. The combination of Agni and order. : These stanzas have the character of a charm but it is very unusual to find such verses prefixed to one of the hymns of the 1-5. but the association of Indra and Soma with Agni is contrary to the practice of the Rigveda. 9. i 27 13. stanzas 1-8 Stanza 16 shews a trace of epic rhythm. and it seems better to assign the hymns to the same group *8 6. In spite of the linguistic notes. confirmed. as i 35. 105 16-18. and the fact that stanza 10 is partly repeated in AY. The metre may be interpreted either as Gayatri with extra verse or as epic Anustubh. Mandala 3 8-1 1. 2. but required here by the order. the general character of this of epic Anustubh metre. ordered collection. all contribute to mark it as belonging to the popular Rigveda. and stanza 16 is repeated AY. The character of disturbs the order of the hymns. Out of order. quoted AY. 105 1-3. III. See on i 13 9-12. The metre of 1-3 may be epic Anustubh. and the nearest parallel is perhaps in the popular hymn *x 173 5c 6a.42 *89 10. II. 170. *97. Uncertain (compare i 27 13). *161. 115. Compare i 35. Stanza 1 is quoted AY. See on i *161.

which also deals with the Cf. 24 *9 *io. 7. *58. *175. but Appended Tristubh stanza. stanza. Appended Anustubh stanzas. These verses form no part of the hymn their meaning much disputed. Stanza 10 is repeated AV.Doubtful hymns 33 1 43 12. 62. 12 6. The position ing it of this hymn is the principal reason for assign- to the popular it is which Rigveda rather than to the cretic period. Appended Tristubh verse. and by analogy with hymns to Brhaspati. i *89 10. isolated in 46 *51 date. See on i 13 9-12. xiv 2 16. A 53 24. 40 4. iv 15 16. is 5. or it may be an appended stanza. The language. *33 *37 is Appended Anustubh Appended Anustubh AV. 8. *39 42 Cf. i *120 10-12. Verse 7a repeated AV. Sadaspati. vi 47 1-25. Tristubh metre. personality of VisVamitra. 13. i 27 13. probably a late addition.ala IV. which appears to belong to the its Rigveda proper. vii 49 1. and Vastospati belongs to the Rigveda proper. 15. AY. *52. Uncertain : cf. in its proper order as one of the series of fragments. 1-7. *83. Appended Anustubh stanza. x *94. Stanza 11 may be regarded as a contamination of the Cf. V. Mand. iii 9 9. repeated stanza. 2. An independent hymn. 82 1. Uncertain : cf. 14. 6. 6a : stanza 8 AV. A mythological This dialogue. 53 1. with connected by subject and language. 61 6-8. Repeated x 126 : : 8. as to which it is only clear that they are of the same period as iii 33 1-12. ir. Apparently a fragment of a cosmogony cf. vii Mandala 5 8-1 1. Indra hymns. 53 2-16. iii 53 9-16. and may indicate later date. but they seem to have the character of the popular *13 hymns. *30 15. The metre indicates late date. The language indicates late date. iv 15 unsupported. Repeated 15. . and being closely parallel to iii 28. *44 14. This hymn is out of order. the hymn is not out of order. Appended Tristubh stanza. is The reference to the Pravargya ceremony the Rigveda. 57 1-3. and perhaps the subject. may this is indicate later 78 4.

x 131 6. xviii 1 48. 53-57. AV. are Mythological poem. xix 11 1-5. Appended Anustubh stanzas. stanzas 19. ii 12 6. AV. *Q6 16. v 62. It does not seem clear that these hymns belong to the vii 91 1. On the other hand stanzas 15-18 are late by metre popular Rigveda. 1-5 (Soma). : *52 52 1-3. : : character. later addition. remainder of the hymn is in If its stanzas 1-3 are a proper order. The language alone indicates late date cf. vii 77 1. 10 are repeated 60 72. Stanza 8 repeated AV. and language. vii 9 3. indicates late date. vii 77 2. namely 1 in AV. r6. 11 AV. 41. Uncertain: stanzas 1-10 stanzas 11-15 in AV. 21 give no certain indications. 92 1. 11. Appended Anustubh stanza. These stanzas appear to form a separate hymn. really separate hymns. rightly 15 i6-r8. and late by the linguistic notes. Uncertain : *49. Uncertain : repeated AV. 7. Out of order. Stanza 9 repeated Unmetrical and probably late. except that several stanzas are repeated elsewhere. This such as *33. vii 86 i. Uncertain. vii 76 6. stanza 3 repeated i 33 AV 2. 9. This hymn consists of a number of fragments. 7. 55 59 : cf. The metre see. 17. linguistic notes. Uncertain. 610 (Indra). *48 *51 22. 4 which stanzas large proportion of the popular 9. viz. Apparently a charm. Stanza 2 repeated AV. 59 9-1 1. placed in the arrangement. . x 126. but the metre is lyric 1. Appended Anustubh stanza cosmogonical. 48. 20. The metre alone Mandala VII. Out of order. Although out of order. 8 AV. repeated *35. and AV. but do not appear in AV. and stanzas 22-25 appear rather to be early in 47 1-25. though they seem to be The first three fragments are roughly arranged amongst themselves. 40. Compare the preceding hymn. and iv 57 1-3. 12-13 RV. it does not appear to be late. 14. xix 15 4. 1114 (Indra) they shew no indication of late date. the 4-6. : *16 47. iii 16 1-7. *15 19. 6 AV. 15. hymn is a play upon an Apriya hymn of seven verses. The Pusan group shews a especially 54. AV. of . xix 10 1-10 . 8. *55 2-4. 2 8 iii 4. indicates that these stanzas really belong to which vii 1 17.44 Doubtful hymns Mandala VI. all of which are out of order in the collection. ii 39.

stanza 20 is repeated AV. Mandala X. Mandala VIII. 6. and hymn 13 is repeated in AV. Mythological poems (Agni and the gods). ix 10 9. *3 Pankti. 2 5. xix 64 3. xviii 3 7. 12 are repeated in their entirety in AV. 2. 6-9. though it is probable that some fragments of earlier date are included in it. *1 34. xiv 1 46. 13 repeated AV. *47 13-18. of this hymn between two stanza. 15. same side. AV. 1-12. as : *102 *102 19-21. 2. as following stanzas 19-21. 45 Repeated 102. repeated AV. probable. Appended cosmogonical Perhaps late Uncertain. The stanza 55 and 56 1 in AV. in 10-11 the subject plainly *100. v 61 : 6-8. *28. 6 1. xviii 1. Probably late following stanza 14. Probably late. xix 59 3. Mythological. Mythological poem (Agni in the waters). 101. late date. others that are clearly late. The language and the epic rhythm in 15c Stanza 17 is repeated AV. *66 69. xviii 3 39-41. 16. 22. 5 is repeated AV. *101 Cosmogonical stanza. vi 51 3. *101 15. i 5 13. : verse. x 8 3. 54. contains some elements of uncertainty: verse 11a for the phrase pitre mdtre see on i 24 1. Repeated *11-13. In the other sections the probabilities are on the indicate late date. Appended Anustubh stanza mythological. Hymns 11. The dialogue form contributes to make late date *27. 2. 54-56 3. *32 40. xix 57 1. 24. The metre is probably *33 *33 47 16-18. 12. suggests late date. *48. Stanzas 3. AV. *51-53. but cf. . 19. 4 are repeated 9 1-5. Appended Anustubh This hymn : shews epic rhythm In the section 7-9 the metre.Doubtful hymns 89 5. The metre seems The position to indicate late date. *1 33. vii 57 2. Uncertain. 14. Stanzas 10. AV. vi 46 3. Out of place. 17d point to *59 15. The Parjanya hymns shew late linguistic notes. but they are in their order in the collection. Probably late. make There are also other marks of agreement in the collection 1014 which it undesirable to separate its parts. and with late linguistic notes.

early date but the position makes this doubtful. indicated by the and by the mythological is *108. *175. more probably : : late. x-ir. repeated AV. The metre indicates late date. : Mythological poem (slaying of the Asuras). *x 146 verse 2c suggests contamination. *191 Repeated AV. 2. See on *x 94. date: cf. v This 124. *139 4-6. x 8 42. *120. to the similarity of position and metre. 62. 3 are repeated AV. Uncertain stanza 3 is repeated AV. hymns must be considered together. of order. In the absence of more definite indications to be of the same late date as those attached to them. 2. vi 63 4 . : *179. *142 *149. Stanzas *1 5 containing the dialogue between Indra and Agni appear to be late. . owing Late date 2. Apriya hymn. 4 in AV. The position confirms the presumption of late date. poem (Pururavas and Urvasi). See on i 13 *119. 123. xviii 3 66.46 *59 these 1-4. cf. Mythological These *101. Out Decisive indications of date are wanting. as of *iv 58. repetition of 101 3. Mythological poem. The date of this ceremonial hymn. *x open to some doubt. Uncertain for the subject v 61 17-19. : vii 84 3. *187. vi 34. 1-6. hymns may be presumed Uncertain. hymn as a whole is out of order. *167. The reference to the slaying of the Asuras suggests late *170. 110 9-12. Uncertain. *139 13. *87 *94. 1. *95. Uncertain: verse 2c suggests contamination. is late date probable. but 127. AV. iii 17 subject of 102. Mythological poem. *x 157. *102. it is repeated AY. Uncertain the position favours late date. wanting. 22-25. *157. 185. For the subject cf. *168. Stanzas 2. v 12 i-u. Repeated AV. The dialogue form suggests late date. The position amongst hymns clearly popular makes 175. Repeated Stanza 6 is AV. *141. 180. vii 72. On the other hand the Vrtra myth in stanzas 6-8 has the signs of For the 'swan' myth in stanza *9 close analogies are early date. *153. Doubtful hymns *60 1-6. The metre suggests Repeated 1.

six of the groups being roughly equivalent to six of the Mandalas. and the remaining four being of a more composite character. It is nevertheless desirable to revise the groups as presented in the Samhita text. as well as those which belong to those parts of the Samhita which are not homogeneous. and that these parts can on general grounds be arranged in three More precisely. REARRANGEMENT OF THE RIGVEDA PROPER. five have the characteristics of the bardic period two are regular in metre. in order that at least those individual hymns which are in most striking disagreement with the groups to which they belong in the text may be separated from them.CHAPTER III. 90. the greater part of the Rigveda may be divided into ten groups of hymns as shewn on the next page. these ten groups. to those These groups to which on the whole they bear most resemblance. or at least those which clearly belong to a different period and then to attach : these hymns. may find some appropriate place attempt brought under in the grouping. processes cannot of course alter the general character of the . Of periods. and thus be review in our examination of the metre. and also that these hymns. and those not yet included in the grouping. first We those shall therefore to detach from each group hymns that are not homogeneous with it. first It is not practicable to define the limits of any of the nine groups with the same accuracy with which the popular Rigveda has been defined. It has been indicated above ( 89. 57-62) that large parts of the Rigveda proper are on the whole metrically homogeneous. and therefore to be assigned to the two have the characteristics of the cretic or normal period : : transition period : and the last group comprises the hymns of the popular Rigveda.

48

The

ten

homogeneous groups
:

respective groups as already defined but if carried out with any success, they will greatly help in removing difficulties which other-

wise might prevent us from obtaining a clear view of the less
striking characteristics of each group.
91.

Table shewing the homogeneous grouping of THE KlGVEDA.

Group

Indications of relative date

49

The criteria available in individual cases have already 92. been summarized in 59, and must now be considered in more which have any of the characteristics of the detail. Hymns
popular Rigveda must generally be referred at earliest to the cretic period, and unless there is some reason to the contrary, they
will

no single

be classed with the hymns x 35-84 in Group IX. Otherwise in other words, such change as criteria are very certain we observe in the Rigveda proper is slow and continuous. On the
:

other hand the concurrence of two or more indications in the same

hymn may
evidence
is

Generally speaking, stronger generally be trusted. required to justify the detachment of a hymn that

belongs to a homogeneous Samhita group than is sufficient to guide us in the corresponding attachment and it will be at least
:

prudent

to err

on the side of caution rather than unnecessarily to

disturb the existing grouping.

The whole grouping being merely a provisional aid to the 93study of the metre, any full discussion of the position of individual hymns would be out of place here. It must be sufficient to state of a close congenerally that there are numerous small indications nexion between particular hymns which cannot be included in a Consequently the argument in the remainder catalogue of criteria. of this chapter must be regarded merely as an outline. As an indication of the value of the respective criteria, the instances in which their indications are not accepted at present are enclosed in In the shorter lists the references are extended square brackets [ ]. to the popular Rigveda, for convenience of reference later.

The most important indications of date are those furnished of by metre. The following features appear to be characteristic
94.

the bardic period:
(i)

All lyric metres

(

24)

;

Usnih, Atyasti, Kakubh-Satobrhati

characteristic of Pragathas, and mixed lyric metres being specially and Brhati- Sat obrhati Pragathas of the Kanva collection Group I,

(Group V).
as common Hymns entirely composed in Brhati seem to be quite in the normal period, and occasional verses in Brhati and even BrhatiSuch cases are Satobrhati Pragathas seem to be found still later. vi 59 1-6 (Brhati hymns): iii 16 (hymn in B.-Sat.) iii 9 i-8, 44, 45; i 105 8 (Brhati with refrain verse), *170 i (B.), *179 5 (B.) * *14 1? (B.), 33 3-3 (B.-S.), 62 6-7 vi 47 19 (B.); iii *53 18 (B.);
: ;
'

(B.-S.),

i

*101 5 (B.), *102 1, 3, 11 (B.). Other instances of lyric metre outside the bardic period are *23 19 (12.8.8), *164 42 (11.11. 8. 8), *191 13 (8.8.12.8.8);
4

50
iii

Decasyllabic Tristubh

10 (Usnih), 21 4 (11.11.11.8), 5 (Satob.), 23 3 (Satob.), *28 3 vii *55 2-4 iv 1 1-3 (Atyasti, etc.), *57 5 (11 8 8) (Usnih); x *17 13 (11.8.8.8), *18 11 (8/8.8.8.4), *66 16 (12.8.8);
.

.

;

(12.12.8.8), *170 4 (8.8.12.12). For viii *3 24, ix 67 30, *x 85 34, see the Metrical Commentary.' In several other cases in the late Rigveda the metre seems to be confused rather than lyric.
'

(ii)

in the stanza.

Dimeter verse containing a varying number of verses This is particularly common in Group II (Atri

hymns).

hymns
:

The most important class of hymns consists of the Anustubh Mandala v, in which the last stanza usually shews an extra verse hymns 9, 10, 19, 52, and 86 shew further variations. Other examples are viii 69 7-1 r, 12-16: and in a less regular way *viii 91, *x 145, *164, and other hymns in connexion with epic Anustubh. Gayatri hymns with an additional verse in the first stanza, or (b)
(a)

of

in the first of each triplet of stanzas are [iii 24]; viii 3 21-23, 9 19-21, 63 1-9, 68 1-12, 74 1-12, 92 ;
(c)
i

v 28

ix 101 1-3

4-6, 82 1-3 ; x 20. ;

43
ii

viii

Gayatri with an additional verse in the final stanza is found vii 94 10-12; vi 16 25-27, 45 31-33, [56]; 90 6-9; *x 60 1-6. For ix 66 16-18, 67 25-27; 31 5-9, 10-14, 79;
7-9,
p. 42.

8 6 see
(d)

Otherwise the mixture of Gayatri and Anustubh (or Pankti) stanzas seems not to be specially characteristic of the bardic period, vi 53; iv 30; iii 53 the instances being i 187; 12, 13; x *19, 176. viii 2, 55,56;

For the use of Mahapankti (8.8.8.8.8.8) Anustubh or Pankti see below, 95 iv.
(iii)

in connexion with

of

Group

Decasyllabic Tristubh ( 49) variations are characteristic III (i 165-190, vi) they are found occasionally in other
:

groups

of the

bardic period.

Hymns
also

in

the

Dvipada

Viraj,

Viratsthana, and Bhargava metres
to the bardic period.

appear to belong chiefly

the

The extent to which decasyllabic variations are admitted in Rigveda depends greatly on the view taken of the proposed restorations indara for indra, -aam for gen. pi. -dm. The occasions for the former restoration are found almost exclusively in hymns which also shew undoubted decasyllabic variations, and therefore the restoration seems at first sight On the other hand the unnecessary. restoration -aam is suggested in different parts of the verse, and in numerous hymns in which all other verses have the normal number of and this restoration therefore seems to correspond to the syllables intention of the bards generally. If we admit the restoration -aam, but not indara, we find that the following Tristubh or Jagati hymns have at least two decasyllabic variations, being at least one in every
(a)
:

Iambic hymns
ten verses:
i

51

60, 61, 63, 77, 104, [145"|, 148, 153, 167, 169, 173, 174, v 33 ,- 7 , 41, 51 11-13 20; vi 20, [iv 21]; x 23, 49, 50, 99, 148. Also the 21, 24, 30, 33, 35, 63, 66, 68 i-8; i 120 x 22, 93, 105, following lyric hymns: 1-9, 127, 129, 135; 132 and i 149, in which the stanza consists of three Tristubh verses.

178;

ii

4,

19,

;

;

In the long hymns
variations
(b)
is

i

slightly less

122, v 45, x 61 the proportion of decasyllabic than as above.

Hymns

in

56

i-ii, ix 109.

The hymns x 1, The metre seems (Mandala vii); several hymns stand in the text in association with hymns in normal Tristubh, and are therefore probably later. The only hymn in Viratsthana metre is ii 11, in which almost (c)
half the verses are Tristubh. This hymn will naturally be associated with those in Mandala ii already noticed as having decasyllabic variations. i 61 is nearly allied to this type.

vi 44 7-9, vii 34 1-2 1, [i 65-70], iv 10 is of a cognate type (5.5.5.11 .).] 6, [46] are partly in this metre, partly in Tristubh. most characteristic of Group IV but

Dvipada Virdj are

[The

hymn

Hymn

The Bhargava hymns x 77, 78 ( 52) most resemble ii 11 in (d) their general character, about half the verses in each hymn being
Tristubh or Jagati.
(iv)

Hymns

in metres

trimeter verses in the stanza are

which contain more or less than four most common in the Vasistha group.

Occasional extra verses are most

common

Hymns composed

in stanzas each of

in the Bharadvaja group. which contains two or three 31 10-12, 68;

trimeter verses are i 149; [iii 25]; viii 9 10-12; ix 110 4-6, 7-9, 10-12;

vii 1 1-20, 17, 22,

[*x 157].
;

Occasional extra verses are found in [*ii 43 iv 17, v 2, 41, 27] ; " vi 2, 10, 15 4-6, i 3 - I5 17, 31, 49, 63; 1-17, 43; [*vii 50]; x 115 6-9. In vi 15 1-3 a dimeter verse is appended to a Jagati

42

,

stanza.

In trimeter verse iambic variations ( 46) are characterthe bardic period, and in particular of the Vasistha group. The principal iambic variations are the rhythms y y v , ^ h ^ at the break, but we also include the rhythms n ^, w, which u are generally found in the same hymns. The occurrences of these rhythms are to a large extent found in the hymns already marked out as belonging to the bardic period by the features noticed under (i), (iii), and (iv) above.
(v)
istic of

least two

in classes already noted have each at iambic variations, being one in every eight verses i 36, 40, 77, 79 4-6, 88, 92 i 3 -t 5 120 1-9, 122, 127, 128, 131, 133 6-7, 135 4-6, ii 11, 19, 20; iv 10, 27;] 25 149,153,167,169,173; [iii v 24, 41, 53, 87; vi 2 11, 10, 15 1-3, 13-15, 31, 35, 46, 48, 63, 66, 68 1-8; vii 1 1-20, 14, 17, 22, 31 10-12, 68, 81; viii 4, 18 1-22, 19, 20, 21, 22 7-12, 23, 25, 26 1-15, 29, 33 i- I5 35 1-21, 36, 46, 60, 70 x 6, 23, ix 108, 110 7-9; 7-12, 87, 90, 97 11-15, 98 7-12, 99, 103; 50, 61, [62 5-10], 77, 93, 99, 105, 115 6-9, 132, 140, 144, 172. very
:

Thus the following hymns
,

;

,

A

42

52

Dimeter verse

large proportion of these hymns are in lyric metre, which suggests that the rhythm may be due to the influence of dimeter verse.
to the in Tristubh (rarely in Jagati) metre shewing this rhythm i 55, 56, 57, [59, 83], 85, 87, 91 i- 4 , 92 9-12, 158 1-5, 180, 181, 184, 190; 11 31, [ii 9, [117], 155 1-3, 156, iii 14, 15, 19, 26 4-6, 35, 51 4-6, 58, 61; 17, 26, 31, 33; v 8, 31, 46 1-2, 7-8, 49, iv 4, 6, 12 6, 29, 37 1-4 ;

Hymns

same extent are

vi 4, 11, 12, 13, 15 10-12, 16 46, 23, 26, 29, 37, 38, vii 1 21-25, 3, 4, 7, 8, 19, 20, 44 10-12, 13-15, 50, 60 1-3, 73; 40, 21, 23, 24, 26, 27, 30, 34 22-25, 36, 37, 38, 39, 40, 42, 43, 56 12-2;,

77];

57, 58, 60, 61, 62, 64, 65, 67, 70, 72, 73, 77, 84, 85, 86, 87, 88, 92, ix 70, 76, 79, [85 9-12, 86 4-6, 34-36], 88, 93, 94, [96 17-20, 95, 97 x [4, 32 i- 5l 39 14], 48, 73, 74, [76, 79], 115 r- 5 97 .9-21, 52-54]
; ;
.

are to be added several hymns in which this rhythm is found side by side with other features which indicate late date, with *i 93 4-8, which we include all hymns in the popular Rigveda iii 21, 22 1-3, 53 111 ( 111 i), 116 ( 95 ii, 110 iv), *133 1-5 ; 9 -it,
:

To the above

and 57
*viii

95 i, ( 100 1-6;

ii)

v 28 1-3, 36 1-5 ( 95 i) vi 64 ( x *59 1-7, 62 1-4 ( 95 i, 98 viii), 65 ( 95
;

;

95
ii),

ii)

:

*95,

*120, *123.
It appears therefore that iambic rhythm is very much commoner in but it is not rare in Mandala iii the bardic period than elsewhere is sometimes accompanied by indications of later date, and 110), (see Its use therefore occasionally found even in the popular Rigveda. extends considerably beyond the groups I-V, and the appearance of this feature in an individual hymn affords by itself only a slight
:

presumption of date.
(vi)

Anustubh and Gayatri verse are most commonly found
metres
:

in association with the lyric

but both are

fairly

frequent

The changes in in the groups belonging to the normal period. are on the whole too slight to be used here as a criterion. rhythm Catalectic and heptasyllabic verses (elsewhere than in the semicadence of epic Anustubh) indicate early date the only hymns which shew at least two occurrences, being one in every ten verses, are viii 68 14-19, x 26, and the following which are wholly or partly in lyric
:

metres:
(vii)

i

120

1-9,

175

1-5,

x

22,

105.

41) in

The iambic re-opening (opening of padas b and d, see Anustubh verse, if regular, indicates the bardic period.

This rhythm is regular in the Anustubh hymns of Mandala v: in addition the following hymns have each at least two such re-openings, 28 1-6], 84 1-3, 4-6, 176 1-5, 187 1-7; being one in every four [i 10, ii 5 ; viii 9 1-9, vl9; vi 2 1-10, 14, 44 1-3, [*51 16] [iii 13];
: *

;

16-18, 19-21,
(viii)

95

;

ix 98, 99, 100, 101 7-16

;

x

21, 26, [*58,

*87

22-25].

Trochaic Gayatri appears most commonly in the bardic
fragments mainly composed in Trochaic Gayatri are

period.

Hymns and

Contamination
i

53

2 7-9, 3 4-6, 27 i-6, 10-12, 30 i 3-i 5i 41 7-9, 43 7-9, 90 1-5; ii 6; vi 16 25-27, 61 10-12 j viii 2, 16, 71 68, 70, 82 7-9 ; 1-9, 79, 81 ; ix 62 4-6, 66 16-18, 19-21 ; x 20, 185.

v

Occasional Trochaic stanzas are found in the following hymns 2 1-3, [22 9-12], 30 7-12, 38; iv 55 v 19; [iii 11, 41; 8-10] ; vii viii 3 21-23, 5, 7, 11, 55, 56, 94. [89] ; In connexion with other metres trochaic dimeter cadence is veryrare but we find Trochaic Brhatt in iii 16, a trochaic Anustubh stanza in viii 95 7, and trochaic rhythm in uneven lyric metre in i 120 1-9.
:

i

:

In striking contrast to the variety of metrical form which 95. characterizes the bardic period are the few and slight indications which specially mark the cretic period, but are also not uncommon
in the

normal period and in the popular Rigveda.
Contamination in Tristubh and Jagati verse.

(i)

External contamination, including the combination of Tristubh or Jagati stanzas with Gayatrl or Anustubh, seems seldom to be much earlier than the cretic period. It is however difficult to define, as in some cases we may be dealing with composite hymns. The appearance of one or two Tristubh stanzas towards the end of Jagati hymns is not included. Probable examples are i 35, 125 ; ii 3 1-7 ; iii 8 1-9. 22, iv 7, 40 1-4, 57 1-3 ; 53 9-ri, 15-16 v 28 1-3, [36 1-5], 60 ; vi 7,
;

vii 41, 44; x 44, 69, 124 6-8, 131; besides [47 22-25], 58; examples in the popular Rigveda.

numerous

Internal contamination of Jagati stanzas by Tristubh verses is found in at least two instances in each of the following hymns i 110, x [23], 62 i- 4 n, 63, 66, 75, [115 1-5], 122. Similarly of 114;
:

,

Tristubh stanzas by Jagati or hybrid verses: ix 85 9-12. Both variations are 72
;

i

35, [180]

;

common

in

vi 47 15-18, the popular

Rigveda.
(ii)

Cretic variations,
(

i.e.

occurrences of the

rhythm - ^ -

after
:

an early caesura

46), are not rare in

any part of the Rigveda

but

they are proportionately

much more common

in the cretic period.

of any length in which this rhythm is frequent probably cretic period, unless indeed they belong to the popular Rigveda but in short hymns the same relative frequency may be accidental.

Hymns

belong to the

:

Hymns which shew
five verses, are

at least two cretic variations, being one in every
:

found as follows i *24 1-2, 30 16, 51, *84 16-18, [91 22-23], *93 12, 94, 96, 100 1-15, 103, 106, 109, 112 24-25, 113, 118, 30 i-jo, 39, ii 3 8-n, 125, 140, 146 i- 4 159, 160, *162, 182; iv 11, *42 iii 5, 6, 9 *52 5-8, 53 2-6, 54 9, 17, 20, 36, 39, 49, v 4, 34 9, 46 3-6, 60 7-8, 76, 78 4, 80; 34, 50 1-6, io-ii, *58 1-10; vii 5, 9, vi 6, *15 19, 52 4-6, [61 1-3], 70, *74; [22, 27 8, 32], 47 20, 10, 13, 18 1-21, 54, 59 7-8, [69], 78, 79, 80, 101, *104 1-7, *8-i 7 ;
,

;

;

54
viii

Cretic variations
40

ix 80, 90, 97 1-3, 34-36, 37~39> 40-42, 43-45, 49~5* ; 12, *58; 8 7-9, 30, 36 1-12, 13-14, 37 1-10, 38, 44 4-9, 47, *52 1-5, *6, *53, 54, 65, 70 1-7, *87 1-21, 106, *107, 111, 112, *164, *169, 178,

x

5,

*179 2-3. The following hymns have both iambic and extent indicated above: i *93 4-8, 111, 116;
64.

cretic variations to the
iii

22

1-3,

57;

vi

In lyric or decasyllabic hymns cretic variations occur only as i 48, follows, several of the hymns having also iambic variations 134; iii 21, 23; viii 10; ix 110 4-6; x 78, 105, 140.
:

Although a few of the hymns, as indicated by the brackets, probably belong to the bardic period, the following table shews in a striking way the extent to which the iambic variations are and the rarity of associated with the lyric and decasyllabic metres whilst on the other hand in the popular cretic rhythm in them Rigveda iambic variations almost disappear.
1 ,
:

Metres

Subject-matter of the
The combination
J

hymns

55

of the two features is therefore found in i 105, 8 1-9 ; iv 37 5-8 ix 101 4-6]. Thus [viii 8 ; a tendency in the direction of the later rhythm can perhaps be observed, not only in hymns of the normal and cretic periods, but even in some of still earlier date. See further in Ch. vn.

142 1-8

iii

j

(iv)

Hymns

in

Mahapahkti belong

to the

normal or

cretic

period.

For this the presumption is only slight. The hymns that shew this metre are *i 191 10-12, viii 39, 40 i-n, 41, 47 1-12, *i3~i8, x *59 8-10, 133 1-6, 134 1-6, *166 5. In x *59 8-10, 133 1-6 the number of verses
in the stanza varies.

The extreme regularity of the metre distinguishes these hymns from the Anustubh hymns of Mandala v, in which also as many as six verses are found, though rarely, in one stanza and it deserves
:

notice that nearly half of the

hymns

are in the popular Rigveda.

larger collections of the Rigveda shew a general their subject-matter, consisting usually of three parts about equal in bulk, of which one is composed of hymns to Agni, the second of hymns to Indra, and the third of hymns to
96.

The

uniformity in

other deities.

On

different picture.

the other hand the smaller collections present us with a In most of these collections (such as those of
is

which Group I
dominate
:

in some, as in

composed) the hymns to Indra greatly prei 65-73, 140-164, and the Anustubh
to

hymns

of

Mandala v the hymns

Agni are most numerous.
were often

It is therefore probable that the earliest collections

devoted to the honour of one or two favourite deities, and that the uniformity of the large collections has been reached by a process of
This process is most easily to be traced in the which are so slightly represented in most of the Agni hymns, smaller collections thus the Apriya hymns, as well as the hymns to Agni Vaisvdnara and Agni Jatavedas, all of which are included in the Agni sections, appear to be unknown to the earliest groups, and in consequence no hymns of any of these types are found in supplementing.
:

any

lyric metre.

instances in which the subject-matter seems to afford a clue to the position of individual hymns are given in the next two

The

sections

:

but these indications are on the whole

less

trustworthy

than those of metre.
97.

The

deities

which appear

chiefly in the collections of the

56
bardic period

Subject-matter of the
are

hymns

Mitra-Varuna (and Mitra-Varuna-Aryaman), Vayu (and Indra-Vayu), Visnu and SarasvatI: and in the same of thanks to patrons for period we find almost all of the songs
These features are unevenly their munificence (ddnastutayah). distributed amongst the different collections of the period.

7

_ 9)

(i)

Hymns
,

to Mitra-

Varuna are found
ii

63],

90 t- 5 136, 137, 151-153; vi 62 64-68, [69], 70-72
;

i 2 7-9, [23 4-6], 41 1-6, v [62, 41 4-6; 62 16-18]; [iii vii 60-66 ; viii 19 34-35, 8-9, 67;

25 1-9, [47
(ii)

1-12],

101 1-6;

Hymns
;

to

x 126 1-7, 132, 185. Vayu and Indra-Vayu occur i
;

2 1-3, 4-6, [23 13],
;

134, 135
viii

ii

41 1-3

[iv

46-48]
[i

;

v 51 1-4, 5-7

vii

90-92
vi

;

26 20-25, 46 25-28, 101 9-10.
(iii)

Hymns Hymns

to Visnu are

22 16-18, 19-21], 154-156;
i

69;
61;

vii 99, 100.
(iv) vii 95, 96.

to SarasvatI are

3 10-12;

ii

41 16-18;

vi

The ddnastutayah are [iii 53 7; iv 15 7-10, 32 22-24] j v 27 (v) vi 27 7-8, 45 1-3, 4-6; 30 12-14, 33 8-10, 36 6, 52 15-17, 61 5-10; viii 1 3-3^ 2 40-42, 3 21-23, vii 18 22-25 ; 3i-33) 47 22-25, 63 9-1 1 ; 4 19-21, 5 37-39, 6 46-48, 19 36-37, 21 17-18, 24 28-30, 25 22-24, 34 16-18,
ix 58.

46 2124, 29-33, 55, 56, 65 10-12, 68 14-19, 69 17-18, 70 13-15, 74 13-15; mock danastuti seems to be found in *i 120 10-12. (See R. Pischel, Vedische Studien i 1 seq.)

A

98.

The
:

deities Brhaspati, Vastospati, Ksetrapati

and Sadasas-

pati, and Dadhikra appear to be most prominent in the normal period whilst hymns to Agni VaisVanara and Jatavedas, DyavaPrthivl, Surya, Parjanya, and the Apah, as well as the Apriya hymns and most of those which have a ritual character, are found Here too chiefly in the cretic period and the popular Rigveda.

belong a few hymns praising liberality (daksind), and easily distinguished from the earlier ddnastutayah.
(i)

190] 98] ;

;

14, 15].
(ii)
(iii)

18 15, [40 1-6, vi 73 ; [vii 97, x *42 n, 67, 68, *182. To Vastospati vii 54, 55 1 ; [viii 17 To Ksetrapati iv 57 1-3. To Sadasaspati i 18 6-9.

Hymns

ii

23-26

to Brhaspati or Brahmanaspati iii 62 4-6 ; iv 49, 50 1-6, 10-11
:

i

;

;

:

:

:

Hymns to Dadhikra iv 38-40, vii 44. Hymns to Agni Vaisvanara i 59, 98;
: :

iii
:

2,

3,
;

26 1-3;
iii

iv 5

;

vi

7-9

;

vii 5, 6, 13.

To Agni Jatavedas
:

i

99

17

;

v4;
(iv)
;

x 188.

Dydva- Prthivi i 22 13-1 5, 105 (refrain), 159, 160, iv 56 vi 70 vii 53. *vii 66 16; iv 13 1-4; Hymns to Surya: i 50 1-9, 115; (v) x 37 1-10, *189. [viii 101 n-12]; Hymns to Parjanya: *v 83; vi 52 16; vii 101, 102. (vi)

Hymns
ii

to
;

[185]

32

1

-3

;

;

Authors and refrains
(vii)

57
;

Hymns
*17
10,

to the

Apah

:

i

*23

16-18, *ig-2i

vii 47,

*49

;

x

9 1-5,
(viii)

30.
;

ix 5 1-7,

iii 4; v 5; vii 2; Apriya hymns: i 13, 142, 188; ii 3 x 70, 110. Other ritual hymns appear to be i 28, 105 4-6, 9-12, *162, *163, [187]; iii 8 iv 23 8-10, 1-9, *io-n, *29 viii *66 15, 40 5, *58 1-10; x 62 i- 4 76, *94, *101, *175. [72]

*8-n

;

;

;

,

(ix)

62 5-10,
99.

11,

Praise of liberality *107, *117 1-7.

:

i

125, 126 1-5

;

[vi

27

8]

;

x *42

9,

It has already

been noticed
is

(

59) that the mention of

the family
only.
article

name

of the poet

The subject has been

characteristic of the bardic period discussed by H. Oldenberg, in an
:

Ueber die Liedverfasser des Rigveda (ZDMG. xlii 199-247) and he has shewn that as a rule the singular and plural forms, e.g. Vasisthah and Vasisthasah, are used side by side, and therefore both alike usually indicate the name of a family, not of an individual
poet.

The

always easy
family,

interpretation of the occurrences in the Rigveda is not several hymns contain references to more than one
:

later hymns of the Rigveda frequently refer to the ancient poets (or supposed poets) by name, whilst in many cases we cannot be sure whether we have a contemporary record or a

and the

mythological reference.

The use of family names is dealt with in detail below ( 104 110) in connexion with the separate groups but the treatment makes no claim to be complete. In some collections, as for instances those of the
:

name

of family authorship is very complete, and the in others, as for instance in occurs in almost every hymn Mandala iv., the evidence is so slight that the use of the family name to describe the collection can only be regarded as conventional. To the traditional lists of authors it is impossible to attribute any

Kanvas, the evidence

:

independent historical value.
100. Refrain verses concluding each of a series of hymns occur in Dimeter and in Tristubh metre, but the latter occasionThey are chiefly found in the bardic ally close Jagati stanzas.
period, but their use occasionally extends to later hymns. refrain distich is found in the cretic collection i 94-115.

A

The maksu

refrain
i

verses
viii

58-64,

regularly used in the Rigveda 80 10, ix 93 [tad no mitrdh
:

are
i

prdtdr 94-115,

i 165-190: ii, ix vidy&mesdm [brhdd vadema vi 86 47-48]: mddema satdhimah td tarema vi yuy&m In Dimeter metre: vii; [ix 90, 97 13, 4-6; x 65, 66, 122]. pdtd utaidhi prtsu no vrdhe v; vi vo made .vivaksase x 20-26 [ndbhantam

ix 97 57-58]:

:

:

.

.

;

any ake same viii 39-42, cf. x 133]. The following are used in a few hymns only

:

suviraso viddtham d

58
vadema
50 sidma
iv
i

Linguistic indications
117 v 55
25,
ii

48 14: vaydm sidma pdtayo raylndm *48 13, *x 121 10: suviriasya pdtayah iv 51 10, vi 47 12, ix 89 7, 95 5, x 131 6 rdyd madema brhata rdksd (trdsvd) utd nas tantlo dprayucchan tuotdh vi 19 13, cf. 49 13 x 4, 7 (cf. i 143 8, iii 5 6, 20 2): devoir dydvdprthivl pravatam nah Of i 31 8, x 67 12: isam urjam suksitim visvam tibhdh x 20 10, 99 12. The refrain verse these only two belong to the bardic Rigveda. indrdyendo pari sravd in *ix 112-114 is an artificial addition, borrowed from ix 106 4b. Concluding stanzas common to two or more hymns are found in many parts of the Rigveda, and usually indicate a close resemblance between the hymns thus connected.
12
15,

*viii
12,

6,

10,

viii

40

:

:

101.

their strict order
in

Many hymns even in the Rigveda proper such hymns are often foreign to the
:

are out of
collections

which they stand, but are not necessarily of late date. For the discussion of the order of the hymns reference has already been made to H. Oldenberg's Prolegomena, ch. II.
In numerous cases there is a slight displacement amongst hymns which are obviously cognate: but the following cases seem to be of more importance: i 2, 3, 10, 24 3-5, 27 13, 104, 189; ii 9, 10, iv 15 710, 48 5 ; v 61 5-10, 84, 87 41 10-12; iii 53 2-16; vii 17, 44-55, 59 9-1 vi 47 1-25; viii 1; in addition to numerous hymns belonging to the popular Rigveda.
;

1

;

102.
or of

The recurrence of more than a single stanza of a hymn, an independent stanza, in the Atharvaveda, most commonly

indicates the cretic period at the earliest.

Most hymns of this kind have been already assigned to the popular Rigveda. Those that remain are i 22 1618, 19-21, 32, 50 1-9, [154]; iv 2 ; v 46 7-8 vii vi 47 6-10, u-14, 54, 68 9-1 [32], 41 [viii
;
1
;

;

24]; 103.

x

2,

9 1-5, 40, 110.

recorded in
:

In the hymns of the Rigveda proper the linguistic forms 84, 85 greatly preponderate over those recorded in

but in those of the cretic period the later linguistic 86, 87 forms constitute a tangible proportion of the whole. Conversely, if in any hymn the number of late linguistic forms approaches the

number

of early forms, there is a to the cretic period. belongs

presumption that the

hymn

Such hymns are i 13 9-12, 22 13-15, 23 13-15, 28 1-6, 29, [41 7-9, 43 4-6], 83, 84 13-15, [135 7-9, 136 1-6], 145, 182, 188; ii 12, 13, 28, iii 21, 33 iv 12 6, 42 8-10 ; 39, 40, 41 10-12 1-1*, 45, 53 2-6, 55 v 5, 47 1-6, 62, 81 vi 9, [16 16-18], 47 19-21, 52 4-6, 54, 59 1-6, vii 46, 59 7-8, [60 14] [76], 101, 102 [viii 17 11-13, 34 refrain, ix 85 1-4, 86 28-30, 97 7-9; x 2, 91, 106. 41, 55, 69 12-14, 77];
;

;

;

;

;

The short
110 8-1 1, 126 8, 127 94-115, x 35-84.
;

collections

59
i

besides numerous

hymns

in the groups

31-35

104. We proceed to revise the grouping of the hymns, with such assistance as is given by the indications noted in 93-103.
I.

The

short collections of

Mandala

viii,

arranged without

preference for the Agni hymns, and composed in metres almost entirely unknown in the time of the popular Rigveda, not only

bear the most evident indications of early date, but have been least disturbed by later additions. On account of their small bulk
they are included in one group.
All the collections in bk. viii

which have no special affinities to other groups are here included, and also the collection i 127-139, which has Atyasti as its distinctive metre.

Group
(i)

I.

The
12-18.

collections included are

:

viii

Hymns
The

chiefly to

metre, either Dimeter or Trimeter. metres.
(ii)

Indra, and mostly in Usnih Attach ix 102-106, in similar

viii

19-22.

author's

name Sobharl
viii

Attach
this

Sobharl);

hymn

102 1-18, 108 (Kak.-Sat.); x 115 (8b 9b Upastuta, connecting with viii 103 see also 94 iv, v).
ix
:

distinctive metre is Kakubh-Satobrhati ; the occurs 19 2c 32a, 20 2d 8a 19b, 22 2c 15c. 103 (st. 8-13 Kak.-Sat., 8c Upastuta, 14c

(iii)

viii

2326.

The

distinctive

named
(iv)

as author in 23 16a 23b, 24 22a
viii

metre is Usnih. Vyasva 23b 28c 29b, 26 9b 11a.

is

27-31, in lyric metres and Gayatri, addressed to the

Adityah and Visve Devah.
(v)
viii

47 1-12 see

43-46, in similar metres to Agni, Indra, and Vayu. 95 iv, 109 iv.

For

viii 95-101, in Dimeter Usnih and various lyric metres, (vi) addressed chiefly to Indra. i 127-139, in Attach ix 111. Atyasti and similar metres. (vii)

II. Stanzas consisting of a varying number of dimeter 105. much the most verses are specially characteristic of Mandala v common form is Anustubh with an extra verse in the last stanza,
:

but we also find Gayatri with an extra verse in the first stanza There are also several lyric (28 4-6, 82 1-3) and other variations.
characteristics of the

The trimeter hymns mostly shew the normal or cretic period, and are therefore probably later additions. A similar collection is found in x 20-26.

hymns

in this collection.

Several of the short collections in Mandala viii are more akin to this group than the last, though a strict line cannot be drawn.

60

The early Anustubh groups
The second Mandala contains a nucleus

have features in

common both with

this
:

hymns which and with the next. group
of early

The hymns included are These include all the lyric, Early hymns of Mandala v. (i) Anustubh, and Pankti, and at least those Gayatrl hymns in which the number of verses in the stanza varies aLso those Tristubh hymns which have decasyllabic variations or extra verses, and all hymns The following list will be roughly correct containing ddnastutayah.
II.
:
:

Group

16-25, 27 ( 97 v), 28 4 -6 ( 94 ii), 30 i-i 4 94 v, 97 v), 38-40 3 41 ( 94 94 iii, 97 v), 35, 36 ( 97 v), 33 ( ( 94 iv), 45 ( 94 iii), 50, 51 1-13 ( 94 iii, 97 ii), iii-v), 42 and 43 ( 94 viii, 97 i), 71 and 72 52, 53, 56, 61 ( 97 v), 64-67, 68 and 70 ( 97 i), 73-75, 78 1-3, 79, 82 ( 94 ii, viii), 84, 86, 87. ( The family names found are Atri in 2 6, 7 8, 10, 22 1, 4, 39 5, 51 8,
2
(

94

iv),

6,

7,

9,

10,

,

72 1, 73 6, 7, 74 1, and Sydvdsoa 52 r, 61 5, 9. The appearance of cretic variations makes us hesitate to ascribe 4, 78 4, and 81 to the original collection, though the proper names reappear in them, as also in 15 and *40 5-9. The refrain stanza sdm asvinoh is only appropriate to 76 and 77, and would therefore seem to have been added later to 42 and 43.
67
5,

Attach ix 98-101, on account of the similarit}7 of metre

(

94

vii).

In the collection x 20-26 the prevailing metre is Anustubh, (ii) with several variations. In 21, 24, 25 half verses occur which form a refrain vi vo made .vivaksase, containing the family name Vimada, which also occurs 20 10a, 23 6a 7b [*24 4c]. In 20 we have Trochaic in 26 a large proportion of the Gayatrl, in 22 uneven lyric metre verses are catalectic. 23 is in contaminated Jagati, but this hymn is not to be separated from the rest, for it also contains the family name (w. 6a 7b), and the frequency of catalectic cadence finds a parallel
. . ;

in 26. Attach x

94 iii), which has the refrain isam urjam... in 20 and, on its analogy, the other lyric, early Anustubh, and decasyllabic hymns in bk. x, and those with which they are associated, viz. 93, 105, 132 (all in uneven lyric metres), 126 1-7 (lyric), 143 (Anustubh; Atri is named la 2d 3a), 144 and 172 ( 94 v).

common

99 with x

(

;

(iii)

The

the

name

addition

collection viii 35-38 is connected with Mandala v by Atri in 35 19a, 36 6a 7b, 37 7b, 38 8b, usually with the of the name Sydvdsva, and by the use of a lyric metre

somewhat resembling that of v 87 and x 132. This metre however appears to glide into Mahapankti, and so far it is associated with a
later period.

Other groups in the eighth Mandala which are generally (iv) similar are those which extend from 60 to 79, and 92 to 94. For resemblance of metre see 94 ii a, b. The proper names that occur are Atri 73 3a 7a 8b, Priyamedha 69 8b 18b, Purumllha 71 14c,
Virupa 75 6b.
Attach x 73, 74 (Priyamedha 73 lib
:

cf.

94

v).

Nodhas in 61 14d. 61 16b. ix 98. 61. v). . : ( 98 iii). (iii) The name Mana or Manya occurs 165 14b decasyllabic variations. The name Grtsamada occurs in 4 9a. 19. and x 147. 99 see also 97 iv). period. and the ascription to Agni Vaisvanara seems to indicate a late date 58-64. 20 ( 94 iii a). in which stanza 94 v). 3 (110 ii). is regular: in 165-168. Attach viii 88-90 (Gotama 88 4d) and ix 88. which has iambic rhythm ( 94 v). 183-184 respectively it concludes stanzas which are common . 16-18 (Anust. ( 83 and 84 13-15 103). cf. 175-176. and Pajriya 120 5b. may be denned as follows : contain a very large proportion of Hymns (i) The name Gotama occurs in 60 5b. 62 13c. tf3 (Mana). . III. 41 18c. 94 iii a. with many i 165-190. 85 lib. the name Pajra occurring i 122 7b 8c. 78 la 2a. 92 7b. 184 4b 5b. 79 10b. 20 as concluding a single final stanza which may have been attached to these hymns later.2 o ( 94 iii c).. : and Sunahotra in [18 6c] and 41 17c. 41 1-9 ( 97 i. 177 5b. 63 9a. and has verses 2b 2c in common with ii 1 1 (4d 5a). and decasyllabic and iambic variations are frequent the family name Bharadvdja is common. and many decasyllabic verses. 6 ( 94 viii). 78. of which the metre appears to be parallel to that of ii 11 ( 94 iii d). 148. 64 lb. Detach i 59. and the family name Bharadvdja in 7b but it clearly does not belong to this collection. 182 8b. 19. and the verses lie lid in common with i 121 (5c 5d). only in 184 5d.The decasyllabic groups : 61 The nucleus of early hymns contained in Mandala ii may be (v) 4 ( 94 iii a) roughly denned to consist of the following hymns 5 ( 94 vii). 171 5a. ( Attach viii 8087 (80 has the Gotama Nodhas refrain).. [39 8b]. 15b. Group i III. 62 13a. ix 93. The refrain brhdd vadema is wanting in ii 4 it occurs in 11. 1 1 . 74-93 (Gotama). 189 8b: Agastya The refrain vidydmesdm . x 61.. 22 (lyric)'. decasyllabic variations. The collection is distinguished by a small group of (ii) i 74-93. With the collection sixth in Mandala we existing first meet a text. any collection of the Soma hymns Detach i (91). refrain verses are occasional. 8 = i 91 3 (see also The prevailing metre is Tristubh. 94 94 v 93 has the Gotama Nodhas refrain). 121. 169 8a. This collection has a greater variety of metres than same length. : A similar (% and 122 nucleus is found in hymns i 120 1-9 (uneven lyric). This hymn has iambic rhythm.... x 77. i collections Much the same is the character of the 58-64 (Gotama Nodhas). and Byh. of which the latter is decasyllabic. ii). The principal metre is Tristubh. 106. most of them indicating the bardic The name Gotama occurs 77 5a. 19 8a. Attach i 2. large homogeneous : the Samhita. 165-191 The collections included 60. All the hymns except 59 contain the refrain pratar maksu . 88 4c 5b.

6 ( 95 ii).62 to the The Bharadvdja hymns hymns thus The collection is remarkable in conconnected. they are employed with a regularity which contrasts strikingly with the variety of form exhibited in Groups I-III. 9 6a. rhythm is The seventh Mandala is also on the whole homoThe predominant metre is Tristubh. and of 174 9 as A vi 20 12. Besides Tristubh hymns. 31 4e. and brings this collection nearer to the normal period. and generally from 68 9 to the end of the Mandala and the Pusan group 5358 is marked by late linguistic forms. connection with Mandala vi is a single Agni hymn. 13.. 7 7a. 103). . the latter seldom being quite regular. 10. 68 9 -73 ( 95 95 i). 16. 17 14c. Attach i 104. 59-62. both lyric and trimeter. 103). 17. in hymns 47 and 52.. 35 4d. The refrain verse td tarema . The 94 i. The name Vasistha appears as follows vii 1 8b. and partly in 6& 8d. 58 ii. These characteristics are wanting in the Agni hymns 1. 52 4-12 ( 95 ii). It would therefore seem that considerable additions have been made to the collection at some later period. 103). 23 10b. ii. 188 ( 98 viii. 14 6e. have occasional extra verses. geneous. and in the Group IV.. 16 5c 33a. 15 3e. Some hymns of a similar character are found in the group i 140-160.. 50 15b.. Indra (30-41). 102. 12 3b. 7 ( 95 i) and 8 ( 98 iii). 9 ( 98 iii. which have often decasyllabic Mandala contains several hymns in Gayatri and in Brhati-Satobrhati Pragathas. The name Bharadvdja occurs 10 6c. the resemblance between these verses seems to point to a common The collection is marked by a large group of hymns to authorship. ( 98 iv. but variations decasyllabic are rare : there are found also several Brhati-Satobrhati Pragathas of a very regular character. in which iambic far more marked than elsewhere in the Rigveda. 15 15e.. : see especially 95 107. 5-9. taining only indicated by the recurrence of 174 2b as vi 20 10c. 47 25b. IV. this . (iv) vi. The long list of is iambic hymns in Mandala vii is 94 v: the collection also marked by : several hymns in (94 iv). 5. ends hymn 19 and recurs 49 13d. 65 6b. 53-57 variations. 24. the verse rayti madema . 59 1-6 ( position of several other hymns is doubtful i. 63 10c. The name Vasistha is common both in the singular and plural. i. given in Viraj metre (i) ninth Mandala. 48 7c 13a. 103). 103). Several hymns. 103). ii. There are a few ddnastutayah. Although these metres are quite strange to the later periods of the Rigveda. 47 1-21 ( 95 i. v). Detach 1. 12. the verse mddema satdhimdh . which are marked by decasyllabic variations 94 iii a. occurs 2 lie. ( 102. 17 ( 98 vi). 102). 25 9d. x 48-50. of which each consists of tive Tristubh stanzas. 51 12b. ends hymns 4. ( Detach i 182 ( 95 ii.

Hymn 69 occurs in the Asvin group. The hymns . Four of them are Agni hymns. 89. Many hymns shew cretic variations ( 95 ii). 60-61. 39 7a. but it does not include the refrain verse. in that iambic rhythm is common 94 v). 76 and 79 ( 94 v). 26 5a. 57. is usually found at the end of a stanza common to two consecutive hymns. though short. 98 vi. 18 1-21. vii). 34 and 56. The name Mdmateya. a metre strange to the collection as a whole . and in association with the names of other bards in *x 181 Id. and have the usual refrain. 9-13 ( 95 ii. like the four Agni hymns. makes it likely they hymns are probably the work may have been composed by members are also later additions. ( 147 3a. 15. 122 8a. (9 6a. 97 1-3. and x 65. and out of order ( addressed to Parjanya ( 98 vi). 150 5c. The characteristic iambic rhythm is wanting in the whole series of hymns which extend (out of their order) from 44 to 55 ( 101) and in all hymns outside the Mandala whilst hymns 82. and this also wants the refrain. 82 and 83 (Jagati). 86-88. wants the refrain. from which there The historical hymn vii 18 1-2 1 is not sufficient reason to separate it. Attach i 55-57 ( 94 v). vii). 90 7b. and 6 ( 98 iii). 59 7-8 95 ii). 75-79. 150 (lyric metre. vii 18. 59 3b. group. 45-48. 54. 95 6a. 99-100. 26. 5 of a later period.93. 41 ( 102). 53. The refrain yuydni pdtd . 23 lb 6b. ( Outside Mandala vii there appears in the collection i 140-160 (ii) a nucleus somewhat resembling it.Elder and younger Vasisthas : 63 22 3b. hymns makes it difficult to think that any of these hymns are really homogeneous with the rest of the If this is a right view. 42. 24-25. 65 15a. and the absence of any mention of Vasistha in them. 67 and 69. 96 Id 3d and outside the Mandala in x *15 8b. Another (54) is addressed ( 98 i. will most naturally be explained as late additions. Vasistha 5c). 80 la. 84-85. *33. Detach 2 ( 98 viii). Another (101) is to Vastospati. 44-48 ( 98 ii. 27. 43. 103). and 158 6a. : : : them The general regularity of the Gayatri hymns to Agni and some others. 101 and ( 102 ( 95 ii. and in *x 95. But the change of rhythm. and in the other through three successive. 78-80 ( 95 ii). 23. 66 have the same concluding stanzas in each The refrain is found independently in hymns vii 9. 4-6 x 122. 83 are in Jagati. 86 5d. 98 iii). Also the hymns vii 28-30. 62-63. 9-11 ( 101).. 92. . 22-25) which has the usual iambic rhythm the Usas hymns 78-80 form a group of three. 90-91. 64-65. 1 1-20 and 21-25. 70 6c. 58. 101). but it is accompanied by a ddnastuti (st. is also a connecting link. 41 and 80. nor the refrain verse is a sure indication that a particular hymn belongs to the original collection. 88 lb 4a. 152 6a. viz. 97-98. and of these two are addressed to Agni Vaisvanara 98 iii) two of the four want the refrain.-Sat. 11-14. 72-73. neither the mention of Vasistha collection. in the one case running through 21 stanzas. though some of of the Vasistha family. and Viraj). 70-71.. 42 6a. 37 4d. It also occurs in i 112. 77 6b. 76 6a 7b. 66 14a. ix 90. 3-4.95. 51-55 1 Q 98 iv. 18 passim). 7-8. 19. 22. 20-21. 107 (Brh. but not as an author's name. 51. 35-37. 73 3d. they have a concluding All these stanza in common. ix 70 ( 94 v) with 71. 109 ( 94 iii b). These. 39-40. x 1 and 6 94 iii 6 and v). 110 ( 94 iv). and hymn 149 is in Viraj metre.

102). V. 2 40b. 30 20c. 55 4a: also x 115 5a. . viii 1 8c. Tristubh. 32-34. 44-50. Detach i 140 ( 95 95 ii). and Jagatl stanza but the third Mandala has certain features almost peculiar to itself which make a separate : treatment desirable. See also ( 97 iii. 39 7d 9b. 4 2c 3d 20a. 9 3c 9d 14d. 67 25-27 ( 94 ii). 30 4b. and the hymns 43 and 44 to the Asvins. Divodasa 26 3d. 2 16c 22a 38c 40b. 50 10a. 97 i). 33 4b. viii 3 9d. 25a. 33 3a. and may be that of the author of the hymns if we are to interpret iv 4 lib literally. viii). These collections are i 36-43. 34 lb 4a 49 5d 10a. 95 ii. ( not very marked. and connect them with others contained in Groups I III. . x 140 (cf. 8 The names Upastuta. 26 lc. and by the family name. 159 and 160 is though their rhythm 142 ( 95 iii. 6c with i 45 7c see also 94 v). 3 16a. 54 8d. for which see 104 ii. Common concluding stanzas are found in the Agni hymns 13. 146 Also hymns 143 and 157. Detach i 50 1-9 ( 98 v. 7 18b 19c 32b. Usana 26 Id. The principal collections of the normal period are 109. The Kanva viii. : All of these shew a very close adherence to the normal forms of the Gayatri. 49 9c. collections. viii 1-5. 10 2b 32 lb. Medhyatithi or Medhatithi i 36 10c 11a 17c. Trochaic Gayatri. 52 8d. 47 2c 4c 5a 10c. 68-97. 51 2a. Further we have the names Praskanva i 44 6c. 185 ( 94 viii. ix 43 (Medhyatithi 3c). there : : . 7c 8d lie 15b 19d. Attach i 14 (Kanva 2a 5b). 37 lc 14b. 8 20a. 9 lb 6c 15d. 17 and 19 to 24. 5 4c 23a 6 3a 8c lib 21b 31a 34a 43c. the author of that hymn is a son of Gotama other proper names that occur are Kutsa 16 10b 11a 12a. 51 lc. Kakfivat 26 lb. 8 3d 4c 8c 20a. Mandalas iii and iv to these are to be added the larger parts of Mandalas ii and v. The name of Vamadeva appears iv 16 18a. 62 4-6 ( 94 viii). the Indra hymns 16. 147-156 and 158. 66 16-18 ( 94 ii. 6-11. 44 8c. Atri and Priyamedha also occur in Kanva hymns. Virupa. would seem more probably to ii). 108. 144. 58 ( 97 v). 11 7a. 19-21 ( 94 viii). and of the collections i 1-30 and ix 1-67.64 The Kanva groups i that agree in character are 141. : VI. viii 1 30d. 49 4d. dalas i and are easily recognized although scattered in Manby their distinctive metres (Gayatri. 48 4c. 45 3d. 49-57. 98 viii). 98 iv). 46 9a. belong to the later additions. 14. and ix 43 3c (see below). but : are several Gayatri hymns. Sobharl. 145 ( 103). 45 5c. Vimada. The name Kanva appears i 36 8c 10c lib 17b 19c. Group VI. (i) The hymns The hymns included fall under five headings in Mandala iv are chiefly in Tristubh metre. 27 1-12 and 30 1-15 ( 94 viii). Group V. and Brhati-Satobrhatl Pragathas). and Vatsa viii 6 lc.

Both iambic and cretic rhythm are rare: lyric and other early metres are occasional. 27. is found.. and the favour this conclusion.. having Mahapankti metre ( 95 iv) (iv) with the refrain ndbhantam anyake same.. which is also appended to the earlier hymns 11. 81. 20 its refrain verse is further found in the hymn-pairs 1 and 2.. 98 viii. comparatively large number in JagatT. further In ix 89 the refrain suviriasya .) 47 1-12 (also Mahapankti).. 29. But.. and x 131 (refrain suviriasya . The group viii 39-42. 28 1-3 ( 95 i). as in iv 51 lOd. and therefore a great number of them probably belong this period.. .. sdm asvinoh of : 31. 97 1-6 106 anyakesaam ( i). 90 (111 ii). 19. 34-51 ( 95 ii). Further the concluding verse viii 40 12d is identical with : iv 50 6d. 55. 37. 44 ii. 11-15. 78 4 5 ( . 19. 10-12 ( 101. 58 have in common the refrain stanza haye ndrah. On the 21. 34. 58 ( 97 v). 43 ( 108 v). The indications of date in the Soma Pavamana hymns are (v) to very slight. as in : iv 51). 103).. 98 viii). 61-67 ( 110 ii)..). 103). 32. 62 103). 46-49.. 16-18 ( 105 v). The greater part of the hymns of Mandala ii seem to form a (ii) similar collection. 85. and sometimes take forms peculiar to this Mandala. Jagati being again common and here we may naturally place those Gayatri hymns in that Mandala which we have no special reason to assign to an earlier date. ii. unlike Mandala iv. many hymns which are not homoii . 17. yet the regularity of the metre stands in great to them. 12 and 13 ( 103). Detach 3 ( 95 i. Attach x 126 8 = iv 12 6. The Indra hymns 15-18 have the concluding stanza nundm sti te . 41 1-9 105 v). 77. perhaps belongs to this period the punning refrain reminds us of the Vimada hymns although contrast (x 20-26). 93 70. 8. as in 1 1-3.. * .. x 91. Amongst these hymns 57. 40 4 80. ( 98-111 A. 26. whole there seems little reason to suppose that the Mandala contains either a nucleus of older hymns or a supplement of later additions in any case only a very small part of the Mandala is not homogeneous.The Vamadeva groups 65 and again Gotama 32 9a 12a. and 76. 63. 55-58 ( 111 1). 20 and 22 ( 105 v). (lyric metres). 92 (x 91 io = ii 14 Of similar character are the majority of the Trimeter hymns (iii) Mandala v. Attach viii . 76 and 79 ( 107 i). Detach v 4 ( ( 95 98 iii). and 94 111 ii). 88 ( 106 ii). 71. 77 the stanza The hymns thus included are 1. except that Jagati is common. containing a play on the for family name Nabhdka (40 4a 5a). The great number of hymns in Gayatri metre. 54. 14 (connected with 13 by the concluding stanza asmdbhyam 95 ii. 103).. though it is only appropriate in 28. Detach 5 1-7 ( 98 viii). and in the stanza which concludes each of the hymns 27-29. 57-60. 76. : geneous appear to be included in the Mandala. 3. 4-6 and 11 1-20 ( 105 v). x 133 (refrain ndbhantam and 134 (Mahapankti). 1-3. 10. 80 ( 95 ii). 13 and 14. 39 and 40 ( tdd.. 69.). ( Attach ix 86 46-48 (refrain brhdd vadema.

ii. The metres used are Tristubh and The family name Kusika appears occasionally. The third Mandala has a metrical character h : of its fairly own. 5-7. The majority of the hymns are in Gayatri triplets. in that iambic variations of the type ^ common. but we do not find elsewhere hymns so regular in this metre. since the name only occurs in i 10 11a. i 65-73. 4 ( 98 viii). peculiar a collection cf. Detach 2 and 3 ( 14 ( 108).^ only are so that the seem to be intermediate in character and those of the cretic period. the triplet being specially common this metre may perhaps be connected with that of the third Mandala. 11. The family name Kusika appears 26 Group VII.. 36. ( Attach x 89 and 104 (refrain suuam huvema. 42 9c. in a hymn out of order. 43. 95 ii): thus they correspond metrically more nearly with the standard of Mandala iii than that of any other collection. The series 65-70 in Dvipada Viraj might suggest (iii) the bardic period. : hymns between the Vasistha hymns The hymns included are as follows The refrain stanza ildm agne connects hymns (i) 1. 23-25. 160 (5d vaydm indra tud sundm huvema). : : (ii) i 1-30. 8 1-9 95 i. 44. 15. 29 (g 98 also composed in Gayatri triplets.66 110. 17 ( 98 iii). 48-50. authorship.. ii. 22. 41 may be of earlier date than the rest. 16. The number of stanzas in the hymn corresponds to that If the hymns 7173 are regularly employed in the series i 4-9.. .. 50 4d. There are numerous proper names (iv) . It is very doubtful whether any of these hymns can be rightly attributed to the Kusika family. 98 viii). viii. 22 i 3 . 103). 28.2 i 102. . 30 1-15 ( 94 viii). v. The Kusika groups VII. 101). though one series hymns each containing ten 94 viii. but as they are used in historical references they supply no clue to the As to the refrain suvirdsah . . 94 viii). in 117 25d see 100. Detach 2 and 3 (Jagati hymns see also 98 iii). 26 1-3 (as 2 and 3). 13 m ( 98 viii. 39. 23: and the refrain stanza sundih huvema . 62 18a. 53 7c 9c 12c 13a. of iambic The hymns i 116-119 are marked by the extreme frequency and cretic variations side by side ( 94 v. 30 20d.). 33 5d. 13. The collections which extend from i 1 to 30 are almost entirely composed in Gayatri. iv. and Jamadagni 53 15b. 34. Gayatri. Hymns in Brhati (9 i-8. iv. . 97 i. Id 3b. : Mandala iii. 38. viii. and in an unusual metre (Anustubh). 18 4c. side by side with cretic variations . 103). but of this we cannot be sure in so 94 i.. Irymns 30-32. 27 1-12 ( 103). it follows from the regularity of the rhythm that the whole group must be assigned to the normal period . Attach ix 61-67 30.. rightly associated with 65-70. (except as in 108) (4-9) consists of stanzas. 53 9d lOd 11a: we find also Visvdmitra 1 21b. 45) appear to be characteristic of this collection: such hymns as 10. also the stanza i 73 3 is almost identical with iii 55 21.

. IX.. Although in many tinguish the collection strongly cases these two features are combined. of (ii) Kutsa 51 6a. 26 1-3. 68 9 -73 . 4-6. a refrain which itself bears the stamp of comparatively late date in its reduction of the Pantheon to a catalogue. except so far as they are occasionally contaminated by Anustubh and Gayatrl stanzas. which many have . 6. Attach ix 97 55-58 (stanza 58 has the refrain as above). vi 1. 97 34-51. 52 4-12. 17. 59 7-11. not so The following hymns may be assigned to much on account of the rhythm (though. and the inclusion in it of Sindhu and Dyava-Prthivl. x 33 (2ab. especially See 18 i-2T. VIIL. 53-59 6. 160. The hymns included are The collection i 94-115. .The hymn cretic group : 67 Detach 120 1-9 ( 105 v) and 122 ( 94 iii a. 105 19. In any case some association between the younger Vasistha hymns and the third Mandala is indicated. 106-115. The numerous additions the refrain i yuyam pata 94-115. ix 80. 101). 98. . 143. 78-80. which have some traces of iambic rhythm. 3. they 123-126 5. 51-55 i. but a general cretic rhythm (95 ii). 100-103. Group VIII. x 65. yet we may disi 94-115 as being on the whole more its metre. iii 2. 159. 146. Also i 51-54 (see 95 ii. constitute another collection very like consisting of the following hymns: 2 1-7 \ 5. 82. The following hymns. but it is connected with the rest by its refrain. and the collection i 31-35 and the Mandala x.. in which the hymns are connected by (i) the refrain distich tan no mitrb vdrunah. 44-48. which occur also in iii 4. 122. 16-17. 3ab = i 105 8). The refrain is found in 94-96. 15. which is markedly iambic . are earlier than the stanzas 8-1 1. and compare the mention of Vimada in 51 3c with that in 112 19a of KaksTvat 51 13b with that : : in 112 lie. Of the hymns with which we still have to deal. 52 ." The hymn 111 differs in a startling way in its rhythm. 98 ix). 140. 53 10c with 112 9c^ and of so forth). 83. ii. 111. and to Mandala vii. marked by hymns remaining in are almost the only metres used. detached from the groups with which stand in the Samhita text for reasons already indicated: i 59. (iii) Group IX. 97 1-3. v 4. 89. 145. the comparative frequency of late linguistic forms. 5-8. 157. it approximates to the cretic standard). 9-13. Detach i 104 ( 94 iii a. Attach ix 90. 41. Q6. 95 ii. as being more regularly disIn both groups Tristubh and JagatI tinguished by the language. and other notes indicating an approximation to the popular Rigveda shewn : 1 It seems not improbable that the Apriya stanzas vii 2 1-7. the cretic as will be as on account of later. period. 107 above. a large number are marked by cretic rhythm. and again a large number by the comparative frequency of late linguistic forms. all having the Vasistha refrain. 105 v): the colourless 121 perhaps belongs to these also 123-126 5 ( 95 i.

and these may therefore. 61 ( 94 iii). 29. 102). 171. 73 and 74 ( 105 iv). 138. divided into nine groups. but even here the is naturally the nearest In the notes are three times as many as the late. 50 1-9. vi 9. 28. i 13. This linguistic argument must not be unduly pressed. 116. 47 1-21 13. 84 13-15. x 106 is further connected with ii 39 by the numerous duals in each hymn which are combined by Sandhi with iva. On the other hand. 96. 5 ( 95 ii). The small groups extending from x 35 to x 84. that of remaining groups : the bardic. in In the hymns metres. 31. 110 ( 124 6-8 ( 95 i). it must not be expected that the statistics now to be given are in themselves sufficient to establish an order of time throughout the nine groups. 22 13-21. ix 5 1-7. 7. 8 ( 95 ii). 188. corresponds to the slow but To this general steady relative increase of the later features. the normal Gayatri. 102). late forms are extremely rare. 156. 139 1-3. both in external and internal form. 123. 3 and 4. The following hymns in Mandala x. normal. statement there are two important qualifications. a regularity which corresponds sufficiently with the and also because the groups concerned have one metre in common. 187. 41 10-12 j iii 4. 111 and 112 ( 95 ii). viz. ( The hymns of the Rigveda proper being thus roughly 112. claim to be the earliest hymns in the various Rigveda. 98 vii). because the BrhatlSatobrhati Pragathas of the Kanva family shew. 103). for which (where other (iv) indications fail) the position in that Mandala affords some presumption of late date: 2 ( 102. 127 ( 103). and the relative nearness of each to the popular Rigveda. 62 : : . 28 1-3. 29. 103). 102. character of the Tristubh of the normal period : . 12. 118. it becomes of interest to enquire into the linguistic character of these groups. which compose Group I. 178 ( 95 ii). 182. early linguistic the order of time provisionally adopted. 188 (98 iii). containing as it does most of the single hymns which shew a large proportion of later linguistic notes. (ii) Detach as above the hymns that have early metres. The ninth group.68 (i) The transition group Hymns i 31-35 (see also 95 i. 106 ( 95 ii. chiefly lyric. 32 1-5. the appear to approach linguistically the hymns hymns of the Kanva family of the normal period. 142. The conclusion last indicated seems probable in itself. 40. v 5. and cretic periods. 30 ( 98 viii. 81-9. 83. 39. On the other hand. The following isolated hymns in the first nine Mandalas. 9 1-5 95 ii. 14. for (iii) the reasons indicated in the lists of hymns detached from each group ii 3. 113. 100. vii 101. so linguistic far as the argument goes. 103). 48-50 106 iii). 77 and 78 ( 105 v). 176. ( and on account of the refrain 65 and 66 ( 107 i). 102.

Contents of the ten groups 113. 114. Group . this provisional arrangement of the hymns in a chronoJust so far as the hymns which are assigned to one logical order. and will be further tested this way by the metrical enquiries in the following chapters. 69 The treatment of individual hymns in this chapter has necessarily been summary and inadequate. 54 and 69. In it will be possible to verify. there will be good reason to believe that the original grouping really corresponds to the historical facts. but the general results are confirmed by the tables on pp. period reveal points of agreement other than those upon which the grouping is based. Table shewing the contents of the Kigveda as divided into homogeneous groups. or perhaps necessary to qualify.

In the Rigveda the vowels are sometimes combined. Thus in the verse vii 40 3d nd tdsya roZydh pari-etd it is asti (text paryetdsti) clear from the metre that the final syllables of each part of the word parietd retain the character of distinct syllables. 115. and the historical importance of these exceptions. and of final syllables ending in h and ra. for there is no reason to think that any final consonant does not contribute to make position. IV.' latter. By a final vowel first word or the : . object of the present chapter to determine the practice of the Rigveda in these points and others of the same character. The discussion in this chapter is limited to the consideration of the treatment of final vowels and diphthongs. it can be restored with certainty in a great number of passages by the aid of the metre. The Sanskrit Grammar includes under the heading of Sandhi the treatment of the final vowels of stems this subject is here reserved for the chapter on 'Syllabic Restoration. According to the rules of classical Sanskrit a vowel in such a position is regularly combined with the vowel following. so that the two syllables coalesce in one. so far as syllabic combination is possible with the is meant a vowel occurring at the end of a part of a compound. In classical Greek and Latin verse a final vowel is usually elided or omitted from calculation if the next word begins with a vowel. whatever it may be ' : but ' final -i -I -u -u It is the are only combined when followed by similar vowels. and so with the other final The treatment of final consonants does not affect syllables concerned. 116. in : which case hiatus is said to occur. the metre except as included above.CHAPTER SANDHI. the extent to which exceptions are permitted. Although hiatus is rarely shewn in the Samhita text. In the Rigveda final -a and -a are usually combined with an initial vowel following. as in classical Sanskrit but at other times each retains the character of a separate syllable.

whether the next verse begins with a vowel or 119. if not combined with a vowel frequently reckoned as short this subject is discussed in the chapter on Quantitative Restoration. without hiatus arising. it is considerable.Hiatus subject to rule 71 long vowel or diphthong. is : ' text rarely admits hiatus. in the Rigveda is an independent metrical Sandhi a final vowel may therefore stand at the unit as regards end of any verse. As the metrical character of the Rigveda is almost entirely 103). as in the hymns described in becomes more probable that the particular verse is also shorter. : There are many verses in which the restoration of hiatus is 118. commentators both native and western claim a general permission to restore hiatus in all cases. common at the end of verses a and c as at the end of verses b and d. destroyed by this method. but follows the the later Sanskrit (Whitney. reasonable to give the verse that interpretation which brings it as a whole most nearly into agreement with the metrical character of the hymn to which it belongs and the general rules of Vedic prosody. The consideration of the metre however shews that the Vedic poets were guided by principles and that deviations from these principles are 117. Each verse : not. but if in a particular hymn the number of shorter verses 94 iii as decasyllabic. which will be noticed in their order. Sanskrit Grammar. . doubtful. Where the Samhita text shews hiatus. unknown at the end of the Similarly Sandhi combination is almost verse. In a few cases. which involve any serious doubt becomes very small the consideration ' Metrical Commentary of these cases in detail is reserved for the the interpretation implied in the lists included in the text is always : ' : that which appears to the writer on the whole the most probable. Where the required number of syllables can be obtained in more than one way. Thus it is generally desirable to assign to each verse a normal number of syllables. or pair of trimeter But the appearance of hiatus seems to be quite as or dimeter verses.' final A following. Within the verse hiatus caesura than elsewhere. but there are many occurrences at the caesura. and very much commoner at the end of the verse than at the caesura. the text uses diacritical signs to denote hiatus. it appears often to be guided by a true appreciation of the metre but even so the text is practically of no value to us as evidence. The Saiiihita artificial rules of comparatively rare. there is usually one method which is in better agreement with In this way the number of cases the rules of prosody than any other. either because the normal number of syllables can be obtained in more than one way. is relatively more common at the The Samhita text takes as its unit the distich. or because the verses in question may be interIn all such instances it will be preted as decasyllabic or defective.

and -u or -u with either -u or -u.72 Regular Sandhi combination Instances of exceptional hiatus or combination at the caesura are Instances of combination separately recorded in the lists that follow. and also of evd. sd. syllables in -* or -u. -a named above. The readings the instances are therefore given as exceptions in sdh. sd in the Samhita text seem all to result from later theories they are often used even when there is really combination. Final -a. especially nd 'as/ and u or u: and less regularly prd. -u are usually written in the Samhita text with hiatus .' and sd he is comparatively rare. but -a stands before u. Duals in -a usually appear in the text as ending in -dv. ix 97 43bc. Duals in -a. The combination of nd as. -u are also regularly There appears to be no instance of the combination of the particle u or u with a vowel following the writing iim iti in the Pada-patha text seems to assert the principle that the long vowel is uncombinable. v 7 7cd. and 122. Nevertheless there are many instances of the combination of duals in -a. *164 24cd. in viii 13 15cd (Dimeter Usnih). prd u.. but these verses of epic Anustubh are regarded as forming Somewhat similarly we find combination may be regarded as a single trimeter verse. Duals in -I. so. as savdyam i 113 lc. The only probable case of Sandhi combination at the end of the vm verse is in *ix 113 7cd : tdsmin mdrti dhehi pavamdn | amfte lokd dksite Here the two dimeter verses one verse of 16 syllables. combination being on the whole more common. as in v 4 6d. md u. below. In the cases of the other words in -a. -%. it appears sufficient to note the instances of hiatus in the general Hiatus is somelists. pro mo iti) are never combined with a vowel following it is occasionally necessary to restore for them d it. forming Other cases in which metrical difficulties would be removed by the hypothesis of Sandhi combination are i 110 9ab. d. in the Pada-patha text as in the example prdcl iti prdci. sma and other monouncombined. etc. but combination of iti. : ' ' ' : . evdm v 6 10a. or diphthong following: and final -i -% -u -u are regularly combined with similar vowels. at the caesura by ordinary rules are given in Ch. pro. mo (where the Pada-patha has 6 iti. and at the end of odd verses combination appears in the text. sdsadandm i 123 10a. vi 27 5ab. with hiatus before it. : Similarly such forms as 6. that is -i or -I with either -i or -I. -a are regularly combined with an initial vowel 120. indicating that the final vo\vel is not capable of combination. But many monosyllables are uncombined. and some of duals in -%\ see Combination of duals in -a 128. times denoted in the Samhita text by the sign m. is usually given correctly in the Samhita text.

62 la2 66 9a 9b 12b. *91 5c. we have no means of determining. p. 173 12a. . iv 31 13c. 19 26a. and The text also gives -a before u. 22 4b. 92 5a 6b. 98 2b 5d. 65 5a. most usually when a vowel follows. 9 3b 5a. 20 la 37 2c. in the Veda. 58 lb. 134 lb. evd id is never found with hiatus. Oldenberg points out [Prolegomena. x 20 8a 22 7b 8c. 129 lc 4d 5c 9e. were so used by the poets themselves. 10c. 16 2a. *viii 91 lb. . 75 8b. 101 5a 103 10b. 54 lOd 4c 4d 3 21 12d. 133 7f. 27 7d 40 9d. at the caesura: i 31 8c. 128 4b. combination being eight times as frequent : 27 3a1 30 3b 9a. 12 3a. . 32 10c. or are due to reciters and editors. 81 7a. . . that final vowels are combined with xm 59 times when a vowel follows. 18 15c 41 6a. 52 10a. syllables : . but there are only four cases of combination with a following vowel. 46 19b 28d 52 9b. 74 7d. (a) in Dimeter verse 2 2 3 79 10b. *59 10a. *33 1 viii 1 3c 8c. 98 10a. whilst there is only one occurrence in which a consonant follows. 22 15b. .(Lanman. 36 4c. . 174 7b.' ' 121. 44 27c. 105 9c 10a. Hiatus is found after -a in the following passages. Again though evd is often used with hiatus. 2 2 ix *5 lib. 55 3b. 4 13c. 5 v 33 3a 4b. 24 3b. 66 lib 16 3b 42b. 33 2a 3a 4a 5a. namely in vi 12 2b. 24 lb 25a. 47 4d. . 51 12b 6 6a. 89 8b. . 63 22a. 16 46b. 119 9b. no doubt as a result of the later theory indicated by the Pada-patha reading. 23 23a. 16 lb. ii 4 8a. 78 7b. but the combination eved is Whether these particles found. . 29 la lb. 22a. 16 4a. . 13 5d. 18 7a. 96 17c. 59 2b. 11 lb. 435. iv 4 lb. 180 2a. 52 2c. 15 3a 3b. 190 5a. . 74 lb. . 61 in Trimeter verse. 101 13a. 116 18b. 120 5c 127 3g 8e. . 35 8a. . *x 101 lib (text -a). 7 lb 1 7a. 70 12a. 23 7c 28 5b. Thus Professor H. 6 6c. 2 80 5a. v 6 4b. . p. 41 8a ii 5 7c 176 5c. 1 viii 1 21a. 61 68 17b4 69 lb 8a 2 9d 18a. . 51 3b iii 32 4c . 89 3b. see the chapter on Quantitative Restoration. 2 3 1 53 13b. 185 2b 2 2 ix 102 8d 46 18e 2 viii 12 31d = 32d = 33d verses of four : i . 4 lb 8a. (b) lc. . 67 4b. 17 Id. 50 2b 1 51 5c. 24 6b. 18 lb. 117 la 8b 11a 21a 177 5a. 114 2a. 82 3a. 19 la 7a. 48 13a. 98 3c. . 4 3 1 26 4d J 8c. Nounvii 39 3b (text -av). The compound which appears as saptarsi in the Samhita text has in the Rigveda the value saptarsi. 5 lb. . ix 76 4c. 88 4d. 95 2c. 6 3d. 33 la 13c. vi 14 Id 5c. 4 13a. 2 3b bis 7d 14b. *166 5e. 386): but in these cases it would seem right Inflection to restore -av rather than postulate hiatus. 28 la 4a. 72 7a. 126 2b. 83 7a. *85 41c 1 *86 2c *90 10b *13cS 93 12d 13d 3 In *97 10c2 *103 13a24 105 8b. . Note). . 14 5d 6b. Locatives in -a are frequently found before consonants within the verse. 14 2c. 25 3c. It is evident that particles such as %m and id are frequently used to prevent hiatus. *170 5d. 1 vii 1 20c. adyd. . x *10 13a. 90 Id. 71 14a. 181 9a. 62 3a. 31 3a 4b. 43 lie. . 61 19b. 91 22c. 120 4b. 19 11a. 10 4c. 25 8b. . 11 12a. according to rule. . 33 7a. 1 . 80 9d. iii 13 la. 4 18 21a. For the quantity of final vowels in such words as evd. 77 la. 21 4a. 67 lc 5a. 44 6c. and it is therefore reasonable to accept the text as it stands.Hiatus after -a 73 duals in -I is more often ignored. vi 3 2b. 41 8b. vii 14 Id. 40 3a2 8d 3 41 7b. 70 4a.

x *86 16a *16c *17a i 31 15d. 168 3a4 9c. 66 7a. 106 7c. 66 8c . 133 6c. . Words not combined . 1 1 . *120 7b. 7 33a . 6 2c. 68 9a. *121 4a *4c. 55 17b. ii 9 6a. . 22 la . . *17c. *97 23c. 75 2b. : . . 8 7 1 1 7 4 9 13a 10 5a 16 7a. x 21 la 26 Id 9b. . 3 11a. . 36 la. . *95 10d 2 96 9d. chiefly in hymns of the normal and cretic period. . . 1 1 . being again eight times as frequent : 1 1 21b 1 38 6a 2 120 6c . . 61 3a 2 79 4d 6 87 lb 2 97 9b. *94 5a *8c. . 104 4a. 190 6a24 7a. 32 Id. 99 lOd. . 96 10b 2 . *145 3a. 61 9b 9 17a 7 65 5c 1 70 12d 4 98 4a . iii 5 2a v 4 4a. 5 4 the hiatus 8 is dtra or ydtra. 156 2d. 3 i 57 3c. . 127 i 8 la . 17 6a. 10 4b4 18 5c. 79 2a. 22 lg 2g 3g. viii : 15c. . 37 6a. 122. 115 5b. . iii 4 2c 10c. . . as follows (a) : in Dimeter verse: vii iv 8 4a. Hiatus after -a occurs in the following passages. . *88 9a *14c. . no one of the instances being quite sd 'he' is found with hiatus about 150 times: with combination about 50 times. 96 22a 2 . . 26 lc . 115 lb. vii 59 5a . 12 la 2 33 lb 10c. 5 29b. 38 9a 9c 3 32 16b 3 . *57 7b. 120 2b. vii 1 14a 15a. 61 5b 8b 8d 15c 2 62 Id (c) in Trimeter verse elsewhere 3a 1 4b. 180 7c 12 8a 12 183 2d. 1 . 178 2a. 27 6c. . 77 lb. iii 13 lb . iv 4 7a. . . *42 10a. . . *27 la. 24 15c 3 viii 4 4a. . . . 4 vi 4 4b. iv 52 2a. . 44 5c. . . . 18 2a. ix 74 2c 7c x 2 3c. iii 10 3c. 47 2b. 1 * 3 nd negative. . . 1 1 . . . . 15 6a. . *124 5d. 93 11a. 49 2b 1 61 3d 14a 14c. : nd only in certain. 40 Id. . . 21 9a2 24 9b 2 34 Id 4 37 2a2 2a. . ix 29 6a 1 108 Ud . *53 la. v 6 10a5 25 9a5 *51 15c. . 35 7c 10b 13c. 17 lb 20 17c. 13 2d 3d 4d. 141 12a 4 149 3c 4 155 lb 1 167 la. i 14 lie. . 34 la 1 1 : . 1 . 1 . 19 3b 20 4d. *12 3a *6a5 *8c 5 *13 5c. * i is found with hiatus about 60 times with combination 104 5b. 40 lid. lib 4 46 10b 29b. 1 1 vi 16 27a3 42b 11 46 5d x .74 . 33 13a 13d 15a 7 . . 63 5a 6a. 120 3c 2 122 4d 13c 4 127 3f 3 129 6a. 94 8c 11 117 lb. 129 5a 4 *162 13c. i 53 3c 10 5d 8b. : . 180 2d. 16 7c2 20b. 15 6a. . *88 12c. ix 81 la. . x 46 5d. augment with at the end of the hiatus. .' sd l he ' is rare. . . 129 7e x 9b. *60 7a. combination 123. 105 4a lib lie. 105 6a2 6a4 *108 5a6 115 4a. viii 1 23a . 53 14b. *53 4c 4 61 24d' 25a. 86 6a 5 . 186 10a 2 35 6a. 67 8c 4 72 3c2 vii 1 3a 7c 22b 25 la. . 41 3d 2 5a4 19 7a2 21 7d 2 38 5d\ 40 3d. 19 23b 35c. 45 9a4 . . *33 10c. 89 la 91 5b 24 92 5a. . Combination of nd as ' ' as. 16 5a 9a. . . 30 17a (a) in Dimeter verse 4 6c 13 lOg. prior element of a compound. 92 7c. 20 18a 2 70 10c. . 174 2c 5b 6a4 2 4 2 ii 4 6c 17 8b. at the caesura 62 8c. . 61 4b. 11a. . 39 2d. viii 43 9b. . 40 3a. 31 v 4 6d . 24 la Id. . 1 . ca. 61 5a 6 . 42 |6b. 133 7c4 148 5b. 74 4a. (6) in Trimeter verse. 20 la 6c 2 9b 12d 14b 14c4 19d. 22 2a4 23 24c. 73 9a 10a. 37 2b. x *10 6a *13a3 *15 13a *17 9c 4 22 11a 4 23 7b 29 5a 2 *32 7b 2 40 8d 4 49 7b 2 50 la lb 4 6d 7d. . . 167 10a 7 169 3c 12 173 4a 11a 12 174 la. 49 7a4 8c 4 63 2a 2d 3 3d 6 66 3b 4 24 3 Id 4 5 la2 13 la 2 22 lb 4a. 46 lb4 59 lc. . . . 102 9c. prd. 32 (6) in Trimeter verse 12 5a 5c. iv 2 18d. . ii 41 17b.

13 7a. and hi as in i 102 5d. 140 13c. plural in -td (-ta). 10 7a. 132 4d 8 6a. *83 7d . 110 3d. -u. . 15 14 dthd (dtha). 11 35 2b. Trimeter verse elsewhere: i 48 7c 16c. 1 . 34 5a 60 lc. 1 x 61 ix 97 49d. So nU u should probably be restored in i 64 15a. . viii 96 19d. 132 4d. -u before similar vowels combination is and more evenly balanced. yd. 21 la. 190 3c . 44 5a. 158-163. . 55 8d. *94 10b. 104 lc 10 5a 6c 2 108 4d'. 64 5a 9 73 7d. 87 4b8 93 7d 8b viii 4 20c. . . 16 5a. . 1 compounds beginning with su-. in -a (-a) from stems in -an. 1 d. . 65 15b. 58 6b 65 la 66 12b 68 9a . 109 la4 x 1 7c 14 *15 4c 11 23 5b. 33 10a. 11 vii 3 7a. 3 at 5 variant quantity see below. 44 3d. ii 34 12b. . 10 9 vd. 127 6b. 51 lie. . . . Id = 2d = 3d. imperative 8 7 6 sd. 1 . 53 9a. . . vi 13 3d. him ii 30 3c. *121 2c. 75 184 2c 3 14 186 6c. *91 5b. 20 2c 7 24c ix 86 44b 45a. . . 21 8d* 2 4 11 4 6d 6 7d\ 8 Id 22 6c 2 32 14c. and compounds beginning with su-. 112 at the caesura: i 36 lc (6) in Trimeter verse. 1 . 50 5b. *x 86 7d. . (c) 1 vi9 7c7d. ix *5 8b. the examples are viva *vii 55 2c. 103 13a 6 11 9 61 9c 14d 18c. kd. 1 . hiatus are After final -i. . 38 6a ia 8a 12 9a iv 2 18b 3 13b 12 v 29 15b. 93 6c. ii 20 3a 7a. gerund in -yd (-ya). lie. 49 6d. . 93 27a. . in Trimeter verse elsewhere: i 52 7b. . . . plural adya (adyd). . . . in -a (-a). iii 57 5b. 58 9c 4 60 4b. 70 12c 39 5a 40 la 3d 4b 6a 42 6a 69 3a. . 148 4a . 15 ii 13 10a. 64 5a 6 *75 I8d. 17 5a. *169 la. 1 . 20 3b. . 1 ii 17 8c. iii 6 6c 19 3d 2 12 8 16 la. 1 . *100 lie. 122 14d. 69 9a. 57 Id. -l. -a . 21 8a. 186 6a. 1 . 61 16c 113 Hd 120 6b. On the other hand longer words. 45 6b 7d. 61 13b. (c) 62 8d. 99 3a. 26b 116 7d. 30 2b 96 9d 101 13c. 24 5d 9a. 97 4a 11 107 2c. . 96 20a 21a. 34 la 4a vii 1 7a 67 6d\ *75 3d 2 5 4 viii 1 16c. 4b. 174 8a. . regular. vi 6S 8a. 139 3b. 8 13 21 9b 25 lb 7b. 189 4d 2 . The exceptions are as follows are generally combined. . . 1 . 62 5b. iv 16 21a. 41 16b 3 16c 3 45 2d vi 4 4b. 70 4b 7a. . 19 26c. . vii 1 3b 29 3a *49 v 29 6c. 17 lb. 169 6a. vii 62 6a. 173 4a. 34 lb. 12 13 11 smd (sma). nltah *x 161 2b. 1 . . evd (evd). 16 18b. After monosyllables hiatus particularly after vi as in x 32 2a. : 1 l 53 11a. . 11 5b. 1 1 1 . . 19 9b 20 8d. *48 14b. vlddm vi 9 6b. (a) viii 21 3a. 2 md. -man or -van. v 58 3a. . 74 6b. 105 4b. 18 4c 185 4a. 77 la 6 5a 6 87 4c. . . 173 5a. . . 34 5a 11 35 3b 4 40 2d. . 2 pers. 52 15a. . 24 3b iii 54 12b iv 2 13 lid 10 la 7 5a. 100 la. 1 . . 1 1 . *88 4c. vi *16 47c. 124. 19 6c. *iv 18 4d. . 1 .Hiatus after . 47 9c. 46 32d. For the true value of the final vowels of dchd (dcha). . 49 5b iii 55 12c. aorist imperative in -sva (sva). v 3 9d. 171 6a. . 144 5b in . 36 5b 41 5a. 49 15c. 4 end of the prior element of a compound. . *95 16a. . 24 22a. ix 109 18a. . *95 4a *101 3b *7a". vii 28 3a. 120 2c. 50 2d 115 7a 5 *129 5d. Combination is rare . vi 5 6a. 12 5a 40 3a 12 43 4b 3 23 4a. 31 12c. *117 7c. 53 3d. 70 5b 134 5b. 74 6c. 40 4b. 46 4d. 71 3d 3 14 4 ix 71 6c 49 10b x 5 t5d. . v 10 6a. in Dimeter verse: iv 47 2c. 48 17c 2 64 5a 6 66 3d 6 6c' 2 6d'. 127 6f 12 130 la 133 6b. x 50 3d. . 1 1 1 . 21 6a. *35 10b. 47 If 2f' 3f (in refrain).

1 1 1 . *162 2d *7c. 43 6a. *121 6b *8a. viii 7 24c. *164 25b. . ix 15 la (ydti). (b) in Trimeter verse: i 117 8d. 3 pratydc. x *9 9a. 182 7b. . *95 4b (vdsti). (b) in Trimeter verse: i 34 Id 122 12b. 58 3c iv 4 lib. 101 16d. *109 2c *116 4c. . . x 4 6b. *103 3a. ations in which a disyllable is the first element are also fairly common. . . ix 96 4b (svastdye) 1 prior element of compound. not pradisy For i 110 9 rtdsya as in the text). . 139 9a2 156 2d. *121 8c (devesu). 1 . *71 3b. 96 16c. viii 69 lid *v 40 9c. . in which it is regular. 116 above. . *113 3c. 80 2a. *86 8d *llc *22c *87 24a *25a.76 Combination of semivowels : 125. . . 78 3a. *13 3a *5b. 99 4b. ix 68 8b. so that the phonetic type. are used Other combinfreely in combination from the normal period onwards. . Dimeter verse: . *142 lb (asti). 31 5d vi 38 3b. *51 2b. 43 24b ix 6 5c 1 13 lb 17 3a 73 3c. v 11 6b. . . . . iii 26 8d. 80 3c 3 97 10c *27 6d. . (svasti)-. Combination of words other than those which consist of two short syllables is comparatively rare. *53 6a. 1 1 . 47 25b. *146 6b 1 *179 Id. . viii *1 34a. *119 5b. 38 3c 5d iii 9 4c. *52 3b. Words as below (a) in : consisting of two short i syllables are also found in combination v 19 la. 110 lie (pradis'irtdsya. *iv 18 3a *lla. Before dissimilar vowels final -i -i -u -u are regularly used with hiatus but disyllabic prepositions followed by the augmented tenses of the verbs to which they belong. . *128 Id *3b *6c 3 *129 7c *130 7d\ *142 lb. 118 8b. viii 16 lib (svasti). *85 38a. 17 2d 4b. 118 7d. 123 2d 6b. *109 5c. 45 4a 5b *67 31a *32a. *23 23a. as ddhy atisthan. 110 6b. *159 Id. *163 2b *9d. . 31 4c. 90 5c x *14 lid (svasti). vii 81 la. . 79 5c 116 12c 2 118 3c. *139 4c. . 2 dadhydc. 86 8c 17c. (Jiamsi)'. 1 1 . compounds : rtvij. . (b) in Trimeter verse : *ii . *141 lb 3 *145 4a. *157 5b. *135 2c. appears to be the essential condition. 68 8a. 116 15d 17c. 1 1 1 . . *81 4d. . *179 Id *3d 3 ii 3 lb 15 9a 26 lb. 52 6c. 8 8b. 99 12c (svastim). *98 lid. ix 26 2a. * svitydc. 146 4c. 49 2b. *72 3c *5c. . 1 1 1 . 120 8a. 135 6d. 124 3a. except in the svdhd and svid. x 1 4c. viii 51 4c 32 4c (sivyatu) (tu). *90 Id *5c. *58 4b vii 76 4c. 1 . gdvyuti. 32 8c 12b. 113 7a. in *191 5a. . 51 168 9d. 47 lb. see On the other hand sudhd occurs in iii 32 15a. . *161 12a. . 33 10b. 61 13a (nu). x 134 le If in refrain verses (janitri). . lb. . viii *59 6d. . 69 4c. *157 5a 3 1 1 . 30 12d *15b. Disyllabic prepositions are combined with an augmented form of a verb following as below (a) : Dimeter verse: i 11 6b. *146 4b (ddru). *161 8a. not the grammatical connection. 50 5a 3 5b3 80 12c. v 28 lb 3 76 2d *58 10a vii *33 la 4 65 3b vi 27 5b 8c 83 8c4 *104 2a. ix 72 3c. ii 12 lb lib. 80 9c. 106 lib x *18 lid. Combination is found exceptionally as follows v *51 14c and *15a (a) in Dimeter verse: *i 191 3d (ni) *vii 55 5d (sdstu). 123 7b 3 *124 4d. *87 15b3 *17d 3 *88 13d *16c 3 89 13a. *100 5c.

70 4b. *127 lc. But duals both of nouns locative tu4 (tve). 35 22c. vii 1 19b iii 59 2d . ii 23 16b. 186 8c. v 61 2a. *95 6d. final -ah becomes -o. -e y -o lose element before an initial vowel following. 122 la. x 61 7c. 43 2c. and final -e is unaltered. 116 6b. are unaltered. 1 . 1 1 . 51 3a (c) 89 6c. *108 5b. 127. 168 9c. and the These rules are so far in general -a. the Rigveda proper. 53 2c. After -ah and in -e initial a. 47 22b. of combination of duals are found. 59 3b v 29 10b iv 16 18a. . 33 13b. iii 12 4b Dimeter verse i 47 9a *x 85 9b. 99 7c 7d. if the next word begins with This result is comparatively rare in a-. 1 forms in -e. is agreement with those of classical of the dual endings. in Trimeter verse. . 190 3d. viii 5 2c v 73 6d . . . . 85 7a 5a. . x 7 5c. 1 22 4d. it seems that the loss of is the initial vowel was not regarded by the poets as equivalent to combination of the syllables. 55 Id *iii 29 3d 186 11a. 77 their final Words ending in -ah. 1 . but the initial a. 71 5a. : yusmd see According to the rules of classical Sanskrit. 126 3a. -I. Dimeter verse: i 79 11a 81 le. -e are very frequently combined In the instances are given in the next section. 167 2a. at the caesura: i 59 2c. 103 7b. . viii *58 lc *3c vii 18 7c. 63 6c. . 1 . 57 5c. 1 1 . 61 9a ix 86 96 20c. see above in Trimeter verse. 1 . 59 4b viii 2 34b 40c. *U1 4b. 8 6c. : 1 . addition the following examples chiefly in the normal and cretic periods 128. Sanskrit. (a) *x 72 4c. and of tue recognised in the Pada-patha text by the addition iti as to asme The unalterable character 170. *33 lib. 31 3d. 44 19b. 49 5c 56 3d. 50 5b *90 3c *4b *12a. . but the popular Rigveda it is occasionally found in all parts of it : in considerably more common. ix 9 4b. and are therefore all alike treated as ending in -a with hiatus and similarly words ending in ~ai : and -au are treated as ending in and of verbs ending in -e. 38 2b. Duals in following iva: : (i) Combination of duals in in : -a . and also incorrectly at the end of odd verses. 22 5d. : (a) 16c. . ivll2d. elsewhere: i *24 8c 30 16d. *103 lc.is lost. 31 9b. As many of the instances are found at the caesura.Abhinihita Sandhi 126. 50 10c vi 9 2b 30 3d 1 10a. 23d. *145 6a *161 5d\ *166 3a *4d. 89 13d. The Saihhita text usually shews this combination correctly where it occurs. with -a. 27 22d. . *85 17d.is lost as follows 1 . *190 lb. 1 . 118 7a (b) 61 3d . At the beginning of a verse of four syllables viii 13 15d 1 1 9. 52 9d. : vii 66 5c. 1 1 1 . 75 2a 79 6c 92 lib.

2 iii 6 10c. of a dual in -u. 44 la. 90 3a. 41 2a. that the simplest explanation is the presence of the parallel form va. such. 7d. ix 88 x 4 6a. 108 3d. -Ih. viii 19 14d. *52 2c. ii 16 5c. : : : (b) (c) in Trimeter verse. 182 4d. 62 5c 5d bis 6b. (b) ii iv 32 23a 2 . . 119 2d. 60 4c (a) in Dimeter verse 5b 7a. 185 Id . with dissimilar vowels following appears to be found also. 44 3d 4b. -ih. v 7 8b. Words ending in -ah. vii 87 2c. vi 46 4a. 63 5d. *43 combined with iva following in i 141 lie. have already been taken account of the forms which indicate late date. 57 4b. 35 5a. -aih appear to be (ii) ii 6 7c. 39 152 3b. 73 3d. *146 2c. it will be well to give the list here. i (iii) Combination of duals in -e is very rare . As however they amongst may also be regarded as examples of a peculiar form of Sandhi. 68 lib. -eh. 76 lc 2b bis 3d 4b. 102 2c. vii 9 5c. 40 2c 3a 3d 4a. 21 9a. iv 56 la. 129. In the dvandva duals as indrdgni. v 43 8c. 1 duals in -?. 46 3a. 166 lc 2 Id 2 173 4d. regularly used in Pali and in the verses of Buddhist Sanskrit. v 64 lc. 54 6b 6d. 69 5c 6c. 183 2b 3c 5d 8c. namely prdcl 'dhvard in iii 6 10c. viii 40 la 3b. ndktosdsd. 116 3a 8c 9a 10a 14b 19c. 86 5d. combination is (iv) the more usual but the following examples of hiatus are found v 86 la 4b 6a. 112 Id. There is no instance of the combination iii 34 Id. 83 Id 2d 3d. 106 lib. 2 duals in -e. 106 27 times. without regard to the form of the ending. *173 2b *2c. . 131 5a. and rodasl 'bhe (for rodasl ubhe) in i 33 9a. 71 4c. 35 7a 7b. *104 6d'. vii 56 x 62 9b. v 31 6c. in Trimeter verse elsewhere : vii 93 lb 3d x 65 2a. 48 18a. *59 lb. 64 5b. 175 6b 2b. 3b. *87 3a. 110 8d. viii 5 21c. the occurrences (i) As Duals in -a -I -e are combined with iva following 1 : (a) in Dimeter verse: i 28 7c . *166 2b. -ah. *125 lc*ld. *75 14a. 30 4d 58 2b .is found in (ii) 121 8a. : i . 63 lc. 8b. The particle iva is so frequently found in combination with preceding words. *91 3c. 31 la. 34 9b. *84 2a. V18 3C 1 . 86 3a. at the caesura : vi 60 13a. 177 4d. there is perhaps an instance in vii 72 3c. *149 4a *4c. . *97 10b. 74 9d. the combination being Combination correctly given in the text in the last two instances.78 Combination with iva (b) in Trimeter verse: i 34 9d. 64 2c. in Trimeter verse 3 6b 2 39 21 times. *128 7c. 117 Id lid 12d 13d 18b. 39 3b 10a. 4c. 73 4b. vii 69 6a. iv 41 5b 5c viii 26 15c. iv 2 4b. vi 30 Id. x 22 5a. iii 36 6b. viii 22 10c 12a. Combination of duals in -i with following i. 184 3a2 3d. 72 5a. ix 96 15d. -uh. vi 49 5c. *161 7b. vi 59 2b 5a 6a 8a 9a. : . 118 9b.

iii 30 21c. *90 93 lOd. *51 15a. In Group IX irregular hiatus 3000 trimeter verses there are only 20 examples. i 14 3b. *87 15c. and Groups IV. In other cases in which it is suggested by the metre it is not always possible to determine whether the irregularity is in the metre or Sanskrit in the combination. -e with following vowel other than a (i) v 52 14c. -im : 131. *iv 18 5a . and IX so little dimeter verse that no comparisons can usefully be made in these particulars. in is unaltered. (ii) final -ah: i 177 4a. x 49 6b. iii 43 -ah. This result is . . *114 4a. Probable instances of irregular combination are ii 20 8d. vi 17 7b. About one half of the variations are in favour of hiatus. 4a. VIII. but combination hardly exists. vii 19 5c. iv *18 2a. viii 1 26c. vii 86 4d. : : 5c. a substantial confirmation of the view that the hymns of the bardic period are earlier in date than the rest of the Rigveda. v 17 3a. The occurrence which it is is in positions in of Sandhi combination in the Rigveda not admissible by the rules of classical proved in the case of duals ( 128 above). 130. . vii 39 3b.Irregular Sandhi Words ending in -am. (iii) 32 6b. vi 19 3c vii 41 6d. Of those in favour of of those in hiatus two-thirds are found in the bardic period favour of combination only a quarter are so found. -um appear to be combined with iva 13 4b . : ii 20 2b. The whole number of variations is about 1200 : the instances in which the rules are observed may be estimated at ten times that number. x *85 40b. The Vedic rules of Sandhi combination distribution of the principal variations from the is shewn in the table following. and one half in favour of combination. x *51 9a. 79 following in *i 97 8a viii 49 4d *59 12c j ii . -im. *161 8a. and combination all . iv 34 3d. 46 28c. -dm. *149 lc. 49 6d. before. 75 3a. It then III than appears that hiatus is far more common in Groups I In Groups IV and V hiatus is only half as common as elsewhere. 48 4d. x 20 2a. In the popular Rigveda some reaction is to be noticed: hiatus is more common. iii au (iv) final -am. final -ai. The number groups : history can be pursued in more detail by considering the of occurrences of each kind in relation to the bulk of the except that Group V contains so little trimeter verse. v 46 2b. and of these 13 are at the caesura: combination of kinds has greatly increased. *166 5d.

-i. and is -u. Table shewing the growth of Sandhi combination in the Rigveda. they 132. Group .80 of sd. -u. History of Sandhi and with iva is rarer. but combination of final much more common than in Group IX. -e is a close approximation to the standard of classical Sanskrit reached. In trimeter verse the occurrences of hiatus at the caesura are common VI Groups equally I to : with the occurrences in in Groups VII to IX all other positions in are twice as common. -ah. of duals. -I.

154). In the present chapter it is proposed to deal with all 133. shall we it must not be expected that the be prepared to find similar phenomena within words: but historical developement will be precisely the same. (i) Syllabic restoration appears to be justified as follows: When the proposed restoration is more suitable to the metre in a majority of the occurrences of the particular word or form. at the end of words. group the facts.CHAPTER V. great majority of cases the result is to restore hiatus within the word. and regularly semivowels before dissimilar vowels. and thus to have a starting-point for the further enquiry. as to small. light 134. as 6 . as in the last chapter the result was to restore hiatus at the end of the word or prior element of the compound. The historical problem under investigation in the two chapters is therefore essentially the same: and the Sanskrit Grammar has reason when it treats of the combination of a stem with a case or As however the personal termination as a problem of Sandhi. Now that we culties have found that the Vedic poems admit every kind of hiatus.' treatment of the interior structure of words involves many diffi' and complications. the instances in which it seems desirable to correct the Sarhhita text so as to alter the number of In the syllables in a word. SYLLABIC RESTORATION. after the occasionally after -a -a and between similar vowels. We shall endeavour first to ascertain and which the margin of doubt is comparatively and afterwards to interpret them as far as possible in the of our general view of the history of the Rigveda ( 153. it has seemed better to deal first with the simpler problem of the combination of words.

dual of -i stems. as in asidma (/) the noun-endings -yai. either as ndasatya. so that it is unlikely that the occasions can be explained as chance irregularities of metre. paantu. hdrioh (e) pronoun of the second person.in composition tui. in (d) as sdmid. -yah. trayidhd (text tredhd). prdyistha (text prdstha). frdyini (text sreni). as sdria (c) same suffix in the tudm. as in prthividm (g) a few words in which y v are the second element in an : initial consonant-group. dadsvat (text ddsvat). ddityd. as ksdam.thematic flexional forms of monosyllabic nouns and verbs and their compounds. abhttid. In connection with these we have to consider the difficult question of restoration in the suffixes -ra and -na. bhads. sudya. in the forms tudm. cyautand and four-syllable forms of the words : ndsatyd. loc. dayisnd (text desnd). ndyisi (text nfei). : ddsndrtis. ruda. urdhud. Syllabic restoration is also justified as a conjecture in words and forms which are not very frequent. the instrumental singular and gen. tvd (e) the optative suffix -yd.rd. and the corresponding forms with superlative and other suffixes. as indara. as in the : suffixes containing v (less often y) heavy words aiigdhue. as ytijia and generally after a heavy syllable. and tud. -bhiah after heavy syllables : (d) after numerous endings and syllables. as jydyistha (text jyestha). vdata : the ending declensions (b) : -aam (text -dm) in the genitive plural of all (c) the endings -bhidm. -yam. Such cases are (a) non. dditid. aadityd or as ndsatid. : : in the declension of : nouns in -man and -van : as dhdmand (text dhdmnd) and (/) in a few isolated words and forms: as bhadsvat (text bhdsvat). if it is suggested in one or two passages and if reasonably near analogies can be found in . ddhie for ddhye : grave accent (b) in many nouns having : : the noun-suffix -ya after a light in nouns having the syllable. as diatih. tud. mdntara. When the proposed restoration is more suitable to the (ii) metre in a substantial minority of the occurrences.82 (a) in all : Syllabic restoration words containing y or v followed by a vowel with as suar for svdr.

4c. uv. *95 9c. vasavya for exceptions see For kvd we must sometimes read ku see 151 iii. indara. (vlrya). -u i *162 20b (tanvah) . *85 30c *31a. rudriya. hj. In addition to the above restorations. and even by some 84 b): and effect is given of the texts (Whitney. *163 lc (b) *2c. x *87 22b. In very common words and forms restoration is not justified by an occasional opportunity in the text. *155 2c. *90 8c.Accented y and v : 83 any of the sections above but the application of this principle is a matter for individual judgment. dhdmand probably represent with fair accuracy the pronunciation required. This principle is recognized by all commentators. vasavya are regularly of three syllables in the Rigveda it is therefore necessary to correct the accent. 62 . *98 10c. but also da'isnd indra and dhdmnd.). : in the declension of stems in 4. (manusyd). yajniya. Sanskrit Grammar. 135. written forms i u are therefore used in preference to the Indian forms has been necessary. and in these rarer cases it seems better to follow Indian methods of spelling. sing. *151 (kanyh) : . 150 4b (manusya). and it is therefore unnecessary to discuss such suggestions as the restoration of -aam -aah in the ace. In any case two-thirds of the instances are in the popular Rigveda. x *51 2b *4c. 68 8d (staryctm) . *58 9a. respectively of the suffixal -a -a declensions. and nom. and 151 ii. The to it in all the texts in the case of words like dhiyah. such forms as dayisnd. in order to indicate that restoration Restoration is also required in the vocative case of nouns if the as sahasia i 147 5a by the side of sahasiaih other cases require it v 29 9a. Exceptions are rare. all of which are con- nected with the history of hiatus and combination. To represent the restorations systematically.). -ya: vi 47 16d iv 1 13a (manusya. Also svdr ii 35 6a (probably). iii *53 17c. : : : : Otherwise the exceptions are as follows ii 20 8a (tavasyd). ddhie for adhye. we must read stiar for svar. 98 6a (pasavyd). read throughout asurya. *130 5d *6a. we should write not But only girdam siiria tudm. the text gives yajnyd (or yajhyd) seems to be required. vi . ix 56 3b vii *55 7c. (a) suffix -ya. 46 12c (tanve). The words asurya (neut. ix 96 21c (camvdh) (tanvam). pi. and pathye *v 51 14b (probably). where in v 61 16c and vi 52 14a. : except in the popular Rigveda. 30 10b i *162 4a *8b. *75 16b. Syllabic restoration is required wherever y or v is followed by a vowel with svarita or grave accent for instance. viii 76 12c vii *50 4e. fern. there are a few instances in which the text appears to be incorrect in other ways.

In divyd and sakhyd the syllabic values predominate. durya i 91 19d. 39 id. 2d 3d. ii 4 4d. 89 2c. ddbhia. but we find the compound yavydvatl. x *34 9c. gddhia. *vii 35 lid. -id regularly required. *x 53 3b. 97 5a. x 106 4c. sdsia. 62 9a. 138 2e 4g. vii 18 12c. 64 6b. hdvia invocandus and havyd 'drink offering' by side. In words included in (i) and (ii) the syllabic are often found side by side. is so marked that it is necessary to that primitive suffixes -iya or -ia and -ya existed side suppose The particulars are as follows : in which y is always syllabic in the Rigveda are (i) uksania (prop.84 136. 94 Id iv 25 7a. 100 3d. x 25 lc. For yavyti the metrical value seems to be yavid or even yaviti. *124 2d. *163 8c. ii 23 13a. but (ii) consonantal exceptionally in the passages quoted dpya ii 38 8a. 44 20c 22c. except in sdntya 'good. The -ya suffixes Syllabic restoration is required in numerous nouns and adjectives ending in the suffixes -ya. yujia. iii 9 3d. 107 24b. iv 38 2d. 79 3a. -yd. 86 lc 4b. v 33 5d carkftya i 119 lOd. samid. gopaydtia brahmanid. 29 6b. pusia. as rc/miyd. viii (iii) . -tya. 28 la. 124 9b. 86 9c. jdmaria. abhriyd. *88 2d. No methodical distinction between these words and others in * ' which the restoration is not required has been established: but the difference in metrical value between e.-ranya) 129 6b. viii 10 3c. 97 33a. syllable. which occur vi 59 9b. -nya. 82 8c. etc. vi 19 13a. the 137. 62 lb. ix 56 2c. hfdia : : Words dvyathia. 97 10b. ksdmia. nary a iv 25 4d . In the suffix is -ya. in which the suffix is -a ydvia. : In the following words also y is regularly syllabic. 37 2b. tujia. 116 6d. 63 30b. 33 2c. 64 16d. *71 5a. the gerundives also dnia. rdthya i 35 6c. 49 2a. 36 5b. jdnya ii 6 7c. *98 5d. name). 107 19b 19 30c. 180 4d. being about three times as frequent as the consonantal values. : : : : 24 23c. 71 9c. mddya ii 14 lb. -yd following a heavy restoration of -ia. 20b. ddmia. Gerundives guhya ii 32 2a. The fact that sdntya is always disyllabic suggests that satyd may .' In many words the syllabic value is represented in the text by iy. hdvia. 13 21a. 60 3a. *vii 35 lid. dfsia. irya vi 54 8b. vi 50 lid.. himid. gdnia. vihdvia.g. aucathid. 40 7c. as follows: divyd *i 164 46b *52a ix 14 8a. vii 1 lie. 68 8b. iii 56 5c. preceded by a light syllable. dhdnya v 41 8c (and in jlvd-dhanya) . v 5 10b. *139 5b: sakhyd i 15 5c. -drutya x 134 7c. ranya iii 55 7c (the hdvya i 108 6b. 109 3b. compounds have always ranya. 110 6c. *103 2a. ix 49 2c (the compounds have always -janya) . 31 7d. as in prajdvatlsu duridsu durya vrnand dtra sakhidya sakhydm and consonantal values vii 1 vii lie 18 1 2c.

*85 lb. *59 6c. (iv) Other words: . vii 8 4b. 32 8d. (iii) : : . *5 lie. 79 8c. viii 62 3c. but the occurrences are It does not therefore seem practicable to distinguish between -la few. 44 2b. x 65 lid. *40 5a *6c *8c *9a. 18 6b. paprksenya v 33 6a vdrenya v 22 3c saparyenya vi 1 6a. *16 3a. (ii) 92 12d. 84 Id. v 55 4b. *72 7d. 29 10b. 62 3a lib. aghnyd *x 87 16c. 65 lie. *156 4b. *151 5c. 56 5e. 30 5d. 66 2b. 31 8d. 93 5b. as mdtsya and mddhyathdina. 68 lOd. *178 3b.-ya after heavy syllables be 86 the true reading. i 27 12b. 6b 110 7a. satrutdrya vi 22 10b. 89 7b. ndndsurya *ix 114 3a. dmartya ii 11 2c. 99 4b. vii 8 4d. 41 5c. x 105 lc . 62 lb. *139 4d. 59 3b 5d. kdvyd i 121 12c. tudm. 69 6a. 54 10b. x *10 9b. 135 ii 24 9d. 63 4b 7d. 39 5d. siirya i 7 3b. 33 4c. *90 13b. hdstya ii 14 9c. as are also the instances of gerundive forms in -ya. 91 6d. 69 2d. vi 18 7b. 137 2e. viii 18 8a . surydcandramdsa i 102 2c. -yd are found with y consonant after heavy : syllables exceptionally as follows daivya (i) v 5 7b. 54 2a. *128 3c. suryd x *85 6c *8c *9c *12c *13a *20c *35c *38b. viii 37 la. *113 3b. 6 10c. agdstya *vii 33 lOd. *x 87 21b. *130 . rdndya vi 23 6d daksdyya i 129 2b. *88 6b. 54 19d. *24 8b. One or two other words. 92 12c. 10b. 33 lb. 7c. 72 3b. anindyd i 180 7c. vi 13 lc. 45 10c. jyaisthya i 5 6c. hotrvdrya vi 70 4c. saubhdgya *x 85 33c. 124 lb. 72 lc. 68 9b. *33 8a. vrtratarya i 106 vi 25 3d. 81 4b. have also always y consonant. ii 4 3d. 12 9a 30a. 115 Id 2a 4a 5b. 102 15c. x 104 9c. 101 2d. 107 7d. viii 3 6b. x 3 4c . viii 24 27b. *x 190 3a. v 27 6d. 49 7a. vaisya *x 90 12c. 122 2d. 101 12c. The exceptions : to the rule hardly number one in every ten instances. vitantasdyya vi 18 6d dhhvsenya v 55 4a. ii iii 4 7a. sdryarasmi x 139 la. suvirya i 184 4d. sdpya x 48 vaatdpyam (probably) 9a. 98 2b. 48 7b. iv 38 12a. dptyd v 41 9c. 37 5c. 43 32b. 86 22d. . 27 5a. auumddya ix 107 lie. i 121 8d. 27 19a. 17 5c. 3 7a lOd. 51 4d. and -ya but suffixes after heavy syllables. *164 14c. 35 2c. 100 18d. 2 drya i 103 3d. ix 1 6b. *191 8a *9a*10a. 7 8b. In the declension of the pronoun tvdm thou the forms tudm. pravdeya iv 22 5b. 34 9a. *58 8a. *35 8a. 2b. iii 30 9f. 44 23b. kdvya i 96 lb. 31 15d 17b. 34 5c. didrksenya i 146 5a. 47 7d. Gerundives: il_ya iv 24 2a. 66 13a. dsvya iv 41 10a. 14 9a. 72 16c. *v 51 15b. *90 6c. 112 13a. and tad (instrumental) ' ' 138. iv 42 x 65 10b. 117 5b. 97 3c . surydmdsd x 64 3c. 37 4a. *130 3b. *x 87 16b djya x 79 5b. The suffixes -ya. vfsnya venyd vi 44 8d . 85 2d. tue must generally be restored. 25 21a. 138 2d 3a 4c. 93 lc. 23 27c. 32 23a. *12 7c. ix 93 5b. . iii 16 la 3b. 17 3c 5a. viii 22 18a. 28 5a. *58 4c. 63 7b. 64 7c. vi 12 Id. 5 2c. they are much commoner in the words daivyd and sdrya than elsewhere these instances are therefore given separately below. *27 21b. marmrjenya ii 10 Id. 83 5b.

*16 13a. 47 14a. : vi 1 iv 31 6c. 35 9b. 79 5d. 16 6c. 12 8a. tvdyd i 101 8d 9b. 54 6a. and especially in the instrumental singular and gen. indra-tvotdh i 132 lb. 38 5a. 20 23 la. where the text gives y v r or n immediately followed by a caseending beginning with a vowel. *86 3a. *97 18c *23a. dual forms: for instance.). 122 5c tvdmkdma viii 11 7c. 21 8a. 34 la. 13 3c. . 94 3c 6a 16a. 6 12a. x 22 9b. 32 lb. 30 4a. tvdr in composition: tvddatta ii 33 2a. 11 4c. 26 20a 25a. 8 5a. 43 4a. 112 7a. For other forms of tvdm see below 148 iii. more often than not. 97 43d. 22 2c . -yd are about equally common in the . 9 3d . 90 4c. 80 2c. 53 10a 10c. 91 8c. 36 2c 6c. 100 la tvdvrdha i 56 4a. 24 26c. 54 8b. 13 4a 6b. tvdydt i 53 3d. *x 98 10a. 36 7d. 46 2a. iii 19 4a . 18 3c. viii 19 16d. 14 5a. tvdm i 9 4b. 33 4a. 35 3b. x *10 13c *14a. 15 8a. ii 18 6d . 91 21d. 21 la 5c. *145 5b. 105 8c. 3 2a. 48 14a. 44 5c. 144 6b. 5 2a. tvotdsah i 8 3a. vi 1 lc. mddhud. dual forms of nouns the measurements -id. 31 10a. 67 10a. 93 33a. vi 2 2c. *95 18d. pitaroh. 31 7a. 36 19a. 45 6a. m&dhva. 3 5a. vii 1 21c. 16 5a. The exceptions number rather more than a quarter of the occurrences in the Rigveda proper. 11 lc (probably). 134 5f. 9d. 92 14c 22c. 11 3a. 60 lc. 37 6b. 32 12b . 64 3a 3b. tvdvat i 81 5c.86 always. 13 26a. 76 5c. 62 11a. 19 10a 30c. and about one half in the popular Rigveda. 170 ii 1 la 3a 4a 4c 5a 5c 6a *5a *5b. viii 1 32 12a . whether the meaning is instrumental or otherwise. *84 3d. 49 4c. 2 2a 10a.as the prior element of a compound. 16 12a. *43 3a v 2 lie. 99 2b . iii 16 3a. 1 14a. tv&hata vii 32 7c . 129 la (probably). 29 lie. x 69 9d . tvdyu vi 47 10c. pitroh. iv 12 la. svadhdvne respectively. *59 5d. 82 7c = 8c = 9c. svadhdvane. 11 la 3a. iv 1 5a. 22 10a. 169 la. 44 5c. 189 3a (prob. (ii) iii 8 la *llc. 50 9a. loc. 66 6c. 16 13a vii 11 2a. where the text gives pdtyd. 104 6a. 60 12e 14c. vii 15 12a. vi 21 lOd. 4 lc. 92 32c. . 45 17c. 23 12a. lc . ix 4 5a. x 38 5d. *128 6b. 18 13c 13d. *15 12d. 139. 174 lb lc 10a. 12 2a. *55 4a. 19 5c. 75 4c. 96 17c. 91 3c. *28 3b. 3 lc. viii 2 16b. 29 4d. tvdduta v 6 8d . 165 9b. *85 26d *45a. 15 2c 3a. 37 4a 5a 6a 6a 7d. 41 7c 10b lid. 32 lib. 104 9c. v 13 5a. *18 12b. 134 4a. 10 *14a *14c. *142 la. 21 2c. 125 2c. 71 5a. (i) in -i -i In the instrumental singular and gen. (iii) tve ii . 139 7a. 28 5a. 43 20a. 61 16a. 118 8a. viii 92 18b. *98 9b. 17 13a. tvdddta (iv) i 10 7b. 39 lb. 91 10c 22c. 98 4a. viii 4 16d. 80 7d. ix 45 5a. v 7 10b. 43 15a 32a. *86 4a. we must read pdtid. 130 6c. *100 x 1 4d. *18 9c. loc. . 17 7b. 110 7c 2c. 48 8b. viii 19 18d. vii 32 23a. 54 5d. 30 5a. 3a. 33 2a. 88 lc. *75 lc. . 52 12a 15d. 113 5b 6a 6a 6b 6b. 67 26a. 81 5a. Syllabic restoration is required. 61 21c. *108 9a. iii 40 6c. 9 4c. (i) They occur as follows : tvdm i 27 4a. 66 12a. 99 Id. ii 1 13c. 123 lie. 97 Id. The pronoun tvam We must also generally read tua.

x 61 3b. v 42 4d. ndvyasl viii 51 3c. ii 1 9b. Restoration of -id : -id. -ioh is required as follows : masc. 44 24c. i 100 16b. tmdnidx 110 10a. vdjavatyd i 31 18d. sfnid i 58 4b. dkutyd *x 151 are found almost exclusively in the popular Rigveda. ii 2 3b. *71 10b: fern. x 93 13a. tvisid x 89 2d. 56 3d. vii 67 9b. and perhaps iv 48 lb. pavyd i 88 2d.Instrumental* of Rigveda proper after light syllables. sdmid i 83 4b. sdcid iii 60 6b. pdtyd *x 85 24d *27c *36b. raydyah i 167 lc. 31 4c. In the masculine may have been in almost all instances been . as -i . istya *x 169 2b. mddhumatyd i 157 4b. devdJmtyd x 63 lie. 78 2c. vasatyd i 66 9a. *35 2a. sumati (text sumatyd) i 31 18d. Nouns in -vl have regularly -vyd. nouns. -yoh: ardnyoh vii 1 la. and -id seems to Ikj due to the influence of the masculine declension the original form in -I seems to have been altered in the text to -yd in several instances. mahyd iv 1 9b. namely prdmati (text prdmatyd) i 53 5c. pdtid *x 85 22d. vi 56 2b. 104 3d. vi 17 6b. 173 i. mithatyd vii 48 3d. and hiranydyl viii 1 32b. required by the metre. x 29 8d and on this analogy it is quite probable that we should restore (in accordance with the metre) : ii 14 2b. sustutid viii 16 3a (probably). : prthivyd. asdnl y The exceptions viz. ndmyd i 53 7c. 4c. hdrioh i 7 2a. rayyd *x 19 7d sdkhyd vi 21 7c. prthivyd. and there is no reason why this form should not be In the compounds -matt is always restored for matyti v 58 5d. navyasid vi 22 7a (probably). sam'icyoh . nouns. v 52 9c. iii 45 2c. divdsprthivyoh v 49 5d. it lias In the replaced in the text by -ind. vi 8 5c. dsvavatid i 30 17a. ddvidyutatid ix 64 28a. 35 5a. iv 16 lib. x 50 2a. For the gen. -ioh are (iv) regular after heavy syllables. used side by side. 26 6d. viii 33 4c' The declension of rai wealth is hardly to be separated from that of ray i a trisyllabic form rayind (rayid. rdjid x 100 12c. rohinid viii 101 13b. 96 20b. and similarly ray dye [v 41 lib]. hdrimd ix 111 la. x 6 2c. to the injury of the rhythm. loc. sddlidraiiyd i 167 4b: probably in dsvavatyd i 53 5d. sdcyd iv 35 5b 5c. the original form but if so. *viii 48 10a. simid i 151 la 3d. viii 96 13c 17d. feminine nouns. x 3 7b. -yoh in accordance with the metre as follows : -yd masculine nouns. feminine -l is the prevailing form in the Rigveda. viii 43 14c. drjunyoh *x 85 13d. rdyid) is required in i 129 9a 10a. brahmaitid viii 6 33a. sdkhid i 53 7c. 151 iii.stems 87 The two forms are frequently sdcid 'kartd pitdrd ytivdnd sdcyd 'kartd camasdm devapdnam iv 35 5ab. gdtumdtyd vii 54 3b. viii 71 4b. : ' ' : (ii) : The text has -yd. as -id . x 3 7b. On the other hand the text regularly gives mati as instru(iii) mental of mati. -ioh ardnioh *iii 29 2a. dual forms of rbdasl see below In the same cases in the declension of nouns in -i -id. tindnyd i 188 10a. iv 20 9a. sumatid v 25 3b (probably). somavalyd x 113 8b and always in . yuvatyoh vi*49 2b.

mahind. Ibman. dadhikrdvan. apparently such as were brought into use later. as krdtund. sutapdvne i 5 5a. reads mdtrvbh (A. iii. -an. 38) this value is given in the text. and sporadic instances of internal hiatus : may be regarded as (i) reduplicated participles in didhidna . -van following heavy an must regularly be restored for n before case-endings beginning with a vowel in the case of ddvdne and some other words (vii) syllables : 84. differentiation of gender to that : required. sy&ma which appear in the text.) i 65 4a. -ana. mddhvd (fem. -van which follow a light syllable. pipidnd . (v) sdhutl. sidma are all commoner than the forms desnd. mddhund. viz. Cf. krdtvd. where it is very possible that the forms in -una have replaced an early -ud the -vd form is the more common. Br. : is required in the dual forms of the nouns of pitaroh (pitrroh) for pitroh of the text. suddvne i 76 3d. -van last instance ii but one devdhutl seems a probable restoration. so that these words are better considered as belonging to the second group and in the same way bhadsvat and dadsvat are most conveniently treated in connexion with bhads and fdads. memiat. and so mdtardh and svasaroh but mdtrbh is found vii 3 9c. The masculine and neuter nouns have the instrumental either in -una or in -vd. iv 335. Restoration is exceptionally required in krdtud (text krdtvd) 151 ii whereas the consonantal value is exceptionally required in panvd (fern. : In the declension of nouns in -man. Restoration is seldom regularly required in isolated words and forms. 33 4b. The instrumental of mahimdn seems to have the three forms 151 ii. prdyistha.) *ix 5 10b. mddhvd. In the dative singular sahdsra-bdhue. are In the case of stems in exceptions. Ludwig. Although dayisnd. i 113 lOd. didiat. . The following forms and words remain. Rigveda. prdtarydvnah i 45 9a.) vii 25 lc. and those in -man. Consonantal value is found exceptionally in dhindmndm ix 88 4c. -an. mahnd. In the declension of nouns in -u there appears to be a similar which is found in the -i nouns. bhuriddvnah ii 27 17b. pipidna. dididna . Certain words. -man. Stems in In the cf. (ii) -at. grdvan. (vi) Restoration viz. except once in rdjand *x 97 22b. other words of the same type have the shorter forms more : often. In the feminine nouns it is usually necessary to restore -ud (or perhaps -uyd on the analogy of the adverbs such as sddhuyd) In the dual the restoration -uoh is regularly for -vd of the text. -itva. 151 ii. In v 1 1 3a the Taitt. vi 247). restoration is not required. svadhdvne v 32 lOd. sutapavnah viii 2 7c.88 *x 24 5b.prestha. memidna except didhydnd : gerunds in -tva. 142 iii b. bdhvolt (masc. except kartvd *i 161 3b. relationship. didhiat. mahimnd: see treated as infinitives ( A 140.

*x 124 Id. For dvih see 148 v. In a large number of instances the verses can be alternatively explained as consisting of fewer syllables than the normal number. and in decasyllabic hymns of the types described in hymn x 26. *x 166 3b jiakd always. as being the more common in these hymns: and the restoration of the full number of syllables is therefore not the catalectic suggested. viii 72 7b. dud etc. namely : jyd bow ': jid. where it is the participle of su stir. except in vi 69 8d. and at the same time be restricted to three or four groups of words. always siond. we must accept that as the simpler explanation. vydntah i 127 5f 5g. except in iv 27 3c. vidntu. dvddamkrti *i 164 12a. but dvd. But in 94 iii bed. and ( in particular as being of the decasyllabic or catalectic types 20. In the groups now to be dealt with the restorations amount ou an average to about one-third of the whole number of occurrences: but in a few individual words (see 140) the restored forms are more numerous than those in which the text is correct. : i ' : vyantu vii 19 6d. *x 87 lOd. vi above that decasyllabic variations are only common in about 50 hymns of the Rigveda. v 62 6d. except in *vi 28 3d. words. to some question. 83 3a. svargd heaven always suargd. a verse occurring in a decasyllabic or catalectic equally well be explained as decasyllabic or catalectic respectively. vl 'go': vidnti. are the more usual forms. the preference must always be given to the shorter verse-form. . It has been shewn in 94 iii. On the other hand svand is always to be read for suvdnd of the text from su press ': but not in vii 38 2d. jyok 'long': jiok. it is not credible that a license of metre should occur in a great number of hymns. iii 2 9d. and catalectic verses in only six. tredkti trayidhd. although Restorations belonging to the second group are all open in most cases the doubt is very slight. but dvyan iii 49 lb.Restorations not regularly required (iii) ' 89 various initial syllables. If then we find that the opportunity for restoration occurs largely outside these In other limits. : ' always srdyiui. 35 6a. 21).' syond soft ' : ' ' sreni and derivatives : * 141. 33 5a. dvau etc. *x 17 2d: dvddasd *vii 103 9a. iv 30 19a. a real doubt arises as to the particular case: all Where however hymn may such instances are enclosed in square brackets below. dvd two are found in i 28 2a.

1 46 2c.) v 16 3a. vi 50 15c. gndah (ace.) v 43 13c gndah-pdtih ii 38 10b. jnd: rtajiidah x 65 14b. pi. sd: ksetrasdMiti iv 38 lc. snd ghvtasndah viii 46 28b. vi 6 4b. rathesthdam viii 33 14a. v 46 8a. The form didm is evidently permissible. au have disyllabic value go: gdvah (gun.) [i 61 10c]. s. 181 8d gdvamdn ix 107 9a: gavapithiaya x *95 11a gdvajdta *vii 35 14d. vi 48 21a. -gopdah v 38 5c 5d. vi 47 24b (preferably). For ablative forms in -oat see below 151 i. with which Lanman compares the Zend ghendo but in v 46 8a gndah is metrically preferable. prd: kak$iaprdam viii 3 22b. ksdah iv 28 5d. [174 7b]. where they appear in the text in combination with a noun or verb termination. v 52 16b. 98 9c. gnd: gndam v 43 6b gndah (nom.90 142. and may be due to a real or fancied derivation from this stem.) i 61 12c. abjdam vii 34 16a. s. sthd: girisfhdam ix 85 10b. and occasionally in those in -% and -u. -prdd i 10 3b. For the ace.) v 59 2b (the only occurrence of this case). Examples. maghdvanoh For nidam. ix 32 lb. 180 5b. navajdah iv 6 3c. 15b. 67 6d. kdstliaam vii 93 3c are metrically probable. This restoration is frequently required in the nom. Closely allied With the above are the following : of the text appear to : : . vi 45 gdah (ace. viii 46 32d 32e tanupdam viii 71 13d. f. 127 2d. maghdvan: maghdvanah (gen. x 23 6d. are as follows ksd ksdam i 67 5a. 4 21b. arranged in the order of the stems. rathaprdam viii 74 10a. The forms [kdfthaah i 63 5c]. -gopda i 120 7c.. s. since diauh often occurs in the nominative and dydam is (ii) ' ' ' : therefore only suggested when it is metrically preferable. viii 68 13b: pdnthaah viii cases. in the declension of radical stems in -a. m. da: draviuoddah vii 16 11a. ace. 62 4b. x 31 9a . . all : : . nail: ndvdh (nom. gopdah viii 31 13b. Grassmann [68 4a] suggests the readings gandm etc. yd: evaydah [v 41 16b]. The examples go: are : gdam : *x 59 10a ix 87 7d. *x 53 5b.' and path are frequently disyllabic. vakmnextMah v 19 5e.' div 'sky. viii 39 6c. 72 2c. . pd: gopda viii 25 la. in which o. 151 4d. i 127 6g 6h. : path Some of the corresponding cases of go 'cow. v 10 Id. jd jdaspdtih vii 38 6a. -a. sing. or with noun suffixes such as -as and -istha. of (i) numbers. tanupdd vii 66 3a paswpdah iv 6 4 c. giristhdah ix 18 la. : : : : . of div sky we may read either didm or dydam. iv i i 7c. pathesthdam v 50 3c paruthdah *x 97 10a. div: dydam path: pdnthaam 31 13c. Radical stems of nouns Restoration is frequently required in radical steins in -a. vdam see below 151 i. viii 89 5d. viii 1 2b.) [i 61 8a].

dhdyistlia *i 170 5b. bhu 'be': bhuuh (bhuvah) vi 15 3a. of which the meaning is often obscure: i [122 la]. x 80 4a da-idm (text deydm) viii 1 5b. prd fill prdah vi 46 5d. 46 lb. bhd shine yd 'go': yaasi vi 12 6c. 61 17d]. especially in the conjugation of the verb bhu 'be. yaati ii 30 lc. ydaidam x 40 la. jyestha: jydyistha i 100 4c. v 29 8c. In the case more probable (iii) of the nominative dyauh the restoration diauh seems 148 iv. pa drink pdah iv 20 4c.) vi 65 3c: or inaghdownah. 155 la. v 18 4b. see u\ maghdvanih (nom. 86 4a. daat 63 9c]. viii 23 23b. : (iv) Restoration -I. [vi da 'give': [daam x 49 la]. [x 148 4b]. 127 2b. ydanti i 37 13a. 10a. iv 43 4a. vi 45 31b. yaatdh i 141 8a. didiatam iii 27 15c.Radical stems of verbs (gen. -u : dl 'shine': dldie iii 55 3b. pdantah ix 98 8b. hit 'call': dhuam etc. vii 65 lc.is found with hiatus in the conjugation of verbs In similar forms from verbs in -I and -u iy. 46 19d. viii 92 la. i *24 12c *13a. dha-idm v 64 4b. 52 2d 4d. [167 8a]. bhuuta vi 50 15d. bhuutam viii 22 16c. ix 65 28c 29c 30c. *x 103 8a. x [50 4d. paantu iv 4 12d. x 122 8a juhue *x 149 5b. iii 56 4b. iv 41 3a.. [v 33 5b] ya-iydm (text yaydm) v 64 3b. vii 93 Id. ii 18 8c. dhi 'ponder': didhie v 33 la. dhd 'put': dhdah vi 19 lOd. 56 la. pa 'protect': pdanti i 41 2b. anayitd ix 108 13b. abhuut viii 46 24d. bhdasi ii 2 2d. sihd 'stand' asthaat i 74 8c. . vii 16 3a. dJiaatam x 93 . dhd 'shake': ddvldhuat viii 60 13b (text ddvidhvat). Here may be included the occurrences of the form pdantam. prauayitdr [i 169 5b]. : * ' : ' ' : ' ' : . vi 50 4d. destha : ddyistha : dhktha viii 66 6d. viii 19 37d. x 23 lc.' which seem to fall in radical y : under (a) this heading. *120 la. 102 lib. du. 67 3d. viii 46 4c pdat iv 55 5c. *x 149 2c: bhuut i 173 8c. : for the instances see Similarly a. 78 5a. pi. iv vi 48 21e. fern. is often required in superlative formations from : stems in -a or as follows 1 2c. 74 4b. Verbs in ' ' -a : g(L go : gaat viii 5 39a. respectively are generally found in the text but a few cases remain. etc. nl 'lead': ndyisam [x 61 4d]. [dhaama i 122 12a]. M.ir-a. For dididna see uo - For didhidna see 140. 22 9a. ddah viii 2 15b.) v 86 3b: 168 i. prl 'please': prayitdrah [i 148 5d]. ndyin i 129 5d. nayitdr v 50 la 2a 5a. 48 9a. si 'sharpen': siat i 130 4b 4c. viii 23 4a. 97 3a. bhuutu i 94 12c. iv 6 9d. *x 88 la. : (6) Verbs in -i.

176 2b. srestha: srdyistka iv 1 6a. 81 9c. yestha ydyistha v [41 3a]. 47 2a. 185 9c. 39 4a 4b. bhadsvat 92 7a. x 37 8b. 32 15a 17a 19a. 28 lc = 2c = 3c. iv 7 3c. 31 2a 13 3a. similar . 51 6a. 175 4d. iydm esaam \\ amftdnaam gih urjdatn viustisu x 74 76 3a (Tristubh). even if it could be supposed that the large proportion of instances was not a sufficient proof of the reality of forms in -aam. 30 2a 11a lie 15b. 53 la 10a 11a. 46 la. 3 lib. 74 8b. 48 2d 3b 4d. 58 4d. vii 20 7b. 187 5c 6a 7b 8a. *158 lb. 35 2c. 16 Id 4d. d va rFijasa \\ Restoration of -aam is very commonly required at the end of (i) dimeter verses. 97 3a. 16 2a 4a. 88 la. 26 9a 9b. 37 3a 9a 13c 15a. 9 5b 5c. i devdnaaiii yd in mdnah 31 15c (dimeter). 7 lc 3b 6c. 46 2b 4a 5a 7a. 134 6c 6d. 9 3c 4b 6d. 173 10a (Decas. 41 3b. 3a 3b 4c 5a. viii 84 la. x 63 16a. sndads. 37 3a. vi 16 26a. 45 4c. 48 5a 5b. [i 174 5b]. 64 4c 5c. 44 9a. 76 2a.92 The genitive plural ending prestha: prdyistha i 167 10a. 36 5d. : i bhds: bhadh viii 1 28c. 143. 169 Id. v 9 2a. vi 33 lb. Somewhat 180 6c. 113 4a. 10 3a. 62 6a. 30 20a. 184 Id. 103 10a. 129 8c lie. 36 la lb. *133 2b *3b. 8 8a 8b. bhaasd vi 10 4b. 86 2b 6c. 17 2c 4a 4b 5b. 39 3d. 84 2d. v 43 7c. cittir Such examples are the following apdam dame vlsvdyuh \\ : vispardhaso \\ -nardam iid sdmsaih stotfndam vivdci maghonaam visvesaam suddnavah \\ 67 10a (Dvipada Viraj). vii 32 10a. 68 5c. 23 4d]. 158 6c. 65 6d. 4 3a 3b. 97 4a. 24 3b. 186 3a. Tristubh). of the instances are at the end of lyric or : dimeter verse. srdyisthavarcas v 65 2a. viii 19 34c (Usnih). 84 la. v 82 lc. 25 14b. 25 lc. vi 45 29b (catal. as follows i 1 8a. : . vii 56 6a. 61 3a 14a. vi 26 8b. dim. 38 10a 12b. : The restoration of aa is also required in the stems dds and (v) bhds (Greek <dws) as follows dds: dadsvat i 48 Id. vi 51 10a. 75 3a 4a. and others fall near the caesura in trimeter verse but a sufficient number of examples remain to shew that a metrical explanation is inadequate. 11 lc. iv 2 7d. x 144 2b. vii34 14b. 86 4a. 18 3b 5d 5e. *191 4c *13a *13b. 74 7a 7b 8a. 52 7c 9c 15a. 93 4d. : The following case seems very desiid : similar dayisiid vi 63 8a. [63 Id]. srdywthasocis viii 19 4b. 181 la. [x 22 4a 5a. is 186 10c vii vdata for vdta 'wind': ix 97 52c . The restoration of is of all declensions A great number -aam for -dm in the genitive plural required in about one-third of the occurrences. verse).i [63 7c]. 40 6d . 127 2e 7d 7e 8d lOg. 67 2c 5b. la (Jagati). 128 7d. 7 9a 9c. 23 lib. iii 10 lb 4a. v 6 7c 7d. 66 3a 3b 4c 5b. [68 2a]. 5 2a 2b. x 3 lc. 88 6c. 56 5a 5b.

5 13a 37b 37e. 41 10b. *88 6c. 78 lc 6a. 93 16b ix 1 3c 33a. 19 8d 37c. 31 10a 10b 14b. 1 . 50 7b]. 40 2c. 128 5e. 28 5b. 61 10a. 181 la. 18 16b. 69 2a 2b 2c 18b. 93 5c 9b 9dl4d. 23 8a. 67 10a. 1 . 1 at end of lyric verse. 47 5a. v 10 4d. 4a.The 87 2d 3e. 94 12c. . 63 la. 187 6b. 105 5a. 101 6c. 66 5b. . 74 la 3a 3a. -bhyah following heavy frequently required in the sufMany of the syllables. . . 103 4a. 46 16a 16a 60 9c. 22 lc 13c 13d. 115 5d. 3 13a. 68 7d. 74 13d. [93 5a] 12b *103 8a. 17 73 3a. 108 13a 13c. 76 la. 2 18c. 32 15a 15b 19c 20a. 18 3c vi 16 lb 18b. 149 4c. 14a 18a. 75 8a. 94 la lc 8a. viii 1 4b 30b. 52 3c. 58 2a. 64 10b. 122 3b 4c 10b. 71 13a. 33 8a 8b. iii 51 10b. 48 8a 12c 12c 52 14b 15b. 29 6b. 65 23b. 6 28a 28b 44a 46c 47b. 133 7c. 173 9b] ii 4 2a. *112 lb *2b . 1 1 1 . vi vii genitive plural ending 93 16 la. 64 3a 3b 3c 4b. *103 5a. [49 2b. restoration of 1 1 . 56 3a 3b. 15 10a. 102 2a 2c . 53 la lb 8d.16c = 17c. 3b]. 7 9b. 20 14b. x 20 f2a. 107 8b. 144. 94 8b. 53 3a. vii 9 2c. 134 lc Id. (ii) Elsewhere in Dimeter verse occasions for restoration are much it is less i common. [33 3d]. 83 7a 7b. 95 3c. 23 7a. 1 1 1 1 . 101 6c 29 Id. 103 10c. but required as follows : 30 5a. [22 10c 12c 14b]. 60 17d. 96 lb 2d. 65 10a 10b 11a. 24 19a 29b. 26 lb. 15 13c. 31 15c. fixes Syllabic restoration is -bhyam. 102 10a 10b. v "[33 2b. [174 10b 10c]. 1 1 1 1 1 . 45 9b 10b 16b 5b. 68 4c 4d 6c. 70 la 12a. [63 10b]. ix 15 5c. 69 2d 3c. 74 2d viii 5 37d. 17 14b. 104 5a. 60 13c. 41 Id 5a 7b. 66 3b. 43 5c. 33 12d. 24 2c 3a. 16 2c 7c la. ix 108 10c 13b lc 12c. 45 3b 8c]. In Trimeter verse occasions for restoration occur chiefly (iii) either at the end of verses in lyric metres (not in Jagati). 24 Id 2c 4c]. 1 . 31 2c. 101 6b. 44 2b. *97 8a*8c. 8 12b 18c. 92 3b 30b. 32 lid 25d. 41 2b. 71 13a. 20 3b 8a 14c 14c 23 2c 24 23c 25 23c. Many of the latter are found in hymns in decasyllabic Tristubh metre. 48 2d 8b 16 2d 7d. 103 6b 7d 10a. 64 4d 5d. . 126 6d. [167 lOd" 168 2c 5c. 50 5a. 23 7b 25a 25b. 187 lb. 23 12a. 7 14a 15a. 27 15b. 102 la 4c. 74 6b. 1 . 67 9a 13a 14a. 70 lc x [23 111 2a. 87 3c 25 3c. x 9 5a 5b. 29 2a. 176 3b. 98 6a. 18 la 2a 2b 16a. 47 9c. [20 3d. 71 lib 13b 15d. 25 14a 23a. 17 14d. 53 iv 1 20a 20b. . [68 2b 2c 4b 7c] 14c 14c 19 7b 33c 34b' 34c 36b viii 1 21c. 45 29a 29c 31a." 129 8b lib 134 6a 6b. . 24 4b 17a 24a. 46 22c. 67 13a. 10b 56 lc 5c. 75 4c. 46 12b. 19 37e. i [61 5d 12d]. 53 5a. [21 8b] vi 3 8c. 2 34c. . or in the earlier part of lyric and Tristubh verses. 67 1 32 5a 7a 1 lie 56 24b. 47 2a. 40 3a. 92 Id 3a 6a. 34 3a 5a. . 51 5a. 25 7a. . *57 3c. 93 3a 3b 4c 13b. [77 4a]. 45 12a. 26 16a 18b. 64 27a 27b. and these may also be the few cases in which interpreted as decasyllabic Tristubh verses : -aam would produce in such hymns an irregular rhythm are omitted from the list. 52 4a 4b 5b. 20 10a. iv9 2c. . 61 lib. 13 3d. 22 13a. [148 4b]. 127 8b] 10f. 10 6a. 20 3a 11a 14b 14d. 46 lc 2b 18b 18c 18d 19a 22d 26b 29c. 45 2a 7c 28a. 54 7b. 93 31b 33b. 1 1 1 . 39 2b 4c 5c 6a 6b. 78 Id 3b lb] lb 3d 4a. 46 8b 9b. 176 la. [169 lc. 133 lf = 2f = 3f. 19 3d. 186 5c 7d. . 25 23b.

6d. 134 4f 2 4g\ 139 7b2 7b 7c 7d\ 142 6b. . 100 10b. . ^d . 47 7d. syavd. . 1 . *104 25a. 48 11a. 33 3b 1 34 2b 42 2b. 11 3c 19 5a 97 lb . 59 2b 2c 61 12b x *13 4a. 1 1 1 . -ud are found as shewn in 151 ii. . 9 16d\ 32 5b 28 2b. ix 94 5a. 53 16a . 1 ] . 60 5d 6a. *163 4b 1 3 49 3d. 124 13a. 3 tdnua ix 14 4b. 78 lc always in yddua. iv 1 2a 3a. . 41 3b. 39 10a. . *104 4c *20c *20d 25d viii 1 17d'. 122 5b. 54 2d 48 4d vii 59 4b x *87 2d. (b) [x 61 20c]. 38 8a. where a heavy syllable is produced by external or internal Sandhi dsva i 175 4d. 1 1 1 1 1 . *135 4b. vi 8 5a 2 2 68 4c. pi. .94 Other suffixes and comparatively occurrences are at the end of dimeter verses: but (unlike the gen. 85 4d. vi 47 23a. -tva: ndvavdstua vi 20 lie. 4 20a. 1 . satdviii 4 19a.v 61 5c. sdmbhrta. common Resolution required as follows . After a light syllable -dhuam appears to be found as shewn in 151 ii. vi 19 3d. : : . 1 .viii 34 12b. . 76 5a2 5b 2 5c 2 5d 2 77 7b. 62 9b 20c. -va (rare): urud v 30 4d. 34 13a 23d. 116 16b. -sva (i) i 37 14c. 85 8c 90 3b 102 3d. *161 8d. ix 67 10a. 1 1 1 . is lyric verse. 109 21a. 65 2b 3b 20b 99 7b. rjrd. *145 3b *3d 156 4c *158 lc *4b'. . 63 3a 1 88 5b. . 1 tdnua iii 31 2a. 58 2a. 64 7d. 39 4c. jird. 103 6b. *163 3a *5b\ -bhidm : i . 32 5d.i 119 lb. : : 11a. 113 16a. 34 6b2 43 t6c 2 2 2 64 lb 80 2c. 1 J . 59 6b 62 6a. iii 2 *191 9c ii 1 lb lc 2 2 iv 26 4b. 1 1 1 1 1 . Much more commonly in dsva. and a few The forms. aghd. 1 .i 100 16b 17c. . 32 2b 41 12a. raksastud viii (a) For the gerundives in -tva see 140 ii. chiefly those which contain v. 146 5c 2 5d. 18 13b. and -vi. 112 5a 21c 2 122 8c. 1 . 157 3b. . 32 14b. 30 3b. 188 10b. . kdnud viii 2 40b.i 116 6b: ajd. 2 tdnua x 93 15b. . vi 55 3b 4a. 52 5c 5d\ 54 53 4c 2 54 lc 2 2a2 2d 2 5a 2 2 2 2 9a 9b . . iii 41 8c. ix 3 9b. -dhvai. . *60 6a 1 63 2c. . . . -vdms (ii) (perfect participle). . l : 20 3a 1 136 lb2 lc 1 6a 2 viii 6 36b 1 x *14 lie. . 145. 55 5b 2 58 6b2 -bhiah: i 7 10b 13 lib. 6d 3 11a. -dhve. 1 . 37 7a. . 2 2 4 5b. - at end of Jagati verse. 23a. 105 9a. 30 4a 36 8a. . v 5 lib 1 lie. . 1 at end of dimeter verse. x 62 8c. 117 18d. . 15 17d'. -aam) they are rare in in Jagati. petua vii 18 17b. . Numerous other initial syllables. v 55 6a viii 1 bis.i 138 4b 4c. 131 5d 132 4f. 46 9d\ 47 3d. require resolution from time to time provided that a heavy syllable precedes. . . ii 41 17c. x 22 5a. vii 32 26b. 64 2c 2 70 2d. 3 170 iii. Resolution of v in the suffixes -tva (substantives). -va. 34 7b. 57 3d. . 39 6a. 31 4a 166 12d. After a light syllable -ua. *85 17a 2 2 2 2 94 lb 2 *7a *7b *7c *7d 110 4d 5d. . principal cases are as follows Resolution of v in the endings -dhvam. 5 8b 1 23 17a. 11 Id. urdhud iv 6 2c. 36 7a. 1 1 1 1 . viii 35 19b 20b 21b. 8 41 lb lc 44 30a 30b. ix 87 8b. *14 15c 2 *15 7c.

and occasionally 198. vi 21 6d. 32 4b. 104 9a. 43 2c. 82 3a. rdtridh *x 129 2b. ix 61 23b.Feminine case-endings : 95 -vdrhs ddsudms i 407c. ddsivdn). 91 11a. 48 3a. tdsidh ii 13 la. devidh iv 1 17b. vispdtniai 96 3d. *91 Id ix 8 9a. *113 2b. gen. light syllable: *29 4b. dudrd sudn 'dog' *x 14 10a *lla. : . (also *86 4c initially). v 21 3c. 76 7a. r tvl. Thus we find anid for anyd in *vii 33 8c (nd'niena). *le. *161 5a. -yd -tvd very frequently in the Atharvaveda avdsia i 140 10c. sudhd iii 32 15a. 138 4a. (d) -vi (rare): mddhul iv 43 5c. 71 4c. chiefly in the later part of (vi) the Rigveda. *x 72 8d. *85 22b. 54 5d. the author's Historical Vedic Grammar. 52 6c vii 16 4a 4d . *x 85 25c 45a. vdmsam). x 75 3a. 91 8d (nd 'nidm): aridh v 33 2d (prd 'ri4h) and iv 51 2c. 187 2b. See parts of the Rigveda. samcdksid i 165 12c. ddsudrhs i 127 la. susyantidh *v 78 5b. ix 55 4b. 58 6a. *158 5a. 23 17a. vii 'M (c) 92 3a. 130 Id. iv 2 8d. -sya d 'diah (da (iii) bind ') ii 13 9a dsiat (2 as) iv 30 20b. ii 32 2d iii 47 84 6c. -yah. 7c. abhitid iv 32 10c. [vi 63 8c]. *57 6b 3c. vii 67 7b. 26 3a. In the feminine case-endings light syllables. So in i 129 llg it seems necessary to restore raksohdtiam tud vaso. of i is 'hian). 160 5b. ksesidntah (ksi 'dwell') ii 4 3b. in -idh drdtidh (i) iLslndrdnidh *x 59 10c. bhtimidh After a i 80 4a. suadanta ii 1 14c. *162 15a. This resolution is found side by side with the forms with y in all also in the Atharvaveda. *137 7c. staid i 109 2b. vicdksid viii 13 30d. viii 25 14c. abhicdksid *viii 1 34c. 96 4a 4b 4c 4d. . 65 6a. prthividh i 39 *3c. . mVhudrks i 114 3b. x *14 3d (svdhd 'nie). 44 16b. 36 5a . 140 6c. 47 8a. 71 2d. x 2 7a. 45 14a. 44 lOd . 52 4a. dpasia iii 24 lb drianti (dr) viii 16 6a. fem. and occurs after both heavy and -yai. Hid x 66 14c. . viii 39 4e . -tvd. -yd. 45 5d. viii 3 22 3a. There are several other instances. similarly didvd iii 46 5a. : (ii) dat. 11a. . ix 8 8b. fa ii 11 5c we should probably restore ftastambhudrhsam (text tastahh 4c. . nicdyid i 105 18c. fiwruih ' x 68 12b. jiraticdksid i 124 8b. abl. urvidh i 146 2c. jivantidh *v 78 9d. viii 57 4d. -yam. Resolution of y v in the gerunds in -tyd. abl. 33 8c. 102 18a. iii 26 la. suddanti viii 50 5c. 85 4d. iv *18 lib. ahian ix 26 3a : (medhdyd 146. 150 la (SV. \ Resolution of the enclitic tvd is regularly dependent upon a (v) preceding heavy syllable it occurs i 40 lb. 57 4b. abhigtirid ii 37 3c. in -iai : devdhiitiai ii viii *ii 32 After a light syllable srutiai 2 7b. : Resolution of y in the verbal suffixes -ya. ix 79 3a 3b. 31 2a. viii 36 4a. 100 lb. 15 12a. iii 8 3b. 60 16a. where resolution of y v seems to be due to a preceding heavy syllable. siendh iv 26 |7a. restoration very frequently necessary. ~tv%: iwstui v 53 14c. 30 3a. x 144 5a. fem. (iv) -tyd: apitia *ii . 6 18a 20a. x 111 3a.

jagmidtam vi 50 10a. i 35 8b. 95 5d 53 7b. 89 7d. 19 25b 26d x 20 8b. 50 6d. pdrusnidm v 52 9a . vi 12 5d. for which see 151 ii. The instances of id. 6 8c.viii 1 19d. yajddkiai viii 39 lb. : 1 1 . namely in the optative suffix -yd and in the conjugational suffix -u -nu.96 and in the older The optative suffix infinitive forms irddhiai i 134 2d. aJidma iv 4 14b. Similarly ksnavdmi 12c] . : Resolved values are found in the suffix -u -nu as follows tanu2d 6b. 40 Id 41 3d 4a 4d 5b. 77 3b. 66 13d 1 87 7b. v 4 8c. 34 24d 25c. viii 40 60 lc. 52 la. 1 1 1 . 2 after a light syllable. 66 12a *121 lOd. x *14 6d.in the optative suffix are as follows : asidm v 64 3a. In two parts of the verbal system we find occasional 147. resolution of y v respectively.v 30 6b hinu. v 64 7a. 50 9c vii 34 21b1 viii 14 lc 2c. (ii) sidm. 79 la. sayddhiai ii 17 6c. 1 . 65 5b 70 2c. dhanu. 27 16d. . 38 lOd. 30 18c. vii 1 20c. After a light syllable prthiviam viii 49 7b. occurring after both heavy and light syllables. . msahidt v 7 lOd lOe ( 169 iii). iv 41 6c vi 44 23b. 97 3c 16d 17c 18d 19b 20c 26b. 4 4d 7b. sunu. . sdcidm x 61 lb . . 94 13c 15d. 53 Hd 15c. 48 2a.iii 20 lc. 29 2b.vi 47 17c. jahndvidm iii 58 6b devidm ii 41 17b. 31 Id. etc. 24 iii 1 21d. 42 17a. Resolution of the -nu suffix before vowels in the is most often found Soma hymns which we have attributed to the normal period. vididtam viii 5 37b. 41 10b. 56 24d. 22c. ix 75 5a. 132 2d2 148 3c. vi 19 13b. 47 12d 13b. ii 2 12a. 54 2c. dhunu. 17 9d. i . except in the word sidma. There is one occurrence of a loc. [v 41 151 i. 37 4c. .iii 53 4d. 50 7b. 121 15d. gamidh i 187 7d. 12d. 73 8b. 28 2a. 38 2c 64 lie. rdhifima (i) iv 10 Id. In the optative resolution appears to be most common in the bardic hymns. 54 56 4d. 98 la. namely svasrudm. 92 4c. 17 7a. 52 5a. 55 iv 8 5a. 168 5c. 39 7b 15b. 31 13d. 18 3d 19 7d 2 71 2b 6d. : dsiknidm iv 17 15a. vdhadhiai x 22 5b. 13b. 62 9d. 73 9c. uhi&the iv 56 6c. . 1 1 J . 45 lie. So too in the noun dhdnvan and its derivatives. . *48 12d *13d. . srnu. 22 12d. viii 20 25a. 1 . . 120 7c iii 1 23c 38 9b. 1 1 . 150 1 1 3c. being more frequent than sydma. fem. x 49 9b. . 36 12c. 8b. in -udm. ucchdntidm loc. 105 4b. pupuridh v 6 9c. 16 21d. 14 3c. 55 lOd. 180 9d 12d 13a 18 8d. bhumidm i 39 4b. . . sdmidm x 31 lOd. in which it becomes increasingly common in the normal and cretic periods. 11 lb *164 40b. 1 beginning the verse. 51 lie. . 59 3d. 51 15d. 126 4d 131 6d 7b. sidma: i *24 15d. 70 5b2 . . (iii) i 17 6c 38 4b 4c. in -idm (iii) 184 lb. ix 86 38d. iv 45 . risayddhiai i 129 8d. fem.

As the suffix and -va occasionally becomes -ya regularly becomes -ia after a heavy -ua. 37 11a 1 52 la 88 5a. [41 9b ]. 75 4a. 1 2 . . . *x 16 3b dvih twice duih i 53 9a.Initial consonant-groups 148. 56 2d *6c 1 *7c\ 74 2d 2 *83 5d'. . 54 3d. x 30 lie. (iii) : tvdm 'thou': tudya i 53 lie. . 44 22c. . 53 8d J . *72 tudt J 9c. iii 14 6a . (iv) div 'sky': iii 1 diaidi iv 1 1 i 6d ! . is suggested : whereas 7 . . . 44 4a 1 16a. 1 1 1 8 5c 1 52 10a 1 65 3b. . 58 2a. 4 7b 5 7a 7a (viii) 1 ' ' : sua- i 1 8c. . lOd . 79 9a 1 x 8 4d. 1 . *59 3b *85 lb *7c' *10b\ 132 la 6b didm (or dydam 142 ii) i 52 lid 67 5b2 141 8b 174 3b 1 ii 17 5d. . viii 2 7c 11 10c 18 13c 44 12b. . 27 4a. 105 10c. . 1 . syllable. immediately after an early caesura. 102 4a 2 ix 85 8d 2 *x 84 4c. 187 4a2 v 32 3a 4a 5a 6a' 8a 33 10a 2 vi 2 9a. . and a very so large proportion of the instances are in decasyllabic hymns. 1 . viii 1 30c. *98 2b 1 112 9c. . *85 42d. : The more important instances are (i) jyti 'strength': jidydn i iii 38 5a. in which such restoration is probable but if we consider the two : words for which this step most readily suggests itself. ' ' 1 : . 21 3c viii 24 x *18 13a. 1 . 2 vi 3 8b. 1 1 1 1 1 : . vi 36 1 vii 7 5c . 1 1 . 149. . iv 22 3d. paramajidh 2 2 . 2 after a light syllable. . . 23 4a. that an alternative metrical interpretation a. 6 3a 1 . 82 6b. ii 2 iii 10 2d v 4 6b. . . 75 5c 94 14a. the case of indra the restoration is almost always suggested in a single position. ix 111 2 2a. 4 9b. 1 1 1 . 07 In several words occasional resolution of y and v is is the second element in an initial The syllabic value does not appear to be consonant-group. it is clear that it is attended with special difficulty. viz. 138 6a. [v 45 la]. . ii 4 21 Id 1 22 4b 1 . 1 1 . ix 95 5b. vii 3 3c. 41 lc. 58 7b. 48 3c. . 1 . there is no antecedent improbability in the restoration of -ara for -ra. *28 2b 40 5b. . 5b 50 13d 51 5a 68 4a ix 86 9b *100 12b x 36 2a 1 44 8b 45 8d. 97 14c . 1 1 1 i 1 . -ana for -na under There are in fact a number of passages the same circumstances. . *x 71 4c *lla *llc. vii 36 5b. tid. 178 la si 'bind': vi sia- iii 1 1 1 . (vi) 'that': sid. x[26 2a ]. 118 lc *124 2d 144 3b2 svd his 2 lid. 56 lc\ . *75 19a. 91 8d. 80 2a. 119 8b . 1 1 . x 50 5a. . vii 8 2a (after fu (vii) syd 160). . vi 30 4b. . lib 1 12a 15b. 1 1 . 89 4b. v 41 lid 1 . . *18 lb. required where one of these dependent upon the position in the verse. . . 1 . indra and In rudrd. vi 66 2b *x 120 3b 1 (v) 1 . viii 7 26c 20 6a. 1 beginning the verse. iv 11 3a 1 3b 1 3c 4a 4c 4d x v 25 7c2 1 vi 7 3a 1 3b 1 13 la1 31 2a 1 vii 5 3a 7d 11 lb. 1 . . . 1 1 1 . . (ii) tva 'several': tua- 147 2c *iv 18 2d bis.vi 65 la. . . 84 6a. 64 5c 87 4c. vii 75 viii 10 3a 3a 64 5a 67 la 1 92 7a 94 10a 1 11a 12a .i 30 22a. . iv 16 10c . .

viii 90 6c . instances which can be alternative! y exseparately. (c) The extreme rarity of the form in dimeter verse. 100 17a. srotra *x 85 lie (dimeter). -rd. 31 3a iv 16 21a. 29 la. 21 2a 8a]. 23 3a. mantra [x 50 4d 6d] mandrd vii 9 2c rdstrd iv 42 la. the : verses can most readily be explained as decasyllabic. 46 2d 4a. [63 la to 9a]. 61 22a. vdsrd [x 99 lb]. : . 96 20a. where there are twelve instances. and (iii) generally in decasyllabic hymns: omdtrd [x 50 5b]. of occurrences in which restoration is suggested is relatively so small in all words of this class that we can hardly not found. On number think of such forms as indara (indrra) or rudard as the primitive forms. indaraya would be Hence the first real evidence are never found in trimeter cadence. 105 4a in Dimeter verse (probably). 105 10c] . 14a. 23 la]. viii 66 5c. ix 88 la. vi [20 3b [21 10a]. vii 37 4a . if they occur in decasyllabic hymns. x 22 7a. and can most easily be explained as an imitation of that of indra it is almost confined to the normal and cretic periods. indra (a) after early caesura: i 33 14a. [178 la to 4a]. 25 lb. candrd [i 135 4f]. x 92 9a. ii . 50 11a. 114 4a 6b 8c rudrd *vi 28 7d v 46 2c. *x 180 3a. 33 13 times . 17 la. 22 3a 10a. . [19 3a 8c. especially as even in this group such convenient forms as indarasya. . 129 la 4a 7f 8b].98 Supposed resolution of the rudrd the usual condition is suffix -ra for restoration. indram&dana vii 92 4a. [24 la 10b]. the ceding heavy syllable. 36 la. It will therefore be necessary to consider these two words As before. 126 5c. ddtrd iv 38 la. 177 5a. 165 3a. seems to shew clearly that the word was always a disy liable The immense majority of the occurrences to the poets of those groups. vii 84 2a . 169 la to 5a. 49 la 6a 7a]. [33 vii 19 2a 6a. raudra [x 61 la 15a]. 22 la. ii 17 8d. . 174 la to 10a]. and is generally found after a late caesura. patrd i 121 la. *vii 35 6c. vi 44 16a. *139 4c. -ri. 23 5a. bhrdtrd iv 23 6b. v [33 4a 5a]. 53 11a. . 36 5d. 26 7b. 39 4d. (ii) i 100 5a. 47 9a. 25 la. . are enclosed in square brackets. almost always after the caesura. 130 2a. . 24 2b. The restoration is necessary. iv 16 15a . [104 2a 8a. tuvimdtrd viii 81 '?c (dimeter) tvdstrd x 76 3c . [51 13d] : . [49 11a. 133 6a. being in Group III. 19 la 2b. [148 la 2a 4a 5a]. a pre- in the case of the other hand. 97 x [22 la 2a +lla 12a 13a 15a. 173 5a 7c 10b 11a 13a. (b) 24a . 186 6c. la 8c. [167 la 10a. 130 10c [ii vi 26 la . . 35 2a 3a]. : suffixes -ra. (i) 89 6a. for resolution occurs in the Vasistha hymns. 30 lb 4a. 20 2a. where the decasyllabic variation prevails. [vi 20 7d] . 11a 13a. 21 5a 6a. [x 50 6c. plained as decasyllabic verses. in spite of the great number of Indra hymns in Groups I and II in dimeter and lyric metres. iii 53 elsewhere in Trimeter verse i 62 3a. 29 3a. 20 4a 5b iii 32 12a. 20 2a] . 93 11a]. iv 3 6d . viii 2 7a. 44 15a. 50 2b 3a 4a.

48 la (dimeter). 72 la 2b 3d. 66 28a. vii 39 4d. 183 3c 5d.Resolution of nasatya and aditya 99 sutrdtrd [vi 68 7a]. in the later periods at least. heavy syllable favours the view that. caritraat *viii 48 5c. hotrd [i 122 9d. 41 2a. 182 4d 8a. sometimes On the whole it seems probable that both forms to the other form. After a light syllable only rjrd see 151 ii. vii 60 9a. . x 22 6c . 113 4a. -nas may be read with resolution in the case (iv) of cyautnd vi 47 2c. 47 9a . and therefore open But the fact that almost all possible occurrences follow a to question. 71 4c. iv 37 8b 8 15a. 101 7c. Restoration in place of a long vowel or diphthong nouns antdriksaat *x 158 lb . occur. 10c. 132 3b]. but there is little reason to consider that either is a primitive form. viii 18 2b. 9 15a. stotrd [x 105 la]. iv 2 netrii 92 7a. 180 9d. (i) : -at in abl. nutana in the text in agreement with the metre shews at least the possibility of the former resolution. 57 lc 4b. : i 34 9d. resolution occasionally took place. x 73 4b and indra-ndsatid viii 26 8b. [x 50 4c]. sing. 85 la 9a. . x 61 Both examples are doubtful: but the occurrence of lie. viii 5 32c 4d 7b 10a] x 39 3c. 25 10b. the normal number of syllables may also be obtained by the restoration ndsatid and the rhythm is sometimes favourable to one. 49 5c. 35c. 151. 9 9a. 129 7a]. 118 4d ii 41 lib. Similarly bhaak viii 80 8a. the instances that follow restoration is sufficiently probable to make it undesirable to treat the verses concerned as metrically irregular. syllabic restoration In several forms and a large number of individual words is suggested by the metre in one or more In passages. -dr in 2 3 sing. sadhdsthaat viii 11 7b. but not with sufficient frequency to constitute proof. [173 4d]. pardkdat i 30 21b. Most of the instances follow an early caesura. of the s aorist dksaah ix 18 lb. nutna. 184 lc 3b 5d iv 3 6c. vi 11 lc. 116 2c 9a 10a lib ndasatyd is metrically preferable (i) 13a 14b 16c 17d 19b 20c 22c 23b. 19 16b. 54 20c. 107 9a 9b. In connexion with these restorations it seems best to 150. 26 2b. of -a : : 72 . ii ii 1 13a. except that the occasions for restoration are relatively fewer : (i) iii aadityd dditid is metrically preferable: i *i 24 15c. [63 lc 7a. 44 4b v 74 2b. The case of ddityd is exactly similar. In all cases consider the proposed restoration of ndasatyd for nasatya. 27 6c. Half of the instances are in decasyllabic hymns. 98 2d 3a. : : . pardstaat vi 54 10a. 117 Id 6b lid 13d 23c. bhadh i 128 2g. 70 6a. (ii) is metrically preferable: 45 lb. [vi 20 7c. (ii) viii ndsatid is metrically preferable i 20 3a. 29 la. viii 5 31a. : The suffixes -na. and reknah i 121 5c. 43 7d. : .

ix 67 25c. yds 'tire': aydas [i 167 4a. sdkhiuh viii 69 7d. gdvia i 131 3b.) ix 89 3a.(proper name): a-urvabhrguvdt viii 102 4a. *x 90 2b. ksdvadah vi 1 7 1 2a. viii 5 35b. viii 98 8a. *x 95 9b. sravasid ii 19 7b. ix 76 lc. 49 6b 7d] ddasl [ii 20 7b. ksoni (proper name ?): ksavani [i 173 7c]. vi 20 10d]. -diotmdnam 4 la]. suggested in pur 'burg': puuh i 189 2c. viii 11 10b ninid i 95 4a. ddsa 'barbarian': ddasa [i 104 2c].'conquer': jigiuh viii 19 18c. vi 14 3b. pi. sing. v 64 2c. arid iv 1 7d : . *52 4a. v. 20 18d. stu 'praise': stavisam i 187 la. viii 74 14a. vdr 'protector': vdah [x 93 3b]. aurvd. iv 3 16b.100 Miscellaneous resolutions -dr in monosyllabic noun-stems: vdr 'water': vdah iv 19 4b. hsbdah stream ksnu 'wipe out': ksnavdmi x 23 2d. jidydn. [173 5b]. x 105 lc. -yah (ending of i. arksd (proper name): a-arkse viii 68 16b. 93 Id. vii 8 la. sura 'lord': Mura i 122 lOd. -dhuam after a light syllable mahimdn 'power': mahind iv 2 lc. [x 23 2d. su-diotma i 141 12a. cf. x 3 4b. vdm 'you': vaam iv 42 9a. -yuh (ending of i. *v 40 7a. . *145 6c.nouns) -va (noun-suffix) after a light syllable: dsua i 175 4d. astaut vii 42 6b. -vah in the u declension: mddhuah (gen. vi 'bird': vaydh (gen. vasavid (see 135) vi 60 lc. vr 'choose': avari iv 55 5b. *v 30 15a. dyut 'shine': diutandm vi 15 4a. sprdh 'strive': spuurdhdn vi 67 9a.) i 130 3b. 49 2a. nu 'now': the restoration nuu is an alternative to nu u. tva: tud after a light syllable: iii 23 4a. mam 'me': mdam iv 42 5a 5b. var. 80 3d. 74 10c lOd. the reading 124. sura 'sun': suura or suria i 71 9b. 1 ' ' : . 148 i. viii 5 29a. r. [122 15d. vdtdpya(!): vaatdpya i 121 8d. 149 3c]. sugdvia *i 162 22a. (nom. dtia ii 34 3a 13c. *x 108 11a. diotanah viii 29 2a. vi 48 17d. vi 59 2a. *x 167 4c. ix 93 5b. viii 46 15a. viii 7 2b 14b. vi 48 17c. x 74 3d. dura 'far': duurd iv 20 la. bhavia i 129 6a. 85 5c. ndvia 'young' v 29 15b. 52 17d. mdria i 77 3a. vi 63 7a. s. vi 66 5c]. 51 2d. vi 26 5c. tu 'be strong': samtdvituat iv 40 4c. reknah: rdyiknah vii 40 2c. -ya (noun-suffix) exceptionally after light syllables: ajurid vi 17 13b. vii 15 14c. ji. pddid *x 102 7d.) ix 24 7b.nouns): aridh iv 48 lb. x 48 lc. spuurdhdse v 64 4d. [ii i 87 2a. x 5 lc . (ii) Resolution of y. 26 24 c.suffix) exceptionally after light syllables kulid x 43 7b. *vii 50 la. rjrd: : : -ya (verb-suffix) after light syllables fjiantah vi 37 2b 3c. rjard i 117 14d. ix 111 lc. x 160 5d initially *i 191 10f=llf=12f. x 132 3c. rjarasua i 100 16b. : : . -yd (noun.

cf. sing. 2a bis: but in v 74 la read kua for kd. tvdm: tdva for te *x 10 13 b. vi 19 12c. x 64 10c. : : : : *135 7c. 115 7a. 168 lc read rodasoh. ii 1 7d. s'mdsru: smdsdru v 7 7c. 100 la. i 67 5a. 25 2b. iva: ukthd : : ddma (few : : for for : pascattdt 4b. v 61 la (1). read sdnubhih prthivi: prthvi : . in the following passages i 33 10a.' vii 66 8b 8c. cfs&ra restore dsakrdh vii 43 5b. bhisdj for abhisdj. iyam: yam i 186 11a. svaraj below. vi 15 mdrta: mdrtia i 38 4b. 99 3d. ix 19 6b: cf. ii 27 1 2b 1 3a. x 168 Id. masc. 63 5b. vii 4 3b. 86 4d. vii 3 4a. 3c. *191 6a. x 96 8a. sumdd. [x 61 8b] and sumddudhnlh i 73 6b. kvd: ku i 38 2c.): svasrudm *x 85 46b. 45 26b. ix 94 viii 11 4b. 8c. 129. which appears in the text ix 22 5a: otherwise rodasioh or perhaps rbdasioh ( 173 ii). 12 4c. sdhlyas for sdhiyase i 71 4c read sdhyase. -vdm -ve (iii) (loc. uttardttdt uttardt x 36 14 b. sing. *x 129 6b. : : : vii 88 3c. 93 7b rdthas or a locative rathe in ratha-ulha x 148 3d. also metrically advantageous. iii 14 Id. Cf. sahasdvan: read sahdvan i 91 23b. viii 79 The restoration is prthvi is rightly given in the text vi 12 5b. svar. sucidan read sucidantah vii 4 2c. viii 2 30b. dame read ddma d devdndm read devan pas'cdt x 36 14a. ucdtha v 4 7a. dtmd apud *x 103 vi 162 19a. : : : : : : : . pauruseya: paurusyena *x 87 16a. va see above 128. stu for stdvante read staunte vi 26 7c svd for svdsya read suydvasd ii 4 4a cf. rdtha: the stem rdthas is to be restored in rdthasas pdtih v 50 5b. 71 7c. dsveya (in Aufrecht's text only): dsvyena *x 87 16b. 73 8a.Miscellaneous restorations -vd (noun-suffix) after a light syllable: *i 101 12b. *173 4a. mahimdn: for mahna read mahimnd vi 66 5c. brhdcehravas: brhdcchravdh x 66 la. for stavdma stavd ii 11 6b. -va (instrum. ii 23 7b. fern. avaydj avayajah i 173 12b. svan for svdnit read svauista ii 4 6b cf. vi 51 2b. vii 34 7b. 118 6a. : (dat. snu: both forms are found in the Rigveda. smdd: both forms are found in the Rigveda. bhiyds: bhydsam ii 28 6a. rodasi: for rodasyoh in i 151 3a. but uncertain. viii 1 26c. 'bhisrindnn ix 97 43c cf.): sahdsrabdhue Miscellaneous syllabic restorations abhi loss of initial a is not improbable in 'bhydvartim vi 27 5b. svar for asvdrstdm read asvaristam ii 11 7b. sdnu. it is advantageous to read sumdd vii 3 8d. [vi 24 7d]. sddh: for sadhantah read sddhdyantah x 74 3c.) : krdtud iv 28 viii 3c. brhaspdti: brdhmanaspdti *x 103 8a. 77 2c . x 24 2a. but smdd *i 162 7a. bhisd. [x 23 lc]. *v 83 9d. 16 46a x 63 13a.

103 5a (vajam).e. The text passages : of the Rigveda rightly records such forms in the following (i. ii 11 6a [indra). 31c (ddha). (dsvdh). In the Rigveda we find dvandva compounds in process of formation.sthdvirasya ghfsveh (text tuvidyumndsya sthdvirasya) mitrdvdrun-d. dmrtdndm ca mdrtidnam ca) i 26 9b utd medhid.mdrtidnaam (i. svdvas: read su-dvan in the Pp. 152.sukrdvarcdh (text pdvakdvarcdh sukrdvarcdh) 5b also). mitratithim utd medhidtithim) i 36 17c. pataydn- mandaydtsakham i pataydtsakham utd mandaydt- sakham) mitrd- 4 7c dmrta. 46 17c (visvdmanusdm).e. *ix 113 3b {tarn) *x 10 12a (tanvdm).vayiyoh (text prayiyoh vayiyoh) viii prdti-ardhim devd. *85 34a (kdtukam etdd). the appearance in the Rigveda text of glosses and of resolved forms and derivatives from dvandva compounds. viii 46 28a. *87 13c (mdnyoh).Hithim (i. x 140 2a. *128 9c (adUytih). In other passages the Samhita editor has given both words in full. *x 15 14c: adhirdjd *x 128 9d. 98 2d. ix 13 6a. v 19 5b {vdyund) vi 10 lb (agnim) .102 : Glosses and compounds svardj for svardt read svardjdh vii 82 2a. vi 47 12a 13c. namely. 9a (indra). prayati yajiie [agninij adhvare dadhidhvam viii 40 9a. . vi 10 lb. . as the : throw out some word as having been added as an aid words in brackets in the following sdm bhdsmand [vdyund] veviddnah v 19 5b. . . 86 3a. 20 2a (dgnim). text. and therefore not only declined irregularly.devasya mahnd x 1 5c (so probably in (text devasya devasya) pavdkd. but also (ii) used as bases for derivatives in lax combination. (i) Where to a verse includes too many syllables it is a ready method verses to intelligence or devotion. vii 41 6d 82 2a (vdm) viii 29 5b (sucih). 40 6c (vdsu). 78 8a (devdh).e. 93 14c (pdnca).utd dhdrmavantd vi 18 12a 35 13a 19 37a viii (text mitravdrunavantd utd dhdrmavantd) utd me prayi. iv 1 2a (ague). *121 7c (ekah). /i 'incite': for hiydnd read hydnd viii 49 5b. thereby destroying the metre so that we need to restore as follows : : prd tuvidyumnd. Lastly we may advert to two kinds of syllabic restoration which border upon the sphere of ordinary textual criticism. purvis te [indra] upamdtayah Probable instances of glosses or other additions to the text are to be found in i 129 llg (jijanat). 150 4a {devdh). 39 6b (apiciam). *164 5c (pdpdh).

but tends on the whole to become rarer in the later parts of the Rigveda. v r n before certain vocalic case-endings. the verbal suffix -nv from -nu. The use of indra after the caesura decasyllabic verse changed to a conventional use of a form indara in the same position after the disappearance of that metre and other reso: in and -na were occasionally employed. the optative suffix ya-. -yam. we . namely Groups I III. In the earliest parts of the Rigveda.. and semivowels is rare and uncertain but there are indications that the text is not entirely to be relied upon either : in these points or in its reckoning of the number of syllables in other ways in a number of individual words. and in the genitive plural ending. the noun-suffixes containing v. and restricted to the same position in following lutions of the suffixes -ra the verse. chiefly after the caesura but the nature of the : resolution is uncertain in each case. except that the fuller form of the genitive plural remains much longer in common use in dimeter cadence than the other forms. As these results confirm those reached in the last chapter. -ya. diphthongs. tvdin. adityd were also occasionally employed. Outside the limits described the resolution of long vowels. 103 history of the changes discussed in this chapter appears to be as follows. and the enclitic tva. Resolved forms of the words ndsatya. the same model. an The additional syllable must frequently be restored in the flexion of monosyllabic root-stems. resolved forms following heavy syllables are fairly parts of the Rigveda. etc. of some -ya suffixes after light syllables. -tv%. -bhyah. Generally. and of y In the noun-endings in -bhyam. and of all after heavy syllables. the gerunds in -tva.History of syllabic change 153. and in some initial syllables resolution is not uncommon. Syllabic restoration of accented y v. the consideration of these changes supports the view that Groups I-III together form the earliest part of the in Rigveda. of v in tvdm. and that otherwise the general arrangement described 57-63 holds good. In the later groups this restoration decreases rapidly in frequency. is the rule throughout the Rigveda: but in the popular Rigveda exceptions increase in number. common in all In the noun-endings in -yai. -yah.

By taking the numbers for the periods instead of the groups we are able to adopt a larger unit of bulk. which lays stress only on the resolution of and v. has therefore failed to grasp the true proportions of the phenomena. : .// We take up the different sections in the order just given. contracted even in classical Sanskrit consequently it appears to be matter of accident that such forms as bhut dhve appear in our text side by side with bhuvat dhuve. In the later periods the use of the ending -aam in dimeter cadence has clearly become a convention. Resolution of long vowels and diphthongs. correIII. (i) In the forms derived from radical root-stems the progress of (a) contraction is very uniform. Where the figures are reckoned separately for trimeter and dimeter verses we take 5000 of each as the standard of bulk. yet for our present purpose it must be considered a real archaism. The whole number of variations with which we have to deal is : nearly 2600. and 200 resolutions of r and n. and so far exercises a misleading influence on modern criticism. in -aah Forms from stems quite commonly (Historical Vedic Grammar. thus bringing them into line with the discussion of Sandhi. 800 The irregular resolutions of y and v. (ii) the syllabic value of y and v is the more common. Groups IV and V. The consideration of the Table in 155 will enable us to justify these conclusions in detail. as a rule. which corresponds very fairly with the number of Sandhi variations ( 131): of these nearly 900 are resolutions of long vowels and diphthongs. 175). 700 irregular consonantal values of y and v. and no longer corresponds to the ordinary pronunciation but in the preVedic period the two forms must have been at least equally common and though the longer form is not necessarily the earlier. For the genitive plural in -am we find in Groups I to III that (b) rather more than half of the occurrences of -aam are at the end of dimeter verse but this lends no great support to the view that the measurement had its origin in the constraining of catalectic dimeter verses into a normal measurement. namely 5000 trimeter verses = 7500 dimeter verses. but is not complete even in the popular Rigveda indeed the Atharvaveda still shews the nom. in -i and -it are not. without departing much from the actual facts. The parallel of the Greek flexional forms with hiatus goes to shew that we have on the whole a genuine historical developement. : : : Irregular consonantal values of y and v. the presumption must be that this is the original form and it would then appear that any tendency to contraction worked more slowly if a heavy Where : . Indian tradition. . sing. The Table is drawn out so as to call in many attention to those features which seem to be instructive particulars the number of instances is so few that they give no practical guidance.104 are entitled to History of syllabic change make a further provisional division of the hymns (i) hymns of the archaic period. of the bardic period into sponding to Groups I corresponding to and (ii) hymns of the strophic period.

It is remarkable that contraction seems to be equally common in all parts of the Rigveda proper. and the gerund ending -tud is quite common in the Atharvaveda though not found in any part of the Rigveda. -ua. In the case of the verbal. : (iii) Irregular resolution of y and v. the order of developement being sUra. and it may be that a few such cases are included in this section. The the Rigveda.suffix -u would seem to be primitive. -yah. and am therefore inclined to distrust the text. and that the order of developement was tu. siama and the word tua shew progressive vocalisation in the other A : cases it is diminishing. In the case of surya the doublet sura (p. -bhiah. as mdrta by the side of mdrtia it is therefore likely enough that &&ra stood originally in many passages where surya is now read. where we find original -i -u retained before dissimilar vowels throughout the Rigveda proper. large proportion of all the instances follow heavy syllables. the values -bhidm seem to be distinctly increasing in frequency throughout the Rigveda. and due . the exceptions being almost always in cases where a light syllable precedes. usually written kva and rodasyoh. The feminine case-endings in -yai. such as tvdm and I find some difficulty in thinking that these forms were really siirya. Since single passages in the Rigveda have retained for us such old forms as ku and rodasoh. Although this change as a whole is of equal importance in all parts of it is not necessarily so in each separate class. and this may be the case also with the other forms. that the original v cases have been slightly influenced by the u cases. tvdm. the optative (6) in -yd and the verbal-suffix -nu before vowels shew only occasional in the optative forms (except siama) this is not found vocalisation later than the normal period. : With regard to the initial groups it is at least clear that such (c) forms as tudya and tudt have a different history from tvdm and tvdm. It is possible that a similar history underlies other -una. and the corresponding datives -ie. -ue may be concealed from us under the later forms -ind. it seems not improbable that an old nominative td may be concealed in many cases under the tvdm of the archaic period. syllable preceded. tudm. in -id. It follows that the form is of metrical origin. Similarly the frequency of such forms as the instrumental masc. -aye. but in single words which are particularly common. (iv) Resolution of r and n. It seems clear that none of the poets recognized in the abstract a word indara. -ave. The differentiation thus depending upon the previous syllable might clearly lead to the resolution of original y v where they follow heavy syllables. suria. history of the enclitic tva is the clearest proof of the real (a) occurrence of resolution of original v due to a preceding heavy syllable. forms. not only in these words as a whole. but there are so many instances at the beginning of the verse that it seems very The form doubtful whether this is really a cause of the resolution. 36) is sufficiently common in the Rigveda. -yam. established as common variants so early.History of syllabic change 105 This is quite in accordance with the rules of Sandhi. surya. In this case it seems probable in which the vowel values predominate.

The Homeric dialect differs just in the same way from classical Greek. Our investigations however shew that this heroic ' ' ' dialects. The occasions proper : this may for these rapidly diminish throughout the Rigveda be explained in individual cases either by the gradual disuse of archaic forms or by the increasing strictness of the metre. pi. since that condition is fulfilled in the great majority of cases in which the r an is vocalised rudara seems therefore to be due to the fancy of isolated poet. In the popular Rigveda the number increases. That the accent tends to preserve the individuality of a syllable The accent may 142. although of a different type. seem to be of the same artificial character. The doubtful cases included in this section number rather more than one. (Altindische Grammatik. syllable must have contributed to justify such a pronunciation. In view of the antiquity of the literary records in the two ' we may well believe that the resolved forms which characterize them are on the whole earlier than the corresponding contracted forms of the same languages in their classical periods.tenth of the whole. and the the case than is syllables therefore lighter and more musical. 135.106 History of syllabic change to the constraining of a common type of decasyllabic verse to fit in The preceding heavy recitation the standard of a full Tristubh verse. may have begun with words like naram. Wackernagel 45) and others. in classical Sanskrit. pura?n. heroic' has already been used ( 76) to describe the of the Rigveda proper. and perhaps also from appears from also help to account for a single syllable developing into two. on account of its resemblance to the Homeric dialect. shews us a dialect in which the vowels are relatively more 154. The text of frequent. apam in which that syllable bore originally the circumflex accent. the metre being in fact disordered. It does not fall within the scope of this book to trace the history of resolved forms according to the principles of comparative grammar. But it does not appear that the Rigveda makes any distinction in this respect between accented and unaccented syllables. number considered in this chapter. i . of -ra : (v) Miscellaneous restorations. and thus the resolution of -dm in the gen. in accordance with the martial vocabulary The term ' This term 85). That forms like pdanti and ydam are not necessarily older in themselves than pdnti and gam has been pointed out by J. when metrically restored. the Rigveda. dignity and vigour which distinguishes it (see be used to describe the Vedic dialect as a may appropriately whole. ' ' general principle admits of many exceptions. and a large proportion may be cases in which no restoration is really required. The resolutions of ndsatyd and ddityd.

Refer to 107 Table of the occasions for Syllabic Restoration. .History of syllabic change 155.

in accordance with the admit : principle of quantitative evidence explained in 9. a long syllable being one that contains a long vowel or a short vowel followed by two 156. metrical laws of the Rigveda recognize only the syllables. it must be because its quantity was regarded as indifferently long or short. and can only be discussed by the help of a criterion not directly dependent upon the text. and therefore no safe conclusion can be of exceptions drawn from the occasional appearance of a particular syllable in a position in which long or short quantity is generally required. it contains implicitly a complete commentary on the quantitative value of syllables: and any question that may be raised implies a doubt as to the correctness of the text. a symbol. such as those given in 31 47. and also denotes each consonant by a distinct symbol or (in consonant-groups) by a distinct element in consonants. Short syllables must on the whole be placed in positions which favour short quantity or are indifferent and long syllables in positions which : If any syllable is used favour long quantity or are indifferent. and (ii) the quantity of final long vowels before hiatus. we are also concerned (iii) with the value of final -o (Pp. The principal questions that arise concern (i) the quantity of variant final vowels. final vowels not uniformly written in the Samhita text.CHAPTER VI. All the rules which prescribe the quantities of syllables in different positions in the verse. that is. QUANTITATIVE RESTORATION. Incidentally iti). -o . The distinction between long and short Since the text distinguishes throughout the long vowels by special symbols. But the general use of a syllable which recurs with any frequency must necessarily reveal its quantity. indifferently in all positions in the verse.

probabilities will be a matter for individual judgment. fifth The chief positions which favour short quantity are (i) the and seventh places in dimeter verse. as is attempted in 175. until a broad view has been reached as to the metrical standards of the But as to the second possibility there poets in the Rigveda as a whole. or the editors of the text have introduced unsound As to the first alternative it is impossible to pronounce alterations. it is seldom that any real difficulty arises in words and forms which are of frequent occurrence. and those of the latter when they are of interest. however. in the third place except as described above. there are fairly numerous cases of forms and words which are somewhat rare in the Rigveda.Position the evidence of quantity loo The principle of evidence here asserted is precisely the same which has been followed in the discussions on Sandhi and on Syllabic It reduces to secondary importance all questions as to Restoration. exactly as in the preceding chapters. the second place in either kind of verse. and the eighth and tenth places in . whilst the readings of the former are noted throughout. and to an extent much less marked (iii) the (ii) second and fourth places in either kind of verse except as just described. in view of the results recorded in the last two chapters. if the second syllable is short. whether early or late and (less regularly) the first place after the caesura also. . Syllables containing short final vowels are. is no real room for doubt. and in the fifth place of trimeter verse before late caesura. Still it should be of value to indicate the probable causes of the most common errors of the Samhita text. the question is whether the poets have regularly used words in positions for which they are not well suited metrically. and generally accepted by critics of the Veda. Although the principle adopted leads to changes in the text which are in many cases considerable. and the seventh place in trimeter verse after an early caesura. For if the text does not correspond to the usage. eleventh) places in trimeter verse (ii) the second place after the caesura. and the ninth (and . and the third place. 176. if the third syllable is short. unless the fourth syllable is short. The quantity is indifferent in initial and final syllables. But. : : 157. and in which therefore the bulk of evidence is insufficient to establish a Tn such cases the decision between alternative rule as to general use. The place trimeter verse in chief positions which favour long quantity are (i) the sixth dimeter verse. the readings of the Samhita or Padapatha texts and accordingly. the general discussion of them will be reserved till the end of the chapter for only in cases in which the metrical evidence leaves reasonable doubt will it be justifiable to pay regard even to so ancient a tradition as that of the Samhita text. admitted .

and occurrences in Sandhi combination. one position before or after the caesura . 7. and may be denoted by 7g. and in trimeter verse by 2T 3T but in either case if the other syllable of the two is short a special position results favouring long These positions may be denoted 2d 3d 2t 3t respectively. which will be the subject of discussion later. these positions again favouring long quantity in a very marked : : way. and the same positions in trimeter verse by these numbers followed by A or B : according as the caesura is early or late the eighth to the tenth (or eleventh) place in trimeter verse by the respective numerals. quantity. of which use will be may be illustrated as follows : made throughout the chapter. These general rules. if the other syllable of the two is short. as in the fifth and sixth places in Trochaic Gayatrl. There are some other positions which are comparatively rare. Fin the fourth to the seventh place in dimeter verse by the numerals 4. : being unoccupied. are altogether omitted from consideration occurrences of long final vowels before hiatus are considered separately : ( 172-174). the seventh will usually be short ( 46) this The third place in the reopening of position may be denoted by 7a. In decasyllabic verses the syllables will be numbered to correspond with the Tristubh rhythm. follow directly from the description of the Vedic rhythms in It will however be of advantage 31 to 47. The second and third places in dimeter verse may be denoted by 2D 3D. Anustubh is very often long. since the most important questions have to do with final vowels and also to express the results by symbols which will readily lend themselves to the numerical calculations upon which we : have to rely. and to the eighth kind of verse when the third syllable place in trimeter verse. Similarly before a late caesura the fourth and fifth syllables may be denoted by 4b and 5b respectively. In 2T 3T 4A 5A 6A 7A 6B 7B 8 9 10 Fin i pari dydvd-prthivi jabhra urvi In 2T 61 8c dvitd In 2t U w 3T 4B 5B 8 9 10 Fin i vavre sandjd sdriih 4A 5A 6A 7a ^ 62 7a viii 8 9 10 11 Fin asydma tad ddityd juhvato havih 27 22c . to base them upon direct experiment applied to undisputed long and short final vowels.110 more freely Symbols denoting position than other short syllables to the second place in either is short. This system of symbols. with the exception of the last. : The various positions in the verse may be shortly denoted as follows initial and final syllables by the letters In. and Thus if after an early caesura the yet require to be discriminated. Occurrences before consonant-groups are denoted by Gr. 5. Occurrences in positions in which the metre is difficult to determine. sixth syllable is long. this position may be denoted by epic Anustubh is usually long ( 41) The seventh place in Trochaic Gayatrl or the semicadence of epic 3e. 6.

taken random from hymns which contain the same proportion of trimeter and dimeter verse as the whole Rigveda. nd titdcm viii is i 65 9a Fin putro nd bahup&yiam The interpretation of the symbols p.Normal In 2t 3t 4b distribution of final vowels 5B \\ I 1 1 IB 5 7 8 9 10 Fin svdsiti In 2i> apsu 3D 4 hafiiso 6. we find them distributed amongst the various possible positions as follows at : Table shewing the distribution of long and short final vowels. and then 1000 occurrences of undoubted short vowels taken in the same way. A . 27 22b in a tabular shewn form on 148. If now we take 1000 occurrences of undoubted long vowels.

as the In most cases however the same value of the final syllable is optional. -I (instr. pi. in which only those particulars are given which are practically required. -t% -ui . (-d. see be seen at once that the general distribution of the syllables concerned closely corresponds to the average for long 157 and the only doubt that arises is whether syllables as shewn in the few occurrences in group E are due to metrical or quantitative For the forms used in the it will 164 : irregularity. 3D are much more nearly equal in number to the short syllables which occur there. and 2 pers. -md. the inference to be drawn is that the quantity was regarded as. 1 pers. end. No precise inference is to be drawn from the distribution of final In the first place. 177 i. except when they occur at the end of the verse. p. vowels are more common than short. 158. nil in the combination nil cit. a late caesura. and smd. -a) (ii) in the 2 sing. From the table in text.. Long final vowels are regularly found in the following (i) in the perfect.112 If Variant final rowels proved long on the other hand a word is drawn by its grammatical function words occur very commonly before the caesura also. and those in the positions 3T. and in end (pronoun) and (v) in the following adverbs dccha. long final vowels as to the rhythm of the verse. evd. tdm vdjam citrdm rbhavo dada nah : There is one exception. visvdhd (including visvdhd).. 4. These various classes of words are illustrated in the Table in 164. Mid. sing. We may then infer from the table that if the final syllable is long. ghd. before will be most common before an early caesura : to the end of the verse.). although the text now and then gives short quantity. the occurrences if it is short. exceed the short syllables in a much larger proportion than this table indicates. optional. the occurrences so far elude the test. 2D. imperative of the root-aorist middle (svd) (iii) in the gerunds in -tyd. Further the final syllables which end with consonants are in a majority of instances made long by position. -yd (includ- forms : : : ing -dyd) : (iv) in : the case-endings -tari. symbols see . visvddhd.endings of the 1 pers. anaha For index to the *viii 48 5b (9). dthd. Where the results are intermediate between those which are to be expected for a long and a short vowel respectively. appear in every position with relatively greater frequency than the table suggests. and therefore in a given quantity of verse. to a greater or less extent. (i) The following are examples of the regular quantities viii 45 35a bibhdyd hi tudvatah a yahi susumd hi te viii 17 1a iv 36 9d. pi. purudhd. 148. Thus the long syllables in the positions 2t 2d.

Variant final vowels usually long is 113 In the disyllabic imperatives of the root aorist the final vowel (ii) always long. 18 18b. *x 71 9c (9). imperative in -a. of which the first only occurs in combination. iii 15 6a. 148 3a. trdsvd. ix 107 6c. viii 18 15b. *87 25d. in (i) and in the vowel in -d. (iii) -%. 43 passages certainly short in the following *x 97 5d *9b. *viii 100 4a. ii 23 Id. as in krsvd. 11. iii 51 lib. Also probably in x 116 i 102 3d. certainly short. There are a few exceptions. nidtsvd (mdtsud). *85 35c. : For index to the symbols see p. -van. but short quantity is a fairly frequent second person plural in -thd. 5. *x 9 9d. the hypothesis of long final vowels meets every metrical necessity. -a occurs in over 1 300 passages in 40 *i 23 23d. gha (text gha) in viii and visvdha (text -ha) i 25 12a. and -man). 63 22c. pi. vii 13 3c. 7d (6A). -td. vi 51 13c. -tand . 9. the neuter -u. end (text end) in vi 20 10b (9). viii 51 3c. *3d*4b. -ta) occurs over plural in -thd. -a (ii) the 159. these case-endings (iv) are rare in the Rigveda for instance -ul (locative) is only found in camici and tanul. *95 la. and -tha 5. : . 122 2a. in the following positions 42 4c. x 38 lb. 66 30c. 114 viii 44 2c. : sd yajndndm dtha hi sd tdsmai visah svaydm evd namante 13 3b In the following forms long quantity is regular in the variation (i) the Rigveda. 67 17c. yahi suno sahaso ydsya nu cit Exceptions are few and very doubtful but we find perhaps dtha in viii 10 5d (5). 8 . Examples of the use of the adverbs named iii are : iv 50 8c vi 44 18d indra surin krnuhl sma no ardhdm vi 18 lie. 102 3a 5c. *47 29d. i 56 2d. *x 18 lib. 1 30a (7 in irregular dimeter verse). vdmsvd as for : : instance (iii) nU no rayim tipa mdsvd nrvdntam The following are examples of the use sd pravolhfn parigdtyd dabhiteh td utsndyd rayim abhi prd tasthuh ix 93 5a. iii 32 12c. 19 34b. in -a (from stems in -cm. -td (-tha. (ii) The second person is : 800 times. almost all of which are in the popular Rigveda: -tya *x 97 21c (5). vi 15 16b. The imperative is : : 57 4d. : of the gerunds ii 15 4a ib. 61 23c. viii 32 7c. 7B. *103 5d. and But at any rate the second rarely except before consonant groups. 148. Except the instrumental forms in 41 and -I. *87 9a*10a*17d. -ya i 67 10b (9). ix 49 4b. 7. -thand. 5c. 144 7a. *130 7c (9). Examples are : : vi tdm duhre aryamd kartdrl sdcd i jathdre somam tanul sdho mdhah suarvatlr apd end jayema (v) ii 139 7f 16 2c v 2 lid.

namely malid ix 88 4a (6A). ix 67 27c.in ordinary compounds see probably be restored in viii 32 10c (2D). vi 20 10c (7B). and as an adverb *i 191 9b (2D). 62 18c. iii 33 12d. 6B). as in puru. ix 46 4b 4c. : t maksii and mithih have always long quantity. : For index to the symbols see p. 106 3b. -tana occur 151 times. x 61 13b (2t) and also as the first element in a dvandva derivative. vii 62 lb (2T). v 51 2c. 59 3d. as puru. puru. for which see 176 ii. The exceptions are all in the 5th place dimeter. 9. The occurrences in D may be taken to be those of the short vowels but as in all cases but one a consonant-group follows. passages. purii. aha. The adverbs may represent an instrumental singular or neut. puru-. viii 60 2b. 164) that we are likely generally to be in the right in assuming long quantity not only in occurrences in and B. 174 2b (7B). *15 6b. 16 7b (4) and in a For separated compound. Further purtt is found regularly in puru cit. In all these endings the occurrences in groups A and B so greatly outnumber those in D and E (see Table. -man. *161 7a. In trsu the long vowel must be restored throughout. 18 10b. but of these 100 (iii) The remaining are at the end of the verse and 22 in combination. iv 36 4b. Only in six passages are plural forms with short final vowel fairly certain. *x 13 4a. favour the long vowel. 72 13c.puruhutdh viii 2 32b (4). stidhih should 166 iv. -van). 2b. but also in those in C. and the forms may be singular. as uric vi 20 5c (2t). *23 19c*21a. though it is favoured in one or two passages. 3 13b. plural form the evidence is insufficient to establish a neuter singular in -u. *175 i 85 6b. 172 3c. 47 . This implies a very considerable departure from the practice of the text. 132 6d. purd vi 44 14a (7B). balm *x 52 4b (6A). 160. 56 9a. plural forms in -a (from stems in -an. although the forms with long vowels do not often appear in the text. 32 17c. occur about 225 times in many instances the interpretation is uncertain. and x 91 7b (all 2t). 7. iv 7 11a. viii 4 lc (4 A). 80 9b. . and the metre is not affected. The Pada-patha has correctly kept the old value of the vowels in tri. 51 8b. (iv) -t The neuter and -ii . namely iv 37 7b. and elsewhere at the end of the verse.ydc chdmsam i 166 13b (2T). viii 18 15a. vii 34 6a. slrsti. Of the occurrences of the short vowels one third are in the popular A : Rigveda. 188 lb. 3 8b. x *19 lb.purubhujd v 73 lc (4. 148. The endings -thana. in ii 3 6a (2t) before hiatus. *x 175 2c. vii 3 4b. Disyllabic adverbs have regularly final -u in the Rigveda the particles u tii nd sd are used side by side with u tu nti tedt. ix 15 2a (4). with three exceptions. 93 15c 15d 15e. sdrma i 58 8b (9). 66 12b.114 -ta: 5. which is also (i) given in the text. i Adverbs in -u. -u vii ii 41 14c. 6) and vi 63 8a (2T. viii 18 21c. *97 19d. this is not certain. the occurrences that have weight being i 58 4c. *85 33b. 97 7d (2t).

vi 48 9d . vi 39 3c. B. iii 41 la. 3t. iv 1 10a. v 31 vi 8 la. i 59 6a 64 13a1 165 10b 2 13a. 111 7d. vii 19 9a. 4B 2 i 148 viii 51 7c. 46 lie 10. i . 62 6c. 16a: 2T 2 i 69 8c. 166 la. 100 6a. vi 23 7b Gr. 82 . 3T 7 la 5B 2 ii 11 3a 15a 16a 17a. : : . iv40 1a. : . x 72 viii 40 9e. (ii) In i 6 5a (2d) vl[it x 147 5b (2t). vi 59 la. (2T). 2 Text tit. As doubtful all occurrences of u in such forms as dtho. . v 2 7d. viii 21 10c. *x 175 4a 9. *18 3b. . 2t. 15a 4 172 3a 2 6\ vi 9 6d. *168 la: 66 3b. 46 28d. 37 5c. viii 27 14c. D . or -ah (171 v). 45 37a 5a. x 69 5d: 89 9a. *x 59 4b. 10\ i 56 2c. of the particle nd. 52 5b. ix 72 8a 9a. as it is not possible to 'determine with certainty the quantity of the third syllable. but also other cases in which -a u stands for final -ai. 63 10c. 28 9c. iii 49 2a 186 9a 4A2 i 32 la. vii 39 *x 27 7b. and distinct from the In the last class are included not only the well-known u of particle. Occurrences of u ml in the second and third places are included 2 here. 171 iv. 4A. . 7 4 5 3 167 iii. : . . . 6d. 22 8b. iii 31 9c. 163 vi. 1 : . iii 38 2d. 33 8d.^ 158. plural. iv 22 6a 169 101 10a 1 C 2D. : : : 1 6 Text 175 ii. Text nu. . Ft su must be left out of account. nil. . tu are : A 1 . ilpo are omitted. 55 18a. 3t2 i 72 (iii) 5 3d 2 i 80 ii 33 7d 8d3 167 9a 4 3 viii 12 4c i 17 8a. 61 lib (2t). -u 1 1 r> He (2t). : . . v 41 la. . -au. 97 38d. 7 times 85 35d 11. : . 4a. vi 29 5b. 3 3 iv 6 la iii 20 2c. 66 3b v 73 8a i . 66 9c : . The occurrences . : . i 132 le times 7. vi 18 8d lie. iii 55 18b. iv 6 7b. B. . *x 94 5d rid should probably be read as neut. as well as the occurrences of an element u which is of phonetic character. viii E 2 5. 2D C. The occurrences . v 32 9c. vi 17 9a. : : 1 Text tu. 3a. vi 18 3c. 1 . u lokd and that in the suffix -tava u for -tavai. 62 2a. i 177 4c. . For index to the symbols see p. viii 101 15c. . : . . 7. 51 2a. : . . . x 132 3a: vi 48 16b2 7A 2 vi 24 3c. In the case of the particle u u many actual and possible (iv) occurrences must be left out of account here. viii 27 9c. 2 2 v 30 3a 2 2T2 i 139 lb. . 85 6a. 25 lc 27 7b 51 3a.Adverbs in -ii. 2T 1 iii 36 9a. 22 5a. *i 164 32b. 15 5c. x 61 5b 7 2 D Before consonant-groups 29 51 times 3D 22 times. puni iv 31 8c (2D). 25 times 42 viii 77 la. v 41 13a. vi 8 lb. 8. ix 92 5a. the text giving u unless A. : The remaining occurrences are otherwise stated : as follows. . 22 5b. : 1 . 2 1 v 67 i 25 17a. viii 2 22a: 3T. . *x 88 6c. 148. The occurrences in decisive positions are: For nu cit see A. iv 20 4b. hiatus are considered separately ( 173 iv) but the somewhat frequent occurrences of u in the second place in the combinations u nil. viii 7 lie. nil are 4 4 4 viii 25 23c vi 27 3a . 107 24a: 2d 1 i 10 11a. 69 x 1 6a. Initial. ii 8 la 8 iii 58 6d. A. . 8. 2 2 E 2 5. 1 . x 79 6b. viii 21 7b iv 16 21a. except when Occurrences before hiatus has to be postulated before u (171 ii-iv). vii 66 5b. i 17 8a. iv 32 la. *x 6 A. 50 5a. : . .. of the particle tfi. 6B. 29 lc 2c 3c. 4b. 112 Id = 2d = 3d3 184 2a 4 3 3 3 3 vii 29 vi 15 la 12b. vi 47 22a. 13 14a. : .

iii 30 6a. 53 4a. 23 7c. 148. 2t'. 133 3a: ii 41 viii 18 3a 12a. 3t viii 103 lc3 : 3d. 32a 3 67 15a 3 93 21a . 9. 86 8c. The following are the occurrences 1 . iv 55 10a. 4 3d. vi 48 *x 179 2a. : . 105 3a. 40 la. v 67 5c. vi 15 la: iv 6 la. *114 4d: viii 3 20a. 1 . : of the particle s% su (v) 3 4 2 3 iv 31 3a viii 7 A. . iv 21 9c. -u. *59 4a. 5B. 4b 2 i 138 4a. v 55 7b: 5. 4. 45 8a. 1 1 . 42 3c. 178 5c. ix 61 13a3 2 B. *81 4a. 6 *x 16 14c. vii 26 3d. 1 . 4 Text u nu. 1 . . 165 14c. v 35 2d. 110 la Id. . i 30 2c. 22 13c 30 3b. iv 7 9d. x 77 4c. i 113 lie 7A. vi 51 10a. Adverbs in . . *167 lc: 21 9b. *v 44 14b. -u . ix 81 3c. x 126 3c 3d 6a: Final. v 29 13d. 2 2 2 vi 68 8a *x 94 8c 1 viii 66 10a 16 5a 17 5a2 2d. iv 22 lOd. 53 6c. . ii 35 10b. . *vi 74 4b2 C 2 2D. 1 . i 27 3 4a3 30 4a. l . 56 4c. viii 82 3a. Many due to restorations which of the appearances of nu su in the positions 3t. ix 45 4a ii 2 5b. viii 7 17b 22b. *vii 35 2a *3a *9c. 36 times. : . : . 94 3a. 139 If 7A. 82 6d. 1 . 4B. E. 40 11a. iii 31 11a. . . vi 45 33a. but see 3 171 iv. iv 5 3d. 1 1 . : . *x 86 13b. . 7B. 82 la 4 i 10 lie: 6 8 . 27 7b. . vii 96 la. 126 6a: viii 6 32b. 34 12a. . 6c. 92 7a: x 56 la 6. i 108 4b. . : 2 v 52 15a2 vi 54 2a. viii 61 12c! 8. *x 18 12a: vi 21 7b. 42 13a. 2T. 120 2c 66 10b. 184 2a. 91 18a. iii 4 10c. *164 26d*48b. 143 7d. 93 6a. 133 7a. iv 32 6a 3 iii 24 2c 7a. 38 2b. . i 34 6b. *x 59 4c: 10. i *179 5c. 5 6 2 1 Text su. 84 5c. iv 2 viii 18 22c. 2 Text mi for ni u. *v 83 vii *35 2b *7b *llc *12b. 2D. x 39 lc. . . iii 46 5c. . . 44 9a.. 82 6a. viii 24 7c. vi 22 2a. : . viii 66 13a: 3T. . 40 la. i 32 15c. 24 16a. 1 Text u. i 76 2b. *x 86 3c. 74 10c lOd. *85 47d: 4. . *i 93 la 3d. 8c: Gr. *164 16a *19a *19b. 73 17a. x 32 2d. 2t. viii 81 8a. vii 20 2 2 100 la 2 *viii 48 3d: 5b. ii 28 7d. i 113 4c. 1 . 33 9c. vi 25 lc. iv 55 4c. . ii 9 2a. 42 3a. 33 la.116 2c3 95 6a. *x 59 8f. 61 6a. D . 5a. : 3 viii 24 lc ix 110 la3 x *10 14a3 178 la3 3d. viii 1 19a. 154 times: 6A. *52 3a. 182 la. i 52 8a. *173 2d: iv 8 4a. 7. 139 4a. x 2 3c. 55 2a. x 75 la 2 *101 lid: v 30 7a. x *15 13b. i 84 3c. . 3 Text ii sil. . 93 15a. v 10 6a viii 2 13c. 10. 31 14d. included here. 39 5c. . 38 2a. 73 8a. viii 20 1 19a 3 61 5a 3 63 8a. . 61 5a. 22 7a. vi 9 6d. Text sit Occurrences of u sil in the second 198 ii. ii 18 2c *1614b: x 61 24b 4 B. 36 13a3 ii 41 7a vi 16 16a3 44 4a. iii 37 2a. 31 7a. ii 35 3c iv 2 4d. viii 23 5a. iii 36 2c. 35 times: 6A. i 36 13a. 43 4 vii 44 2a: 2b. 11. 124. 139 7a 8a 2 148 3c. 37 2a 2b. 27 Id. 37 2a. 40 2c. 173 12a. 1 . 26 la 15a 23b. 40 3c. vi 38 la. . 31 7a. ii 5 3b. C. . . Gr. . 169 5c. 34 times: 3D. 11. ii 35 15b. 45 9a. 26 4a 2 2T 2 i 76 3a. viii 27 3a. D. 6 39a. 54 la. viii 18 18a. 7. . 58 la. x *16 llc*13b. . iv 21 9d. ii 20 la . . Perhaps 7e. 6 times: E 2 5. x 160 2a. 35 6d. 20 19a. 62 6a 93 6c 4A. i 129 *v83 10a. v 63 *191 6d. 81 2 3T 5 29 times: 3D 5 22 times. *94 14c 1 12 9a 4A2 ii 34 15d. 20 4a. 10a. . iii 30 21d. . 32 19a: 2d1 vii 89 la v 62 2a. i 37 14c . iii 24 4c. 3d above are will be explained later in this chapter but : For index to the symbols see p. . 52 10b. and third places are 4 167 iii. ix 49 la: 9.. 77 2b iii 1 3c. 178 la: 4B 2 viii 24 lc: vii 29 2c. 13 25a. x 100 2a. 39 la.

2 vii 18 21d. 113 8a: viii 1 18a. viii 84 6a. vii 20 5c. viii 3 8c. 50 4b. 1 Text ddha. iv 2 15a. 36 11a. D . 26 8b. 1 . . . v 74 lb. 1 1 . iv 30 23c. iii 5 10c. and yadi 117 even apart from these the long vowels are amply justified for all four particles. 7. ii 41 3a 20b. 33 lc. 180 10a. l . iv 21 i 56 4a. 54 2d2 63 8d: 4A2 i 125 3a. vi 24 5a. 29 3d. ddha (ddhd). x 1 6a. 12 19d. 38 6d. A 1 B . ix 44 6a. iii 4 9a. 1 . i 42 6a. 3 Text ddha ydd. 2D 1 i 13 6c. ix 99 2c: ix 72 2c. . i 153 lc3 ii 28 9a. 35 2d. x 115 lc: yddl. viii 26 8c: 5. 31 3c. 184 la. . 83 9c. i B 1 . 180 7c. adha. *50 11a. *95 12d 3 *14c: 2d. *163 2d. . *29 6a. 5. . 4 ix70 2d. 74 5a. 2 Before 46 31a3 75 16c. i 13 2c. 3a. 27 5a. x 35 Id. 2t. 61 74 7a. 5 14c. 2 . iii 5 8b. 45 9a. 3 *xl35 5c: ix 67 22a. viii 22 6c. *104 15c. 10 2a. . and yddl the short and long vowels appear side by side in the text. 36 2d. v 22 2d. 38 lc. x 25 lc 2c. 6 7c. viii 101 10c. 26 5a. *v 40 6a 3 iv 1*7 10a. C. 6 2b. 74 times: vi 36 lb. 6 5a2 19 12c. v 16 4a. viii 1 16d. D . ii 13 8d. 27 4d* lc. 98 11c. ix 97 11a x 61 21a 23a *83 7b. 8c. 142 lb 8d. 6. 115 . . B'. 82 7b 1 2 Final 17a 1 94 8a. 27 3b. 129 llf. Perhaps 7e. *81 7b. 167 2c 3 iv 2 14a 16a. iii 38 2d. viii 22 lb:* 4 2 i 25 19b 28 8a. 182 13c. For index to the symbols see p. 55 la3 6a. 20 3d. 2 2 132 3a. C 2D. vi 42 3a. 2 20a. . . . 104 7a. x 55 5d: *vii 104 15a. . *11 4d. . v 48 4c. . 98 7a. 10\ *x 12 3a. (yddi) which closely follows the guidance of the metre. 30 2a. 56 7b. 93 12a. 27 14c. v 1 11a 2 i 113 17c. iv 31 6c. 3 Text ddha. 2 2 2 iii 31 6a* 13a. 44 la. 2T. . viii 5 18a. viii 1 10a 16a. iv 25 8. . 156 ii 17 4a. . vii 15 14a. ix 48 5a. v 52 3c: 2t. A 1 . . 2 2 vi 25 6b. *i 161 13d. ddha: lc. 31 2c. E 2 7B. 5b i . vi 30 3a. : A . 161. 61 22a 24a. viii 19 23a. 65 28b 1 *x 127 4a: 5B 2 39 times: 2 33 times: 3T 2 14 times: 3D 2 11 times. 44 9d. *vii 104 14a *15b iv 21 8d. i 57 2a 102 7d. . 8 times. . vi 56 6c: 2T. and it appears reasonable to restore the forms with long vowels in all occurrences in groups A and B. 92 29c. ii 31 4d. 4a. 82 4a. iv 7 2c. *18 9c. 8c. x 35 13a. 53 12a. 1 . *85 27d. 6 3c. vi 10 4c2 vii 4 169 6c 3 186 9c 3 ix 110 9a3 x 6 7a. 2D. 26 3a. *x 11 4c: 2d. but see 198 ii. 52 11a lib 16c. . . . vii 56 lb: 10. vii 66 12a. 90 3c v 85 8d. 136 4e. 2t. The occurrences (i) are as follows : iii 36 3d. 2 2 2 1 ix 14 3c 15 3c 1 i 11 3c ii 5 6a viii 13 21a 32 6a 61 10b A. . 7B. 2 v 79 la. ii 29 2d. In the adverbs adya {adyd). 8. . . C. 30 3b. 71 6a. 44 Id 3a. x* 129 7b *16l 2b x 61 25a. . *viii 100 2d. . consonant-groups. 66 vi 2 7a. . v 17 4c. 97 22a 6a. i 168 8d. 2 x 25 3c. 45 3d. *97 2c: 2T.adya. 55 5c. 148. Before groups. 86 6c. x 143 lc: 6 173 8d. . . 88 2a. as well as in the positions 21) and 4D. adyd: 1 v 51 13a. 6 A. . 30 10a. . yddi: (iii) iv 41 3c. E . 22 6c. x 30 2d 3c. . *i 93 2a 1 2 3 The text has adyd. . *18 13d. v 13 2b. i 115 6a. vi 18 14a. 1 . ix 86 46d. 114 lOd. (ii) ddha. 33 3d. 2 *10 14d. 101 9c. 52 lie. The text has adyd. vii 78 5a. x 92 14d. vi 1 2a. adyd. i 188 la. iv 44 3a v 56 lc.

. 14 s villi. . . 6 always long) occur as follows: A. viii 96 8c (CyA). -dhi. 9 times. E. -hi the occurrences in positions (iv) favouring short quantity greatly preponderate. but adyd and yddi. *x 10 10c 8. great change is required in the text. . -hi 2 ix 14 2a. 3 2 3 14 vii 1 3a ii 26 2a viii 60 6a 10. srnuhi. i 27 13c. 163. should perhaps be rearranged. i 27 3c 36 15a 129 9b *133 2c 13 9 93 31b 7 11a 46 25b 7 60 9a5 9d 5 78 10d For viddhi tu vii 31 4c viddhi tti is a probable reading. . 9 times: *i 161 8c. vi 25 6d. viii 53 4a (7B). and probably in the positions 2D and 4 also. vi 53 4b: vii 25 2c. ix 89 7c 10 lie 8 vii 1 13a viii 3 12a 12c 12 5 7 5 7 iii 45 lb iv 48 Id 7 viii 3 2d. . 10 3 dldihi. Forms in -uhi (except srnuhi. vi 25 3d'. n dhehi. jahi are exceptionally found as follows (ii) krdhi vi 47 lOd (6A). 2t. The imperative forms in -dhi -hi are regularly short in Rigveda: but srnudhi. 5. long final vowels. Final. and The quantity of krdhi.118 47 4c 1 : Imperatives in 4 1 v 74 . but the prevalence of occurrences in the position 7 compared with 8. . sonant-groups. Distinct evidence of quantity groups. chindhi. the forms in -ahi is uncertain. vii 82 8b. . . 13 7 yaJu. . *x 16 1 3c. s 7 ii 11 17d vi 2 B. 6 addhi. The Atharvaveda has almost always ddhd. The forms krdhi. both according to the metre and the text but they appear frequently in certain positions : For index to the symbols see p. 18 times. : . *161 lc*2a. i 54 9c: D. 178 3d. . There remain for consideration a number of final vowels of which the quantity is regularly short. A Of other forms in -dhi. The long vowel. Before C. . suggest that the vowel was generally regarded as short. jahf most commonly. sagdhi. 7A. but the long vowel should be read in all occurrences in groups and B. *104 15a. . in which the final vowel is (iii) 1 vi 45 14c. No A 162. srudhi have regularly. 4A. 148. 12 5 9 gagdht. see 173 iii. 1 The text has -ulv. 4 dhdihl. 3d. *129 7d. . 2 4b. i 129 11a 5 *164 40c 189 4a5 22d 4 10 5 9 *x 51 5a 1 96 10d". *x 95 6A. : . . however. may reasonably be restored in the occurrences in group A. for which otherwise these forms are equally suitable. . viln. D E *viii 100 3b. . 8. i 13 2c. iii 54 5b. 2 barbrhi. ix 91 4b. 2d. 7B. . . srudhi occur very commonly followed by (i) the word hdvam. 10 times. 44 9c. x 22 10c. 2 Text yddl. Before con5c. . The forms srnudhi. viii 65 5c 1 the positions 2t. : . pdhi. where the words cases. pRrdhi.jahi vi 44 17a (7B). . 1 ilu. but the value of the final syllable is the same in all The only exception is srudhi viii 66 12d (5). 3d. and also in Such occurrences are: A. . v 3 10b. : : . is as wanting. Text yddi. vi 22 4a. and the rather frequent occurrences before consonant-groups. but the restoration of a long vowel is very probable in the positions of group A. viii 45 22c: B.

the final vowel cannot ordinarily occur in any of the positions of group except 8. The special positions may be termed positions of protraction. 148 Id Also 2t. and the special to decide to endings described as capable of protraction. Protracted forms of the first person plural in -ma occur 28 times (i) in the eighth place. tena com- . viii 63 10c The corresponding dual form in -va is never protracted. 165 7c. iv 10 Id. and (iv) the (iii) the thematic and perfect imperative in -sva . caesura and at the end of the verse. protraction fall Endings capable of into two groups. instrumental of nouns in -ena. 2D. It is however clear that there are certain positions in the verse to which long final vowels are admitted more readily than other long syllables. 80 7b. as syllable is short. vii 27 5b\ 57 4c. and assume the final vowels then to be We A short. 94 lb Id 2d 3d 4a. 63 lOd 14d. and are usually found at the beginning of the verse: they appear therefore as protracted in the second place. as in the next section. 45 5b 6a. 1 : . The protracted group forms are (i) the first person plural (other than of perfects) in -ma. tdtra . 1 : 1 . 1 The text has -ma. 1 . The text only occasionally gives long quantity in the positions 2T.Short finals protracted in 119 It is difficult which the long quantity is otherwise favoured. as in indra-vat. i 73 9b. what extent these variations are the result of metrical or are due to some reminiscence of archaic long quantity laxity. 33 4a. . The first than two group consists of words which are usually of more syllables. as yena. The second group consists of words which are disyllables. x 2 2c. and those which belong to certain endings more readily than other long final vowels. the second and third persons singular of the perfect in -tha. The regular* short quantity is however easily demonstrated. 111 25 22c vii 20 8d x 2d. especially if the third The forms are (i) paroxytone adverbs in -tra. as ii 2 10b. being in the proportion of almost 50 in 1000. as nearly all the words have long penult. iii 33 10a. observe at once that. sumna-yu. or other phonetic cause. and in order to simplify the discussion we shall disregard the occurrences in these positions. (ii) and (iii) pounds disyllabic stems appearing as the prior elements of and derivatives. *58 2a *2b. ydtra. follows 8. Group I. v 3 6a 42 6b. 132 lb If. For index to the symbols see p. is (ii) -a . pronominal instrumental in -ena. lc. and are most commonly found before the The final syllable in this in the eighth place of trimeter verse. 1 viii . 148.

*x 42 lOd. ' 2d. . *x 72 7c. i 25 8c 9c. : 1 1 1 The text has -tra. *88 4b. 138 lc. 116 24a. 24 24a 4b. 84 5d. 56 3a. 2c. 19 20b. ix 70 2b. 17 10b. . . 1 are ii 17 6c. vi 23 8a. The instances are: *vii 103 2d. 1 . *71 2c. . 45 6d. have always -a. *163 4d. x *10 8d a *52 1 1 1 1 : . 165 2d . 8. 32 2c. x 112 3d. vi 49 5c 1 2c 1 iv 36 9c 43 6d\ 51 4c. Also once in the position 4b. *101 3d . viii 29 2a 2 ix 2 viii 61 70 9d. kena. vii 18 24b. vi 41 5c. 148. *x 180 2b: *viii 48 9b. 1 The text has -ena. vi 16 17a 17c. viii 15 12c: 8. 183 3c . -a. 97 44b 44d. *164 3d *21a *33d. Group II. The instances. 53 5c 6c. 8. largely accounted for as in the last section. 44 9c 9d. being in the proportion of 25 in 1000. i 31 5c. The pronominal instrumental yena. The paroxytone adverbs in -tra occur in the positions of (v) protraction 38 times. 4c. . *28 8d. *vii 55 7c. as paprd i 69 lb.) and -a (3 pers. *53 8c. i 122 9d 12c: 2d. iii 31 iv *18 5b. vi 16 3a viii 1 : . vi Protraction in the prior elements of compounds and deriva(vii) tives is not always readily recognized. 41 5b . 87 5c. 111 2d 4c. There is also an occurrence in the sixth place of epic Anustubh. *149 2a: 2d. x 5 5b. The instrumental forms of nouns in -ena are protracted in the (iv) 8th place 24 times. ixlll2e. 37 3b. viii 66 12c. 21 6c. 1 Id 1 *102 9c *114 7dS *121 5b 2d. 80 lc. Perfects of verbs in -a. 34 8c. ix 61 19b. iii 60 5b. 42 4c. x8 17c. and are not included above. 20 26b. i 22 4b. 2 1 The text has -a. vii 37 3a. iii 53 5b.120 Disyllables with protraction The perfect forms in -tha (2 pers. being in the proportion of 20 in 1000. *18 13d. *88 17a. and 7 times elsewhere. x 126 2c. being in the proportion of 35 in every 1000. The occurrences are: 1 . 90 4b. 67 6b. being in the proportion of 140 in every 1000 but this is largely accounted for by the grammatical function. 24 lOd iii 60 2t. 83 2a 2c 2d. 1 1 . iv 26 7c. are as follows. *vi 75 8c*llc. being in the proportion of 200 in The instances every 1000. vii 18 17b. viii 34 3a. i 117 2d . 6a. The instances are i 32 5c. 96 17c. 1 . *181 Id: 2t. ix 108 4a. v 1 5d. The occurrences are ii 33 2d. 2t. viii 3 9c 10a. -tha. vi 1 6b 9c. including a few instances cussion in For index to the symbols see p. 186 5c 1 v 54 15b. i 42 5c 50 viii 12 2a 2c.) are protracted (ii) in the 8th place 18 times. yena nu may be restored in i 72 8d. v 61 14b. 9c *139 4b. *180 lc. vii 1 4c. 145 la. 33 13b. i 115 v 41 iii 23 Id. 67 *vii 55 7c. : . viii 12 1 The text has -ena. 39 3d. jahci viii 45 37c. vii 1 24c . vi 32 2a. 111 2c. ii 23 13d. The thematic imperatives in -sva (with which we include (iii) perfect forms like dadhisvd) are protracted 9 times in the eighth place. 16 48c. viii 20 6c. te'na occur in (vi) positions of protraction 37 times. 6c. as many of the words are of comparatively rare occurrence. ix 80 4d. 6b. ii 9 3c. vettha. 117 21c. *164 30d *38b. 1 1 . according to the dis165-169 below. .

. 15 166 iv. : . 166 devavi. . and the position before an early caesura twice as common as before a late caesura. . . which indicates that protraction is not a mere theory of an editor. . but several shew a preference for All the remaining forms in those that favour short quantity. . their general use agree with those that have short final vowels ' ' : yet there remains a substantial difference. devdvant. 1 166 vii. iii. 5 3 abhi. 7 sramayii. . n vdjasdni. utb in the text (see appear remains in the distribution between the groups. only slightly from that of 159 appear ordinary long final vowels: the forms included in much more frequently. The table that follows hardly needs explanation so far as the first three groups are concerned. . vii. signs disappear. and similarly those in B than those in ~D . and indicates that the vowels capable of protraction were also not considered very suitable for the positions of short quantity. A 1 ' For index to the symbols see p. . 17 ib. ii 41 10b 9 ix 100 la2 *x 97 7a3 viii 38 2b . . vllii-. 8. Thus the use 164. although they can (as a rule) only appear in the eighth place The treatment of such forms as ydtra. iii 51 2c 58 6b' 4 10 3 3 M 10 3 vii 41 7a 72 lb 100 2d viii 20 2a 71 3b vi 1 7b 17 lid 9 15 2 2 15 8 15 110 lie' x *15 9a 36 8b . . . . tatrsand. indravant. The conclusions reached in the preceding sections are confirmed by a close examination of the statistics of the generally use of words and forms of the different groups. . 161) all these A In all the 'forms capable of protraction' ( 163) we observe that the long third syllable and the position before a late caesura Yet the are greatly favoured.B ix 74 5b 97 26a 49c 51a 104 2c 2 15 4 u 160 lb 13 91 15c 11 *101 Id *109 lc 40 5d 3 48 7b 76 4d 15 78 lb 2 13 2 17 2 9 vi 16 14c iv314a v 35 5c 61 13b vii 32 24a 2d. . . 3 prdti. but still only occasionally. . . mmnayii. tvesdratha. . . v 9 7a 8. 12 sukhdratha. In the group of adverbs ( 160. 166 vii. utd. 168 ii. . ib. and 10 6. . 2t. . . 8 vrtrahdn. 18 167 v. 166 vii. . . . 4 Srutdratha. 9 167 6 iii. . . .Quantity proved by in the positions : : statistics : 121 2 2 . 148. *-. and so far short vowels are indicated. 167 v.B ix 101 3a 6. . ahihdn. ytna comes out most ( 163). as ordinary short longer words appear ten times as often in group vowels. 30 lb 13 36 ed iv 42 9d 2d . Even when we include amongst the latter the forms that 171 iv). . sarvarathd. i 48 2a 5 6 8 i 31 7c viii 60 13c 14a 48 2a3 72 2c7 83 la3 12c3 10. The adverbs named in 160. ib. but corresponds to some special characteristic of the forms concerned. . 1 . ib. also the quantity of a following third syllable is indifferent. 168 ii. . 169 10 iii. in the positions of the forms included in 158 differs which favour short quantity. . ydsya (with kdsya and tdsya) which have much the same form and grammatical function. . 1 . . . a marked difference as dpo. 161 are found in all positions. . dsvavant. : . 13 168 168 14 iii. 16 sumnaydt. . In the first two the occurrences in group are far more numerous than those in group E. 166 vii. 3 3 2 ii 32 117 9d8 118 9b8 121 12d 9 122 7c' 8d 123 12a 140 13a 9 11 10 v 8 7b'. clearly when they are compared with other words like dpa. .

122 Table of the variant final voivels Distribution of variant final vowels in the Rigveda. .

In compounds of which the prior element is a noun (including adjectives. is long (i) if it is a case-form. as aksnayddruh. . voc. etddrs.words some effect in the direction of lengthening. vii 66 9ab. and occasionally (iv) in stems in -. as i 122 7a (7 A.v 62 3b (2T). and occasionally the pair varuna mitra in the same shape. and doubtful cases are rare. it is not easy to classify the occurrences or to treat them in a systematic way but it may be recognized that the quantity of the final vowel depends partly upon the historical character of the prior element. For index to the symbols see p. and numerals). it is possible that y r I s and m produce in certain groups of . mitrdvdruiid. The text : gives regularly indrdvisnu. text is in geueral agreement with the metre. Owing again to the complexity of the conditions. : besides any other influences that may be at work. in composition 123 In the quantity of the final vowels of the prior elements of compounds and derivatives the Samhita. pronouns. as is clear from the fact that it never stands at the end of Other Tristubh verse. but also in : the oblique cases. uMkhala. . On the other hand we always find the triplet varuua mitra aryaman with the singular vocative forms. or in any other position in which a is favoured. and perhaps (v) in stems in -an. and so forth restored. It will be convenient to consider separately: (i) compounds in which the prior element is a noun or numeral. 7A). The phonetic influence of v following will in each class claim special attention (iv) : in addition. 9): cf. Here must be included the dvandva duals. the final syllable 166. as vrsdyudh. 148. 66 6b (2d). as lopdmudrd. and the augment and the vowel of reduplication. (ii) compounds with prepositions and particles (iii) vowels preceding derivative suffixes . form. . The prior element (vi) has always a long vowel before -vrdh and perhaps in some other cases before v. purii- tdma. . and indrdvdyU must in all instances be sUrydmdsd. and (vii) is occasionally found in some other words in positions favouring long quantity. sendni (iii) in most proper names. and partly upon the phonetic character of the initial sound of the posterior element.Variant final* 165. jit nadttama. tudvrdha (ii) if it is a feminine stem in -a or -% as urvaraCtmdjur. are often found side by side with others that shew the compounds short final of a stem as sandju but sanajd. ace. as in evdvadd. the prior element having usually the long vowel not only in the nom. visvdmitra . . mitra . mitrd. vdruiiti i 151 6b restorations favoured by the metre are mitrd and iv 1 18d (2T. of that element y } . With the compounds of case-forms are included those of (i) Such adverbs which have flexional endings.

The long vowels may be connected with adverbial gatuvid. ix 101 lid (6).rsdgird. 104 4a (6). But puriwdmi has always the long vowel. In i 43 4a gdthdpati should be other compounds give dhdra-. visu. 101 3a: and 41 4a. ndrdmrhsa. but correction visvtinara is not justified in view of the general usage. (ii) viii 25 4a. viii 20 2a (2t). iv 40 5b (4b). 72 17a. v 64 4a. for which see subsection via.the compounds have regularly u the metre favours u in madhupfcam ii 10 6d. see further 174 ii. we is not certain. in the single occurrence viii 98 9b (6).visuvdt and visuvrt are regular. For index to the symbols see p. (iv) composition there are many traces in the text. vii 41 (2t). where u is favoured but In purutdma the text follows the metre if rightly. Cf.is regular. 148. madhu. probably even in vi 22 4d. 160 above. as urunasd. 170 iii. see metre suggests the restoration of u regularly or occasionally. dasyuhdn. but visuanca should probably be restored for visucind in *i 164 38c (6 A). 61 5c (10). vd. but u should probably be restored in most (if not all) occurrences of vdsudhiti. viii 8 17b (6). x 42 3d The reading of the text is only supported by *i 164 49c.' vilu. The words purutdma. and vedudd i 140 la (2t). the following v being an accessory cause. where the word follows an early caesura. or the end of the dimeter of Jagati verse. purubhuj. or in viluharas *x 109 lc (2t) these appear rather to be instances of protraction (163 vii). as final -u in Of always confirmed in : follows : uru- : urdyuga perhaps urukrt in the single occurrence viii 75 lie (2d). 67 5ab. vibhu. ix 96 10a. 48 3a (6). as isuhasta. 10 6a (10). purui'dvas. namely i 128 8a (10). Further there are many cases in which the forms.the compounds have commonly u. restored in the single occurrence ix 110 6b (2t). should restore purubhuj in i 3 lc (2d). : (iii) Words such as ugrddeva. viii 23 16a (6). and it is therefore hardly safe to restore vllu. 6c. and in vasuvid the long vowel should certainly be restored in the majority of instances. mddhumatl iv 57 3a and madhuvfdham x 75 8d.in vilupavi v 58 6b (2t). iv 8 2a (6). though read in ii 41 16ab (2d) ambit aim devltame corresponding to ndditame. come under this heading. sometimes after an early caesura.124 Final u in composition ii varuna should probably be restored in vii 66 19a. vii 38 Id (both 6A). independently of their presumed derivation. namely in i 46 2c (6). v 73 lc (6). vibhdvasu. . ulukhala. are sometimes found at the beginning of the stanza. 60 12d (6). sometimes ii. ix 98 Id (5). The text probably gives dhdrdvard correctly in ii 34 la (2t). jardbodha. the metre. vasftjti whilst other words have -u. In iii 18 4c the metre suggests visvdmitra. 181 Id (2t). vasu-: most of the compounds have u. : and puru.should be restored in vibhusdh for vibhvdsdh of the text v 10 7c (3D). puriitdma. as visudriac for visvadriac : : : : in vii 25 Id (5A). in which positions u is favoured . the other occurrences are iii 31 17a and vii 90 3d In vasuruc u should probably be (both 6 A).

v 36 6d sarvarathd v 35 5c (2d). iv 42 9d (2t). (vii) In the following compounds there is not sufficient evidence to ' : establish the existence of a long final vowel. i 122 7c (2t). surdtha the short vowel is usually favoured. In the forms pdvlru. and nine times in positions in which a short vowel is required. pdvirava. Probably janasdh should be restored in i 54 lib (6A). (c) and The preceding vowel is regularly long. 167. In dasyuhdn the short -ratha: tvesdratha v 61 13b (2d). dlianvacard v 36 lc (2t) . 76 4d. rtdvfdh. See also 174 i. -vrdha. bhurisdh and a few ( 178). The preceding vowel is regularly long. 78 lb. As to a possible reading devaviyam. ix 110 lie (2t). The preceding vowel is always lengthened. 2D. -magha or -sah. : A be restored. (b) therefore tuvimaghd v 33 6d (6A) should probably be corrected. 118 9b (2t). girdvfdh. parvatavfdh but mahivfdh. with In compounds of which the prior element is a preposition which may be considered certain combinations of For index to the symbols see p. 97 26a. as in virasdh yajndsdh. j)dv%ravat.Influence of initial consonants 125 The final syllable of a stem in -an seems to be represented by (v) -a in vrmyHdh but more usually appears as -a. syumagfblie vi 36 2c (2t). when it also follows the caesura (see subsection iv). are only doubtful exceptions. tugridvfdh. as in the compounds : suggests the restoration of -a in the following words: ddsabhuji i 52 11a (10). but ii 13 5b (6 A). see 172 (2t). etc. In x 75 8d (10) madhuvfdh may rayivfdh are equally well attested. A long vowel appears more is element or less regularly when the second one of the words -vasu. carsanlsdh: but short in nrsdh. but the syllable in question appears with some regularity in positions of protraction devavl: the second syllable occurs six times in the position 2t. i. . 148. but i 108 3b (6B). -vrdh. vii 71 3c (2t).. satdpavitra vii 47 3a (2t) and satdbhuji i 166 8a (2t). rdum'dh. tuvlrdvat there is no lengthening of i before r in a compound. ahutivfdh. x 91 15c (2t). 17 lid (2t). tuvlrdva. In candrdratha. viii 38 2b (2d). -han ahihdnam i 117 9d (2t). vajasdni iii 51 2c (2t). vii 15 14c (6). sdrira. wtrahdnam i 121 12d (2t). x 36 8b. six times in the positions 2T. but is unsupported. long vowel is found in annavfdh. -v/dh. The metre favours kavlvrdhd viii 63 4a (6). -sah. -vrdha. sukhdratha v 30 lb (2t). (a) other contributory causes: sahdvasu ii 13 8a (6A) and puruvdsu. The metre however and syfimarasmi (vi) viii 52 2d (2D). other words. vi 16 14c (2d). ghrtavfdh. but a stem development such as is found in gabhird. 19 3b (6A). x 160 lb (2t). namely ix 74 5b. of brahina-. often with -vasu. -magha. 104 2c. irutdratha : vowel is usually favoured. sydmagabhasti i 122 15d (2t). or particle.

iv 31 13a (6). The general use hardly justifies us in keeping the iv). Otherwise the use is that of a short vowel. viii 93 21a (2d). vii 32 24a (2d). pratlvi. abhi sdt : (iv) the negative a- is regularly long in dvrta and frequently in arista and cognate forms. vii The metrical use of abhiyiij is the same as that of purubhuj and it may be right to restore abhlyuj in iii 11 6a. abhi nardm ix 97 49c (2t). (iv) The metre everywhere supports the i for instance in 133 7e : value is optional restoration dvrta for dvrta. 148. for prepositions the reasons given in 171 iv. which occurs i 7 6b (6). vii 27 2d (2t) and viii 23 29c (10). prdvand. and should perhaps read drwvrta *iii 29 6c (10) and suvlvfta i 10 7a (2d). . namely in abhi dud x 48 7b (2t). 164. but we find dpdm-ti viii 66 3c (10). Thus (i) many -vrta prepositions appear to have long final syllables before and some other words beginning with v (ii) dpd is found : in the combination by the metre in dpd vrdhi (iii) abhi is occasionally supported abhlyuj. The compounds dpi-. prdvaryd. updvasu. : has apdrrktd. as shewn in the Table. abhi sat ii 41 10b (2d). and is found in the text in combinations : such as abhi nardh. The text has further adhrvdsd. trivft subsection iv. dnapdrrt vi 32 5c (8). abhi nah i 140 13a (2t). upo. In derivatives of ris the subsection i. and prd-vrta always occur (i) where the long vowel is favoured. saho of the text. auprdvargd we should perhaps restore anapdvrjydn i 146 3c (3t) and parlvfjam viii 24 24b (6). There occur also (v) some other irregularities which cannot be classified. d-v?*ta in In cognate words the value varies thus nivrta. : : Long quantity is given in the text and supported by the (ii) metre in the phrase dpd vrdhi. the negative is short in v 42 8a (9). ix 97 51a (2t). ix 101 iv 31 4a (2d).126 Prepositions and prefixes prepositions with nouns and finite verbs. x 89 3a (8). Before derivatives of vrj the long vowel is less certain the text . (6). viii 45 On the other hand abhisdh must be restored in ix 21 2a (all 6). prdvfs and prdvrsina quite consistently with the metre and updvidd viii 23 3c (2t) and parlvim x 62 10a (6) are restorations favoured by the metre. so that the questions that arise have to do with occasional long quantity. and in dsat. ddsdpravarga. Amongst the occurrences of are included such forms as dpo. abhi su are probable corrections. . (iii) ( 166 8a. abhlvartd *x 174 la (2D). reading abhi where it stands in the text as a separate word. abhi navante. abhi navante ix 100 la (2d). For abhi nUb ii 33 7d (2t) and abhi su iv 31 3a (2d). the phonetic character of the second element seems to be of chief importance. *x 59 3a (2d with hiatus) abhi nd. vi 19 4d (9) and (6) : cf. except vii 27 2d (6A) cf. and suvft are regular. pari-. All the prepositions and particles with which we are concerned have ordinarily short final vowels. abhi nardh v 9 7a 3a (6). 4 8d (6A). ii 2 7b (10). For index to the symbols see p. abhi-.

in several compounds as sundra. indravant. 25 2a (8) and regularly in aristdtati in the popular Rigveda. 148. (ii) unless the prior element contains itself a long vowel. which may be ranked amongst protracted vowels ( 163 vii). oplju ii 31 5b (10) and nlhdrd *x 82 7c (init. . ii 25 3a (7B). (i) and the corresponding Long vowels appear to be found regularly before the suffix : -van. uparuh. The metre favours pratidhfse viii 60 13c (10) 14a (10). In tisat the metre agrees with the text in supporting the long vowel in v 12 4d (2t). 89 5b (4B). Amongst miscellaneous which may iv). sumndvdrl such as maghdvanah as antecedent and the metre suggests to maghonah etc. vlrudfi. and should be restored in vi 21 3b the But d.). and even in asthmhtt from asthdn. devdvant have regularly a (7A). The words prasdh.Influence of snffixes 127 elsewhere. in agreement with the metre. sdptivant vii 94 10c (2D). v 56 3c (9). *vi 28 7a. 78 3c (10 in irregular ending). (cf. variations we may notice a long vowel in the text before ruh.svavant. occasions being v 16 3a ix 32 lb (6). 168. visiwdnt. *97 7d. vi 24 9c (8). and perhaps rank with the protracted vowels (163 vii). subfiarva. 160 also be correct. *vii 104 12d (10) and perhaps *13c (v) (7 A). i : : For index to the symbols see p. In the derivatives of feminine nouns in -i or -7 the short vowel must frequently be restored thus mktirant v 31 6c (2T) and *vi 75 9b (6A). simivant i 141 13a (7B). -yd. and sii. 86 Accordingly matsardvd (from a stem matsardvan) rection in ix 97 32c (10). Before suffixes we find long vowels always before -van. : : 53 5d (6A). vii 18 4a. rudh in anurud. as is the case in words like tdvant. x 8 2b (7B). namely in *x 60 8e = *9e = *10d. except where the ending is -yavant (-idvant) long and short vowels almost in: differently in derivatives in -ydnt. ii 27 13a. suydvasa have in the text pra-. i 42 8a (3d). su-. x 106 lOd (all 5A). *137 4b (all 3e). svadhdvant.. rnavdn. -yu verbs in -y. quite consistently with the metre. indravant iv 27 4a (6A). but short vowels regularly before -mant. forms as rtdvan. for both text and metre favour the long vowel vayunavant again is correctly given in iv 51 lb (8). but the long quantity is inadmissible except in prdsahdm i 129 4b (8) and suydvasa Thus prasdh is required in vi 17 4d (10). sunfta. a probable cor- On the other hand a short vowel is regular before -vant. silmdya. occasional a of the text needs correction in dsvavant i 30 17a (5). devdvant (iv 26 6c. vi 65 3c (6). Amongst derivatives of nouns in -a sutdvant is an exception. but long in i 63 5a (8). -vari. vi 27 7a (8). and visvasuvidah i 48 2a (8).fi. -vant. 10 but perhaps misplaced) the other occurrences are chiefly in the second place. and suydvasa i 190 6a. is 3b (6). the (2t). pdtnlvant.

satruydnt in vii 20 3d Less (6 A). and (iii) of some other words. As to derivatives in -mant it need only be noticed that vdsimant : must in all cases be restored it occurs i 42 6b Cf. is the short vowel : metrically preferable. In the formations in -ydnt. i 31 7c and *x 15 9a as instances of protraction.128 x Vowels of reduplication ii 6 6b (7B). . with vi 15 5d (6B). and in the intensive verb-forms. but pipdya must be restored throughout. sumndyu only where the vowel occurs in the second it may rank with the place protracted vowels . v 57 2a 170 iii. as vdvrte but vavrtvdms. perfects vdvasdnd. 148. The length of the vowel is irregular. vdvrdhijJi (ii) of words beginning with r. -yd. dvidhat. The metre seems to support the latter form. : by the metre as an From lean cdkan.). syllable follows. v 31 13a) though not favoured by the metre. ii 33 12b (3t). iv 25 2a (2T). x 122 5d (10). vdvasuh. ii 32 Id (6A). for v 24 3b see The text 192. and the restoration of a short vowel is required in rjuyd i 183 5c (7B it is probably the adverb of rju). as dvar. but the text usually (i) long vowel is more common if a heavy agrees with the metre. . From nam the text has ndndma i 48 8a (5A). so too a possible restoration of sramayuvah in i 72 2c (2T): see 163 vii. vdvrje. vasuyu as optional forms in i 130 6a (7B). dvidhyat and (more rarely) before y and r. in disyllabic reduplication. : In reduplication a long vowel is found (i) in many of words beginning with v. gdtuydnti i 169 5d (6A). vasuydnt. ndndma. (2T) and x 20 6c (6 in Trochaic Gayatri). The restoration of nandma is probable in the first instance. rdrahdnd Also (iv) . ii'4 6a (5 A). but tatrsdnd i 130 8f (6B). From tard (trd) tdtrddnd should perhaps be restored in iv 28 5d (8) and v 53 7a (init. *x 15 9a (2t). in all strong forms this is confirmed by the metre. as tdtrsdnd. (5). 87 6c (3T).is regular cdkdnanta should be restored (iii) 169 4c (5B) (cf. pi the text has plpdya. hrsivant 31 Id (6 A) : but hrsivant is justified in i 127 6f (10). in iv 7 lc (10). 169. -yu both short and long vowels (iii) are found. and (v) in the re-duplication of nouns. A From rue 'shine' rurucuh is suggested (ii) optional form. and only in one case. i From tars (trs) tdtrsuh tdtrsdnd are given in the text in i 31 7c (2t). certain is aghaydnt iv 2 6d (6A) . for the word never stands in Tristubh cadence. : gives sumndydnt. as rdrand. vi 51 12d (7B). 173 lie (6B). vdvrte. From dhar (dhr) the text has dadhdra. The augment is occasionally long before v. etc. For index to the symbols see p. From pi. vdvandhi. rtayu perhaps in v 8 la (6A) as found in the text in viii 70 10a (7). as vavakre. ix 74 2d (5A). iv 16 15a (7B).

The augment also appears as a. as avrnjan. Cf. which sd. nouns in * -i\ and or the opposite.has really the long vowel. and draik occurs several times in positions which prove the first syllable long. (ii) of d for a in the strong cases endings. namely in iv 39 2d. v 67 Id (7). and the adverb visvdhd (text often visvdhd).in the formation of -ni. With very few exceptions the in the quantities of vowels flexional the final text accurately represents syllables of stems before In some cases however the metre suggests restorations. and we should (iv) therefore read ddvldhnat viii 60 13b. namely in i 100 5b (5A). The occurrences which favour the short vowel are as follows 2 8c (7). must therefore be referred to a yuyodha reduplicated aorist or intensive formation.The augment From yudh yuyudhuh vi 25 129 is a probable restoration. -te -thdm. 9 . dsate all is : For A. but not altogether necessary. *x 180 la (2T). dvidhat (text (vi) as dvidhat). pipdya. In the remaining occurrences tdrutra is a tempting correction. vi 22 2c (5 A). and perhaps in the vocative. ix 110 12c (5 A). Nearly instances (i) dsathe i these points admit of some question. From many forms even of the same verbs the augment has its tudm putro bhavasi yds ta dvidhat ii 1 9c x 33 4a kurusrdvanam dvrni But avar must be restored i 92 4d (9). dvidhyat. yuyudhdte i 32 13c (5A): the other occurrences are iv 30 3b (6). as the number of not very large. (i) of a i for d e % in dual verb-forms before the endings -the. dvrnak (text varying). f and of a for d in the declension of compounds of sdh. namely. dceii citrd vi duro na dvah i 113 4b ii 17 6d (6B) the text gives dvrnak with some support from the metre. avrjan. In regular value. verbs with the suffix -na. in -u . besides the two in which the text has sa-. Similarly before r we find drinak ii 13 5b (8). some nouns in -an and -man. on the exact analogy of tdvltuat iv 40 4c. of nouns of usds. 68 7d. In nouns the vowel of reduplication is generally long but endorsed by the metre in i 145 3c (5A). vi 24 2a. sdsahisthdh. 15 6c (7). In disyllabic reduplication i is always long. namely viii 96 15d The forms sdsdhah. 170.before y in dyukta v 17 3b (2T) and dyunak *i 163 2b (10). in the declension of feminine (iv) regularly of ri. inclex to the symbols see p. dvrni (v) : tdturi is . vii 83 7b (10). The augment is long in the words dvar. plpdyat above. in (9).ru. From sah the text has sdsdha and so forth. sdsdlmt. and in the locative case. though we have 5b (9). 113 13b (7B). 148. but sa. 132 lb (5A). v 59 5b (10). dvrkta. 151 8d (11) 9a (11). (iii) of i for %.is preferable in all decisive instances.

viii 31 6b (7). i 136 3d (7). 65 4a (7) pdrijmanam viii 72 10b (7): purutmdnam viii 2 38b (7). vii 99 4b. and usually appear in the text. 12 2c. 144 6d (11). 148. *88 18b: v 28 lb 4b. viii 27 2b: 4B. and in x 89 9d (4b) in spite of the metre the text has vfsanam. For neut. v 1 lb. v 66 2b (7). and is indifferent in cases the following passages : 1 1 1 . vii 97 5d (7a). ucyate (text ucyete) *x 90 lid. usdsd. vi 30 5d. rathaydvana viii 38 2a (7). 2d. (d) chiefly at the beginning of the verse. rdjana x 61 23a (7B?). v 41 lc (9). iii 31 4c. 124 9d 13b. to be an archaism which is In the declension of usds the text usually gives usdsam. 182 2c (6B) sobhate \-ete) iv 32 23c (7). : : s'ubhrayavana viii 26 19c (7).130 i Stem vowels (7). dadhathe i 151 9a (6B) yunjathe i 151 4d (9). iv 30 9c: 2T. vii 62 4a (9). v 74 3b (5). 25 6a cisathe . The regular form vfsanam should probably be restored in ix 34 3a (2D). . x 35 2c: 10. ii 12 7c In addition the compounds usdsandktd. Assuming that the final a is justified (for which point see 158) there is no metrical occasion. . ndktosdsd always appear at the beginning of the verse the text gives the long stem vowel in both the metre favours it in the first compound. : evidence. . 40 3a (7). In the declension of -sah visvasdham is a probable correction (b) in vi 44 4c (5) and satrasdham in viii 92 7a (7). On the other hand the long quantity given in the text is quite admissible in asate v 68 4b (7g) 5c (7g)j dsathe v 62 5d (3T). B. Stems in -van anarvdiiam ii 6 5b (7). (c) Probable restorations. 1 The text has umsam. and vasathe i 152 la (10). vii 41 7a. cakrate viii 29 9a (6B) caksaihe vii 70 5b (7B) (7). 76 4d. iii 20 la viii 96 la: vii 75 3b 7 A. are : Stems in -man: mahimdnam (or mahitvandm) viii 46 3a (7). dhitdvanam iii 27 2c (7). usdsah are favoured by the metre. chiefly in the accusative singular. pi. 134 4a. 65 4b. form visvdha appears somewhat frequently in the text. . by the side of visvdha. 182 3a (11). rdsatham i 46 6c (5). 1 1 . etc. vi 46 lid (7). trdsitham iv 55 lb (9). usdsah in accordance with the metre: but the forms usdsam. 88 4d. 72 4a. 71 2d (9). as there in the stem-ending. ii 28 9c. (a) (ii) usdsd. 42 5c. vahathe (-ethe) i 135 8a (6B). x 124 8c (7B). vii 90 4a x 39 lb. anusatdm viii 8 12d (7). and is required by the metre in numerous other passages. perhaps rdjanam v 54 7d (6 A) 14d (6A). . Stems in -an: tiymdmurdhanah. as iyate vii 39 2b (10). 72 2a. (e) is A no apparent historical justification. 5 13d. in A. 8. asate ii 41 5c (7). v 80 lc. iii 55 la. dadhdte i 185 2b (10). is only For index to the symbols . : : the second. -ami there is no sufficient contrary to the general use. for a The existence of a locative in -av from -u stems see p. The short vowel seems therefore disappearing. i 92 2c. viii 92 8a (7). iv 3 lie. i 123 12d. yunjathdm iv 45 3b (9).

lb. The locative vdstav however. i 126 2 x 116 9b: C 5B. as in hrinati. 51 6a (6B). ii 13 13a (6B). symbols see p. drundnd. and tdviseh for tdvisyah v 29 14d (6 A). l>efore consonants -au. : . The occurrences are as follows: 3d.correctly and (iv) the short vowel must regularly be restored in all similar formations. 97 3a 12d 16d 19b 40c. 5 2 From this point -av is probable. ii 27 16d ix 63 8b. 9. prepare us for the occasional restoration of -i. Similarly from a stem in -i v 41 14d (9). Similarly yuvateh may be restored for yuvatydh x 40 lib (6A). . viii 4 5 E. . appears as vdstor in defiance of the metre in some passages. 1 -/iv is correct. Final -av regularly appears in place of a dual in -a before hiatus (except before u-). it is necessary to investigate the relation 171. . in vasibhih viii 7 32c (7). From ri. 70 7b (5). 4 pdrsav. srinlhi. as well as the relation of the disyllabic ending -a u to final -ai. 2D. vdstav. 2t.-Locative in -av 131 indirectly recognised in the text in the form sdno in the combination In the remaining instances which sdno tivye or sdno avydye in bk ix. To complete the and as a preliminary of the endings -av -o in the text of the Rigveda to final -a. and ah. in agreement with the history of the forms. even when the syllable is short. the examples are given on p. in the phrase vdstav usrdh. *x 167 4a. and If we disregard the elsewhere more correctly as vdsta usrdh or usrdh. varieties of spelling. viii 6 46b iv 25 2b 5 45 41a. 3 druhydv. : B 1 . 52 6a (6B). 92 . ix 97 17d (9). (iii) A in -i and -i. A 1 . ' : u discussion above ( 160) of the particle to the study of the quantity of final long vowels before hiatus. rin%. prindn. Final -o is of various origin: sometimes it must be resolved -au. u. The restoration of an instrumental in -tl for -tya has already been discussed in 139 iii. ix26 5a. we find that the form in -av is considerably more common than that in -av and it may so far be justified historically. x 123 2c. into -a u or -a u. rl flow/ the text gives ritui-. *x 101 10b (2T): see also On the other hand the restoration ndribhiah is required in i 43 6c to account for the form -bhiah ( 144). 6. v 49 3b5 5 x 40 2a lc. can be recognised by the metre the text regularly gives -av. ix 11 5c. ii 39 3c 5 5 vi 46 8a 3 vii 69 5a viii 46 26a ix 86 3c. The confusion in Vedic Sanskrit between the feminine stems and the frequent transitions from the former class to the latter. . 91 45 5b 5 . : that -av stands in just the same relation to -avi as -an to -ani. . 92 4c. viii 25 12c (7B). 29 (A 5). the latter element being the particle For index to the u or u. v 52 7b: 5b. bhrindnti. 96 13b. . viii72 2b. 148. 3D. 65 16b: 10. 170 ii. i vocative in -u in place of -o of the text seems probable in {/) 135 9a (7a). 5. In all these cases the rarity of a long syllable in the second place after the caesura is the justification of the change. which is probable 166 ii. .

incapable of combination or shortening. with the intention of disguising hiatus or consecutive short The combination -a u syllables in the second and third places. we find that in all other instances the vowel in the Rigveda is not capable either of combination or of shortening before vowels and as it usually ends the first word in a clause. which has the effect of making the final syllable exempt both from combination and from shortening where a vowel follows. viii 62 la. and it cannot be equivalent to dtha. representing -ah. for final -au revision becomes regularly either Final -o -a or -a u. : . it is used as a textual correction of -a or -a in forms of all kinds. and also the vocatives. probably also in mo and The genuineness of the form dtho is shewn first by its appearing no. *x 86 2c. Again dtho does not stand for dthd u. viii 7 33a. go (ca u) vi 66 3b. vi 37 2a. u combined in the Atharvaveda. immediate reason to assume the presence of the particle except where the metre requires the restoration of an additional syllable. but this is (ii) no satisfactory indication of its origin or value. If however we exclude final -o as as indo (indav) appear as indo iti. it may be conceived as resulting from the combination of This explanation however owes its final -a or -a with the particle u. -au. without Of a dual in -au becoming automatically -av regard to the quantity. or -ah. for even vocatives in -o. attractiveness to the theory of the Pada-patha. there is no or uncombinable. namely in vi 63 la. for in that case there would be almost as many occurrences of the particle u in this one combination as in all other possible positions. Final -o represents a phonetic developement of a. (i) -ai. this use being almost restricted to the later parts of the Rigveda: and lastly. ix 89 la. in all positions in the In about 15 instances ( 174 i) the final syllable of a dual in shewn by the metre to be short. represents occasionally final verse. 22 la. namely in : 40 la. hdnto (hdnta u) viii 80 5a. viii 62 la. for in no instance is it equivalent to dtha u with hiatus. pro (prd u) i 186 10a. before vowels there is no trace in the Rigveda proper. just as in the locatives in 170 ii e. In i 186 10a.132 Final -av and -o sometimes it represents a modification of final -a or a in an adverb. represented by its writing urn iti. although in the latter case no metrical advantage is obtained. This leads to the conclusion that in an earlier state of the text the duals in -a were rightly distinguished according as hiatus or combination took place. that the particle u has a special character as pragrhya As this is hardly the case (see 173 iv). the ii thus obtained stands with hiatus before a vowel 6 (d u) vii following (iii) ( 124). Only in one instance is such a dual written in combination. certainly in dtho. in the popular Rigveda and the Atharvaveda. whereas all revisions only of the text of the Rigveda have been carried through the whole of it and secondly by its use before vowels and consonants alike. no (nd u) *x 86 2c. usually appears in the Pada-patha as -o iti. mo (md u) i 38 6a. and that a later and mechanical -av is has substituted -av for the ending before hiatus. as shewn below (subsection v).

although nu cit. *59 4a 8e = 9f = lOf) and only twice elsewhere in the verse. *x 18 13b (8). 164): but the words in which -o appears are used exactly as ordinary words with final short or long vowels respectively. and (in the case of -a) in the positions 2t. that is. utb. and before consonants 40 times in 34 instances the next syllable is short. viii 25 23c. 94 6a. times before consonants. Both words occur occasionally before vowels. and in 79 of these the next It is clear then that utb stands according to a system syllable is long. Such a revision must have been the work of a later hand than that which recorded the value of the 'protracted For in the case vowels. namely in iv 21 9c (8). and.Final -o 133 which is contrary to the principle of quantitative evidence. and only twice outside it. *ix 114 4d. and vi 54 3d (init. and upo. seven instances being in the popular Rigveda (*i 50 13d. x *27 20b. Further the general practice of the poets ( 157) does not lead us to expect that utd would be followed by a long third syllable in so high a proportion of the occurrences as 90 per cent. So in one : instance no stands for nd u otherwise it occurs ten times in the popular Rigveda. namely in v 31 13b (6B). 148. where nu may be read. As examples we may take the three words which appear most commonly. in all cases but one in the position 2t or 2d. 2d. tdtra su i 37 14c. dpo. bhuytima sti iv 32 6a. the restoration of final -a seems generally most probable. before hiatus. it is impossible that the later tendency to shorten the word to dtha can have been at work at this period. In one instance mo is to be resolved into m& u (subsection ii) in addition it is used twenty times initially. before short third syllables. It follows that dtho is a real phonetic developement of dthd . but in viii 67 15a the metrical j : upo occurs thirteen easily removed by reading dpa su. The chief exception (besides no difficulty is For index to the symbols see p. usually without metrical advantage. ix 61 13a. In other words ending in -o for -a the usage is the same. indicating that the editor concerned reckoned nu. : Much more frequently. and then are neither combined nor shortened. but the metre can be corrected by reading upa sti in i 82 la.). and though the particle u may in some cases be present. and sahd su viii 7 32a. if so. because we find utb nu vi 47 lc. . The cases of mo and no are not quite so clear. On the other hand utd appears 88 times initially. of the protracted vowels it appears that the words received special treatment from the bards (Table. Further dpo occurs four times before consonants: in three cases the following third syllable is short. Of these utb appears always initially.' and has not as much historical justification. with such consequent corrections as ihd sti ix 65 5c. viii 103 lc. utb ghd vii 29 4a. it seems probable that final -o of the (iv) text represents the result of a metrical revision of final -a or -a. That this system was that of the bards themselves seems improbable and the more so. ghd are the regular forms. but are generally similar. and utb nu cit viii 40 10c lie. 72 6a 18a. and is betrayed as such by its systematic use in positions in which it is metrically convenient. kena nu viii 66 9c. gha as short syllables.

sing. Instances of final -I before a similar vowel are so rare and so indecisive that the rule affecting them can only be inferred from the . 29 2d. requires some similar explanation. The forms given above are more common in the popular Rigveda than elsewhere. it must be clearly distinguished from the u which attaches itself to certain verb and noun endings. vdpurbhir d carata anyd-anyd If we take into account only the instances in which -a with hiatus is certainly to be restored. but not to such extent as to In particular. vii 85 2a. Where -o stands for -a (as in eto. as explained in the last section. *164 5d *28b. vai {yd u). *137 6a. : Since the particle u u a. ii 33 9d. evo. tdpo) either a vowel In the latter alternative u is follows or one of the particles nu or su. or at latest the third when the second word is a similar particle. is generally optional dual forms usually retain long quantity. it will be seen from the list below that the short value on the whole predominates. but as the vowel is long in this case the analogy is not exact. u v 17 3a. 31 4d. vii 68 4a (devayd u). and therefore regularly appears later in the sentence. vdu: 13a. *104 13a. are numerous occurrences in the eighth place. and of final : - -u with hiatus before similar vowels. *x 10 12a. *142 -d asd iv 23 or pi. and the number of If further we occurrences before early and late caesura is equal. except in *i 161 12d. : 162 21a. probably contained in the former it is more than doubtful. which seems very frequently to stand for prd u. with the words asau (asd u). *117 la. Phonetic u iii above) is pro. 3a. v 73 9a. 44 5b. Also the neut. *83 10b. *27 5a. sing. This is the case with the ending -tavai {-tavd u). The reading -sitdsa u jicvah. of nouns in -ar. where prd should be restored. the proportion of instances with the full quantity will be greatly increased. fem. in dnumatyd u *x 167 3b. : iv 21 9d. and other endings in -ah (-a u).s such is regularly the second word in (v) the clause.): i 156 lb (evayd u). sing. It may however be noticed that the short value very much' predominates in the nom. 172. include those cases in which the final -o of the text represents a slight phonetic variation of long final -a. in i 140 4b. 168 lb (devayd u\ 10b (turayd u). *58 9a. So u jananta ii 18 2c.134 subs. u (nom. and u may very well belong to the word following. viii 23 *ix 112 la. v 2 10b. The quantity of final -a before hiatus. 88 2c (adhipd w). katho. *i 62 12a. and with the nom. x 65 lOd (dhanasdu). but the Examples are : viii 46 10b asvayd utd rathayd i 62 8d. there justify the rule vocalis ante vocalem corripitur. sing. nom. The occurrences are -tavd u i *24 8b. vii 4 8b. in bhd u i 46 10a and the gen. sing.

and the locative tue tve regularly retain the long quantity. Initial. *v 51 15c. x 26 Id 9b: 2T. x 61 18c. (iii) vii Instances of hiatus of -i before similar vowels are found 28 3a (5B). iv 3 13b. *x 88 4c (4A). . x 30 2b. . vrso iii 27 14a (2D). viii 17 lb. iv 2 18b. rdkso i 174 3c (2T). 104 lc. For the dual forms see 174. . viii 33 lOd (2D). but a resolution into a vowel and a semi.). viii 46 10b: 6. 185 4a. . (ii) hiatus ( j .). v 45 2d. x *27 9d (2t). 46 29b. iii 13 lb. 16 la. *101 3b. 2t. sing. iv 16 la. The following are probable instances of -a with hiatus A. -I. adverbs etc. iv 52 2a. i 62 8d. *96 13b (2t). . 24 9a. when followed In particular. of noun in -ar. 140 13c. except that duals in -e. ii 13 10a. or in the case of -o for -ah. 48 times: 7A. 34 lib: 8. Gayatrl). vii 1 7a. vi 21 2 2 8d. 186 6a iv 33 10a vii 69 3a x 49 10b 65 15b E. vi 4 4b. 113 13b (2T). the quantity cannot be determined. i 60 4b 104 5a. The vowels -. 19 6c. vi 20 8d. mo *i 50 13d (in. *85 35b (2d). retain the long The diphthongs -e -o become short syllables when followed by any vowel. 5 A. dissimilar vowels. . x 105 lib lie: 9. 5 times 5B. viii 101 13c. v 65 6d (in. 69 9a.vowel. *x 18 13b (8). for considering that there authority even in the text of the Rigveda no real hiatus. B. . . 70 12c. i 48 7c 16c. by duals regularly. 15 times 2D. 96 9d. In the following cases the text has -o representing -a before 171 iii iv) .Final analogy of ( -a. in -o. v 41 16b 16c. nom. 3t. 77 la. viii 20 17c. i 133 6c. 173 4a. -u before hiatus 135 final -a. C. 70 7a. viii 103 13a. x 5 |5d (reading 2 ptisa dsya). : : : : : : 1 1 1 . dhiyam for dhi-am. . but there are many exceptions. ix 39 5b (2D). *19l 2c (2D). 48 times 3TD. D. 148. 132 4d. 17 times also vi 16 27a (6 in Troch. and the particle quantity. x 61 9c: 3d. . *129 5d: 2d. 133 6b. -u are regularly shortened 173. For -u we have only the restored phrase n4 u and as w& is then always initial. i 120 6b: 12d. 87 4c. . 40 4b. and in the vocative of -u stems -av for -o. In all these cases there is is For index to the symbols see p. u most commonly. vi 23 4a2 *75 3d. ii 17 lb. *137 4b (2T) aryamo i 167 8b (7A) eso i 46 la (2d) dvisto vii 34 12a (3T) imo vii 1 18a (2T) cattb *x 155 2a (2d). 40 3d viii 1 16c. dtho i 28 6c (2D). : . *60 8e = *9e = *10d (2D). . 5. 8 times 4. In internal combination we find su-uktd and other words. *ix 5 8b (4). 61 14d 2 1 caesura after third syllable. i 189 4d 2 2 2 6A. The diphthongs -ai -au are regularly represented by long vowels when followed by any vowel. v 29 15b. v 53 14b. 124): always with (i) -u. 6 times: 4A. 24 5d vii 34 4a. ii 19 3d. a retention of the Thus in the declension of dhi the text gives original ending -as.

tdsma p. . vi 71 5a (2t). *88 10c (2t): u (chiefly in the popular Rigveda) *v 30 15d (6A). *149 3d : : (6B). mimihi vii 19 lie (8). iii 8 8b (7A). iii 5 2a (3t). snathihl . : . 4 162 iv. tve : The locative tue. x *27 7a (7B). . In the case of u u the long vowel is most favoured before (iv) instances in decisive positions hiatus. . and yajtianiyarn. viii 21 7b (2t). On the other hand the restorations proposed in this chapter frequently require that nu and sd should retain their long quantity even before vowels. jdnati i 122 9a (3t) (i) prthivi i 94 16d (8). since (iii) the only forms that occur rather favour the long quantity. jahl i 36 16a (8). susdml *x 28 12a (8): (c) in the imperative in -hi: ihl i 80 3a (2d). Exceptions are probable as follows final -I is retained (a) in the nom. For the duals see the next section. v-oi. x *86 2c (2d). etc. sing. *107 6b (all 3t). vii 25 2a (8) final . -u before hiatus Consequently the regular use as the following : may be represented by such restorations iii daksindv&d vdjini pr&ciy eti pibd tu somam gdv-rjlkam indra yuvo rdjdmsi suydmdsas dsvdh : 6 lc vi i 23 7b 180 la. in -o are shortened. such as prthivlydm i 139 lib. which is always found in Tristubh cadence suffix -lya is probable in yavlyd i 167 4a (10). as in viii 22 16c Vocatives (9). namely trsd i 58 2b (2t). sing. vi 27 la (7B). . tdsmai. loc. For index to the symbols see 148. noun-stem is regularly retained in the gen. namely nd *iv 18 4c 1 vi 47 lc 2 viii 3 13c 72 6a2 18a2 94 6a 2 (all 3t or 3d). -I yuyodhl i 189 3a (8). as tu iii 30 12d (9). as in ii 5 8c (6) Final -ai. pdhi viii 60 9b (2d).. *i 161 12d (7a). su i 111 2d (9): but in these words the short forms are also found before consonants. 173 12c (10). -au seem regularly to become -d before a following (vi) There is however in the Rigveda a vowel. 3 167 iii. The forms tu nu su are frequently found before hiatus.136 Final -I. sd ix 65 5c 2 (3d). as ydsma v 7 8a (5). though both forms are common are u ii 24 6d (3t). vii 96 4a (5). always retains the long vowel before whereas asme is always shortened. *102 10b (6B). sddhd ( 160 i) ii 3 6a (2d). 1 175 ii. (v) hiatus. vdji x 56 3a (2t) uti vi 29 6b (2t). for other words in -o see 171. .. vii 5 4a (8). nu i 100 10b (9). (b) in the instrum. as indav ix 40 4b. and the ending -tavai. as prativiam regularly occurs in dimeter cadence. *x 10 5d (8). marked distinction between the group asmai. and tanii *x 183 2b (4A). Other instances are very uncertain. as in classical Sanskrit. *x 59 3a3 (3t). ydsmai and such words as asau. in -dhi. - 171 iv. viii 15 la (2d). (ii) and a dual rodasiyoh. The former are perhaps occasionally shortened. x 73 9c. The text reads devdvyam in six passages at the beginning of the verse. *x 88 17b. and the protraction of d is probably correct. vii 81 la (2d). So also td vii 31 4c (3d) 4 1 . and yusme presumably has the same value. of a : The rule for forms in -u depends largely upon analogy.

29 9a (7a). a dimeter verse begins with indravaruna. 26 7c (7a). 3a (5). but these seem insufficient It seems as evidence of a fresh variation in so common a form. (d) (e) dual of a noun in duals of verbs in : ubhd (ubhe) v 30 9c (7B). it is of the vocative case only. vii 60 12a (4A. vii 93 6c (9). not only in questions of external and internal Sandhi. 66 6b (2d). -u before consonants. vi 63 la (9). x *90 lid (5). viii 26 13c (9). In the part of a dvandva compound have been dealt with in following instances the text has -a. 85 4b (4A). and the restoration of indra vdruna on the analogy of varuna mitra in i 122 7a is not improbable similarly mitra vdruna may be restored in i 15 6b.). admissible in vii 19 4d. but still certain. 132 lc (5). (i) (b) (c) -i : vi : -u i 46 13c 60 13a (7B). v 64 6a (fin. viii 5 31b (5). : : : 175. vipanyu viii 8 19c (5) are favoured by the metre. vlra vi 63 10c (7B) and hotara v 5 7b (5). 148. contrary to its general practice For index to the symbols see p. iv 15 9a (5) 10a(5). 61 la (7A). Similarly the restoration of -I is favoured by the metre for prthivi in ii 31 5c and iii 54 4d (both 7A): and that of -u is In four instances in i 17. *x 94 9a (6 A). vi 68 5b (7 A). Examples of the regular usage are : jrayasdnti dram prthu (text jrayasdndv) patho nd jdnasl ubhe dnu td hi katrdm dhdrdyethe dnu dyun pdyum v 66 5c ii 2 4d vi 67 6 a. 151 4a (7A). namely 3b 7a 8a 9b. 166 i. but the metre favours or at least permits the restoration of -a i 15 6a (fin. ii 27 15d (9). 25 4b (5). are regularly long but to this rule there are many exceptions in the : 174. occurs cf. It appears from the whole course of the discussions in this chapter that the Sarhhita text stands in need of substantial revision. both of nouns and verbs. v 43 4a (5). and the readings of the text. : -a is found wherever hiatus It will here be convenient to deal with the question of (ii) Occurrences in the first possible duals in -a. namely -a before hiatus: i 184 lc (9). before hiatus earlier parts of the Rigveda. : before hiatus 137 whereas the latter develope into -a u. : Exceptions are relatively few. 67 8c (9). Besides we have only deva v 67 la (5). sdvistha vi 68 2b (7a). mamhistha vi 68 2c (7a). 65 Id (both final). For the text see 171 i. In the dvandva compound indrd-agni. All duals. therefore more than doubtful whether a dual form in -a exists before consonants if so. . 143 (5). 35 24b (5). v 74 la (a) 86 5b (5) 5d (5). as shewn in (5) asau appears as as&v x 132 4a. -e -e : (9). -i.Duals asma v 64 2b 171 v. but It follows that isolated quantitative also in questions of quantity.).). 166 i. vii 65 2a (9). viii 9 6b (4).

and which Such inconsistencies it is therefore impossible to ascribe to the author. the number of which might be indefinitely increased We : (yend prthivydm ni krivim sdyadhyai \yena jdnd ubhdye bhunjate visah (krdhi no ray a usijo yavistha \krdhi rdtnam susanitar dhdnandm (avis karta ii ii 17 6c 24 lOd iii iii i i 15 3d 18 5a \jyotis karta mahitvand ydd usmdsi 86 9b 86 10c 1 1 piba {pibd rdye sdvase somam mahatd indriydya huydmdnah is x 116 la x 6 1 c. and probably at an but the more exact methods of earlier date. are illustrated in the passages quoted below in pairs. and from different points of view. 148. is therefore executed in a more intelligent spirit. and even in the same hymn and the same verse. it is at least safe to say that it is better to face them than to treat as a text affording a basis for successive : metrical investigation one that is so full of inconsistencies as the Samhita. which produces an artificial regularity of metre going greatly beyond the regular practice of the bards. must always be received with considerable hesitation and for the purposes of metrical : seems safer to assume in such cases that the investigation words and endings have their ordinary value. and This work partly by the metre. the Samhita is guided partly by considerations of grammar. that is to say. is in itself and under such circumstances no restoration sufficiently probable of the original text can claim to be more than approximative. That the text of the Rigveda should have been corrected by editors. Lastly in the case of final -o we can recognize in some words such as dtho a state of the text so early that it recognizes the historical differences between various parts of the Rigveda. and all modern criticism is agreed in requiring restoration. and in others like ut6 a restoration on purely metrical grounds. constantly find in hymns belonging to the same group. The extent to required in each class of forms is An isolated variation in the text may in some cases be a trace of For index to the symbols see p. which correction discussed in the next section. so that in this point the Samhita revision is precisely of the same character as in questions of Sandhi. the text follows mechanical and entirely disregards the metre. . by the recognized value of the endings in the editor's own time. so far as he was able to appreciate it. differences of quantity for which no metrical or other cause can easily be imagined. In the case of the variant final quantities. than the Sandhi revision rules : literary criticism make it possible in the majority of forms to establish independently rules which are more uniform and at the same modern time more in harmony with the general character of the metre. it In the case of final vowels before hiatus.138 Irregularities of the text general tenour of the metrical evidence. on the other hand. But whatever difficulties are involved.

) and appears to be a quasi-dual on the analogy of the preceding dydvd other forms are found in the text in i 69 3b (8). iv prd-pra i 129 8a (2t) The text also has three times nahi nu.The Samhitd an text 139 earlier recension. i. viii 45 22a (6) 38a (6) -asya 79 4c (8) -esu loc. : : 176. Pp. the bias it quantities even of the variant final vowels. sd i 145 lb (8). as ista-purtd *x 14 8b. and 31 5b (2d). 31 x 178 Id (8): ca. iv 5 14c (8). 9a (8): . in -d from stems in -an. and which appear so rarely in positions that favour long syllables that there seems no reason to look for any other explanation than the usual metrical liberty of the Rigveda. nd aha. and may serve as a valuable confirmation of the But if it only has the effect of removing an indications of metre. : . i 61 *i 162 19a (8). viii 19 24a (8). the most probable explanation is that it is due to excessive regard for the metre. : A : in particles: dva vi 46 lib (6) ihd. x 45 4b (2D). TS. 173 6c (fin. The form bhiimd is found in i 61 14b (5C). whilst ndmd is favoured by the metre in i 48 4d Cf. sing. : 11a (4). iv 2 16d (2T). *142 2a (8). 34 9b. i 77 2d (2T). -van. 34 Id (8). ca d) cand. vii 28 la (2t). (ii) . 167 9a. i 84 20b (6): nd. which were no doubt established in the later period and the short : values are therefore introduced wherever the metre or indifTerent. twice nahi nd at the beginning of the verse nahi nd is probably the correct reading throughout. viii 31 9d genitive (3D) rdksatl risdh ii 26 4c (10) on the analogy of rdksata risdh ii -ti 3 sing. -ihd. *vii 33 2d (8). x 25 7d (final syllable misunderstood for 8) -ta 3 pi. viii 53 4a (10 the text has ca. is favourable Similarly for a. There are also occasional instances in compounds. In all the instances that follow the text gives long quantity to syllables which have ordinarily the short vowel. iv 5 5d (8). pi. *x 34 8c (8) ha. and brdhmd vii 24 4b (2t). *129 lb (8). -man. 176 lc (2D). In the case of the verbal endings in -d. and that chiefly at the end of the verse or distich. or before consonant-groups. namely in i 80 15a. : : . vi 17 16a (8). slight doubt exists as to the neut. -ta. For index nu and su the text prefers the to the symbols see p.). -a vocative. 148. The occurrences are (i) as follows : : s. 166 v. i 147 3d = iv 4 13d (the Sariihita has ndha. v 41 7d (2t).e. for which the poets preferred the long values. but see -ta 3 sing. in flexional endings -d nom. the text prefers the short values. iii 57 5d (8). 62 8a (8). In spite of numerous errors in detail. it remains the fact that the Sarhhita text on the whole gives a true picture of the Further. : . ii 33 13c 167 iv (8). occasional metrical irregularity. correctly nd ha). viii 3 13c before vowels (173 iii). . : : : . shews is readily explained by the examination of its usage in each class of words. tu. vii . (8) *vii 104 13a (init. and in *iv 18 4c. masc. For the long final vowels enumerated in 158 the text only occasionally gives short quantity. vi 27 3a before consonants. (2d).

-yd. without regard to metre. 2D. : The chief practical importance of establishing the rules of the text (so far as it is possible) lies in the presumption that some of the exceptions represent an earlier recension. In the words that are usually short the text only introduces the long vowel in the 'positions of protraction. not only in the positions freely 2t and 2d. and sometimes evd and Before the caesura the perfect forms and evd have the short smd. ir- The Th. On the other hand ddha ydd appears always. 4. 2D. and sometimes in 7A. u nu. and for nu and sti either lengthens arbitrarily a preceding vowel or substitutes a form in -o. 5B. It appears fairly regularly. (iv) For index to the symbols see p. and purudhd. -td In the very numerous occurrences of the verb-forms in -d. sfi the text shews a strong bias for (iii) In the positions 6. the same correction. nit. several positions the text gives gerunds in -tyd. In the forms noted in 158 the text usually gives the long (i) vowel correctly. and probably also in 3TD. 8. gives the short vowel. nu cit. In the positions 3T. all are treated by the poets in the same way.140 Samhitd rides of quantity short values except in certain phrases. 2d. 2d a long vowel is occasionally given otherwise almost always the short vowel. In the case of adyd and ddha the text appears to be in general agreement with the metre. except before consonant-groups and at the end of the distich. The treatment of separate words differs slightly. and in final The neuter plurals are treated in the same way and need syllables. which should certainly be corrected in the very common position 4A. The perfect forms have the short vowel at the end of all verses. In the adverbs u. Zubaty (Der Quantitdtswechsel. In the adverbs ddha.' and there very regularly. but also often in the positions 2T. not In only before the caesura. Vienna Oriental Journal. but also in 2T. 3d (where a long vowel is urgently needed) the text gives the short vowel. adyd. Thus end. and u su are evidence of an earlier and more : correct recension. often in 2T and 2D. but occasionally shortened when final. correctly. though . 1888 1891). end always have the long vowel. 148. and less regularly in 4. 3D the text usually gives the long vowel correctly. Only a very short summary of the results can be given here in many cases more detailed information is given above. . but aorists in -svd are short. as have also dthd and evd. vowel. (ii) the text gives the long vowel correctly in the positions of Group A. -thd. and yddl the long vowel is more recognized. But the apparently arbitrary exceptions nu (initial). visvdhd are only whilst instrumental in -t% are shortened before consonant-groups long before groups. tu. In the positions 2t. Otherwise it in 2t. but in the positions 3t. 10 the long vowel is used the short vowel. but smd is given correctly: the aorists in svd are also short. Gottingen 1874 1880) and J. but -dya. practice of the text has been elaborately investigated by Benfey (Die Quantitatsverschiedenheiten.

4B. Generally. and tu&van i 30 14a. who had yet never made an exact (vi) : : : : study of revision. and occasionally in the position 2D. 2d otherwise inclines to the short vowel. the vowels of invariably given whether correctly or not. It is therefore impossible to regard the authority of the text as final. 5b entirely escapes him. as when in viii 12 17c. vrudhi. difficulties arise. 148. but not elsewhere. nor distinctly formulated his own principles of Occasionally we can recognize mistakes by the editor in the application of his own principles. 2d. In all positions other words have the short vowel. is a question which does not directly concern the metre. krdhi correctly. 2d. position if the vowel is followed either (i) by ch (cch) (ii) by Ih or (iii) by -nn in the third person plural of secondary tenses and in the nominative singular of participles. 177. quantity in the positions of Group E he thoroughly understands but the importance of the positions 3t. or when he applies the rule for consonant-groups before tudrn x 61 22a. the editor shews exactly that knowledge of the metre which might be expected of a man thoroughly familiar with the text. 189 6a. indifferent. 4b. nor whether he thinks that the positions before the caesura and at the end of the verse are preferably short or only In the positions 3T and 3D he prefers the short vowel. and. . 13 7c he fails to recognize the Dimeter Usnih metre. as it appears. The sounds represented by by position. The practice of the text gives us some insight into the editor's The comparatively strict rules for short appreciation of the metre. perhaps correctly the adverbs in -tra regularly so in the positions 2t. kena. correctly. often in the positions 2T. and 7A. but in the positions 2t. its mechanism. tenet have the long vowel in about half the occurrences in the positions 2t. The pronouns yena. 3d. . Of the less strict preferences of Group B he exaggerates the importance of the long vowel in the positions 2t and 2d. and still more in 2T and 2D whilst on the other hand he is unaware that the long vowel is preferable in the positions 4A. but not to such an extent as to fail to give the true value to most of Before a consonant-group a short vowel is almost 158. It is not clear what view he takes of the position 4. but rarely elsewhere. 2D.Quantity by position (v) 141 Of the imperatives in -dhi. Vowels capable of protraction are written as long in the eighth place. -hi the text gives srnuhf. m m h also : contribute to make length The following are examples diaur dsld utd cchadih ndrdyaso rid jdlhavah ud abhrdniva standyann iyarti sbraam srinanti matibhih suarvidam visvarii *x 85 10b viii 61 lib 44 12a 5b vi 31 2d vi ix 84 drljidm bhayata djman d te For index to the symbols see p. In questions not connected with final syllables few It must be noticed that syllables are long by .

: : im i is to be restored occasionally in the popular Rigveda *i 164 7a (7B) *16c (7B). usually. instances in which ch appears to have the value of a single consonant. 148. which has been suggested on the analogy of sddhuyd and other words. tj'lhd is tr[hd. tubhyam. So far as the metre have been proposed on Amongst forms in which -nn represents a double consonant (iii) are to be reckoned aorists like dtann vi 61 9c (2d) the nominative singular in -nil as sadfhh has the same value. iv 16 lib (9). urviyti: this form usually follows the caesura. -inn : : and pusann. The metrical conditions do not allow us to trace this proposed form with certainty. historical which the restorations dflhd. 116 10c (9). in concerned. -in usually for those in but the final syllable of magliavann is always long by -ann. *x 27 lid (7B). namely uch&t 7a (6 A). and similarly we find always This evidence outweighs the few gaccha-. tdsmin must everywhere be restored (10). this The following list indicates the readings chosen book as most probable : (9). and uruya is an attractive correction. for which forms without the final -m are not uncommon in the text. cachadyat x 73 9b (9). -inn of the text. except in the participles drljid. these restorations are unnecessary. 30. dcchidra. sahdchandas *x 130 suchadistama vii 66 13c (6B). asmdbhyam. and must also be restored as shewn above on This restoration is not required in *i 50 13a or *126 6c. for instance in i 94 7a Locatives such as djman. *125 4b (6A). grounds. dyuh: a neuter form dyu must be restored in i 37 15c (5). the syllable occurs in the semicadence of epic Anustubh. and such as occur are of little for the purposes of historical importance. and in isdnakft viii 52 5b (5). as an error of transliteration: it is however convenient from the point of view of etymology. prccha-. h a large number of (iv) instances would be required to throw doubt upon the ordinary rule for position. rdjann. The value cch is correctly given in the devandyarl text wherever the sound occurs medially. (ii) Ih is rare in the Rigveda. iii 32 lc (8). x 144 5c (9) 6c (9). vfsann and sahasdvann instance in ii 28 9b (4A). : For index to the symbols see p. since urviyti is everywhere admissible the form in the text is required in x 92 12a (8). and vocatives in -an. . isdna: isdna is metrically preferable in i 61 15b (9). the compounds ayu-sdk. Cf. which affect There are very few irregularities in the Samhita text individual words. for In such common sounds as ih. iii 49 2d (9). which in all (i) cases has the metrical value of dcchd .142 Textual corrections The most important word containing ch is dclid. 62 15a (5). i 48 3a (5). 178. for -ann. 65 5b (5). 44 6c 89 "9d (9). viii 18 18b (5). Such instances are only found in the datives mdhyam. 90 2b (5). m. etc. where p. and the writing ch may therefore be regarded. yaccha-. vii 4 6c (8). Ih being always treated in Sanskrit alphabets as a double consonant. position.

tri : 155 3d (3t). . ji 'conquer': the metre requires jigivdrhs in ii 12 4c (6 A). brhdnt a form brhdbhih is : which cf. (7B) . which always stands at the beginning of the verse. vi 67 4c (6A). but barhih-sdd in vii 2 6c (2t). in ix 4 4a (2D). ndr. brha-divd are metrically preferable to those in the text except in one passage. for instance in viii 9 lc (11). bhurisdh read bhurisdh ix 88 2a. (6 A). ix 75 2d (2t). the metre always supports purusa. 83 2c (7B). and to be restored in ix 58 la lc 2c 3c (always 5) see below on vdjrin. brha-divd. purusa. das for dasat daslt is metrically preferable in i 120 15a (10) and 139 5c (7a): cf. : mandi i is mandim : 9 2b. i 8d iii trtiya: a form tritiya making position is required in iv 34 4d (2t). vii 1 21d. vi 32 4b (7B). iv 10 4c (9). as found in the Samaveda and the Pada-patha. das: an alternative form das is suggested by the metre in i 76 Id (9). vi 19 7d (6A): but jigivdrhs with the text in 15 4b (10) and preferably in x 78 4b (2T). See also 177. and the forms brhd-diva. satvdta. and never in Tristubh cadence. x 66 8b (2t). cardtha stands in the text in i 66 9a (9). : mahdnt the dual mahdna (or mahda) and plural mahdtah (or mahddh) are suggested by the metre in v 41 13a (6 A). 41 2d (6A). x 1 3b (8). v 44 lb (6A). Cf. mahdnt below : mandin cf. chardis is always to be read without r. purupurusa. *x 27 22b (8). nfndm is to be restored. and cardtham i 70 3b (9) 7b (4b): all the passages need emendation. cf. in the gen. *90 6a (2d). and cardtham is required by the metre in i 68 lb (4b). viz. : metrically preferable in iv 56 4a (6A). : ' ' : : pavakd is everywhere to be restored. purusatrd. mrl_ vi 50 12b (7a). tisfndm must be restored in the single occurrence v 69 2c (10). ix 68 lc (2t). For index to the symbols see p. however iv 21 8a. jvisitanca ksdya: for ksdyasya vi 71 6c (6 A) ksayasi is well suggested by H. mahimdn for mahimnd we must restore mahind i 59 7a (6A see : : also 151 * i). 29 3b (9). 18 v 62 9d (6 A). Except in the forms purusdta. vii 3 7a (9). The only occurrences unfavourable to the long quantity are iv 3 3b (6A). for instance i 114 10a (8). barhisdd the reading in the text is preferable in ii 3 3d (6A). as is made clear in the derivative mfllkd. Oldenberg. ' : pity particular by mfl is everywhere to be restored. nf man pavltdr pavitdr must be restored. pi. iv 22 3b (7B). 148.Textual correct ions 143 uruvydc: for uruvydncam we must restore urusincam in v 1 12d 166 iv. An instrumental mahdbhih stands in the text in vii 37 Id (7B) and must be restored in iii 36 Id (6A). *x 15 3c pavakd : : (2t) for *4a (2t). viii 103 4b (7).

to be read as as in camnan i 104 2c. There are however instances of change in the opposite direction : in particular. hdri-candra. vii read perhaps svdslti for svdsiti in i 65 9a (2t). In these positions there are about 600 variations from the general rules. : i 80 7b virdsdh i 35 6b is supported by the metre. 174 5a (9). and it is also numerous emendations have consestrongly opposed to the metre been proposed. see on mandin above. : ydsmin it is possible that a locative form ydsmi should be restored before consonants in i 33 14a (9). . viii a vocative form vajri is suggested by the metre in 99 lb (5). and always in ds'va-. visvd-. 8U-. s'vas for sasavdrhs we must restore sasanvdms throughout. There seems to be an exception in iv 42 10a (6 A).144 yaj vi : History of quantitative change for ayeje read ayaje i 114 2c (9). iii 22 la (7B). but also in iii 18 4a (10). : han : for hanydma read lidndma viii 21 12c (5 A). and I incline to restore this. but the short vowel becomes more common in the later part of the Rigveda in the remainder the vowel is regularly short. not only in the passages named. 20 5c (10). The rendered difficult historical investigation of quantitative change is by the fact that the metre only indicates the quantity with any certainty in about one-third of the occurrences. but hardly sufficient evidence of the existence of a form in -uni instead of the usual form in -uni. as in san 87 2b (10). (9) . cf. : historical vowel in changes are in the direction of shortening a long final about one-half of these instances the vowel is regularly : long. The form sdtvd satisfies all the conditions quently both of meaning and of metre. in shewing about the same margin of choice which is allowed In the majority of instances the the matter of hiatus ( 131). vi 24 This form occurs in the text in ii 19 5b (10). sc at is the beginning of words and in the latter element of compounds c. x 63 7a (9) and perhaps 36 2a (4B). visvddevdn is probably to be restored for visvdn devdn in viii 10 2c (6A). 8b (10). stavdn. No such word is known in Sanskrit. i 174 sdnuni neut. but traces of the long quantity are found in the earliest parts of the Rigveda. 49 la (9) and *x 95 4c (9). pi. 179. i 155 lc (3T) is quite consistent with the metre. in which stavdn appears to have been falsely corrected to stutdh. the dual and locative forms before vowels For index to the symbols see page 148. See on sdsmin below. but seems to require emendation as regards the first syllable. vajrin (5). *x 125 lb. : : 4a sdsmin: a locative form sdsmi should perhaps be restored in see on ydsmin above.

In the table ( 180) long quantity is presumed for all variant vowels in the positions of Group ( 157). 4b Group E. and also in the positions 5 A. This change is connected with increasing strictness in the Rigveda in the question of hiatus. most common in the cretic period. less than 150 in number: but they are unmistakeable. if used with hiatus. tena remains unexplained we are not entitled to assert that -d is the older ending. in the instrumental of nouns in -end. Whether such forms as usdsam. the two endings being metrically equivalent and there are traces of similar change in locatives of -i and -u stems. and but little more in the text but the Pada-patha forms shew : that at a later period most of the forms had short final vowels. vasu-. in the perfects of verbs in -a. such as the duals. as well as in the verb-forms in -d. 6 A. Quite distinctly the statistics shew that the long final vowel in adyd. -hi. (as paprd. So far as these are evidence of the history of the Rigveda itself. we can trace with certainty the same process within the Rigveda itself probably also in the imperatives in -dhl. and also in the positions and short quantity for all vowels in the positions of 2t. In the adverbs in -u. A : (i) SJwrtening of final long vowels. in the verb forms in -md. and in : of puru-. -thd. retain the long quantity. when (as we shall see) the long syllable in the 8th place The occurrences taken together at least was most rigidly required. -svd (thematic). 2d. and can most naturally be explained as intentional revivals of archaic or supposed archaic forms. end but yena. forms of 158 we have hardly any trace of the shortening in the metre. 10 . but later the long value admits of no exceptions. with the assistance of the table in the next section. however. and -td. whilst the duals and adverbs in -a. are the oldest Yedic forms of the strong cases of usds is uncertain in compounds : any case they do not survive the period of the Rigveda proper. are. -u are regularly short before dissimilar vowels. and 6B. and in the paroxytone adverbs in -trd is (for the Rigveda) a comparatively late development. : (ii) Lengthening of final short vowels. and yddl. changes they indicate that the hymns of Group I are its earliest part. ddhd. (iii) Developement of final -a into -o.History of quantitative change 145 are occasionally short in the earliest parts of the Rigveda. dthd and md. The oldest poems not only shew considerable liberty in the choice between hiatus and combination. but also where a final long vowel stands before hiatus a further choice between the full value and shortening. In this direction considerable change can be traced in the Vedic In the period. -a v. In the popular Rigveda we observe a series of forms which formerly ended in and -a. it is true. The paradox of the existence side by side of the forms end. regularly used before hiatus written in -o or -dv. The accent of the adverb is probably of importance. papraH) and in words ending in -ai : A. But in the greater part of the Rigveda final -I.

an aorist imperative in -svd but a thematic imperative in -sva an adverb end but instrumental forms in -ena. Wackernagel. But that the verb-forms in -a. -dhi we cannot venture to affirm on the authority of the Rigveda it can only be said that the oldest Vedic writers most freely use the long vowels. But even within the Rigveda the shif tings of quantity cannot be systematically connected either with the position of the accent or with the effect of neighbouring heavy or light syllables. -dlil are in any absolute sense older than the forms in -a. are of a phonetic character and the very different explanations given by the writers on comparative grammar are not easily reconciled with the facts as given in detail above. -ta. -i. and it is remarkable that the poets should use as regularly as they do a perfect in -ma but past tenses in -ma. Many flexional endings employ forms with long and short final vowels side by side. and that the long vowels of the Sariihita text are not as a rule artificial protractions on metrical grounds. so far -au. -tJtd. This example serves to emphasize the technical and. with consequent shortening. and is found at work on a still larger scale in the later history of Sanskrit (J. -tha. -u it is perhaps faithfully recording an earlier stage in the language than that indicated by the occasional appearance of -a. to some extent. -td. but imperfect records of an older stage of the : : y : language. and are attributed to the shifting of the accent away from the final syllable. arbitrary character of the sharp distinction usually drawn by the poets between long and short vowels. (iv) Protraction of final vowels of stems. as the evidence of the Rigveda goes. The amount of material available in the Rigveda is too small and not sufficiently homogeneous to lead to historical conclusions. contrary to the ordinary pronunciation. -n. Altindische : Grammatik. Variant final vowels are found in Latin also on a large scale. To the wider historical questions suggested by the changing quantities of the Rigveda only slight reference can be made here a summary of the widely differing views of the writers on comparative grammar will be found in WackernagePs Altindische Grammatik 266 c. Amongst final stem-vowels perhaps a majority have historical justification. According to this view the pre-classical writers of Latin comedy shew the later historical facts. for instance in redeo. 41-43). . Only on a broad view of the whole evidence it may be said that the shortening of long final vowels is a phonetic tendency of early Sanskrit. and are therefore found in the earlier parts of the Rigveda but phonetic lengthening is an increasing force. -l. whereas the classical writers almost everywhere restore the archaic long vowels.146 and Problems of comparative grammar These changes. which have a syllabic increase to -a u. The analogy of Latin might incline us to think that the long vowel is in all cases the older form or again we might look to the accent as the original cause for the discrimination. dma. When the Rigveda indicates a long vowel in the neuters plural in -a.

Table of quantitative changes 147 180. Refer to . History of Quantitative Change in the Rigveda.

Symbols .148 Table of symbols 181. Tabular explanation of the symbols corresponding to the various positions in the verse.

(b) in Anustubh stanzas. and (c) in Gayatri stanzas in all parts of the Rigveda. which make up about one-third of its whole bulk. combined in many different in the other groups of the archaic and which trimeter and dimeter verses are arrangements. For this purpose six groups of hymns are selected . and less often in Groups III and V. chiefly in Group II (Atri and Vimada hymns). which will form the chapter. The Rigveda contains some 15. . DIMETER VERSE.^ s. it is In order to appreciate fully the history of dimeter verse. and stanza substitutes a trochaic Gayatri is found most frequently in the Kanva ' and ' epic Anustubh stanza consists of two pairs of dimeter verses. VI and VII . tions from this scheme being comparatively common in the opening of the verse and occasional in the cadence. The general character of dimeter verse and its principal varieties have already 1841 above its distribution amongst the been described in : 1 ' homogeneous groups made. necessary to replace these general statements by definite measurements. in each of which the cadence of the first verse and the ' the opening of the second approximate to the rhythm ^ -. namely ^ . (ii) the * Trochaic rhythm hymns (Group V) (iii) in the cadence. 182.CHAPTER VII. but especially in Groups V. (i) is shewn in 114.000 dimeter verses. The statements already : may basis of the discussion in the present be shortly recapitulated as follows the general type of dimeter verse follows a metrical | v w variascheme of an iambic character. Verses of this type are found (a) in the lyric metres which prevail in Group I and are fairly common and in strophic periods.

VIII. and again in the odd verses in epic but in the other groups it is distinctly less common.^ -).150 for Typical dimeter groups 185. is represented by all the hymns composed : grouping For epic Anustubh the forms which appear in the odd and even verses are tabulated separately. Consequently Group I is its hymns in lyric metres. and may be distinguished as . but they are quite sufficient in number to illustrate the general character of the different kinds of verse whilst there is an obvious advantage in adopting every precaution for the homogeneity of the groups. In metres other than epic Anustubh the rhythm of the different verses in the stanza is approximately but not exactly the same see below 186 (last paragraph). For Trochaic Gayatrl it is necessary to disregard the grouping of 91. . The list of hymns is given in 197. the iambic form (^ ) : in all of these forms the fourth The 'normal' opening. and IX are altogether omitted for the present. because they contain comparatively little dimeter verse. comparison in the table in These groups correspond 91. predominant in the early Anustubh : Anustubh group. mainly in that metre.^ syllable is long. and . The hymns tabulated in 185 include rather less than half the dimeter verse of the Rigveda. no distinctive metres. Trochaic Gayatrl. but are generally to the 'homogeneous groups' defined in to the further restriction that each of them consists subject most its entirely of hymns in which dimeter verse appears in the form characteristic of its group. table. : however. IV. and the -). is common. in as appears from the every form of dimeter verse. usually including rather more than one-third of the whole number of The 'iambic' opening is almost equally common with it verses. quantity of the first and last syllables being indifferent. The independent. each part In the opening three of of the verse has eight possible forms. Groups VI and VII by their Gayatrl hymns the popular Rigveda by the hymns and fragments in which Anustubh is used without the admixture of Paiikti or Mahapankti. 183. Group II by represented exclusively by Anustubh hymns. : of dimeter verse appear to be and may therefore be discussed separately. In Group I hymns in Brhati-SatobrhatI are not included. because this metre is common to Groups I and IY: whilst Groups III. The opening and cadence these forms are relatively the normal form (syncopated form (. without regard to the previous the larger part of these hymns belong to Group V. on account of the small amount of verse composed in this metre.

except in the reopening Anustubh. on the Even the adjacent contrary. division into feet of two syllables is not traceable in the Rigveda. but is transferred from the second to the third in the and also in speaking of the general iambic syncopated form rhythm of the verse as a whole but it must not be assumed that the ideas which these words connote were present to the Vedic poets. The 'syncopated' opening three times as common. v^ ^. the distaste for the iambic and irregular forms is more marked than the preference for any special forms: but the three most common forms are those in which two out of the three distinctive syllables are long. The division. There is some practical convenience in speaking of an ictus which falls normally on the even syllables. The forms of the opening which contain syllables in the second and third places may be considered as are not common in any part of the Rigveda. without the comparative preference for * ' ' : ' : A ' ' : : In the reopening the syncopated form shewn by the Gayatii groups. The complete change of cadence found in Trochaic Gayatri has no effect upon the opening. fourth and fifth syllables appear to be without effect one on the other for in the comparatively few cases in which the fifth syllable is long. is ten times as common Of the irregular forms ^ w ^ ^. and therefore the former form in the first three groups is not very ^ ^. They least of all in the Gayatri and epic Anustubh groups.^. however. : There are three corresponding forms with short fourth syllable ^ ^ . the fourth syllable is also long in the usual proportion of three-fourths or more of the instances. into members of four syllables each seems to be fully established. with the exception (as consecutive short usual) of the reopening in epic Anustubh. are each about one-third as of these the forms ^. The opening of epic Anustubh corresponds almost exactly to the opening in the lyric metres. in which they are relatively much more : common common the remaining form .Dimeter opening* and in the 'reopening' ( 151 41) of epic Anustubh it is only r. by increased strictness in the cadence.^ ^ is only about one-sixth as as the corresponding iambic form. except slightly to increase the frequency of all the less regular forms whilst the change of rhythm in the reopening of Anustubh verse is accompanied. and therefore the usual terms applied to the Greek and Latin classical metres are unsuitable.udv is rare in the early Anustubh and in the opening of the odd verses in epic Anustubh. and in the reopening of epic Anustubh it is more than found.. and irregular. markedly less common than either of the forms More strictly therefore we should say that ^ w ^ ^ is irregular as^w . common of epic as the corresponding forms above. but in the Gayatri groups it is comparable in frequency with the iambic opening.

there is only one regular form of cadence in dimeter verse.152 throughout Dimeter cadence the Rigveda. and ^ v^ ^ - in the Gayatri and epic Anustubh groups. the most Gayatri. syllables are long. and the statistics indicate that 'heptasyllabic' verses should be treated in the same way. The special cadence of Trochaic Gayatri can most easily be explained as borrowed from Tristubh metre. with Trochaic common that is. In every one of the groups the variation which differs from the iambic form only by shewing a short sixth syllable is about as common as all the other variations taken together. This explanation would also account It is for the comparative frequency of catalectic verses in this metre. In Trochaic Gayatri the most common forms of the cadence have trochaic rhythm.and ^ ^ ^ * are rare. whilst the trochaic form is among the rarest . According to my own calcula- tions. in the proportion of 11 to 10 there is perhaps a slight natural excess of long syllables in the language. ' : ' taken together hardly amount to one-tenth of the whole number of instances in the Gayatri groups they are only one-thirtieth. Trochaic Gayatri and elsewhere. and finally in the full cadence of epic Anustubh only one- hundredth part of the whole. namely the iambic form vy .^ -. is All investigators are agreed that the quantity of the first syllable indifferent under all circumstances. It is Catalectic and heptasyllabic verses are much more common in among the odd verses of epic Anustubh than generally agreed that the quantity of the final syllable is in all cases indifferent. therefore convenient to treat ' catalectic verses as shewing a variation of cadence. after the iambic form in the two forms is ^ *. In the lyric and early Anustubh groups all the variations from this standard 184. but the length of the seventh syllable is of more importance than that of the fifth. and next to it which both the sixth and seventh . The iambic cadence is ^ . as position is indeterminate. Except in Trochaic Gayatri and the semi-cadence ( 41) of epic Anustubh. No exact method of measurement is available. : the rhythm therefore has little or nothing in common But whilst every possible form is permitted. ' . the forms fairly common In the semi-cadence of epic Anustubh the iambic form is still the most common. a long syllable : is more common in this position.

. Table shewing the distribution of the variations in Dimeter verse.Normal dimeter rhythm 153 185.

61. In both points there is approximate agreement between the lyric and Anustubh groups on the one hand. (8) Gayatri of Group II. in i . . 7-16. 13. 120 1-9. : 4-6. 17 1-13. The groups chosen are 27 in number. 48. 18. 7-12. 15. . : Strophic period 33 1-15. 93. i 14. Paiikti and Mahapankti. but some of them are much smaller in this : possible to bring almost all the dimeter verse in the Rigveda under review. and in the sections just named . viii 31 of Group III. and the two Gayatri groups on the other. (14) Trochaic Gayatri (see 197 ii). 87-89 4 90. and also i 84 1-6. 17. 98 7-12. viii 14. i 88. 175 2-5. 30 1-6. 70 1-12. viii 38. (21) lyric hymns and verses in bks iii and iv 94 i). 89 5-7. also . 99. 51 5-10. ix 98-100. 103. 132. The groups contain on the average about 500 verses. as Normal and cretic periods (15) to (20) Gayatri in six groups. 75. 3 1-3. i (11) 172. 70 14. . (7) Gayatri of Group I. . i 84 19. (3) Kakubh-Satobrhati. iii 13. (4) Uneven lyric metres. 25. 49-54. 93. 28. 69 17. 82-85. (6) Brhati-Satobrhati. 102 1-18. (5) Anustubh of bk v. 79. .154 common. 98 1-6. viii 69 1-3. 84 7-9. 31 i-t 4 43-45. 15. ix 111 ix 108. namely so as to be homogeneous both in period and metre. and 7 have one or other of the special rhythms. 60. 1-3. i 2 4-6. 59 7-ro. (13) Gayatri. v 24. 62 r-6. ii 41 1-9. v 6. 74-93 and vi 197 i) (10) Trochaic Gayatri (see i 80-82. 92. and the change seems therefore We have therefore the due to time rather than to metre. 14. (2) Atyasti and allied metres. 78 n. 23-26. of the normal period. 101 . 27 7-9. and ix 107. viii 12. 11. i 127-139. viii 1-11. 44 1-6. in i 36-50. 66 1-14. 10-12. We proceed to tendencies by examining them which together include nearly progress of in several small groups of all trace the these hymns of the the dimeter verse Rigveda. vi 46. 18. 97 1-9. x 22. whilst in the Gayatri of Group VI they are only found in 5 verses in each 100. and in bk v and viii 60-79 (9) Gayatri : The groups are composed as follows i 79 Archaic period (1) Usnih. Of the 27 groups 20 use exclusively way it is the normal dimeter rhythm. 27. shewn in the table. 105. ix 101 1-3. viii 80 1-9. ii 5. 15-18.' v 53. (see : . 72. viii 17 14. (22) Anustubh etc. viii 19-22. i 10. and (ii) the diminution of irregularities. 20. 74 13-15. (12) Brhati-Satobrhati. 158 6. v 56. vi 43. 24 1-3. 176. and that Small dimeter groups (ii) in the early Anustubh hymns irregularities either of opening or cadence are found in 15 verses in each 100. 42 1-3. 101 4-16. 95. 41 16. 78 51 '13-15. vi 2 1-10. 92 13-18. which may be shortly presumption called (i) the increase of syncopation. 26. viii 32. v 40 r-3. x 21. 84 10-12. n. ix 102-106. though there are a few hymns in which the metres are so mixed as to elude classification. vii. and they are selected on the same principle as the larger groups in 185. 71 io-r 5 77 10. 143. of two historical tendencies. v 87.

and in all the groups of the normal and later periods except in ix 1-60. 52 i. and in the normal period the Gayatri hymns of ix 1-60. 42 4-6. 3.Syncopation 21 . 47. of the cretic period. from the early Anustubh hymns. 155 22 4 53 ia. 3 47 1-12. 186. and the Gayatri hymns In the strophic of Group III. x 9 6. iv 7 2-6. Gayatri. then. In all the lyric metres irregularities are twice as common in verses beginning in Gayatri those in the first. 97. Pankti and Mahapankti. in all parts of the Rigveda. 134. (27) hymns in mixed Anustubh. 102 57.4 viii 33 16-18. see 198 i. 9. copated forms decidedly outnumber The occurrences of the iambic and syncopated openings are far too numerous to be recorded separately but more detailed information is the end of the book. 8. In the Anustubh groups the variations in the odd and even verses where Pankti and Mahapankti are are distinguished throughout mixed with Anustubh the additional verses added to the respective distich s are reckoned as even verses. . On the other hand the syncopated forms are relatively common in two of the Gayatri groups and in the Pankti. so far as can be lyric This increase appears to continue in the popular Rigveda. : : With regard to the developement of syncopation it from the table in 195 that in the archaic period the appears Kakubh-SatobrhatI. For other metres this distinction is not made in the table. 19-22. x 131 4 133. do not widely differ In all the groups named. given in the Table of hymns at : * ' . whilst stantially in agreement with the early Anustubh hymns 187. As to verses of four syllables see 192. see : Popular Riuveda i . 105 1-6. yet it is very marked in the hymns of these periods considered as a whole. judged from the small amount of verse in the metres and in Gayatri for in groups 24 and 25 the syn: the iambic forms. 142. and third verses are respectively in the ratios of 5. iii 28. 60 1-5. 7 198 ii and iii. 176. . a. . 175. 37 5-8. Although the increase of syncopation is therefore not equally observable in every group of the normal and cretic period. x 62 5.Mahapankti hymns of the archaic period. (25) 94 i) (24) Scattered lyric verses (see 23 16-18. iii 8 1-9. do not widely differ from them. and 2. . and uneven lyric groups are subUsnih. 93 9-11. . viii 39-41. 101 4 6. 158. 153. 9-18. a distich as in those concluding it second. i 28 1-6. 119. period the Brhatl-Satobrhatl and Gayatri hymns. 189. the iambic forms preponderate largely over the syncopated forms. (23) Anustubh etc. 29. 48. (26) epic Anustubh hymns and fragments. : the Atyasti and Brhatl-Satobrhatl hymns. 57 1.

of the whole number of verses. which bears the ictus is In the opening the short second syllable very often one which contains a final is vowel: in the cadence the short sixth syllable almost always an initial syllable. it is discussed further in 196. normal and ere tic periods the proportion is generally much smaller. As the variation is in all cases between 188.. In Anustubh irregular forms are generally more common in the odd than in the even verses. in the uneven lyric hymns. : early Anustubh. to have some special suitability for positions which call for a long syllable : but if this were . For the cadence of hymns in the uneven lyric metres see below. In the strophic period the average is only slightly less. 191. the irregularities being generally about 3 per cent. But in detail there is a irregularities are again : remarkable difference. In the strophic. again. as constituting a In the odd verses of cent. Pankti. it is of little cadence in this group must be considered special type. but as it would seem. In the archaic period departures from the normal rhythm reach the proportion of 38 per cent. whilst the hymns in Gayatri. importance in comparing the groups. so that the Irregularities in the cadence narrow limits. Anustubh hymns of the of the odd verses are verses. the variations reach 13 per is the general average of the archaic period 7 per cent. wrongly. If language appropriate to the classical metres may be permitted. and Mahapankti are much more regular. The irregularities of the opening^ ^ ^ ^) are substantially of the same type as the chief irregularity of the cadence ^ ^ ^ but the latter occurs the more commonly. shew much greater variation. 196.156 Irregular forms As regards irregularities in the opening the groups of the archaic period shew great uniformity. Later we find that the two Soma groups ix 1-60 and 61-67 shew an extreme regularity. either for some specific reason or merely as final vowels. The final vowels found in the second place have been thought. we should say that in each case a short syllable bears the ictus. whilst the Kusika groups i 1-30 and iii do not In the greatly differ from the standard of the earlier periods. cretic period irregularities in the many times more common than in the opening even In the lyric and Gayatri hymns of the popular Rigveda common. For the 'short syllables which bear the ictus' a quantitative explanation has often been sought. but in the popular Rigveda it increases.

l . 52 lc 3a 6c 1 . . At the same time a reciter would be tempted to give an artificial intonation to words so occurring. *75 12c . 45 31a 47 23c' 24a 82 7c 9c8 87 ed 1 . 2b. 129 3c 8c 130 3g 5c. . and it has been noticed that the same initial But it has not been shewn syllables frequently recur in this position. . and this is probably represented in the text by the long vowel or form in -o for which in many cases we have seen reason to restore a final short vowel ( 167 iii. *93 3c. addition there are about 50 instances of uncertain or protracted character. . viii 2 27a 36c 56 6b 59 9c 61 lib. 105 5c 6a 15b. . : 189. 1 1 . etc. and is not determined by it and by this test all In the lists of the syllables in question are genuine short syllables. 91 13c 15b. occur in the sixth place have been supposed to have some special affinity to long vowels. 52 5b 8b. 61 3c 4c. 30 8b. 6 25b 31c 47a. but of individual instances of it and this may account for the apparently arbitrary difference between the two irregularities. If we consider that irregularities of both kinds are most common in the earliest parts of the Rigveda. ii 6 4b 11 8b. . . as follows : The occurrences may be grouped The second syllable contains the final vowel of a word or of (i) i 4 the prior element of a compound. 187 4a. . . 171 iv. 31 |4a 12 5b 13. one-third of the opening are about 350 in the instances the second syllable in . 1 1 . The In irregularities number. 7 2a 1 41 flOb . 19 5a 6b. or substitutes -o for final -a ( 171 iv). or -a as representing final -ah or -e but only in some 20 instances is the third syllable of this In several instances the text gives a long final vowel In second place. 175 la. It is of the essence of quantitative length that it determines the position in the verse. . 12 27b 31b 13 9a 12a 28c 8 17 26 lib 5 27 15b7 28 5c' 21 9a 15a7 22 9a. . that in any case any very large proportion of the occurrences is found in the sixth place and as in the previous cases. 7 10a. . not indeed of the older practice as such. 12 7 1 vii 32 |24a . attention is called to the special forms and words : which have suggested a quantitative explanation. . 4b 1 7 7 8 27 4a 35 3a 50 2b *51 14c. 14 5 v 6 9c 5 . . 5 8 5 4 lid 8 5 14a. 1 . in the ' ' final vowels is ( 162. 16 25b. it becomes probable that their later use was largely due to reminiscence.Irregular openings so. . 32 lib.*30 20a 37 3c 1 10 4b 4c 9 8 1 1 13a. 134 6d 142 7 12 iv 9 iii 24 2a 51 11a 7 . 157 find them used also in the third place to make up whereas they are openings. and in the sixth place syncopated Again the initial syllables which extremely rare in both positions. . 10 5 1 11a4 158 6a. . 28 2b . 127 10d lid. . it is noticeable that these syllables are very rarely indeed found in the second or third : : we should place of an irregular opening. . . : . or -a representing final -ah or -e 5 6 7 2 3 4 13 11a 22 2a 18a. 74 6a. occurrences. 23 15a *24b. 24 9a 14a 30c 16 4a. . . 20 4c 5 16 6 48 12b 1 vi 2 10a. contains a short final vowel. 163) in the second place where the third syllable short. 1 1 . however. 1 1 . 74 6a 81 2a.). .

68 Id 10c. 23 4b. of -e 1S stem in 8 -an. 41 4a 4d. 7b. 27 3c 1 1 iii 45 lb iv 48 36 15a 42 5c. . *146 6a. The . ix 7 lb. 127 7c 132 2d 2 *137 6b 1 144 3b. Final -a or -o for -ah. 18 ddsa 19 (text ddso). ix 61 38 2b. Id 1 . 20 abhi (text abhi). Quantitative variations in normal dimeter cadence are from the standard increasingly rare in proportion as they deviate 190. . *23 16b 1 1 1 1 . 86 2a. 15 la20 2a 20 3a 26 6a 20 23 3b 27 la 20 3a 20 28 2a 20 4a 20 57 4b 5 61 3a 10 62 la7 11a 20 63 4a 7 9b 67 9c 2 99 4c5 100 la 12 *114 3a. 171 2b 1 *172 3a. 59 9a. 72 8a. iii 5 lb. 12 4a 7c. 96 lb . 63 21b. . 31 5b 47 lb. 62 10a. . 94 la. *158 lb 1 69 lie 93 9b . ha (text ha). 33 6a 14b. . Other instances (iii) 42 3c. 105 ii 8 5a. 46 25b 1 60 9a 9b 9d'. 67 14a. Final 17 before vowel. 52 3c 10a. . 1 Imperative in -dhi. *141 5a. 74 lc vi 16 19a. 69 13a. . 6 46a. . *91 3b 6 92 2c 93 20 20a 34a 1 98 5a 2 *102 19a 19 ix 3 7a 20 8a 9 7c. 130 4d. v 7 2d 3a 7d 8b. . 128 le. 14 6c. 176 lc 2 19b. 12 9a. . 1 l '. vii 66 3b 8b. 4d. 78 lb. . 92 lb 8c. 1 1 . 9 14 prdti. . 87 8d . 46 19b 22e 16 27a7 54 7a. : 1 third syllable contains a similar final vowel i 8 9b 26 5a 9b. ca. 4 14b. 83 7a 6 88 5a. v 35 5c. i 3 reading vdruna. 1 . 26 24b. viii 2 3b 9a 26c. *174 2c 2 *186 lb. 93 14a. 46 12a.158 1 Irregular openings . 1 . 20 26b. . in -a. *145 6c. -hi. vdsupati. 2 Neut. 84 2a. saptd (text saptd). 166 i. 50 6a. 1 1 1 1 . 103 3b 24 3a. x 26 2a 5b. 84 4b. 67 6c. 5 35c. *97 5d *15a 3a. x *72 7c. ddsa. 66 2b 20a. . . x *90 2a1 126 2b. . 36 7c. 111 2e. . 1 1 . 126 2a. 7c. . pari. 37 5a 6a. 37 2a. . . 93 12c. 84 2b. 66 la 2b. 25 . 80 3a 129 9b *133 2c : ' . . 27 4c. 13 2b. 126 2c. 66 8b. 18 5e. 48 5b 14b. 1 Final -ah. 83 3b 3c. 45 17a. 23 3a. 150 2a. utd. 1 4 dva. 127 7g. 38 9c. 61 13b 14b. 39 4c. viii 19 23c 45 26b. 1 1 1 1 . 22 14b 19b. 56 lb 1 61 16b. *iv 57 5b. . 18 7b. 37 7c. * Protracted or other vowels of doubtful quantity are found (iv) in the second place in dimeter verse as follows i 22 4b. *85 35a. *22a. 46 4b 8d. *112 x *59 8d = *9e = *10e. 78 lOd 93 31b . 19 16b 20b. 6Q 7c 12d. 5 utd (text utd). 102 13b *19b. . vii *55 3 11a 12 2a 2c. 63 10c. *136 3a *6a. (ii) esd. : 2 abhi. . 17 4a 8c. 98 2c. 34 3a. *55 5d. . 45 31a 34b. 89 5c. 64 20b. . . *97 3b *17a4 102 lb *119 6a 19 *lla. 25 8c 9c. viii 1 . 11 2a. . 72 4a 12a 74 4d 7b 7 7c. 12 16 3 7 -a neut. 19 5a 5c 14b. 65 5a. 69 5a. . 31 12a 16a. u nahi. 1 1 1 . 1 30 la 19 32 10b 17a 19c. The distinction between final ' be satisfactorily drawn but specially capable of protraction cannot it will be noticed that the number of the latter is not sufficiently large to affect the general conclusions drawn in this chapter with regard to ' : short vowels and those which are the irregular openings. *97 7a. vii 32 7b 15b 20a. *58 7a7 *86 20a 3 93 13c. 101 5b. 13 2a. 25 5b. 62 2a 67 4b. vi 16 3a 14c 17a 17c *48c. 33 13b *17c 5 34 lc = 2c = 3c'. 175 i. 65 4d. . 16 4b. 54 8b. 71 2c 8b. 17 10b. 6b. . 137 2g. 86 7b 1 129 lc. . 24 6b. . 82 la. . *91 lb. 1 . 35 8b. . . 101 la. iv 32 24b. ddhi. . sing. *113 6b. 68 4a. . . vi 44 la. lipa (text iipo). 10 15 2 6 abhi. 142 4a 172 lc 2b v 5 3a 67 lc 2b t5b 3 71 lb 2b 3b. 24 24a. . 65 3a 14a 30a. 79 lb 9d\ 81 8a. 62 4d. 8 23a. sing. .

10 (ii) 7a. 59 8b. 37 6a 6c. 62 9c. 1 . 1 . 32 15d 46 9a lib 5 47 24c. 22 3c 4b. ix 5 7c. *119 lc = *2c = *3c. 44 3a. 40 7a. 9 3b 4d 5d 6c t7a 17 Id 4b 4c. 4 v 5 9c. 66 la 2b 2c 4a 6a 7a 8b 9a 12b 12d 14d 19a 6 94 8a. 3 24d. 150 2a 3 iii 9 la 7b. and (ii) the short sixth syllable ^ ^ ^ * The syllable long seventh syllable is. 3 dthd (text dtha). *175 3c. 13 12c 13a. 59 la lb 11a. 15 7a 10a. occur with considerable frequency. . 30 9a. 8 6b. 127 4a. 22 7b. v 6 la 2d. 45 23a 29c. 27 10c 13a. 105 3a. 9 2a. 14 Id 2b 3a 4c. 132 Id. 9b 9c.-. 151 iii. 84 lb Id. 23 1 . 8 20a 21c. 1 itthd. however. gtihia. dvase. *91 3b. 64 4a 5d 6a 6b. 80 10a. *113 6b. 17 2b. 13 2d 6a. 129 lc 3e 10c lOd. 22 lb lc 9b. 48 3a 8e. 15 la. *153 3c. 37 8a 13a. 32 4c. 67 7a 18a. 15 9b. 28 4a. 48 lb. 16 17c 18c 38a 42b 42c 45c. 19 3a 9c 35d. 53 4a. 28 3b 32 25c. 52 2a 5b 10a 10b. 84 2b 7a 20b 86 lb 8b 9a. 10 5d 22 Id 18b. 39 6a. In the sixth place the initial syllables of arusd. 82 4b 7a. 86 6b. 24 la. 12 5c. 18 4a 5a 9c. 86 la 5b 11 6e 6f. 36 15a. 37 la 5b 11a. hdvia. . a rare variation. 53 12c 16b. 24 3a. 6 4b. 52 lc 2a. 45 la. viii 1 5b 21b 29a. 56 2a 4a. 25 16a'. 3 6b. 54 Id. 90 2c. 87 le 4d 8e. 61 15b 1 64 2b. *158 2a. 105 4a 5c 8a 15a 17a. 144 lc. 62 5c. 42 4c 9b. 11 3c. 19 5a. . 48 3a. 29 7a. 80 4c. 60 3a 19a 20d. 70 3a. 84 9a. 6 35c 42c. 66 5a. 16 la.^ . 16 9a. 43 7a. 16 2b 8a. and frequently in the semi-cadence of epic Anustubh and the allied Long fifth syllable: i . 78 2b. 55 4a. 98 7c. 14 7a 13a. iv 30 21c 24c. 175 4d. 9 lb 20a. 29 la. 7 8a. Short sixth syllable: i 1 2a. 4 13a. 26 9b. 17 2a. 35 la. 20 17c 18d 20d. 93 lc Id 2c 9c. 2 15a 38a. 7 4a. 20 2b. 134 2d. 44 la. *97 4a. The (i) irregular forms of the cadence occur as follows : 10 la 2c. 70 lid. they may in influence of Trochaic Gayatrl. 2 see on rathas. . 126 3c 3d. 94 12a 12b. 74 5a. 38 2c. 22 3a 6a. some cases be due to the most common words like ajdra.Irregular cadence form. 7 la 2b 14b 34a 35b. vii 16 Id 5a. 5a. *97 la *lc = *2c = *3c. 73 4a. 65 3a 4d 5a 6b. 11 2a 6a 9a. 92 7a. 61 3c 7a 9d 3 lie. . 6 la. 28 2a. iv 1 2c. 15 7c. 31 7a. 52 2c. 30 3b 4a 17a 21c. 61 4c. 18 la 3a 9c. 26 2c 5a 6a. 25 la 3b. 38 2b. 51 lib. 2 lb 35b. 27 9b 9c. 24 4a 1 25 16b. 50 5b 2 *51 14b. 62 14b. 43 8c. . 37 la. 47 12a 12c. 82 4a vi 2 2c 4b 4d 7b 5a 6a 9c. and the syncopated variation v^ ^ . 46 4a 5b. sdvase. 18 lb. 61 2a 62 21c. 53 8c. metres. 102 lie. 30 2a. ii 41 4a 2b. 7 2a 2b 3c 4d 5d 7d. 41 8a. 107 8a 26b. . 36 5a 13b. 50 la lb 2b 2d 4a 4d 5a. *52 2a. 54 7a viii 1 vii 81 5b. . 40 lc. 9 9a. 12 14a 19a 26b. 3 la. 28 8b. 48 la. 19 36c. 46 27b. vi 46 5d. 127 3d 3g 4c 5d 5g 9g llg. vdruna. 101 2b. 79 2a.* are both rare . . 187 2b 7a. 67 2b. 3 12 20c 24a. 159 Hence the most common variations are (i) the long fifth ^ -. 38 3b 5a 10a 2 14b. 176 5b. 65 2b. 42 10a. 137 le 3d. but it is not clear that this frequency is disproportionate to their occurrences in other parts of the verse. x *16 14d. The trochaic variation . in which they are the forms. 25 7b 9b. martitah.

105 4b. 1 . 75 5a. v 5 11a. . 61 8d 14b. 105 16a2 7c. 152 i. 132 2c. 53 7a. . 40 9c.(epic -. 6 10c. 66 9a. ix 5 2a. 48 3a. 187 lie lid. 44 21a. 32 10c. 39 lb. 188 5a2 11a2 2e. ii 41 7a . 4b. vi 47 23a. 1 reading in. . 41 5c. 39 3c. 64 lc. 1 1 1 . viii 8 21a. 62 14a. 74 8b. a 24 viii 1 30a 3 19d 24a1 10 3c. 32 18b 84 2c . . Syncopated ending ( *) (iv) 46 3a. 82 8b. 93 3c 6d. *28 lb *6b. 10a. 6 2 iii text carta. 51 8a2 46 3a. 31 5a 12b. 24 t2a 12 25 le. 29 5a. 29 8b. 65 4b. . 93 33b. 26 25a. 134 5d. v 5 7b 10b 2 7 la2 7c 3 10d\ 9 la 4c. 50 4d. 88 lc. 69 11a 2a 85 8a. 78 7a 7b. 91 15a. 187 lb 5a. 30 2c. 40 3a 1 47 2c. 28 la. 90 la. make up trimeter verses with regular rhythm. 80 lOd lid. 51 3c. 105 8b. 13 la 14 2a 2 18 8a 17b 20a. - reading vwrtidndm and omitting apiciam. 52 8a 53 12a 12b. 118 6c. : i Irregular endings ( ^ ^ ^. J . 23 6a. 176 lb. . . 103 74 ix 33 6a. 60 7a 10b. . . 142 2a. 133 6d. and frequently in the semi-cadence of epic Anustubh and the allied metres. 142 see a. . 46 12d vii 16 la. 84 5a. 47 24a.w -*): i 3 11a. . 32 3a *28 2a *3a\ 40 3b2 59 7c. 89 3a. 24 lc. 74 5d . 5c. 143 la 3d. 37 lib. . except where otherwise noted) (v) 4 8b. 127 129 2d . 1 reading 9 somanaam 5 su-draiiam. . . 29 7b. 99 7a. syncopated. 1 ! 1 . 39 9a1 43 11a1 46 21d' 24a. 12 6c 5 18d\ 66 7b 18a. 24 7a 20a. 27 2d. . 27 6b. 61 9b. 45 7b |22a 7 37a |38a 7 46 lb8 2b 9b 10b 24d B 9 29b. reading luiasatyd. 72 2a 74 5d 6a 75 5a. 101 5a 1 *102 22a. 1 1 1 1 . 64 7a. *173 3b. 1 iii 24 175 3a. *113 4b. 27 5a 13a 18b 21a.160 Irregular cadence 4b 11a 19a 26a. 46 29a. . 7 i. 126 la Id 6c. iii 12 5b. v 5 4a 5a vii *55 2a *3a. 27 3c. 70 5b 6b 8d 10a. 11 4a. 84 2a. 67 27a 28a2 98 7b 1 101 6a 7b 12a 13c1 x 21 7c 5 24 3c 5 25 5c 5 6e5 8c 5 9c 5 26 7c 1 93 13a 1 *113 6d. 118 8a 126 2c 1 *158 la 1 *3a *4c 1 *175 la *lc. 4 2 3 . 76 lib. . 187 5a. 34 la. and irregular endings are also used in 199. 9 5a. 33 14b. 3 8c. 94 lb 9a x 25 9e. 77 5a. 14 8a. text avd. . . viii 9 15a. *166 5e. x 126 1 4a. 105 la. 16a. 66 19a 21a. 94 la 4b 5b. 107 lOd. 30 3a. 59 8c . 44 4a. 83 3c. see 166 vi a. *155 5b. 61 la. 25 24a 30 2b 2 33 la 36 2 Id = 2d = 3d. *191 lOd *10f *lld *llf *12d *12f . ix 101 12c. 134 5e. 22 |3a 5a lie 13d. 47 10a. 68 8a. 9 9a 74 2b . 10 167 3 reading mitra varuna. 4 text abht. 89 5b 1 5c. 61 18b 19a. 1 . 32 24a 39 6b2 46 2c. 40 4a. 63 4a 4d 10b. . 103 5b. 24 7b. 5b 7c lOd. 8 142 vrsabha (text vrsabha). 62 12a 15b. with the verses of four syllables which follow. 67 13b. 62 5a 10a 10 *90 2b. Trochaic Gayatrl and epic Anustubh as shewn in . . (iii) 120 2c 6c 7a 7d. 140 3d. . . 72 2c 6c 7b 13c 17a3 74 la 7d. 69 la 12a 16d. 20 4d 4e. . . 70 13b. 105 6b. 93 3b. 98 8b2 99 7b. 1 . see Trochaic ending (. 1 . 98 6a. 38 5a. . 1 . vi 2 2a. 25 2a 3b 4a3 4b 11 10a 14a 18b. 25 2a2 50 3a. *153 3b. 20 8d. 27 lib. . 132 5c. . 10 6d. see ending cadence). 150 la iv 30 14a. 59 3a. 58 2b. 103 10b. 68 4b 11a 18b. . 156 4a. 19 32a. 1 . 43 3a. 142 7a. . 101 3a 3 ix 1 6b. iii 6. . : vv - i 1 1 1 1 1 . 158 5 The trochaic. 67 25a 26a. 18 3b. These verses. 49 7a 51 la. *102 Id *12a. 106 lb 6c. 17 3a. 44 2b. . 158 6a. 92 10a 14a. 51 9a. 33 2a. 151 iii. . . . 1 . . 185 3a. *101 4a*6a. 25 10b 26 2b . 101 4 x 20 7a. 24 19a. 71 14a. 93 13d. 55 3c. 98 7c. iii 11 6b. 85 la 9a1 98 12a. 72 3a. 151 iii. " 12 ir 174 i a. 16 2a. 61 9c. 80 4a. 1 ending ~ v.

x 22. 105 : V are most noticeable in this respect. v 7 10c.Syllabic variations 191. 48 2a. 60 4a. vi 14 2a. 28 4a. In Trochaic Gayatri and epic semi-cadence verses of seven syllables are not uncommon. v 61 1 6c (yajfiyasah. 45 29b. 46 lib 22e. if we read a-arkse for drkse in 16b ( 151 i). Of hypersyllabic dimeter verses hardly 20 are to be found in (iii) the Rigveda many of them may be removed by restorations already 151 iii). 11 . 84 la. but this seems to be due to the general irregularity of rhythm. 39 2d. 61 16d. 187 7d. 64 la. 69 8a. 70 7a. catalectic verses out of 16. 127 lOd lie. 161 the Syllabic irregularity usually : takes form of the catalectic verse. Outside these Other verses of seven syllables (heptasyllabic verses) are comparatively rare. 16a. 103 x 140 12a 1 176 4c. 53 la. hymns catalectic verses are not very common. 3c. iv 48 lc = 2c = 3c (candrd suggested. vii 66 2a. 50 9a. 52 16a. very large if the syllabic restorations discussed in Chapter are not accepted. 172 la. iv 15 7a. 132 6d. Perhaps the only typical hymn in catalectic dimeter metre is x 26 (Anustubh). 152 i). however. 18 15a. 175 4c (ii) viii 24 30a. 175 lc.' and the cadence is therefore the same as that of Tristubh verse. iii 8 134 3d. as shewn in 185. 86 6a. 26 5b. 2a 13b. viii 4 7a. The following instances are 152 ii). 74 4c 8a. *191 9b *12b. *x 119 11a. 105 5a. 10 4a. 84 135). ix 18 7b. Verses of nine syllables are very rare. and this may be directly due to the influence of the Tristubh verses in the same hymns. 25 13b 22b. 25 23a. vi 16 2b. 31 2a. Hymns i 120 1-9. 68 17a. In uneven lyric verse the proportion of catalectic and heptasyllabic verses is very large. 30 9b. 98 8a. 35 2a 2b. 176 5c. 50 2a. this is fairly frequent in in these to mark an attempt at a Any few such hymns belong to the archaic period. . 74 13a. 86 5a. x 20 2a (omit agnini not so easily remedied. in which the eighth syllable is replaced by a 'rest. 75 2a 8b. 98 4b. In normal dimeter verse the instances of syllabic irregularity are as follows (i) : Catalectic verses: i 29 3b. and 11 must be. The number of such hymns is. Here out of 36 verses 23 may be. 72 2a 7a. with the cadence ^ . *164 5c. 71 10b. as *i 191 6a (prthvi. *113 5b. ix 12 3b.- one or two hymns. and seem rather to indicate careless composition : A : a. 128 7g. 46 32a. 13 lb. . viii 19 37a (prayi-vayi-yoh 1). and it seems probable that in most cases some metrical restoration is required. 31 10c. 129 5g 8e. 17 11a. 27 5c. measured as catalectic or heptasyllabic even the smaller figure is too In viii 68 14-19 (Gayatri) there are 3 large to be due to chance. iii 62 Heptasyllabic verses i 6 4a. and seems new metre. 47 23c. 132 6e. 38 3b. v 7 8c lOe. 21 5c. and hardly seem to be intentional. 152 : ii.

iii 10. The occurrences are : : . x 21. x 93 2b. The difference can as a Thus the native commentators do not distinguish the two forms of Usnih metre.9 98 4-6. viii 46 7. and not a prose formula) has trochaic rhythm. The great majority of verses of four syllables occur in this Dimeter Usnih metre. *191 16d. In T ii 22 (in w hich the division into verses is very uncertain) we find there being tuvisusmah. but are considered as forming. nor single instances. . The rhythm seems to leave no doubt that the four-syllable verse is derived from the latter half of dimeter verse. 24 1-3. complete trimeter verses. viii 3 21d 187 11a lib. sdti/a induh . 15 4-6. Verses of four syllables otherwise combined are found in *i 191 13. and other Usnih hymns contain verses which are capable of being analyzed as above. 97 n. 12. These verses are seldom recognized as such in the native tradition. 13. These instances are 30b 30c. Dimeter Usnih is found in i 84 7-9. There are about 150 verses of four syllables in the Rigveda. *126 ix 67 v 7 7d. 25. in vi 51 1 5d gopti amd altogether eight occurrences of these variations. 18 t. ix 102. All of them have strict iambic rhythm (as tdm Imahe above) with the In i 84 7-9 the refrain is indra angd which (if following exceptions. ii 22. 91 17. (iv) i 187 la. viii 12 28c = 29c = 30c. vi 43. which may be regarded either as a ' ' shortening of the Anustubh stanza or an extension of the Gayatri stanza. 18. Dimeter Usnih hymns cannot be quite strictly separated from Thus the hymns named above those in the corresponding lyric metre. *x 85 46c. it is metrical at all. the second that which is regular in trimeter verse and if the former rhythm is regularly repeated throughout the hymn. with the dimeter verses to which they are attached. 106 1-3. ix 60 3. 109 22. rule be readily detected by the rhythm. is it But if we compare \ possible to do so confidently in all verses such as yena hdmsi ni atrinam dturtadaksa y tdm imahe viii viii 12 led vrsaria vrsanvasu first 26 lc is we observe that the has the rhythm which regular in : dimeter verse. 62 7-9. 51 13-15/ viii 12. 126 1-7. there can be no doubt that we have before us the combination of a verse of eight syllables with one of four. *vii 55 2-4. 98 10-12. 13 13c 15c. 8-10. v 51 5-7. not included in the tables. The rhythm of verses of four syllables is closely associated with that of the cadence of dimeter verse. *158 2b *2c. viii Defective verses containing only six syllables are found in 102 7a. contain the following trimeter verses iii 10 6c. and a break also regularly follows the eighth syllable. 172. . 192. *51 15a.162 i Verses of four syllables 6a.

Further the stanza that follows (13) is so irregular that it cannot be brought under any scheme. It seems unlikely that there is any conscious imitation of older forms. the popular Rigveda variations are rather more common. but in the strophic period it is still is only common. as it stands in the text. In Trochaic Gayatri syncopated forms are fairly frequent As however all the less regular forms of the opening are rather more common in this metre than elsewhere 185). 200. 193. found once in every half as In the hymns of the archaic period the iambic cadence five verses. are not Trochaic endings are in accordance with any known Vedic metre.Regular parts of the special types *vii 163 almost entirely restricted to the archaic period. 4. common. whether 194. Verses of three syllables appear to be attached to stanzas consisting of two dimeter verses in v 24 2. though rudely dimeter. Of the irregularities in the full cadence of mixed Anustubh verse in the popular Rigveda a large proportion are found in the Mahapankti stanzas i 191 10. (Table. and the last verse of each stanza. this'is no clear indication of date. the few variations that occur consisting exclusively of occurrences of the In the mixed Anustubh of long fifth and short sixth syllables. If these stanzas are left out of account. so opening must therefore be due to the instinctive reaction resulting from the growing tendency to exclude this form from the semicadence and the re-opening. for the whole growth of the epic Anustubh rhythms is gradual that the idea of deliberate choice seems to be The renewed favour given to the iambic form in the excluded. 112 . except in the hymn 55 2-4 in the popular Rigveda. the opening of the odd verses is similar to that of the early Anustubh. are further discussed in 198. the iambic opening being twice as common as the two syncopated forms together. The characteristic forms of the cadence are discussed in In the epic Anustubh of the popular Rigveda. we tind much the same regularity in the cadence of mixed Anustubh and of pure Anustubh hymns. 197. 3. and the stages of the transition from the ordinary the special rhythm.[2. These verses. both in semi-cadence and cadence. This reversion to a rhythm which appears in the interval to have gone out of use is remarkable. in the opening. The cadence in the even verses is extremely strict in the Anustubh of the cretic period and in pure epic Anustubh. pure or mixed. begins with the unaccented word tua. The forms of the semi-cadence and re-opening in epic to Anustubh.

.164 195. The small dimeter groups Table shewing the metrical character of the small dimeter groups.

11 4. if collected together. v 19 1. 81 x 41 7-9. . ix 62 4-6. to consist of the following : The groups referred (i) hymns . isolated trochaic verses: i 2 (iii) 41 3. of these hymns see 240-244. : hymns hymns of the archaic period .* quarters of the verses being of the type syncopated are exceptionally rare. If we may trust the general view we have obtained of the developement of the dimeter verse. 61 10-12 viii 16. 3 21. 43 7-9. i 3 4-6. (ii) vi 16 25-27. the same. 6 . vii 38 89 4 2 . 79. viii 2. and only one hymn of of For the further discussion 197. . the uneven lyric hymns represent its earliest : Vedic type. 30 13-15. 38 7-9. as suggested above. The in special are shewn 199 for three forms of the cadence in Trochaic Gayatrl groups of these the first two are the : and strophic periods respectively. long fifth and short sixth syllables. 35. : 71 1-9. 66 16-18. 2. vi 47 24. In all the groups the forms with short sixth and long fifth syllable are rare. 30 1. 22 n. 53 12. iii viii 11 5. The difference between the Trochaic Gayatrl of the archaic and strophic periods.Uneven lyric and Trochaic Gayatrl rhythms 165 . 199. and the trochaic and syncopated forms are the most common so that it is plain that the metre is substantially . 7. rests upon a comparatively small collection of instances it can only be regarded as a possibility. In the cadence of the verses of openings eight syllables the iambic form is again found in about threefourths of the verses. v 68. The third group consists of single trochaic verses found in hymns chiefly composed Trochaic hymns of the archaic in the ordinary Gayatrl metre. . 5 32. of the : . 20. *x 175 Isolated verses are recognized by their having in two at least out three verses cadences other than the iambic and those with The table shews that such verses.. 94 2 . exhibit the characteristics of Trochaic Gayatrl. this type (v 24) hymns has been given in 186 i 150 and viii 35 might The character of the rhythm is shewn in the table The amount of matter is very small. In the strophic period the iambic cadence is only found in one-tenth of the verses and the syncopated form is as common : as the trochaic in the so that it would seem that the relative change is same direction as in the opening. 2. 7 33. 82 7-9. of the strophic period x 185. 19-21 . iv 55 8. The in list perhaps be added. the openings of which were considered in 193. 10-12. 55 4 68 16. forms part of any of the family collections. i 27 1-6. 90 1-5 ii . 10. the chief variations being the long fifth and the short sixth syllable these variations do not agree with those : found in the other special forms. 8. 70.

The forms found in the opening of the odd verses and in the cadence of the even verses generally normal dimeter verse. usually only of one or two attached to older hymns or combined with other metres.166 198. Tristubh. standing separately in the text. viii 47 13-18. and 2 per cent. third. 59 8-10. they illustrate tendencies which can also be clearly recognized in the of dimeter verse in the Rigveda proper. In most of these hymns iambic re-openings are fairly frequent but we only find one such form in each of the hymns vi 75. developement The groups analyzed are composed i as follows : epic Anustubh. 166. epic Anustubh three shapes (i) in hymns in which the stanza varies. (iii) in longer hymns. ix 112-114. 14-16. 145. In all Trochaic Gayatri. 91. the groups the trochaic and syncopated cadences are relatively rare. : : . 60 7-ia. Rhythm of The Anustubh . x 58. and Jagati. are analyzed in the table in resemble the discussed : In the semi-cadence the iambic rhythm thirds of the verses in the first in the last group. companion -form cadence. 10. viii 47 13-18. and have been already the forms found in the semi-cadence and re-opening 199. but only in one-third In the first group the variations are fairly : evenly spread over all the possible forms in the second the epic ^ is the most common and in the third this and its form ^ . in verse of the popular Rigveda appears in number of verses in the is combined with Pankti (ii) fragments. stanzas. of the : instances respectively. such as Gayatri. the of the gradual developement of period of hesitation and experiment . 86. combined with Paiikti or Mahapankti verses (i) 191 1-12. in which Anustubh or Mahapankti. ^ are together as common as the iambic Single hymns can be found in which the new forms and these are presumably the latest in the distinctly preponderate. vi 75. ^ ^ syncopated form ^ epic form ^ in the is After the normal opening * the most common in the second group. Here we have every indication new standard forms through a and although the different stages of this developement all fall within the latest period of the poetry of the Rigveda. that is. In the re-opening the steady decrease in the use of the iambic form is noticeable it is found in 18. and therefore there is no direct connexion between the semi-cadence of this metre and the cadence of Rigveda. 164. is still found in two- two groups.

48. i 23 20. 48. 1 In i 93 1-3 we may obtain 2a. Amongst these hymns there is only one which contains the iambic re-opening more than once. 137 3b. 170 3d 4b. and always shewing regular cadence. 28 8. 159. 170 2-4 ii 8 6. 155. Fragments containing more than one stanza. 164. 104 25. 159. (ii) epic Anustubh fragments. The only fragments in which the iambic opening occurs more than once are i 133 2-4. seeing that in the whole group the iambic cadence is still shewn in two-thirds of the verses. 179 lb. 7. 87 22-25. 7. x 58. 7 v 40 13. 154. 163. 170 2-4. viii 47 13b. 90 9. It is therefore probable that the hymns ix 114. 136 lb. but this is not remarkable. 39 6. 51 16. 190 Id. (ii) Rigveda : : 133 2d 4d. 1. x 85 12b 33b. . 190. 12. 15. 6). ^ in one third of the verses in rhythm. 24 4-6. 167 166. 161 i above) and otherwise (see the semi-cadence in this fragment is iambic. viii 91. i 23 20d. x 9 8. vi 75 13d. 126 6. 14 13. (iii) epic tenth Mandala. 137. 184. 137 4b. 103 13. If the first syllable of arista is not reckoned long. ro. On the other hand i 191 10-12 and x 59 8-ro agree entirely with the early Anustubh type ^ and none in ix 114. x 16 14d. 4 This group is remarkably large. namely x 152. 166 are of as late a date as the hymns generally in the third group. 78 5-9. ix 5 8-n. x 19. 2. 97. 9. vii 55 5-9. single fragments. 142 7. 93 2b. These are found exclusively in the hymns 72. But these readings are uncertain : . in which the new rhythm 135-137. vii 104 25b. 151. and commonly in i 191. x 164. . 16 ji-14. 22 vi 16 47. as suggested above ( 167). 32 6-8 iii 29 iv 24 to. 83 9. 8. are i 93 1-3 vi 16 47. additional examples are found as follows: x 60 8e 9e lOd. 113 4b 8b. 51 14. 141. 53 20. 103 1. 191 Id 2b: (iii) in the epic Anustubh hymns. 19 lb 7b. 90. 50 1013. 1 . 162. (iii) x 97 7d. 136. and adyti in 160 v. 85. 90. and at least once in every ten verses. 14. 173. considering the small size of its It is of course difficult to define the rhythm of the separate members. 161 5 179 1. 48 22. (i) ix 112 2d. 16. 6. 37 rr. 57 4. 5-8. this hymn and x 141 agree generally in cadence with the earlier hymns. ix 5 10b. x 87 22-25. is The hymns x 72. 59 i2. 164 51. 174. 66 15. vi 51 16b 16d. with which may be grouped the wedding hymn x 85. 152. : . 60 7-12. x 59 8b 9b 10c. 4. 87 2225. 133 2-4. .History of epic Anustubh ix 112. viii 33 19. being Anustubh hymns. vi 51 16. Many of the single verses are free from irregularities of cadence. 87 22b 25b. 22-24. . 173. The last fragment therefore agrees with the early Anustubh rhythm in both points. 174. 18 j 4 19 (exc. although their general character as constituting a transition to epic Anustubh is clear. most pronounced are The following are the examples of iambic re-opening in the popular in the mixed Anustubh hymns. x 58. 32. 93 1-3. 173 3b. x 145. 135. 191 1. 33 . v 51 14b. 100 7-9. 109 6. 152 lb Id. 164. which contains long sections entirely composed in Anustubh. 78 8b. 86 in the Anustubh fragments. 67 31. 146. The cadence is of the ix 114. (i) epic semi-cadence by reading su in la. 17 14. 166.

Table of special dimeter rhythms Table shewing the special rhythms of uneven lyric.168 199. Trochaic Gayatri. . and epic Anustubh verse.

142 . so that the hymns may belong to a transition period. earliest period. 105. 34 1-15 the rhythm of the opening might be either early or late. : : : : Table shewing the rhythm of small groups of Anustubh verse. namely i 191 10-12. The hymns of the cretic period. iii 8 x 62. Divodasa has only a few verses in Mandala iv the prevalence of syncopation in both openings seems to The group viii 39-42 shew that the metre is influenced by Gayatri. Groups . . In x 87 22-25 and 141 the linguistic notes of early and late date are about evenly There divided. 176. except in the regularity of The Kusika hymns i 10. remains x 152. but viii 47 1-12 appears to be later. but that of the re-opening rather resembles the late metre. shew a distinct approximation to those of the popular Rigveda in all points they are i 28 1-6. which certainly combines the language of the later Rigveda with the early Anustubh rhythm but too much importance must not be attached to a single short hymn. Of hymns in the popular Rigveda with the old rhythm two are Mahapankti triplets. 11 agree in rhythm with the the cadence. whilst the few verses in bk iii agree with the Divodasa hymns in character. however. The proportional figures for these small groups are given below. 49. but the whole number of verses is generally so low that no decisive weight can be attached to them. whilst also favours syncopation x 131-134 agree with the earliest hymns.The The amount between III and cretic period marks a transition 169 of matter in X is so small that Anustubh verse in the intermediate groups we cannot satisfactorily determine : the rhythm used. 9 r6-i8. x 59 8-10. viii 8. The Vasistha family has no hymns in this metre. 29. The Kanvas use it in i 45.

and the iambic cadence strict. lyric metres. Even in the archaic period the iambic cadence was more stanza. The following acceptable : further explanations on these points may be (i) The position that the Atri earliest in the Rigveda is 132). General history of dimeter verse The analysis of dimeter verse in the Rigveda shews that a series of changes take place in the rhythm. the normal and iambic rhythms are equally common in the type is ' in the cadence the opening. and the book are later in date. strictly A adhered to in the even than in the odd verses : a correlaid sponding restriction of syncopated openings to the even verses The the basis for the developement of the epic Anustubh metre. of this new rhythm are found either in fragments or in beginnings hymns mixed with Pankti and Mahapankti time it : but in the course of which developed into the precise but delicate and varied we find in the latest hymns of the Rigveda. and in one or two in other metres. nearly all of which are also found in bk x. Pankti. and : Kakubh-Satobrhati do not greatly iambic opening rather more is differ from this type.170 201. Many Gayatii but the great hymns have a rhythm of the type just described majority of hymns in this metre differ from it in . usually considered an 'irregularity of these hymns. should also be ranked with the earliest hymns of the Rigveda will run counter to the presumption usually entertained that the hymns these hymns (see hymns in uneven in this Anustubh hymns are amongst the confirmed by the great frequency of hiatus in That the Vimada hymns of bk x. Hymns in Usnih. and metre which already shews the essential features of the sloka of Sanskrit epic poetry. all other forms being occasional iambic rhythm is alone regular. and in the uneven lyric hymns. and Mahapankti but otherwise the developeof syncopation does not coexist with variety of form in the : systematic use of trochaic cadence is found in many Gayatri hymns. evidence of their late date. but variations are common as compared with the later periods. favouring the syncopated openings at the cost of the iambic The same tendency can be observed in several hymns opening. is really an argument But the ' . but the rather less common. Atyasti. may be The ' recapitulated as follows earliest : shewn us in the Anustubh hymns of the Here Atri and Vimada families. which correspond These changes generally to changes in the form of the stanza. in ment Anustubh.

found that some groups of the Gayatri hymns. furnished their proportion of Gayatri verse. viii 66. 87. : in other cases are v 56. : the earliest metre 171 is for the verse of the popular Rigveda. all the materials necessary for the construction of the more elaborate metres. the great mass of Gayatri hymns in the Rigveda is distinguished by characteristics which suggest a later date. as in i 84. and the stanza of three verses seems to be a reduction from the normal stanza of four. the verse of four syllables. Undoubtedly these elaborate stanzas presuppose earlier and simpler forms but it is not necessary that these earlier forms should be such as the Rigveda has preserved. and it is impossible to doubt the antiquity of Gayatri metre as used in a have also hymn like viii 46. the rhythm of the trimeter verses in these hymns agrees with the earliest type of trimeter seem therefore led to the view that verse. namely the dimeter verse. but in the later lyric poetry only verses of eight and of twelve syllables are used. 9 and 18 above (table. is in general agreement with that of the Atri Anustubh hymns. the whole balance of the Indo-European structure of metre is based upon duality. complete and regular hymns are rare. shew in the opening the same general character But whilst the archaic and strophic periods as the lyric hymns. 101 fragments in these metres are appended to hymns in a different metre. Further.Anustubh to the contrary in some points. though irregular exceedingly strict in othors and the rhythm of the : uneven lyric hymns. : : Gay atri on the whole appears to be later than Anustubh and (iv) This is first suggested by the form of the stanza for the lyric metres. In particular. We The complexity of the lyric metres has led to a general belief (ii) that they must be of comparatively late developement. as will be seen later. 195). as in the third Mandala therefore the last survivor of the lyric metres. : We . and the trimeter verse. : : of Brhati and Satobrhati metres is found in In the archaic period these metres are variously combined. it is occasionally used in the normal period. On the other hand in the Vasistha and Kanva collections the hymns are regularly composed in alternate Brhati and Satobrhati stanzas. 69. vi 48. 70 1-12. viii 17. this has been felt with regard to the Atyasti hymns i 127-139. several of the uneven and other lyric metres have a stanza of three verses. for instance Nos. the The Brhati metre is still variations being comparatively unimportant. 77. We find however in the Anustubh hymns of Atri and Vimada. and at the same time furnish a transition to the metres of These characteristics are the growth of the the popular Rigveda. and in the uneven lyric hymns. the tenth Mandala preserves for us many hymns which could not find a place in the 'family books' because they did not conform to a proper metrical standard. 97 1-9. But although the trimeter stanza of three verses hardly belongs to the earliest part of the Rigveda. in combination with lyric metres. other metres being often found by their side such hymns (iii) The combination two distinct stages. as well as their vocabulary. 78. The mutual influence of the dimeter verse of eight syllables and the trimeter verse of eleven is sufficient to account for the creation of a dimeter verse of seven syllables and a trimeter verse of twelve in uneven lyric metre all these varieties are combined.

10 are trochaic or syncopated. p. The use of the trochaic cadence cannot be restricted to any one It is systematic in the Vimada hymn x 20. 2. Critics who are disposed to adhere more closely to the text will of course find a larger proportion of irregularities throughout the But as these will be found to occur most freely in the Rigveda. and there are four examples each of the long fifth and irregular forms.172 Transition metres syncopated opening. striking. the metre being characteristic of a Mandala which contains few. and the differentiation of the verses according to their position in the stanza. is not borne out by the On the contrary. 1 Dyaus Asura (Halle. may be on the whole regarded as irremediable by any process of restoration. we seem to have specimens of analysis of the metre. the Gayatri groups of Mandala ix In the longer hymns 1-60 syncopation is not a striking feature. . shew remarkable favour to syncopation. The high proportion of forms with short sixth syllable. for it is difficult to find hymns in which syncopation is not prevalent. that the Rigveda represents the period of decay of the old Indian lyric poetry. belonging to a group which we have reason to think one of the earliest in the Rigveda (v) period. the more elaborate form of the stanza was metres necessarily abandoned when increased attention was given to the rhythm of the single verse. though we have no means of identifying In the whole of the book the regularity of the cadence is these. 1882). archaic period. which belongs to a Kanva group and records the name of more than one member of the Kanva family: whilst in iii 16 we find trochaic cadence in a Brhati hymn. iii. seems to be specially characteristic of this family. as distinct metres. von Bradke 1 . help to bridge (viii) the gap between the Rigveda proper and the late Rigveda. the increased regularity of the cadence. but are very much less regular in cadence than other hymns of the normal period. and it is probable that many of the hymns belong to the earlier family collections. the general conclusion that that period is characterized by some degree of irregularity will rather be strengthened by such a On the other hand the irregularities recorded in this chapter view. (x) The view of P. expected. There are no catalectic verses and only 2 of seven syllables. It is quite clear that the differentiation of odd and even verses which led to the developement of epic Anustubh was attempted in these but as before. : Throughout this chapter the theories of hiatus and syllabic (ix) and quantitative restoration explained in Chapters IV-YI have been assumed. Pankti and Mahapankti. (vii) As might be elude classification. hymns earlier than the normal period. The Kusika groups i 1-30. and therefore specially valuable as evidence. and of trochaic and syncopated forms. in which dimeter verse is (vi) represented almost exclusively by Gayatri. : but the longest hymn in Trochaic Gayatri is viii 2. Of the variations 31 are of the short sixth syllable. if any. In the archaic period these metres seem to have been hardly known.

and are not necessarily eliminated even by wide reading. equivalent to two short. : 202. it must at least be clear that Vedic dimeter verse cannot be explained rightly by any of the methods applied to modern European verse or even by those which are applied to the classical Greek and Latin metres. The essentially syllabic character of Vedic verse produces first is an impression of great simplicity on simplicity is only on the surface. and in the disposition of acquaintance but this consistent with great skill : rhythmical result. or that the poets would at least wish it to be so if they were not Statistics arise ' ' hindered by intractable material. either in the sense that it is more often short than long.Technical skill in dimeter verge 173 composition in various styles. and so regularly is this the case. Thus many Vedic scholars. and assumes an iambic or trochaic rhythm as the necessary All Western scholars are under the basis for almost every metre. long syllables in this place are almost twice as common as short . Whatever difficulties may be felt to attend the explana- tions given in this chapter of particular rhythms. nor can we trace any division into feet corresponding to the standard feet of classical Modern European verse. leading up from the rude early metres to the perfect form of the Brhati-SatobrhatT hymns of the strophic period. perform a useful purpose in counteracting errors which from acquired prepossessions. The sudden disappearance of this kind of composition coincides with a fresh elaboration of the rhythm of single verses and the latest metrical schemes of the Rigveda. even if less attractive to the European ear. temptation to attribute to the poets of the Rigveda at least a wish to follow the same standard but the facts prove that the : types which really floated before their minds were often of a quite different character. in one period . Yet both these suppositions are demonstrably wrong. can hardly be described as intrinsically less beautiful in design or less perfect in execution than their predecessors. is words and syllables to produce a required which may be of a highly complicated character. that in the whole Rigveda it would be hard to find ten successive dimeter stanzas in which the short syllable is Yet the poet had no difficulty in his material. though based on accent instead of quantity. As a fact. for in equally common. Although Vedic verse shares with Greek and Latin the system of quantities. agrees with Greek and Latin in taking the foot for its unit. the cadence he successfully achieves this very rhythm. conclude that the third syllable is preferably short. having discovered an iambic rhythm in dimeter verse. yet there is no trace in it of the principle that one long syllable verse.

Vedic verse would be premature until the corresponding forms of trimeter verse have been examined. and the latter the second in favour amongst possible cadences. yet entitled to assume that there existed at an earlier period verse more rigidly iambic in character than that preserved to us in the earliest Any conjectures as to the character of prehymns of the Rigveda. It appears on the whole that iambic rhythm lay for the Vedic poet in the past it was part of an inheritance upon which he desired The existence of this rhythm in the very earliest forms of to improve. : ' ' : . neither from the classical nor the modern standpoint do and ^ ^ ^ such sequences as suggest any kind of rhythmical effect yet the former is in all periods the favourite Vedic opening. in another in ninety-nine verses out of a Again.174 hundred. Pure iambic rhythm not sought in nine verses out of ten. Vedic poetry seems to set a great gap between it and the apparently At the same time we are not non-quantitative verse of the Avesta.

165-190. the number of verses being nearly 24. To these advantages another is added. and the absence of sharply contrasted types. which is perhaps the most important of all. Nearly the 'homogeneous groups' defined in 91 contain sufficient . Besides historical developement there are two other possible the individual taste of the explanations of metrical variations . and chance. For field historical investigation trimeter verse offers a verse. If it appears that there is a corresponding developement in a number of other metrical features which have not yet been taken into account. . namely that the Samhita text provides us with a series of collections. the variety of the internal structure. within all each of which the rhythm is almost absolutely uniform. the test suggested in ' 113 may be applied. : though clearly marked. iv.000. 94-115. chiefly concerned with their external form. if ' anywhere. 203. vi and vii can be used almost as they stand in the Rigveda as units of investigation.CHAPTER VIII. much richer than dimeter The greater amount of material. trimeter verse to establish their respective rhythmical laws and such groups as i 31-35. The main features of the verse are the same throughout the Rigveda differences of structure. are concerned with points of secondary importance. poet. it can hardly be doubted that we are within sight of the true history of Vedic metre. TRIMETER VERSE. are revised The homogeneous groups and ordered in as finally defined in Ch. and Mandalas iii. In this field. Almost two-thirds of the Rigveda is composed in trimeter verse. Ill accordance with certain striking metrical peculiarities. all combine to make it easy to trace the steady developement of the rhythm. 74-93.

which compels it to neutralize or destroy subject its Variations on a small scale own creations : and this principle is just as clearly indicated in the rhythm of Vedic hymns as at the gaming tables of Monte Carlo. it groups a natural explanation that these variations represent the bias r left in of a single poet. will take place in every possible direction. with the result that the family name Vasistha. containing not impossible that the initial syllable should be twice as often long as short. It will appear that the to this view. and of w hich hardly any that appear immediately to succeed it. to its own law. though it laughs at all other laws. or perhaps of a small group of poets intimately This individuality has often been associated in their work.176 Between the History and chance historical developement of the metre and the individual taste of the poets no sharp line of distinction can be drawn: the history of the Vedic periods is in the last analysis the history of the hymn-writers who belong to each of them. metre of this collection lends some support may be ascribed to chance. But for practical purposes individual taste is characterized by comparatively abrupt variation. If particular features are found in one group of hymns for which we are not prepared by the groups which just precede trace is is it in time. In a short Vedic hymn. If a ball is thrown at hap-hazard on a table on which exactly half of the compartments are red and half are black. suggests to many Western critics also a striking personality. then in 1000 of his verses the syllable is sure to be short in just about 500. If a Vedic poet is really indifferent to the quantity of a particular syllable. this certainty for instance. some 20 verses it is . that causes which are not directly connected with the is to say. Conversely if there is a decided balance in favour of the long or the short quantity. there must be a metrical motive somewhere at work. alone amongst those of the ancient seers of Indian tradition. then in 1000 consecutive throws the ball must fall very nearly 500 times into a red compartment. to In a large body of verse such chance variations sense of rhythm. and therefore they will have no perceptible effect upon the rhythm as expressed in Chance. As however the number is of chances is decreased. is always averages. recognized in the subject-matter and general treatment of the seventh Mandala. Such cases however will not diminished.

Parts of trimeter verse
often be found
:

177

and any theory that might be built upon them would soon be abandoned as the result of further enquiry. In

the present chapter (as indeed in those that have preceded it) we shall find it from time to time necessary to deal with small
quantities of matter, in which the variations that occur may quite possibly be due to chance. But the uncertainties of the particular case do not produce a like uncertainty in the general conclusions
If the survey of the facts be on the whole the errors in detail must necessarily be relatively sufficiently wide, to
led.

which we are

unimportant. The habit of ascribing the metrical variations of the Rigveda to chance is the necessary result of imperfect familiarity with the
details.
is

The

critic of

metre who has convinced himself that chance

a totally inadequate explanation of the facts presented to him will feel bound to look for some other cause or causes. And since
the phonetic structure of the Vedic dialect is obviously pliable extreme to the hand of the poet, he will necessarily fall back upon the conclusion that the variations which occur are due
in the to

changes of metrical taste
or

:

and whether these changes are

unconscious, the product of the time or of the individual, they belong in a broad sense to the region of historical
conscious
investigation.

204.

The general
in

explained

18-35,

structure of trimeter verse has already been its distribution amongst the 42-56
;

shewn in 114. As with dimeter verse 'homogeneous groups' in 183, so now we begin a more minute study of trimeter verse by a general sketch of its most important features, and by giving in figures a precise measurement of the part filled by each in a
is

of groups, corresponding generally to the 'homogeneous 91, but so selected as to represent those parts of the groups' of
series

Rigveda of which the homogeneous character is most assured. The questions to be investigated fall naturally under the
following
'opening,'
(iii)

headings: (i) the caesura, (ii) the 'break,' or rhythm of the

the

fifth,

syllables, (iv) the cadence,

and

(v) variations

rhythm of the and seventh in the number of
sixth

syllables, so

far as

they are not included under the preceding
are the
(ii)

headings.

The groups investigated
period
a.
(i)

the

lyric

hymns, and

following: in the archaic the Tristubh hymns of
12

178

Typical trimeter groups

Bharadvaja (Mandala vi); in the strophic period (iii) the Tristubh hymns of Vasistha (Mandala vii) in the normal period (iv) the hymns of Vamadeva (Mandala iv), and (v) those of Kusika (Mandala iii), in each of which groups the Tristubh metre and in the also found prevails, but a few Jagati hymns are
; ;

hymns, almost equally divided between Tristubh and Jagati metre, (vi) of Kutsa (i 94-115), and (vii) of the small groups of hymns extending from x 29 to x 80. In the eighth
cretic period the

and

last

group are included

all

the

hymns

of

the

popular

Rigveda.

The
that

statistical results for these eight

groups are given in the

Table in

212, and are the basis of the discussion in the sections
follow.

now

Although the general type of trimeter verse is on the whole (i) the same throughout the Rigveda, it is necessary to exclude wholly or partly from consideration here certain hymns which have a very
distinctive character. Many verses which are treated

by the native authorities as of the trimeter type have already been analyzed as consisting of dimeter verses with verses of four (or three) syllables attached. This is parDimeter TJsnih ticularly the case in the metre to which the name of Verses of this type are entirely 192) has been given in this book. ( excluded from consideration here. Hymns in the uneven lyric metres (27) form a special class so far as the number of syllables in the cadence is concerned. Other variations which they contain are included in this chapter. Hymns which contain 'decasyllabic variations' ( 49-53) in any large proportion fall into two classes, which it now becomes important
' ' ' j

to distinguish.

The first class consists of hymns in which different decasyllabic variations are found, in proportions varying from one verse in ten to one verse in five. These we now name as a class hymns in decasyllabic Tristubh metre, though in fact two or three of them are in Jagati metre, and in these the verses which contain rests are of eleven These hymns are treated separately so far as the decasyllables. but other variations are included syllabic variations are concerned throughout the chapter. The list of these hymns is given in 94 iii a. The second class consists of hymns in each of which some decasyllabic variation is predominant, although some of them contain many verses in Tristubh or Jagati. These we now name hymns in decasyllabic metres, with the same qualification as in the last section. Variations of all kinds occurring in these hymns are given in the lists in this chapter
'
' :

within square brackets or in special subsections, but are entirely excluded from the tables. The hymns here referred to are i 61, 65-70; vi 44 7 -9 ; vii 34 i-m, 56 i-n ; ix 109 ; x 1, 6, 46, 77 ii 11 ; iv 10 1-5, 78 i-6 ( 94 iii b c d). The special features which occur in the uneven lyric hymns and in
;

The caesura
decasyllabic
this

179

hymns of the two classes will therefore be discussed in chapter primarily from the standpoint of their occurrence as The hymns in decasyllabic occasional variations in other hymns. Tristubh will also be considered in this chapter as a special class but the consideration of the hymns in uneven lyric metres and in deca:

syllabic metres, so far as their respective characteristic features are concerned, will be postponed to the next chapter.

The complete investigation of trimeter rhythm involves the (ii) tabulation of the quantity of almost every syllable in the trimeter verses of the Rigveda, in connexion with the position of the caesura in In the following points only it has seemed sufficient to each case. take samples of the rhythm (a) for the initial syllable (b) for the regular forms of the opening ( 215); (c) for the occurrences of a As in dimeter verse, we have natural pause after the eighth syllable. no means of determining with completeness the quantity of final syllables, but feel justified in assuming that it is metrically indifferent. Where samples only have been taken of the quantity in any particular position, it has seemed desirable to examine not less than 500 verses in each case.
:

;

1

'

The caesura is the dominant feature of trimeter verse, 205. and its position decisively affects the rhythm both of the opening and of the break. The caesura is a natural pause, corresponding to the taking of the breath in recitation, and occurs regularly in all parts of the Rigveda either as an early caesura, that is, a
pause after the fourth syllable, or as a late caesura, that is, a pause Verses of these two types are after the fifth syllable ( 43). combined in the same stanza. everywhere

The

position of the caesura
:

is

in itself indifferent in all parts

of the Rigveda but indirectly one or the other position may be on account of some rhythm of the break which slightly favoured

depends upon it. Thus in the Vasistha hymns the more often late, on account of the favour shewn to the break J - ^ in connexion with secondary caesura (see below) and in the Visvamitra hymns and the later periods the caesura is more on account of the favour shewn to the cretic break often
is
' '
:

caesura

I,

early,

,,---(207).
In a few cases, chiefly in the archaic period or in the popular the position of Rigveda, there is some difficulty in determining

In 'decasyllabic' and 'hybrid' verses there is always a well-marked caesura, but it is not easy to say in each case
the caesura.

should be considered 'early' or 'late': these verses are 225-230. Elsewhere we appear to find further considered in either (i) a caesura dividing the two parts a weak caesura,

whether

it

namely

122

180
of a

'Secondary caesura

1

compound, or (ii) a caesura following the third syllable. Both forms of the weak caesura are characteristic of the archaic 214. period, and are further considered in Chiefly in the Vasistha hymns we find a variation which we

may term
first

the secondary caesura, being an approximation of the eight syllables of trimeter verse to the dimeter type.
existence
of the

The

'secondary caesura'

in

the
:

Vasistha

hymns may be inferred from the following considerations In all other parts of the Bigveda a pause is found after (i)
the 8th syllable in about 35 per cent, of the verses, which is just the proportion that might be expected if no special rhythm were aimed at. But in the Vasistha hymns this pause occurs in no

than 57 per cent, of the verses. Certain other variations of rhythm, namely the caesura (ii) ^; and - -^ after the third place, and the breaks n w - **, v
less
| (

jj

are found very much more frequently the eighth syllable than elsewhere.

when

there

is

a pause after

The

verses in the Vasistha group which combine one of the

features last mentioned with a pause after the eighth syllable amount to about one-sixth of all the verses in these hymns,

and roughly account

for the higher proportion of each of the It therefore appears that separate variations in these hymns. it is the combination of the pause with some other feature

which characterizes this

collection,

and

in

the Table in

212

the instances in which the combination occurs are considered
separately.
is usually a pause in the sense as well as in the not however absolutely necessary that this should be so and we find numerous examples in the Bigveda in which the caesura separates either (a) the two parts of a dvandva dual, or (b) an accented word from an enclitic which follows it, or (c) the negative particle or the augment a-, when combined by Sandhi with a word preceding, from the remainder of the word to which either of them belongs. As these occurrences seem to have no historical importance, it will be sufficient to give a few examples here namely
(i)

The caesura
It
is

sound.

:

:

(a)
(b)

asmd indra
\\

-varund visvdvdram
\\

vii
i

84 4a

hdstesu khadis

ca krtis ca sdm dadhe dgne tokdsya nas tdne tanundm asm ad pari u vdm isah purucih
\\

168 3d
9 2c

ii
iii
i

58 8a

(c)

sdm

vatsend -srjatd mdtdram punah durvasase -mataye md no asyai
\\

110 8b
1

vii

\\

19b.

Examples of secondary caesura

181

Further examples of (c) are found in i 59 2c, 168 9c, 190 3d ii35 13a; v 11 3c; vii 61 3d; x 61 7c, 68 10b ivll2d; 89 13d, 99 5d, *103 lc *2a.

That the syllable before the caesura, like the final syllable of (ii) the verse, is indifferent in quantity (syllaba anceps) is a theory as old as the Samhita text itself, and finds expression in the systematic neglect to record the long vowels of certain endings in the position 4A 176 ii). Western critics have also often inclined to (see especially this view. There is however no foundation for it in the usage of the indeed the quantity of the fourth syllable is more strictly poets regulated in trimeter verse when the caesura follows than in dimeter verse where there is no caesura and the quantity of the fifth syllable where the caesura follows is better marked than that of the third in the same verses, as appears from the Table ( 212).
: :

(iii)

common

is

Of the forms of the secondary caesura by far the most that which employs the break v, and it is illustrated
s:

by the following examples somah sukro nd
vdsistha sukra
\\

\\

vaydva
(

didivah

ayami pdvaka
|

vii vii

64 5b
1

8b,

the

first

example

having in addition a short eighth syllable.

Verses of this type amount to one-tenth of the whole number occurring in the Vasistha group, and are more than twice as common there as in the Bigveda generally: they may therefore be appropriately termed Vasisthi verses. The frequency of verses of this type in the Vasistha group sufficiently accounts for the preference shewn to a late caesura.
(iv)

The

less

important forms of the secondary caesura
:

may be
:

illustrated as follows
(a)

caesura after third syllable with pause after the eighth

d
(b)

citra
||

citriam bhara
(

rayim nah
navanta
(

vii

20 7d
:

iambic break

u

^ - ^ with pause
u

after the eighth syllable
vii
:

prd dhenava
(c)

udapruto

42 lc

iambic break dtutujim
'

jj

- v with
tutujir
I

the same pause
asisnat
'

cit

^

vii

28 3d.

(v)

distinctly influenced by dimeter entire agreement with it.

Verses with secondary caesura as now defined appear to be rhythm, yet they are by no means in

syllable,

212) 5 per cent, have caesura after the third 23 per cent, after the fourth, and 72 per cent, after the fifth. Of 100 dimeter verses measured in the same way the proportions are 22, 32 and 39 respectively, whilst 8 verses have no break (except within a compound) in any of these positions. Of the verses which have one or other form of iambic break only 12 per cent, have a short eighth syllable. Although this proportion is much higher than that usually found in trimeter verse, it is very very

Of our verses (Table,

182
much lower than

Rhythm of

the opening

that found in dimeter verse, in which the quantity of the eighth syllable is of course indifferent. Again in our instances the fifth syllable is long (as in the last example) in one-tenth of the verses, whereas in dimeter verse a long fifth syllable is seldom found more often than once in a hundred verses. seem therefore to be precluded from using the convenient title dimeter Tristubh for these verses, although it is very suggestive of Neither are we in a position to assert their general character. definitely that this type is derived by contamination from dimeter verse it is quite possible that its leading variety v^ n - ^ was directly derived from the more usual form ^ n ^ ^ (which is only equally common in these hymns) in an endeavour to introduce a more varied

We

1

'

:

rhythm (see 207 iv). In any case it does not seem probable that these verses represent a primitive type for in that case we should expect to find them accompanied by the general freedom of metre which characterizes the archaic
:

period whereas in fact the Vasistha regular as (say) those of Vamadeva.
:

hymns

are on the whole quite as

206.

under

all

In the opening a general iambic rhythm predominates circumstances, as in dimeter verse but the develope:

not only different from that found in rhythm dimeter verse, but also varies according to the position of the
of this
is

ment

caesura.
If the caesura is early,

every group

fall

within the formula

about two-thirds of the openings in * *,-: if it is late, the

proportion is always as high as three-fourths, and in the Visvamitra group it is much higher.

Before an early caesura the quantity of the third syllable is usually indifferent in the groups of the normal period a short vowel is preferred, in the Kutsa hymns a long vowel.
:

If the caesura

is

late the four regular forms are

used almost
:

indifferently in the archaic and even in the strophic period but later there is a marked preference for a short third and a long ^ which is identical fifth syllable, giving a normal form

--

t

with the usual form of the Pentad
therefore be called the

Dvipada Viraj verse, and Pentad opening. may The syncopated form - ^ is fairly common in the lyric of the archaic period, and still more in the cretic and hymns
in
,|

popular periods dimeter verse.

:

but

it

has no such developement as occurs in
little historical

variations of the opening are of very importance: they are discussed in 215.

The other

Rhythm of
'

the break
first

183
have the

In the following stanza Pentad opening
'
:

all

the verses except the

indrd

yuvdm

( |

varund didyUm asmin

6jistham ugrd n ni vadhistam vdjram yb no durevo vrkdtAr dabhitih tdsmin mimdthdm abhibhuti bjah
\\

\\

iv 41

4.

As however the construction of stanzas in which all the verses have either early or late caesura is foreign to the metrical conceptions of the Rigveda, stanzas of this type are only found here and there as chance
productions.

See further

207

ii.

207.
caesura,

The rhythm of the break depends directly upon the and is much more varied when the caesura is early than
:

otherwise

this, as

has before been observed,

is

also the case with

the opening.

1|

The normal forms are |jww- when the caesura is early, and ^ ^ when the caesura is late each of them includes about
:

40 per cent, of the occurrences in the archaic period, and an
increasing proportion in the later periods. The subnormal forms after an early caesura are

v
||

,

r

w v

w,

- ^ ^
II

;

that

syllable.

the remaining possible forms with short sixth Of these the ere tic break is found in about oneis,
fl

seventh of the instances in the earlier groups but in the hymns of Visvamitra and of the later periods it is about twice as common.
:

Occurrences of forms with short seventh syllable are about onethird as frequent in every period as the corresponding forms in which that syllable is long. With a late caesura the only sub-

normal' form

is

^
I,

^ ^

:

in the archaic period this

is

almost as

common
half as

as the normal form, but

in the later periods it is only

common. The iambic forms
|,

w.

w,

^
jj

-

w,

jj

w may

all

be con-

The sidered as occasional forms, at least in the archaic period. form s/ - v becomes rather more common in the strophic period,
later groups but the two forms decrease rapidly in frequency. The occurrence remaining of these forms in connexion with secondary caesura has already

and retains a certain importance in the
'

:

'

been discussed in
are
all

205.
as irregular.
period,

The remaining forms may be considered
relatively

They
rare

common

in

the

archaic

and

afterwards.

In particular the form ^ u v/. It has been necessary forms gave way to the dominant type . when used with secondary caesura. at any rate as used in the Rigveda. the iambic forms (except irregular after the strophic period. the trimeter rhythm tends towards an absolutely rigid scheme.ww. w seems also to rank as regular in the cretic and popular periods. and the cretic form seems to mark the later and this general view was assumed as a starting-point in 94 v. : \j \j The Indian theory of classical Sanskrit metre unnecessarily distinguishes two forms of this verse. according to the quantity of the initial syllable and it fails to take adequate account of the caesura. 95 ii. ^ equal force to all . is used side by side with the cretic form. Still the term Indravajra will be convenient for the scheme just given. for instance. In other particulars also the |j : classification of the forms does not apply with periods . (iii) The use | of the iambic forms jj cretic form ( w- : Although the form v w v is a regular form in all periods. where it is actually . viz. : ' ' w w. holds good for about two-thirds of all the trimeter verses in the Rigveda. assumes importance even in some of the later periods. istic of the earlier Vedic periods. if we may modify the traditional meaning by regarding the quantity of the initial syllable as indifferent.184 Rhythm of the break The rule given in 45. w . as the examples quoted comparatively inelastic in rhythm above ( 206) shew. Although the Indravajra verse (illustrated in 206) never becomes established as the basis of an independent metre. and the late caesura as essential. There is therefore less common than the form v^ ground for thinking that the latter form was encouraged by some distaste for the In the end both these three consecutive short syllables at the break. ( | is ( | v : . and that of the y present the most striking features of the metre of At first sight the iambic forms appear to be characterthe Rigveda. it (iv) comparatively little used in the Vasistha hymns. but is largely qualified by the particulars now ^ is much Thus after an early caesura the cretic form given. that the caesura should be followed by (i) two short syllables. as it has a bearing upon the quantity of the fourth but it seems that even with a late caesura the quantity of the fifth syllable is primarily affected by the syllables that follow. more common than the form |v. It is now seen to be subject to important qualifications.v^. it holds a position of such prominence amongst the various forms of Tristubh verse that it may fairly be considered as the dominant type which has emerged from the competition of numerous Vedic rivals. and to a certain extent in the group x 29-80. which is the most important feature in the verse.^) are really u With a late caesura both the opening and the break are (ii) hence. and in the hymns of Visvamitra.n v to consider the quantity of the fifth syllable in connexion with the opening also.v^. and must be considered as a more regular form and the form **.

The prevalence of short final syllables in this position in the Vasistha hymns is associated with the secondary caesura: as becomes clear when we observe that almost one-half of the verses with short eighth syllable in this group have one or other of the .Rhythm of (v) the cadence 185 Irregular forms of the break are much commoner if the and early than otherwise the most common form is n ^ After a late caesura * w v^. and are therefore tabulated quite independently all other quantitative variations in the cadence are very rare. caesura is : . and .* in 208. The preference iambic breaks u v . shews that there is some arbitrary or conventional element is and almost always an initial or associated with the opposite tendency syllable. Tristubh verses. eighth syllable.in JagatI verses. however. 222. w ^ \* given to final short vowels as compared with final syllables ending in . next to it is more common than This gradation follows naturally from the rule in 45 in ^n both types of the verse the rarest forms of the break are those in which the caesura is followed by two long syllables. This liberty is. not extended equally to all syllables. and are almost confined to the : archaic period. : The regular rhythm of the cadence is .v .v. not only in the archaic and atrophic periods. Final syllables are found twice as often with this quantity as initial or medial syllables and final vowels are found about twice direction is still : as often as final consonants.^ .w.%/ . 209. This rhythm appears to be almost entirely independent of the caesura and the rhythm of the break. is In the eighth place a short syllable employed fairly often. but also in the hymns of Vamadeva in the normal period and some liberty in this : retained even in the later periods. || "jj . There are some important syllabic variations which .j consonants calls for some different explanation the phenomenon is clearly analogous with the similar preference for final short : vowels in the second place. The short tenth syllable is about half as common as the short medial syllable. This fact as is also the short sixth syllable in dimeter verse. and the two are discussed together in 221. shewn in the short eighth The short eighth and the short tenth syllable are associated in the same verse just as often as might be expected by the laws of chance.

rather than as decasyllabic verses. 224.which the Tristubh or Jagati cadence is extended by two syllables. The first two variations may be explained by contamination (ii) ' ' ( 55. For the moment we put aside ( not only these verses. and are occasionally found in all the groups except those of Visvamitra and Kutsa. . several of which develope into the distinct metres found in the decasyllabic hymns. in . as is clear from the great number of such verses found in a The occurrences are confined to single hymn (viii 97 10-15). and becomes frequent is in the popular Rigveda . expect to find the rhythm quite twice as often as These verses are therefore included with the decasyllabic verses in . in which the Tristubh cadence is found in a lyric or Jagati stanza. but also all ( : those that occur in the hymns and consider only those the Table shews that these are relatively common in the archaic period. of the word is short in two instances out of every three.186 Syllabic variations affect the cadence only. ii) whereas if the resolved value were correct we should ( : . and in the popular Rigveda. 56): 'catalectic Jagati' is not uncommon in the archaic and cretic periods. as is in decasyllabic Tristubh metre verses which occur sporadically 204 i). and (iii) the hypersyllabic verse. Of decasyllabic verses there are many varieties 226-228).^ ^.| ( j the Table. by hiatus lyric hymns In Chapters IV and VI I have preferred to interpret verses either or by syllabic resolution. The syllabic variations which affect the verse as a whole are those exhibited in 'decasyllabic verses' ( 49) and 'hybrid verses' ( 56). the archaic period. that most of the verses in which some part of the word indra follows an early caesura are to be interpreted as decasyllabic. These are (i) the catalectic Jagati verse. now finds support For in such verses the final syllable in considerations of rhythm. in which the Jagati cadence is found in a Tristubh stanza. whilst 'extended Tristubh' very rare except in the popular Rigveda. the extended Tristubh verse. The conclusion reached above ( 149 i). in 223. Hybrid verses are occasionally found both in the of the archaic period. These variations are further discussed 210.^ . regularly the case in decasyllabic verses of the corresponding type 226 i. 'Hypersyllabic verses' constitute a special metrical developement.

In the archaic period almost every variation is relatively common. But in all the subsequent periods we find very general regularity. For the intermediate periods we can only expect to trace the history in its broader outlines. cretic periods. 211. so that it would seem that the trimeter rhythm was not at that time established in any very strict form. The evidence of early date is very much the same. . on the other hand.' which are nevertheless common in all periods. may happen predominate merely as the of 'contamination' in the popular Rigveda has been used as evidence of date in Chapter II in other already points the metre of that period is in close agreement with that of the cretic period. and single hymn. Hence we can readily detect the archaic rhythm.Developement of trimeter rhythm 187 whenever the evidence appeared in any way adequate.' and 'the cretic break. whichever be the explanation favoured in particular instances. The variations which characterize the archaic period are both numerous and distinctive they are also generally similar to those : which characterize the same period in dimeter verse. rhythm which In the general picture of the developement of trimeter is shewn by the Table in the next section there appears a broad contrast between the groups of the archaic period and those that are subsequent to them. even in a small group or a The characteristics of the strophic. with special favour shewn to one or more forms. such as the 'secondary caesura/ the 'pentad opening. can only be observed in large : bodies of verse in small groups and to single hymns the favoured types of these periods result of chance. may In the popular Rigveda the variations are all such as be explained by the contamination of verses or parts of ' ' verses of different types. normal. On the other hand the provisional theory : The frequency of an 'archaic period' receives confirmation from the appearance of a great number of new features which are seen to characterize the groups assigned to this period and the examination of these details promises to supply us with the means of defining with : considerable accuracy the list of hymns which should be assigned to this period. In so doing I have followed a principle which commends itself by its simplicity both to Indian and to western critics of the Rigveda: but the present enquiry rather points to the conclusion that decasyllabic variations should be more freely recognized.

Period . Trimeter rhythm Table shewing the principal varieties of trimeter RHYTHM.188 212.

' which includes phenomena con- nected both with the ordinary caesura and with the break. 2 2 x [1] 85 3 86 3 88 3 89 91 93 95 96 97 (lid 21c) 106 107 108 [109]. the first to claim consideration. these variations are most in the archaic period. . 2 For occurrences combined with a pause 3 Vasisthi verse ( 213). The list then is : i *24 2 33 36 39 44 51 54 55 7 56 57 4 58 2 59* [61 3 ] 63 2 71 72 73 77 2 3 4 2 3 3 79 83 85 2 87* 88 89 (8b) 92 *93 100 102 103 104 110 111 112 8 113 4 2 3 2 8 20 116 117 118 119 121 122 125 127 3 128 131 *133 (la lb) 134 140 2 2 2 2 3 2 3 2 141 144 149 151 152 153 154 155 2 156 3 *161 *163 *164 165 167 3 3 2 3 2 5 ii 9 10 168 169 171 173 175 *179 180 181 182 183 184 185.The 213.' The number of occurrences under the two headings must be added together to give the whole number in each 1000 verses. 2 3 v l 2 3 4 8 12 28 (lb) 30 2 34 37 38 39 42 43 44 45 50 51 *58 2 4 2 2 31 32 34 36 4l 42 43 4 45 47 48 49 54 55 57 60 3 76 77 81 *83 87 2 2 2 5 3 3 vi 1 3 4 6 10 ll 12 15 (15b) 16 (46b) 21 22 23 26 27 *28 30 2 4 2 2 2 2 31 33 37 38 39 40 41 44 6 48 4 49 50 51 52 (14d) 60 62 63 3 64 65 2 2 8 4 2 3 3 3 5 vii l 5 2 3 7 4 6 7 8 5 10 13 14 17 18 19 20 66 4 67 68 2 69 72 2 2 2 2 2 2 21 22 4 23 5 24 25 26 27 2 30 31 32 34 36 37 38 5 39 406 42 3 43 45 9 2 4 56 6 [and 11a] 57 2 58 8 60 7 61 62 64 65 4 67 5 68 6 69 2 70 6 71 72 3 73 75 2 2 4 4 2 2 3 7 2 viii *1 77 84 85 86 87 88 90 92 95 97 99 100 *103 *104 2 2 4 s 6 3 7 (34d) 4 15 18 19 20 21 22 23 24 27 33 35 4 36 46 *48 49 53 57 60 4 4 ix 68 70 71 2 74 75 76 3 79 83 84 66 86 87 88 89 96 98 2 99 101 . but are also occasionally used later. [Notes 1 to the Table opposite] after the eighth syllable see under 'secondary caesura. The Vasisthi verse occurs so frequently that it will be suffi(i) cient to give the references to the hymns only. . Verses occurring in hymns in decasyllabic Tnstubh are not included here. for : instance in the Visvamitra hymns and the collection x 29-80 the third syllable is whereas the form with caesura after hardly found later than the strophic period. . . 2 2 3 2 2 5 2 2 [ll ] 12 14 17 19 2 21 23 2 24 27 28 29 30 31 *32 2 (5b) 33 35 36 3 4 5 38 3 40. But the Vasisthi verse ( 205 iii) is very common in some of ' ' the later collections also. is common Apart from the Vasistha hymns. indicating by an index number the number of occurrences in each hymn. 2 2 2 2 3 4 *10 3 *13 *14 *18 23 3 *27 29 30 31 32 (Id 2d) 35 3 36 2 38 39 40 3 2 2 2 2 4 5 10 3 5 3 3 45 [46 ] 48 49 50 55 *59 61 62 63 64 68 69 70 73 74 76 77 2 4 2 3 3 3 (7b 8b) 78 (7b) 79 *82 *83 89 91 92 93 *94 *95 99 104 s105 106 3 7 123 4 124 *125 132 140 147 *168 172 *179. 113 115 116 *117 *120 . . iii l 2 4 5 6 7 14 6 15 18 19 2 20 21 22 25 26 *28 *29 3 2 30 5 31 7 32 2 35 36 38 7 39 2 45 48 50 51 3 53 3 (9b 10b *17b) 54 55 2 56* 2 2 2 10 2 2 2 3 2 2 iv 2 3 4 7 5 5 6 [10 ] 12 *18 19 20 22 24 27 29 57 5 58 59 61 . Amongst these the 'secondary caesura. reference to the verse is only given in the case of composite hymns. it 'Vasisthi' verse 189 In order to apply these results to smaller bodies of will be necessary to record more precisely the occurrences of those variations which are of historic importance. verse.

56 3b. 36 lb. 35 9b 15a. 122 8d. 158 4b. 3 9c 10b. 73 2a 6b lib 15a 19a 20d. 88 3c 3 viii 25 18c. *18 12c. 98 lc. 54 15b 7 lOd. 8 5c. 1 151 iii. . 30 4b. 2 15 10b 12a 15e. 84 2b. . 19 2a. 20 4c 23 7a. 87 lc 4c 8c. 96 2c 17c. 66 lc 8d J 5 1 vii 1 3a 9a 13a1 14c. 117 6d 16b. 59 2c. 113 3c. 23 9a. 27 10a. 18 7a 7b. 140 6c. 3 iii 1 17c. 74 6c. ix 74 4c. 58 7a. 174 9d. 7 lb. 25 lie. 61 4b. 46 2d. 4 Id 12 6c. 33 lc 8b 9a. 122 3c 4b. 118 6d. 88 3d. 71 4c 83 Id. 166 14b. ii 4 3d. 65 lb. . vi 15 12d. 2 7 2 37 lb 2b 6c. . 84 la lc. 79 3a 3b. 68 3a 6b 8c 2 70 lb. 20 7d. 1 . *164 13d. 2 1 caesura following the prior element of a compound. 17 15c. 1 . 87 9a. 41 3a 13b. 2 lie. 63 2c 1 2 iii 16 6c. 160 5d6 1 . 190 4a. 67 5d. 46 7c. 58 7b. *87 21d. 23 27c. 14 6c 7a. . 14 8d. 64 5a 67 2d. 32 6a. 2 121 13d 127 8a. 2 also short eighth syllable. 48 7b. 80 lOd. 22 3c. v 2 9d. 6 3a. 60 4c. 40 2d. perf. . 111 2a. 55 2c . iii 4 4b. 92 3c 4d 95 2a 2b 4a 5b 6a 97 2a 2d 5d 5 2 viii 15 2c. ix 96 4b. 8 5b. 1 1 . 128 4a. 43 3a. 18 17d. 58 9d. . instances is dimeter rhythm wanting. 1 . . 16 46a3 17 lOd 1 20 6c. 61 20b. 29 3c. . 97 3b 9a x 106 7a 2 7d. 166 iv. 2 2 vii 1 3b 4c 5a 50 12a 14c. 42 lc. 60 la 67 5b. 116 Id 7d. . 135 6a. 33 8a. 37 3b. 96 5a. a iv 12 6b. . ( . . 2 2 2 2 2 2 2 86 4b 6b 6c 87 lb 5d. 1 1 . ii 14 4d . 101 8a. 2 3 51 9b 24 . . . *17 Id. 48 17c iv 33 10a . 26 5b. 13 4d. 1 . 33 8c. 180 8c. 93 5d. 40 lb 3c. 15 lb iv 2 3a. only in these 3 4 with 151 i. . 61 2a. 54 lie. 41 7c. 44 10a. 6 2 sing. 53 lib. 86 43a. 26 3c 6a. vi 1 3c. 86 4c. 35 4a 6a. 132 2b. 18 19c 22 9b. 32 3b. 87 6b. 172 2b. 20 3b. 27 16a. 55 18a. 36 viii 25 23c. 189 4a 190 3d 2 4b. x 4 lc 2b. *95 7c. 21 8d. 37 la . 151 ii. . 106 5b. . 36 6a. . 56 14b 19c. 5 2a. *120 9c. 60 6c 7b 12a. 50 7a 12c 12d 14a. 17 Id 13c. 36 lb . 21 6d 4 7a. 16 5b. 32 13a. 19 7d\ 28 3d 34 24a. in -thd (~tha) 7 142 i. 8 108 15b. 53 6a 7c lib 14a. 24 12c. 57 3b 6d. 65 lb 15d. 4 3 in eighth place. *164 29a *52a. . 99 la 2 6a. vi 2 lie. 120 7c.2b = 3b. 61 4c. 90 6d. 88 5d. 51 9d. 49 14d. 4 57 6b. In the following instances the break n w * is combined (iii) with a pause after the eighth syllable: i *24 15b. 20 4d 5d 7b 8a 9a. 26 5b. 17 5d. 67 5a 5c. . . 66 la 73 5a. . 100 4c 10b 15a. 38 Id 4 6a 7b 7c. 75 5c. [11 10b 12d 21c]. 36 4c 9b. 2 J 2b 2c. 57 4b. 167 5b. 46 28e. . 57 4c 6b. *93 5d. 7 19 2b. 22 2c. 23 3a. . 6 7a. 56 17b. vii 2 7c. . 77 2a 5a. 126 8b. 60 4a. 55 4a.190 Other forms of secondary caesura' ' The following are instances of the combination of a caesura (ii) after the third with a pause after the eighth syllable: i 36 18a. 156 2b 3a 3c. . 89 6d. 34 8a. 26 6d. 70 7c. 141 6c. 19 lc 4d 8c. 10 3d. 100 3b. 116 5c. The break w is combined with a pause after the eighth (iv) u syllable in the following verses i 55 3d 6d. 40 5c. . 26 5c 95 5a. 14 3b. 23 8a. 146 3c. 60 10a 96 2b 6 ix 69 8d. 43 3b. 68 3c. 63 2d 7c. 5 170 ii c. 116 21a. 173 lc 1 178 3c 186 3b 5a 11a. 15 5b. 99 8a. 158 2c 5c. 1 . 16 4a. 1 . 104 Id. 105 4b. 85 la Id. 87 6a. 113 13c. 5c. 39 2d 3c 7a. 117 3b 4c 7d 13b. 76 6a 7b. 21 lc 4b. 17 lb 6c. 26 Id 29 5b. 50 3b 5a. v 31 lid. 23 7d 26 3c 6c. 25 fid 4 27 Id. 24 15c. 123 5d. 89 5c. short eighth syllable. 7 4b 7d. 97 3d 6d 54d 2 23 3a 30 2b. 29 3a. 68 5a. : 121 11a. 49 4d. 132 6a. 70 la. 1 . 63 8b 64 Id 3b. 47 lb. 3 10a. 33 5d 12c. 58 5c. 60 5d. 77 5c. *24 9b. . ii 9 3c 25 5a.

*101 7d. even when occurrences in verses which have secondary caesura are excluded from consideration. The consideration which seems decisive in favour of metrical interpretation by means of the 'weak caesura' in the latter instances is the extreme rarity of trimeter verses which cannot be explained in one or other of these ways. namely 1 8d ii tudm sahdsrdni satd ddsa prdti ii 20 Id sumndm iyaksantas tutivato nrn indra svddisthayd gird saclvah iii 53 2d. they are not so common as in : the archaic period or in the Vamadeva hymns. . 91 4a. 214. 45 4b. *120 3c *5d. *103 lie. Undoubtedly the some phonetic character of the Vedic dialect makes a pause of kind at one of the points named almost unavoidable but only a a pause can very strong metrical feeling for the necessity of such account for our finding only three trimeter verses in the whole : Rigveda which have certainly no caesura of any kind. 74 4a. 61 4b 12c' 13b 16c 23b. 30 Id. 1 1 1 . 94 3a. *27 12a. as in ddabdhavrata dnibhrsta \\ -pramatir vdsisthah -tavisir hanti qjasd \\ ii 9 lc 25 4c.Weak caesura 191 70 lb 2c. *85 23b. A caesura separating two elements in a word may be postulated where the rhythm of the break confirms the following examples i 62 6b dasmdsya c&ru -tamam asti ddmsah : \\ confidently it. 110 8a. 39 7d. the absence of any regular rhythm in the break weak caesuras postulated in the following examples doubtful very makes the ' ' : ye dhenuih visva tatakse \\ y -juvam visvdrupdm iv sUrydya cid okasi sve 33 8 b v 33 4c. teristic of Both forms of the 'weak caesura' ( 205) are characthe archaic period. 1 2 127 iv. 22 15d' 23 4a. 126 8c' 133 7c. 93 6a *95 6b. 7 6 159 3 151 iii. 73 5d. It does not therefore appear that this variation is by itself an indication of secondary caesura': but in the cases referred to it is corroboratory evidence of ' this type. 43 4b. There is however some difficulty in determining the extent of these variations. 6. *59 3a. Conversely. doubtful see 159 : 4 i. 158. impv. 74 la 76 lb. 5 with short eighth syllable. -hi in eighth place. x 4 6c. *10 10a. 64 3a 70 lie. 111 3c. Short final syllables in the eighth place are specially common (v) in the Vasistha hymns if however we except those instances which occur in the verses already referred to. 75 3b. . 86 4c 36b. 107 16a 26c. 96 17a 97 27d 34b 53c 53d. 93 4d. 99 5d. 91 7b. 35 4b.

11 4d. ix 72 lc. 33 8b 20 lc 5c 1 18c 4 lie 13d 1 24 6c. 52 9c. follows the third syllable 3 . secondary caesura see . 115 2d. The caesura is at the point of Sandhi combination i 118 7a v 45 9b. The verses i 122 5c. 1 1 . 42 8d. . 4d 5c 3 5d 16a 16c]. *33 11a. . .192 Weak caesura weak caesura separating two parts : A of a ' compound is also found in Virgil see Dr A. x [1 5a]. but as they Occur in hymns which are otherwise in trimeter metre the text is probably incorrect. 1 . 26 5d. 135 4f 186 5b iv 2 19d. 120 9a. 48 8b 50 lb. 60 17c. 1 . 1 . 41 3d. 3 149 iii. 94 Id. 81 4c. 129 4b 1 141 iii 2 19 8a 25 4c. 66 6c 1 . *]67 4d. 34 8d. 148 lb 184 2c. 49 5b. 97 31d x 32 5d. 68 lb 2d 6a 7a. 1 1 . 127 2b. x 50 3c. 26 4b. 62 lb 122 2a 127 5b. Comm. 105 4c lid lie 122 la. . 3 Viratsthana verse. 74 6c. 1 caesura after prior element of compound. 63 2c 2 64 3b 116 3d. 15 4c. 1 . 69 3a. The opening provides very little material which can be used for the historical investigation of small groups of hymns : . The caesura follows (i) derivative: i 30 16c 1 . 44 lOd. 11 3c. . *66 16a 1 viii 21 9b. 68 2a. ix 87 5b. 37 8c 24 7d 2 33 2b. . 88 6c . 1 an irregular 4 rhythm results. v 33 4c. 50 5b. 60 8d. 65 15b. v 33 7a. . 53 16d. 23 19c. are not included ( 205 i). vi 3 4d 6b 8b. 97 13a. 58 2b. . 22 10a 1 35 14d. *vi 75 18b. 186 8c. 1 . 105 5c. *17 5b. 155 4d. viii 96 3b. (ii) The caesura 2 . W. 15c 15d 21c. . [61 2b]. 58 8b1 60 5c 1 [61 3c 6a 6b. 1 . 2 after third syllable. 61 Id. 58 7c 9a. 51 10a. 4 but see Metrical 2 151 iii. Occurrences in dvandva compounds. : i 36 lc . 160 la. *85 37a *98 5b. as indra n -varund. 99 8d\ *101 2b. 7 lid. iv 1 4c 7d. 20 4d. 14 lc. 1 1 1 l 1 . . 35 5b. 140 6c. . 76 2b. For examples accompanied by 213 ii. 4 5d. 215. Apart from the weak caesura irregularities are few and historically but we may conveniently record here the following unimportant : instances (iii) : 7c. 62 2d 5a 5b 5c. vi 2 11a 6d. 8 6d. 1 1 1 1 1 . : The metre is so uncertain that the position of the caesura (iv) cannot be determined in *i 162 16c. and viii 46 20a. . See also 205 i. . 6b 1 8b 19b. 61 13d 108 13b. reading Sasvddibham in one word. note on attached : the prior element of a compound or 36 lc 2 10c. 4 vii 20 6a. those which imply an irregular rhythm at the break having an indication to that effect ' ' the metrical division of in the Classical Review for July 1904. 93 5b. 1 1 . and vi 12 6a appear to be (v) dimeter verses. 99 lc 87 5c. 26 5d\ 29 4a. 86 40c 88 3d. *95 7a. 3 lb 22 3a 23 4 7a. 61 14a . . 33 4d\ 34 2b 3d. 73 8a 122 8c. 16 46c 1 ] . iv 26 7a. viii 46 22a. vii 2 7b. . 1 ii 4 la 9 la lc. 23 5b. 38 2d 5d. 12a. VerralPs ' compound words in Virgil The instances of weak caesura are as follows. 27 lie. ix 72 4a 83 5c 84 la.

or (iii) the caesura is late and the fifth syllable short. and the changes that take place being more gradual. In the two extreme groups there is (except as just stated) very little difference between the openings in the four classes but in the normal group there is a very marked difference between dimeter and trimeter verse.^ syncopated type (-) being everywhere but occasional forms are comparatively rare ' where the caesura is late. namely (i) the lyric verse of group I ( 91). and fourth syllables according as (i) the caesura is early. (ii) the caesura is late and forms this is the fifth syllable long. not twice as common long as fifth syllable before the caesura. and the irregular forms ^ ^ ^ ^ (^) which have consecutive short syllables in the ' ' second and third places but that forms which have consecutive short syllables in the fourth and fifth places are also irregular.^ . In the hymns therefore of this group the dimeter and trimeter a. it is more than fifteen times common. Perhaps the clearest view of the rhythm is obtained by examining the quantities of the second. . It appears at once that all the groups agree in the general contrast between the regular forms ----(-). done in a rather complicated task. : form as the syncopated form. in trimeter verse it is never less than six times as common and if there is a is . and also the corresponding proportions in dimeter table at the The proportions in groups of hymns. (ii) the normal group VI.The opening 193 its metrical character being less marked than that of other parts of the verse. third. and (iii) the popular verse. 13 . To obtain the clearest view we take the groups of which the historical character possible is in other ways the most pronounced. and the attempt is therefore made here to discuss this part of the verse in more detail than was In consequence of the large number of possible 206. Still a history of Vedic metre would be incomplete without some account of its developement. in three selected Rigveda.with that of the syn^ ^ For whilst in dimeter verse the iambic copated form ' ' ' ' . end of this section gives accordingly the which each possible form is found under each of these conditions. The remaining forms may be considered as occasional forms under : all conditions. the ' the most common . which is brought out by comparing the frequency of the iambic form * .

The line of distinction between occasional and irregular forms (iii) cannot be quite clearly drawn. Groups . surdno dive-dive nd ni misati .194 rhythm have developed The opening in opposite directions. It follows from the Table that the 4th syllable is long in three (ii) cases out of four before early caesura. ^^^^^. is naturally very rare: yet we find a few examples of it. ^^^^^. 183.. as ^^^^.. in which the irregularity is doubled. and every trace of the rhythm of the opening lost. and in nine cases out of ten before late caesura.is rather common.u u . According to my is calculations. Consecutive short syllables in the fourth and fifth places are particularly rare. the number of long and short syllables in this position Of. Table shewing the rhythm of the opening. the difference being due to the more strict rhythm employed generally in the latter alternative. and amongst the irregular forms . Amongst the occasional forms ^ v^ ^ and * ^ ^ are somewhat rare. almost exactly the same. The favour shewn : in dimeter verse to a long third syllable may perhaps be connected with the fact that the fifth syllable is always short but at any rate the favour regularly shewn in trimeter verse to a short third syllable is most marked when the fifth syllable is long. ' ' The nine possible types of irregular opening are included in (iv) the formulae The form ^ ^ ^ ^ ^. *iii 29 14c. doubtless because the late caesura is almost always followed by two short syllables. : The following particulars also deserve to be noticed It is agreed on all hands that the quantity (i) of^ the is first syllable in all cases indifferent.

5b. 26 5c. . 31 6c. but. 43 2d 7a 7c 7d. . 118 2d 11 131 f6a 2 140 fl3a3 143 8d. 61 13c. 167 lOd. *29 2b *6d *14c. 12 6b. 177 2b 182 2a 4a. 6 3d. 116 8b tl4c 2 25b. 33 la 2d 10a. 48 lc 14c. as is held. 60 3d. 96 7a. 4 lc.Irregular openings 195 The instances of consecutive short syllables in the 216. 89 3c. the proportion being slightly larger than in dimeter verse ( 190). . 39 f6a . 6 5b. 39 8c. 27 3d. 40 la 2a 4b. 104 7c. [68 2a]. 62 lc 8 Id. Out of about 350 instances of final short vowels in the second place quoted below. 51 7a. . 49 4a. 152 2d. 87 lc 2b. 44 14a 19b 4 47 13a*28d. *170 5c. 21 2c. 21 la lc 2a. . 38 8b. or -a as representing final -ah or -e: for the vowel the text in several instances gives a long vowel. 165 4b 6a 9d. and connexion with the question of protracted vowels. v 29 24 7c. number is too large or too small the general argument is unaffected. 51 3c 8a. iii 1 21c. 5 3c 4c. 171 la. 54 8b 10b 15d. . 30 lb. . 5 8c. 4 4c. Consecutive short syllables in either position are less common in the normal period than elsewhere. In fully one half of the instances the second syllable contains a short final vowel. 153 |4c 2 *162 lb *14c. 186 la lc 4a 5a 6a 7a 8a. 28 lb. 124 14a (a) l compound . 36 2d. 49 14a 15d. 121 4a. 32 2c 4 33 6c 6d. 4 ii 2 2c. 200 examples of consecutive short syllables in the fourth and fifth places. 57 3a 12 6 2a 6a 4 30 6a. 59 7c 64 lid. 12 2 9 4c. 132 . 100 14b. 25 5a. second and third places in trimeter verse are about 650 in number. 30 19c 20b 13 31 20c. 54 2a f4a 2 4 2 4a. . 54 4a. Only in some 33 instances final short do we find a similar short final vowel in the third place and we therefore have repeated the phenomenon already noticed in dimeter verse ( 189). about 35 are instances in which the text gives a Whether this long vowel. : any trace : Similar metrical phenomena are observable in the cadence its and the problem as a whole. . 57 5a. *15 19b. *163 3a *4a *6c *lla. 48 |2c 4 vi 4 3c. 38 fla. 49 2c. . 149 lb. 26 3c 4 6d. 41 2b 7a |7d 10 8a fl6b5 16d 19a. 16 14a 20b 4 *18 3d *4a. 2 2 33 13c. . 35 13a iv 1 6c. 4 54 lb. 20 8c.3d 4b 4d 10b. are further discussed in 221. 13 lOd. . 10 6d [11 t5c ]. 12 8b. 77 3d. or substitutes -o for final -a ( 171 iv). 15 Id. . 5 5b lid 12d. 113 5a 6c. 1 . . 20 la 5c. 129 fSa 10c. 55 4b. 19 6d. 51 12c. 76 la 4b. CoDsecutive short syllables appear in the second (i) places in trimeter verse as follows : and third of a the ictus falls on a final vowel of a word or of the prior element 4 or derivative: i 32 lb 14c 33 9a. 50 14a. In these instances the fourth syllable only rarely contains a final vowel nor is there of any special metrical value of the fifth syllable. . 50 Id. 7 f6a flOc 9 9a. 127 3a. *164 lOd *lld 4 *31b *36d *41b *43a *45b. . on insufficient grounds. that a short final vowel is specially capable : of bearing the ictus if There are less than it occurs in the second place. 1 . *28 4d.

*82 7d. 81 3c. 108 14a. 8 6d. 86 4c 6a. 96 14a 18d. 70 Id. *52 6a. . 110 5a. *161 6d *8a *12c. 107 2b. 53 f3a 2 56 16b. 68 2b 3a 12 69 2d 12c. 48 3b. . *87 19c. ix 70 6b. 48 f7b *59 6d. *35 lie. 62 3b. 30 lc. 42 4a. 31 8a 8b 8c. 16 2d. 134 4a 4f. *164 40a. 30 10b 13a. 11a. 2a *14b *14d *16c. *121 4a *4b *4c. 110 3a 5a. . 36 13a. ii 3 3a. 38 6d. *130 lb. 21 3b |8c 23 3c. 54 6a. (c) the ictus falls elsewhere: i 31 16d. vii 1 2b. . 5c. 108 2a. *102 2b *7c *10a. 45 4b. vi 10 lb. 47 2d *30b. . 61 3b. 88 2a. 21 2d. 181 7d 9b. 35 3a 7c 9c. *71 |4c 2 73 2 4 |9b 74 4b. 126 8b. 141 6b. 105 4b 51 4a. . *179 2c *4d. 10 but doubtful. 23 7b. 111 3d. iii *29 15b. 62 7a 11a. 58 5d. 1 . 22 5d. 147 5b. 117 lOd 16c. 4 24 4b 28 fla 7 *33 14a. 35 5c 5d. 86 16c. 46 lb. 116 2a. see 175 i. *108 5b *7d *8a. 47 2c. . viii 24 [34 10b]. x 4 lc. 72 2d 4c vii 5 3b. 183 4b. 16 6c. 90 5d. 27 3b. 91 20c 22a 22b. *168 3b. 48 2a. 23 la. 77 8c. 151 iii. ef. iii 20 la. *52 2c *3c. . *95 11a. 55 4b. 46 80 2d. 100 4c 6c. 178 3b. 61 9c. *27 21c. 34 9c. 45 8c. 26 9c 12c. ^ 121 c. 15 4b 6c.. 17 7a 10b. 79 4b. 57 3d. 61 9d. Irregular openings 2 63 9a 9d 4 70 la. 89 28a. Id 3b. 131 7a. 160 3b. 19 6c. 38 2d 3b. 9 151 13 i. 34 [17a 18a] 24a. 20 3a. 99 2b 5b 4 100 2b. *117 lc 2 *4a *6c. 98 2c. 37 3a. . *71 . 70 4a. 72 3a. 5 6b 10b. text dpo. 53 11a. 174 10b. . . 33 3c. 62 6a 40 6c. 20 lb. x 4 7a 4 8 4a 5c. 23 17c. 69 5c 5d. . viii 20 5b. *88 18b. 5 3 8 11 166 v. 32 9c 13a. 64 14a. 50 2b. 66 5c 4 8d. 66 7c. 84 13c 88 3c. 71 2c. 6 4d. 49 9a. 33 10b. reading tdva mdnah. - text utd. *18 2a. 139 lib. 45 lb. 116 10a 10c 18c 21c 22d 25d. 49 8c. (6) : 14c. 54 2a 3d 6a 6b. 89 10b. 96 fed 1 lid 15c. 168 ii. -a from stem in -an. 2b. 99 7c 8d lid. ix 70 7a. 1 . 88 7c. 71 8a. 1 . 13 12a. 155 4a. . cf. *162 13c *14a ii 1 lie. *103 9a. 92 3d 94 la. 9 4b. *40 7a. x 55 4a. 26 2c. vii 1 6a f8c 16c 18c. 20 2c 7c. . 39 2a 2c. 66 6d 9d. 106 7c. [46 8b]. 73 5d. 38 4d . 125 2a. vi 4 7b. *103 10b. 37 6a. 140 lb 4c. 4 5d. 28 5d. 60 14a. *20b. 167 8b 8d. 19 3c 6b 7b. 96 2b 18a. 76 4a 8d. 94 lb. *17 6d. 1 text upo. 139 3c. 1 . 60 5c. 27 10a. *15 5b. text hd. 8 2d. 123 9d 10c. 56 v 30 3c 12a. . . 64 lb. 31 3 49 lid. 20 7c. . 93 13a 97 7a. 61 26b. . 16 lc. 12 2a. [67 10a]. . 124 2a 2b. 79 lc 3d. 41 16c. 99 3b 6d. 38 4c. *33 9d 2 *12d. *35 3b *4d *9a *12c *13a *13b *13c. v 41 9c. *120 2c *6b *7d. 35 8b 58 2a. text prd-pnl. . 33 8d. 168 9b. 64 Id. 24 [11 7c]. *14 6a. 39 2a 9c 10c. 51 4c. 48 6c. 3 9d 12d 16c. *94 7d *95 4b f9b 9 96 4a. 95 3d. 32 6a. 4c. 2 3c 4b. 85 7a . 158 2c. 32 lb. 7b. 41 7b. *104 8c 2 4 *19b *19d *24a viii 1 24c 19 33a. 59 4a 77 fla 81 flc 82 3a.196 .. [34 16a ]. 68 lb 5d 12a. < 4 it. the ictus falls on a similar final vowel in the third place i 44 152 4c. 44 12c 17b 22d 23a. 97 |49c 3 f 51a3 53a 55c 55d. 121 8d 9d. 63 3b. 130 6f. 76 Id. 53 6a. 39 5c. 98 lc 4 ix 69 f 2a 73 8b. . 64 2b 2d. 67 9b. 23 5c. 92 lie 12c. 142 i. 48 lib. 73 8d. [65 9a]. 110 lc. . 27 2d 13a. *103 6a. . 175 i. 46 13c. . 127 3f 5f. 57 3c. *103 3a. *34 8b. 61 3d. 109 7a. 97 23b 41d. 79 4a. 41 6d. 61 5d. 1 1 . text brdhma. 67 12b 92 flc'. 54 3d 10a 14b. *18 3c. *16 5a. 30 lc Id 5a. text abM. 45 3c. 57 2b. 20 2b 8b 13c. 39 5a. iv 2 11a 19b. 114 4d. 36 Id. 98 2a. 110 8c. 61 la. 63 16c. 7 la. cf 166 . *10 |13b 8 *13 3a4 *15 Id *13a. *177 3b. *29 11a. 62 3c 5b. *53 4c 14 *4d 4 *5a 4 54 6d 4 6c.

2 vi 1 7b. 160 lb. 46 3b. 43 2b. 124 7b. 52 15b 17a. ix 72 6b. 64 3d. vi 3 5c. *10 8d. 4 177 iii. 172 4b. 1 . 50 4d 6c. 61 12c. . x 8 6b. 148 2d. x 3 2d. 34 5c. 115 2c. 15 5b. 53 5c 6c. . The opening is frequently defective by one syllable. 116 2a. 36 9c. 76 4d. 51 lc Id. . 17 7a. 36 2b. 2 3c. 54 lid 13c. *28 6d. 99 5d. . 40 2b 5c v 2 5c. 2 lie 87 5c. 83 2a 2c 2 viii 3 9c 10a 12a 12c 2 20 2a 6c. 42 4c. viii 9 lc. 53 3c. i. agreeable to analogy if not strictly proved. 46 4b. 21 6c. and in such cases it is a convenient assumption. 20 8d. 189 4a 2 . 59 4a. reading vfsanam. 58 7c 7d 8a 9a. *88 17a. Consecutive short syllables are found in the fourth and fifth (iii) places in i 32 5d. 70 7b]. *149 5b. *18 13d. 18 24d. 183 5b. 110 lie. 31 13b. *149 2a. 121 12d. *107 la *5d *6d' *108 11a. 117 14d. 35 16a 17a 18a. *35 4d. 108 4a. 58 6b. 37 3a. 57 5b 6c. 88 6b. 91 15c. [56 9b]. 217. 62 4b. *109 la. 31 5c 4 35 4b. 10 2d. . 20 la. 30 lb. 168 5c 173 8c. 47 4b 10b. . *163 4d. 24 9a. 131 5a| *139 4b *6a. *125 3b *8b. 39 2b 3a 3d 4a 4c. 96 2d. 48 lid. 17 6c. 53 9d. 7 5b. *179 3b. 48 2a. *130 4b. 60 2c. iii 23 Id. [78 lb]. *29 14c. -hi. The instances are discussed in 228. . 118 9b. 52 8c. 12 2d 3a. imperative in -dhi. 6 4a. 2 10d 2 ix 74 5b. *101 Id. 154 5a 6d. l 1 . *120 3d *5c *7a *7b. 93 8a. vii 1 4a. 62 6a. *28 8d. 110 3c. *71 2c. . *18 12b. *101 2b. *108 11a. 36 lc 1 6d. *163 8a. vii 1 4c 13a2 24c. 123 lc. 31 9a 6 107 10c. 36 15c. 88 6d. 44 [8a] 16b 18b. 41 3b. 75 5d. 129 11a2 140 2 165 2d. 44 9c 9d. . 24 lOd. 142 iii a. be convenient to note here the occurrences of (ii) protracted final vowels and others of doubtful quantity before a short third syllable: namely i 31 7c. 8 2a5 9 3a. *48 2d. 61 3b. 43 6d. 104 2c. 151 4d. *87 14d. 81 la . 1 139 i. *88 2d. iv 26 7c. 32 2d. 89 4a 7b |9d 17a. 72 2c. 105 7c. 97 4d. 25 3a. *164 5c *8a *8d 2 *44a. 68 lOd. *10 8b *10c. *75 8c *llc. 15 3c. [78 5b]. It will ' ' 1 . 144 5b. . 73 3d. 66 10a. 17 lid. *121 5b. that the defect is caused by a rest at the fourth place. 77 7a. 148 3a. *182 lc = *2c = *3c. 53 10b. . *103 7c. 41 17c. 45 8c. *74 lb. *85 43b. 96 4c. 48 6a. 40 5a. 184 2b. 41 5b 7a. *94 lib *95 2b. 106 8a. 80 4b. . ii 2 2d iii 2 5b. 61 2b. ii 17d 2 ]. 63 3d 6d. 41 3c 16e. *164 3d *21a *33d *40c 186 5c. [68 lb. 54 15b. 30 4b. 100 16b 104 lc. . 42 9d. . [6 2d]. 33 Id. 166 8a 183 3c. 76 la. 29 3d. 59 8a. *104 lie. 5 9c. 121 Id. 36 8b. 3 12 8b. 101 3d. 69 5b. 155 Id. *102 9c. [11 51 2c. 95 1 10b. 7 74 3a 6d. 39 8c. 1 147 iii. 174 3d 8d. 13a. *56 4c *6a. 23 5b. 47 4c. 152 lc. 128 la. 60 7b. 31 6b. *53 8c. 89 3c. *51 5a2 *52 Id. 138 lc. 117 2d 9d. 7 la 4c 6b. 5 148 vii. v 8 7b. 1 -a for -an ( 166 2 v). . 49 2b 8b 12c. 27 lc. 89 7c 97 26a. 49 5c. 2 7 6 142 3 147 iii. 99 5d. 122 7d. *94 14a. 5 7d. 47 8b. 170 ii c. iv 1 14a. 83 la. 51 4c. 23 lb f 8a. *95 3a. *32 9b. 189 7d. 45 3d. 47 3a 71 3b 3c 72 lb. 106 3c. 107 73 5b 9c. 112 3b. 69 2c. 48 2a. 49 2b. 21 18c. 160 3a. g 142 i. *114 7d. *109 lc. *121 8c. 96 5c. 40 5d. *15 9a. 38 5c. 19 7d. 44 2c. 122 7c 8d 9d lod 123 12a. 100 2d. .Irregular openings 2a.

i .

23 lb 6d. 173 lid 12a. 139 3a. 33 3c 15c. 16 2a 14d. 20 5d. 153 3d. 112 8c. 88 2d. 17 19b. 22 la . 8b. 91 4b 18d. 33 10a. 24 6a. 96 19b. 60 3c. 26 6a. 55 3c. 85 4b lid. 45 3a 7a. 78 [6b] 7d. 21 18a. 58 3c. 20b. 70 3c. 79 5c 84 20a. 38 lc. 73 2c 7d 10c lOd. 6 la 7d. 2 4c. 34 4d. 39 14c. 76 7d. 39 lb 3b. 168 lb. 8 8d. 66 lb. 115 5b. 28 3b. 19 3d. 27 12b. 68 6a. 56 22b. 4 4c. 16 2b. 2 6b lib. 37 Id 6c. *83 7b. 88 3b. the instance should be added to those on p. 19 2c 10a. 92 3c. 33 3b. 31 10a. 70 3b. The iambic break . 128 4b 7b. lc 4c. 4 4 lb. 47 8a. 156 lb 4b. 26 5a. 23 7d. 5 38c. 34 3d. 14 6b. 97 19a 21d. 70 4a. 15 3b 5c. *165 3c *181 la. 43 14d. . 43 viii 4 5a. 133 6b. 67 5c. 18 13a. 90 4c. 168 3c. ii 1 13c. 61 4c 5a 8a 24a. 73 5b. 31 2d. 26 5a. 21 3d. 48 lb 7a. 52 2d. *33 2c. 177 5b. < 1 - 178. *94 7c *8c. 155 lb Id. 5c 17c. 13 2a. 18 19d. 2 7a. 31 2a 4b. 14 2c. 67 10b. 60 3a. 2 vi 4 2a 3b. 29 lc 4d. [61 lc 14c 16b]. 36 3b. 35 2d. 17 5b. 22 10c. . 46 lc 2a 5c 14c. even apart from secondary caesura. 96 6b. . 30 Id.w. 96 la. [11 9d 13c]. 19 2a 6a. 42 4c. *83 Id . 7 5b 5c. 3 3d 5b 5d. 78 3c. 37 4b 8b. 35 5c. .The iambic break (ii) 191) w. 31 3b. 180 5b3 181 5c 6c. 33 4b. 41 5a. 112 11a. 157 4c. 21 4a 7d 8b. 3 9c 11a. *95 6d *13a. 8 3d. 37 2d. 92 2a 3a 3d. *10 lc. 152 lb. 71 2a 4b. 48 3b. . 28 3a. 56 7c. 23 2a. 55 2a. 4 20a. 66 13c. 6 2c. 177 iii. 5 4 but see 84 A 13. 26 8c 15c. 52 2c. 41 lb 5c 7a lOd. 57 lc. 77 8a. . *13 5b. . 75 5a. 67 6a 6b. 3 *164 24b. 86 Id. 132 5d. 96 9d 10a. 44 lie. 101 8c. lid 4 51 lOd. 106 7b. 38 5d. 53 10b. 68 4a. 131 4a 5a 7a. 55 lc Id 4b 7a 7b. 122 5b 13c. 27 2a. 63 4a. s/ u The iambic break is very 1 much more common. 11 3a 5d. *104 17a. 58 2c. 122 10c. 62 2b. (iii) occurs as follows: i 34 lie. :i 5b. 77 10c. 165 15b. 59 la 3c. 87 3b 5d 6d. 111 3c. 10 5d. 184 4b. 4 5b. occurring as follows: i 31 17a.n . . vii 1 16c 18b 20a 23b. 7d 8a 18c 4 53 14c*. 117 14a. 47 22a. 4 7c. 35 la. 88 6c. 12 4a 4d. 49 4c. 113 158 3a. 21 8c. . 61 3c. 93 3a. 87 7a. 21 6b 8a. 6 10a. 180 6d. 124 4a. x [1 4d]. 30 9a. . 25 12c *100 ix 68 70 5d. 4d 5b. *1 8 2d. 17 14d. 77 5d. 97 15a 15c. 22 4d. 153 4b. *12 6d. 63 9b. 20 lb 5c. 46 2b. 31 2c 3b 13a. 67 7d. 42 Id. apart from secondary caesura. 28 5b. 89 14a 15b 17a. 32 4c. 19 5a 9b. 15 3c. 79 lb. 62 3d. 24 3b 10c. 43 la 2a 2c 2d. reading asmabhya 3 142 : ii. 86 17c 42d. 116 lb 2a 3a 8c 9c 12b. vii 7 5a 6a. 101 lc 6 46 26c. 86 3d 5a 34b. 60 7d. 7 lc *74 3b 3b. 23 3a. 8 2c 7d 8a. 19 ii 2 12b. For instances connected with 'secondary caesura' see 213 i. 37 2d. 22 10c. 108 16a. iv 1 6c 6d. [11 3a 4c 13a 15a 17a 17b]. 86 lb. 34 3a. 17 lid. 7c. 30 2b. 48 6c. 141 10b. 56 2b5 *59 lc *6c. 5 3a. 32 4d. 18 2c. 92 3d 6d. 53 lc 14a. 3 la. 121 14a. 15 2c 14b. 108 lib. 20 6b. 18 15c. 24 4c 6a 9c. 8c. 40 7c. 103 ix 70 3c 5a 9a 9b. 103 3a. [6 4d]. 93 10b 15a. 40 6d. 3 12a. 88 4c. 17 la Id 8c. 36 2b 2c 41 4d 10a 10c 14a. *51 3b *8d. 29 4a. . iii 1 14b 17b 18d 20 la. 4 3a 4d. 144 2c. v 32 2b 4c. 85 8d lib. 110 7c. 105 7c. *164 lid. 111 2a 2b 2c 3b 5a. *121 8c. 99 7a. 173 6b 6d 12b 174 2b. viii 1 85 2a 3d. 16 lid. 7 lb 8a. 8 7c. 60 7d. 169 14c 4d 6d. 123 2d 6b. 18 5c. vi 4 6a. 92 15c. 133 6f. v 1 7a. 50 6a 68 2a. 2 116 6b. 169 6a 8a. 4 iv 6 8b iii 2 5d. 81 4c. 30 (A 13). 25 2a. 94 la lb f3c 1 . 100 4a 16d. 127 6b. 38 3c. 20 10c. 91 20d. 171 6c. 13 9b. 15 8b 8c. 70 12c. 50 3d 7d. . 39 3d 4a. 44 12b. 149 2b. *100 2d *3c. 20 9b 20a 20c. 22 15b. 120 4b 8a. 94 Id.

The irregular break n (vi) itself unrhythmical. 35 2d. 51 *57 5a. *169 2a.. 73 lc. 61 13d 22c. 41 4b 17a. 190 6b. common is by far the most of the break the occurrences are spread not very unevenly over the whole Rigveda. 148 5b. 4 vii 7 6c. 30 3a. 93 12b. 18 lib 22b 27 5b. . 24 5b *9c. 29 Id 2c 4c. *53 21d. 18 4a. 59 2d iv 3 7a. *120 3b. 35 9c. 114 lib. 83 Id. 11a. *53 6b. 42 5a flOa 3 54 3b. *161 5a *9c *12d 2 *162 lid. 28 3b. *1()4 24a. viii 19 34c. . 1 . x *18 5d. 141 12a. 32 5c. 21 6a. 48 2d 11a. 7c 186 2a v 31 12d. 53 6c. 36 5d. 106 lb. 5 145 iv. 16a. *149 4a. 70 5c. *170 4d. 4 la 6c. 30 15c. *14 15c. 3 151 iii. 66 2a 10a. 1 6 169 iii. . 121 12d. . 49 10c. 97 52a. 63 2 ix 72 8b. 62 7a. 67 4b. 91 2b 2c. 169 iii. 12d 55 6b. 127 8b. 29 4c. 81 3c 6c. 67 5d. as follows: i *24 Id *2d. 23 2c 28c. 76 8d. It is . ii 1 9d 184 2d. 26 7a. . 22 18a. 33 3b. 170 i. 53 lb. 6 177 i. vi 13 vii 1 19c. 74 2d 88 6b. 23 2c 5a. 41 12a. 52 4d. 83 3c. 30 2a. *103 2c. 145 4b. 63 6 131 2d. As this proportion is normal. lib. 31 4b. 108 11a. 28 5d. 31 8d. 12 lb. 66 4b. 1 . 27 22a. 46 5b. 91 6c. 23 4b 7a. 123 2b. 103 2c 3c. 37 lb 2b. 2 169 iii. iii 2 4a. It is most common in the : |. 9 5a. but also directly after early caesura. 20 8c. 167 10b. 165 13b 13d. (iv) 2 7 136 ii. 149 3c. ix 69 7b. 1 151 i. 120 8b 9b 9c. 68 2b. 44 15b. 1 cf. . *47 29c. 88 3a. 68 5b 7b 7c. It is possible that some quantitative explanation may be found in the case of the word aditya. 55 2b. 42 7d. *15 10c. 4 3c. 73 2d. *130 7a 150 5a. viii 1 16c. is a pause after the eighth syllable. *12 3b. 144 5a 7. 145 vi. The irregular break (v) 55 7d.appears not only to be in contrary to the normal rhythm somewhat more common than might be expected under the circumstances. For instances connected with 'secondary caesura' see 213 iv. 50 5a. 19 2b iii 4 4d 8c. 12 la. 34 6d 7d. iv 3 3b3 5 6b. 83 4d. 41 9c. 23 4c. 42 6b. 25 5c. 88 4b. 33 12d. 139 8b. 25 6c. 19 29a.200 Irregular forms of the break x *10 13c. 168 ii. 61 8a. 4 but perhaps vadhnmantah. 17 3c. *27 10b. 135 4a 5f. 71 4d 2d. 54 lie. 19 lb 2d 6d. 33 8c. ^ occurs as follows i 36 1 7c. . 73 3c. vi 10 6a. 16 16c 18d. 170 ii/. [46 4b]. 133 7a 7f. 26 5b 7c. 80 7b. 104 5c. *107 10a. 40 4c. 79 4d. v 2 lie. 80 la. a 5 171 iv. 35 5a. 48 8c. : archaic period. *170 5c. . 4 178. 71 8b. 112 9c. but see 178. 22 3c. 116 7c 18a. 3 6c 7d. . 24 4d 10c. 37 2c. 64 16d. Of the irregular forms p : 1 . 23 16c 21c. 20 3d. 41 2d. 186 lc. 39 3a. 20 lib. 11 4a. 45 4a lid. 140 8c. 19 9d lOd. 62 2d. *28 7c. 56 16d. 173 ii 3 5b. it now seems better to treat it as irregular. 40 8a. . 26 13c 5 33 5c. 110 2a flS x 2 7c. the variation appears not to be connected with secondary caesura. 99 6b. 122 7d 15a. 48 3c. 41 2d. *107 6a. 46 16a. |132 lb 134 If. 20 5c lid. which occurs . *120 5a. .. Of these occurrences about one-third are in verses in which there . 143 6c. 24 24c. 46 7c. This variation was treated above ( 94 v) as iambic but as it is very rare and has no connexion with the secondary caesura. 21 oc.

vi 2 11a. 6 6c. v 3 iii 5 7b. *169 2b *3b. 61 7b. 37 7b. 4a. 48 2a 5c. *93 6b. 86 40c 148 lb. x 22 10a. 1 . vi 26 8a. 65 3b.v-: i 60 4a. 122 2a. 28 2d. 25 7c. || Form - iii 2c. *161 6c *7c . 105 8c. The same forms occur in compounds as follows: ^ M ^ . *12 2d *4a.^ . iv 4 12d. 3b 79 ii 19 2a. 83 5c.i 30 16c. *85 37a.-: 23 3d. 48 14c. : (c) Form ^ . 105 lid i ^H [61 4d]. 34 8a 43 4d.' the break are the rests and hybrid both of which are con- . following the prior element of a compound instances are given separately. 115 9d 9e. 169 7a. | Form ^ vy (in addition to occurrences in compounds. x 30 13a. 28 la 1 . 93 7a. 18 4c. 30 4a. 51 2a. 68 2d x 48 8b. 173 8d. ii 19 8a. 22 7b. 189 3a. 56 6b. viii 4 14a. 4 8c. 1 ii 3 4d'. iii 2 3d. 23 vii 19 lie. 3 5c. 115 4a. ix 103 5c . . 69 7c. lie. *104 2c. for (a) which see below): i 62 9d. 25 15c. 127 la . 46 12c. 41 3d. 33 8b. 25 24c. 31 1 . *108 10a. 21 10c. ix 81 lc. 99 5c. 173 9c lb 3b. iv 29 31 lc 25 4a. 24 8a. *142 lb. . [61 2d 9d]. 16 46c. 7c. 29 3d viii 21 lb. 40 6c. 17 6a. . . 61 9a. 112 lie. 36 7b. 13 13c. ii 4 4b. 34 3c. 35a. *164 lib. 1 adityd (10 occurrences). 64 3b. lib 1 . 219. 101 vi 26 5d 8c. 21 6c. : 213 under this heading.Irregular forms of the breal 201 rather frequently in this position. 70 13c. 77 11 3b. 33 8d. 129 4b. 85 5d. . 27 la 3b 4a : 1 iii 16 3c. 27 16d. viii 61 14a. 27 22c 29 9a. 19 7d (=i 174 8d). vii 1 4a. 139 3b. 186 9d. v 41 15b. 174 8d.-: i 3a 1 . (b) 139 2a. 66 3c. *35 13c. 44 3b. v 41 6d. viii 19 vii 46 lb. 60 10c 20a. *95 9c. Id. 33 4d. 103 7b. ( . . 42 15b. *100 12b. 67 6d. 16 9d. All the irregular forms are much rarer when the caesura is a somewhat large proportion of the instances there is only a these doubtful caesura. . 62 Id. 68 2c. vi 8 lc. 2b. 26 22c. x 61 25d. i [61 lb]. ii 4 122 14a. 51 12c! 69 2d. . 18 8c. 70 6d (d) vi 26 [61 12b]. 5a. 3 reading sasvddibham. 20 lb. Form r. 61 9c. . 54 8a. 3 lb. 24 18c 23c.i 60 5c. 19 Id. 74 4b. 20 2c 12a. 96 16a. 3 ix 72 *vii 66 16a. 63 3b. viii 66 6c. *15 9a. 61 . *102 lib. 20 f8c 13d. 30 6c. 6c. 106 13c. 51 lid. 99 7c . 1 1 . 50 12b. 127 5f. 96 15d 45 4c. 30 9d. 58 8b. x 8 5c. vii 88 6c. x 3 . 90 lb. ix 86 42a. The ' principal ' syllabic ' variations connected with verses. but in no other case does any The occurrences are i 94 particular word or form come in question. 100 12a. *107 9c. 18 4b. 189 7d. and are not included in the Table in (vii) In late. 22 3a. a 145 ii c. 108 14a. 116 22a. 61 14d. 4a. 160 5a. 124 8a. 67 3b. 20 5d' iv 1 2d 2c. 37 4c. v 41 14c. 22 18c. x [1 6d]. 71 lc. 1 but see 174 ii. 111 2d. 180 10c. iv 1 4c 6b 8b. 23 12d vi 27 6d. ii 20 8c. *52 4c. [70 10b]. 93 8a. 74 6b. 22 lg= 2g = 3g. 56 23b. 141 8c. 23 30c. 2 *98 10c. 5a. 17 3c. 89 5d.

: 1 . 61 5d. 148 1 165 13d 15a. 3 12b. 6 4a. 32 5c 18c. 16 lb 4d. 120 2b. 46 2c. 44 4b. . 30 2b. 4 6d 7c 12d 15d. 133 7b. 38 4a. 131 la. 129 6f. 51 9b. 121 4d 10a. 103 5d. 87 7b. 27 4b. 46 4c. 51 15b. 66 4a 4b 7b lOd. 1 . 104 la. Examples of a long eleventh syllable are hard to find in any part of the Rigveda. 19 4a. but also in the hymns of Vamadeva.202 sidered below in in Short eighth syllable 225-230. 2a 5a 25 2b. 71 6b. v 1 6c. 18 lb. 40 3d 4c. 50 4b 15c. 100 10a. 101 Id. . 112 19a. A long ninth syllable is comparatively frequent in Pentad hymns. 7 8c. 21 la. 150 2c. 19 2d. 4 8a. 23 29c. 37 2b. 174 6b 8c. 49 4b. 34 3a 5a 5d. 14 167 10b 10c. 12 lc. 10 2b. 127 7a lOf. 36 5d. aud these variations are also continued to a later period. 72 6a. Of the remaining variations again the short tenth syllable is much the most common : but both this and the remain- ing forms are almost confined to the archaic period. 3 2a 5a. and removed by emendation the : \\ ardnibhir dasti dvcise pur a dyasir ni tdrlt dsmai vaydui u ydd vavtina tad vivisma hotaro nd u diviydjo mandrdtamah tejisthdbhir hatxii ddsyun \\ 127 4b 20 8d vi 23 5a ix 97 26d i ii 220. 23 4c 7b. 39 4a. 43 4c. 189 5a. 145 fib 147 4c. 21 8c 8d. 86 3a. 21 lb. 13 2c. 24 la. 31 21c. 40 4a. 1 . 138 3f. 38 5a. 18 2c. 166 12d. 32 lc. *33 |2d'. : Hypersyllabic breaks are very rare the require to be Rigveda following apparent examples occur. The variations of the cadence in trimeter verse follow the same general laws which we find at work in normal dimeter verse. 23 5b. 44 21c. Thus the short eighth syllable in its various forms is not only found in the archaic and strophic periods. 3 7b. 17 19a 14a. 152 2b. 20 2a 8c. 33 113c 4 iv 2 la 6d 7d 12a 19a. 183 4c. 53 2b. 29 2d. 39 3a. 4 21a. 19 10c 18a 124a 1 24c 25b. Short syllables take the place of long much more freely than vice versa. 1 . 51 3d 2 5 f Sd fl4c 5 6 Id 4d. 5d. 43 5c. 33 5b. 2 8a lOd. *162 119a ii 1 5d. 59 2c 8c. 42 3b. 33 lb 6b. 3b. 186 10c. 24 10a. 42 4b. . 22 13b 14a. 1 . 34 lc. 31 5c. 45 4b. viii 1 28c. 41 7a. iii 2 7b 9b. vi 1 2d 3b. 72 vii 1 8a 13b 15b 8 18c 19c. 79 14c 82 9c. 24 4d. 96 4b. 35 8b. 55 2b 5b 6 5d 6d. *161 10c *13b. 67 5d 7c 11a. 57 2c. 53 2b. and marks the divergence of this metre from the parent Tristubh : see farther 249 ii. The occurrences follows (i) : of the quantitative variations of the cadence are as Short eighth syllable (in addition to the examples enumerated in 213): (a) Final vowels: i *24 Ha. 56 15a. 25 lc 3a 7 29 6d. 41 9d lie. 17 7b. 77 lc.

22 3a 8a 15b. 145 la. 4c. 168 iii. 4 9b. 33 2d 4a. 152 6a 6d. 36 lie. 150 i. 34 5b. 144 2c 5a. 93 5b. 17 20b. *164 30d *38b. 122 9d. 48 2b. *120 9a . 158 4a 5b. 3 6a. 61 5b 6b 13d 20c 22d 65 lid. 148 4d. 48 8c. 112 5a 15c. 23 13d. 23 3d. 87 4a. 81 2d. 74 2d. 53 15d. 97 38b. 29 2c. 66 4c. 48 2a 6b 8d. (d) It will be convenient to record here the instances in ' ' final 1 . 37 3b. . vowel of prior element in compound. 75 lc. 51 42 9a 51 6b. 35 7a. 92 14c. 121 122 lid. 56 3c 6c. 50 3b. 3 1 175 v. *149 4a. 126 4c. 115 5d. 116 24a. iv 6 lc 9c. 45 5b 6a 6d vi 1 8c. . 53 5c. i. 49 6b. 42 6b. 19 3b. 84 2c. *84 7d. *179 lb. 31 4b 8d. *83 3b. 124 7b.Short eighth syllable 24 24c 21a'. 53 5b. *101 3b. 20 5d. 32 2c. 83 4a. 38 2a. 128 If 4f. 30 6d. 27 Id. 4b. 70 lc. 4 7b. 60 5b. 96 ix 79 lc 5c. 19 10a. 99 Id. 52 3b. 68 4b]. 29 2b 4c. 203 27 7c 16a. (b) 89 2a. *117 7d. 182 8b. 33 9c lie 13c. 28 3d 4c 5c. 173 12b 2 174 iii 5 2b. 73 9b. 6 6b. 4 lc 14a. *27 3c. 34 iii 31 12c. 70 4c. v 27 vi 9 5a. 26 2a iv [10 Id]. 169 la. 112 9a 15a. 2 151 " iii. 62 6d. 104 10 132 3c 3d. 160 144 7c. *58 11a. 6 2d 3b. v 1 5d. 20 lc. 115 9b 9d 9e. 1 151 i. 29 9a. 114 Id. 41 9c. 223 i. 100 6c. 118 7d. 150 2c. 7 7d. lie. 15 9c. which a vowel in the eighth place is protracted or otherwise of doubtful quantity: namely i 31 5c. 23 6d. [78 6d]. 127 2f. 23 10c. ii 1 2b. 92 2a. 45 7a. 31 4d. x 2 7c. 11 lc 3 12 5d. 54 12d. 49 10a. 49 11a. *161 5b *7c *13a. 41 17a. 51 2a. 151 Id. 61 8a. . *56 110 8b. final s 4 142 175 a. 73 3a. 19 lb 7c. 3 20a. 84 2d. 103 5c. *13 3c. 60 2c. [61 lc. 48 9b. 89 11a. 79 2a. 189 4a. 67 8b. M\ 145 4b. vii 1 la lc. 64 5c. *52 3a. 56 21a. 9 3c. 15 la. *163 2d. 33 10a. 47 lc. 3 12a. 52 lc. vii 4 2c 12 Id. . 97 19c. ix 70 8a. 108 13b. 61 la. 25 7c. 61 5d 6a 7a 7d 13c. *94 4c. 89 3a. 68 lc 2c. 121 7a. 70 6d. 40 4c. viii 1 10c 30c. 92 lc. 38 7a. 84 2a. 98 3b. 9a. 77 5a. Syllables not final: i 36 16c. *88 6c. *18 5b. 91 10b. 73 Id 2d. 21 4a 9b. 32 5c. 45 3c. 31 7c. 111 3f . 227 iii c. 2 Id. 88 3d 4c. *10 13a. 103 2c. *59 4b. 38 6b. . 48 12c. 52 15a. 92 5d. 24 2d. 19 3a. *12 7d. *58 2a *2b . 5 t8c [46 2d]. 105 x [1 6c]. *95 4a *13a. 165 7c. *94 10b. 44 16b. 56 lc 5c. 138 If. 46 13c 14c. 97 14a. 94 4a. 36 4d. x [6 la lc]. *164 3a. 54 6a. *75 7a. 148 la. . 39 10c. ii 2 10b. *129 4c. 91 23b. (c) = 2d = 3d. 16 5a 11a. 130 lOd. 178 fid 2 4b. 86 42b. *164 2c *23b. *120 6a. 226 3 i. 33 lc. vi 3 3a. 23 7c 8c 16c. 40 12b. 50 6a. 33 13b. 60 15c. *48 3c. *82 4c. 132 lb If. 153 3a. 2c. 66 3c. 86 18c 21b. lib. 30 14b. 117 21c. Final syllables ending in a consonant: i 32 6b. 48 4c. 25 3c 5c. v 1 2d. 17 3d 14b. 37 2a 4c. 169 5b. 69 10a. 117 22c. 50 2a. 13b. 19 5d. 25 5d. 189 3a 1 . 3b. 88 5c. 108 10a. viii *1 33d. 96 Id 111 4a. 30 10c. l . *48 5a. *34 6c. 76 8a. 34 2a. 100 5c. 30 3b iv 2 7b 9d lid. 63 2 3b 6c 7d. 24 23c. 110 2a. 87 2a. 26 7c 8c. 115 2d 7a. 15 iii 1 15d. 186 6d. 74 3d. 138 4b. 16 16c. 19 12a 29a. 134 3b. 39 3d. 20 5c lie. 23 7a. 52 ix 81 lc. 69 9c. 20 lib. 21 5d. 61 2b. 99 la 8a. 10 178. 11 la 5d. 19 7a. 94 lb Id = 2d = 3d 4a. *130 7a. *104 12c. 158 lc. *12 Id *8c. 23 9c *49 lb. 154 2c. 53 2c. . 93 8a 14b. 167 lOd. 15 2a. 89 4a. 54 lc. 88 2c 6a. 6 7 s 151 ii. 5 7b. 149 3c. 105 3c. 56 6a 6c. *56 4d. lb. *107 lb. *59 7a. 25 5b. . 40 6c. 93 6a.

1 . [56 5b 6a]. [109 3b 10a lib 15a: 15b. vii 1 3a 18 17b 24b. *12 5c. 169 5a 2 ii 1 3b 10a. 58 2a 1 4d 4 3 5 [61 Id ]. *29 7c. 15 10c. 8 6d. . 68 3a lie. *42 lOd. 126 lc. 108 10c 3 x [1 7a]. 29 4d 3 33 3d 3 5d 3 37 4b 3 3 40 5a. (ii) . 93 5b 9a. 65 12d. 23 4b. 105 6a 42 la. : (v) (a) - - Long ninth syllable combined with some other irregularity * or . *48 9b. viii 1 16c 30c. 52 3a. 100 16c 4 103 4d. . . 39 14b 4 50 3a 14a. 11 4d 13 lb Id 2a 5d. 9c. 39 3c. 1 ajdra. 115 4a 5b. i 48 4c. 41 5c. 57 4c. *139 4b. 1 2c 1 . . 3 lc 5b 6a 6c [6 6c 3 ]. [x 78 3c]. 4b. 6 jdna. 4 7d 3 5 5 8 5c 10 5d 6a. 5b. . 93 7b. 93 14a. 38 6a 7c. ( 4 rdtha. . 105 la 7c 10c *108 8a. 67 6b. *103 10a. 53 lb. 12 4c. 60 8a. 168 2a. . . . 57 6d. 66 10a. 2 5d. 35 lib. . *11 8d. 22 4a. 5b. (iv) Long eleventh syllable: viii 25 17c. 135 4a 140 2 4 4 3 143 3d 145 2d. . 55 2 106 5a. 37 3b. 49 Id 9b. 20 lb 5 Id 3b. 92 la. . 49 12d 15c 5 50 2a 3a 7c 3 11a. 61 4c 6 8c 6 ix 70 3c 5d. . 149 lb 158 5a 166 2a.. 25 27 5b. [34 lib 14b]. 104 3b. 31 5c 4 33 2d 5 3b 5a 5b 4 10c. 7b *8b. *95 3d *4b 9 2 lie. 174 9a. 21 2c. . 73 9b. 74 2d. . . . 46 2d 29 la. 2 ddma a. 181 lb. 2 9d. . *165 3c. 127 3f 4b 5f 7b 9f 129 lb 6f 10b 134 3a. text rdkmti 3 superlative in -tama. . 88 lc. 186 2d 8c 9c. *88 44d. 68 lc. 50 4c. ] Short tenth syllable i 36 10a 12a. 44 lid. . 62 3d 5a. . . 101 2a. 91 4b 96 17c. . 63 3c. 4 3a 3 10c. 4 lb 3d 19 lb Id 3a 6a 6c. 20 5d. 56 3a. 16 lb 20b 17 18a. 2 14c. 87 6a 9a vi 1 12c. 122 10d 8 6 6 6 lid. 34 10 5 v 2 Id 4c 7b. 59 4a. 18 15c. or similar phrase. 17c. 97 10c. 79 lb. 76 id. 24 3a 3c 4b 7c 10b 10c. 1 . 12 lb. 111 lc 2c 2d 4c. 65 2b 66 5a 7b 4 67 la3 3c 5 lid. a . *180 lc *2b. 19 6a. 19 4c 14a 23b 32b6 20 20a. [70 9a]. . 89 6a. . 96 2c 4b 15c. . 90 4b. 57 3d. 61 la. 121 la 9d. 29 2a. 94 la lb 3a. 22 ix 93 lc. 20 lb 4 2b 5 6b 3 26 f4c 7 5 iii 1 5c. . ri. 36 4a. 115 5c. 167 lb 3 2c 13c. . 22 16c 6 25 10c. 84 5d. 68 2 vii 3 5b. 7 dvasc or dvasd. . 13 3a. 30 lc 6a. . *121 8a. 48 4c 6 18a. 27 5c. *181 Id. . . 63 lOd 14d. 25 la 6a. . 70 2a 8c. *164 3a. .204 6b 2a 6a 1 Short tenth syllable 1 1 25 3d 32 2a. 112 11a. 5a 6b. 1 Imperative in : -did. . . *98 3b. 87 5d. 34 24b. .sdh 175 i) : rdkmtlm is a quite vii 2 lc. 89 3b. [46 5bJ. . . *129 f7b 148 4b. 97 44b x 2 2c. -hi. 80 lc 4d. 15 3b 5d\ 17 12b. . . . 99 4c 7b. 63 3b 3 64 5b. 5 5b. 80 7b. 77 3b 85 3c. *50 4a. 60 4c 2 . 76 2b. . 1 taatdmbhat. 28 4d. [44 9c ]. 97 lb 26d 3 104 2c. 60 ix 70 2b 9d. 2 but see 244 iii. viii 20 4a 24a. 24 26c. 93 4a. 74 lc. fid 38 9a 39 2d 41 4d 3 lOd 15d 16e. 26 24c. 78 7d. 148 Id. 112 3d. 42 4c 5d 5 7d . 6 93 5c 6a. 69 7c. 40 5b. [67 2b]. 3 6b. 38 la 5 4 51 lib 62 5a 3 9a. 32 10c. *. 111 3a. 71 15c. 100 8c. 4 iv 1 2b 19c. *87 19c.w v. 89 lc *10c 91 21c. 141 12b 4 3 173 7c 8 8d 5 lib. 1 1 . x 6 2b]. *103 2d viii 15 12c. 63 4a. 46 26c.

. (vi) i Long eleventh ii syllable 168 lb. vii 21 9a 58 6b iv 26 6b. *167 4d. 68 2a . Details are given in the Table below with regard to the second place no distinction is made between dimeter and trimeter verse. 61 14d. Short syllables in the second place only slightly diminish in frequency in the strophic. sixth and tenth places the short syllable generally initial or medial. eighth and tenth places rapidly diminish in frequency after the archaic period. : to the influence of the secondary caesura ( 213). . are now in a position to consider as a whole the of short syllables to bear the ictus: that is to say. (c)u-^or. This is probably due varying degrees.. fourth. *142 lb. employment short syllables in positions in which long syllables are regularly 221. 151 i. [ix 109 18b] 1 17a. We eighth and tenth places in trimeter verse. 11a. and cretic periods. 4 4c. 2 see Metrical Comm. the proportion being much larger than chance can account for: whereas in the fourth. sixth. combined with some other irregularity: viii 20 9b. The relation between vowels and other short syllables remains uniform throughout. vi 12 6c. iii . . and are again more final common short in the popular period.. the short vowel is found in about 3 per cent. i 122 x 74 4c. [ix 122 10b 109 21a] vi 1 . 205 .w. required. . . 46 12c. is In this point therefore no historical developement perceptible within the period of the Rigveda. But even in this period there is a distinction in usage for in the second and : : eighth place the second syllable in half the instances contains a final vowel. 63 2c. 20 4a. vi 29 2c. and which are immediately followed in each case by another short syllable within the same part of the verse. 133 7a. .* or 60 3a4d. 93 7a. Short that is. Short syllables in the fourth.Short syllables bearing the ictus 121 8c. 1 . though in is To this there is only one exception short eighth syllables which contain a final consonant are more common in the strophic than in the archaic period. 115 5a. 11 3a. (b) w . normal.-v^^v415b. and in the second. : It first appears that in the archaic period these variations are about equally common in each position that is to say. 26 7a. x 30 13a 2 61 2c . of the verses. 22 lg = 2g = 3g. in the syllables are thus vised in every possible position second and sixth places in dimeter verse. 48 17c viii 22 2c x [6 lb]. *83 5c.

but much more closely akin to the : latter. . been given in the table in Here again the movement is on so small a scale as to have little practical importance so far as it goes. it indicates that protraction is a secondary developement. The details (proportionate to 5000 verses) have already 180. This reduction seems not to extend to trimeter verse. distinct alike from the shortening of original long syllables and the general license to throw the ictus on final short vowels. Table of short syllables bearing the Periods ictus. 222. apparently Howfor the reason that a short third syllable is more common there.206 Syllabic variations of the cadence ( It has appeared previously 188) that in dimeter verse there is a distinct reduction in the of irregular openings after the archaic period. The table further shews that both in the 2nd and in the 8th place protraction is on the whole more common in the later periods than elsewhere. number ever the movement is on so small a scale that it does not seem to be worth while to pursue the matter further.

v 87 9a. and (b) in the popular Rigveda. instances of catalectic Jagati are classified below as occurring (a) in lyric hymns. 129 5f 8a. 150 lc 3c. are not found in either of these periods. 32 25a. (b) elsewhere in the Rigveda proper. (i) syllabic irregularity and (ii) contamination (a) of Jagatl stanzas by Tristubh. marking the cretic and popular periods. Similarly extended Tristubh verses are classified as occurring (a) in the Rigveda proper. 46 20c 26a. marking the archaic period. which accidentally take the same outward shape. Hypersyllabic verses. in view of the great body of Tristubh verse in which no such variations are found ' : but it cannot be so confidently denied for the ' catalectic Jagatl ' variation. namely of the cadence. ix 107 9c. lyric Jagati not in accordance with Indian usage. and (c) in the popular Rigveda. and in the second verse of stanza b (Jagati) below ' ' (i) : () md no hrnltam N dtithih vdsur agnih puruprasastd esdh viii yah suhota su-adhvardh (b) 103 12. Catalectic Jagati is found in the second verse of stanza a (Kakubh). The question therefore arises whether there is a direct historical connexion between the two former variations as occasionally found in the archaic period. mostly in the archaic period. marking the latter only. In accordance with the presumed origin of the variations. 127 la 2a. vii 16 5c lie. 111 la.Catalectlc Jagatl 207 popular hymns as catalectic Jagatl. therefore entitled to distinguish two sets of phe- ' nomena. But so far as the evidence were goes. 101 9c 10a. (b) in Jagati stanzas in the : Catalectic Jagati verses are found as follows the 'uneven lyric' hymns ( . viii 24 14c. lyric stanzas as follows: i 36 12a. but seems convenient for the present purpose. sdtyam tad indra\\ \\ varuna hrsdsya vam mddhva urmim duhate saptd vd?ilh idbhir dasv&msam u avatam subhas patl yd vam ddabdha term ' \\ abhi ptiti cittibhih ' viii 59 3. 90 6c. 103 6c 12b. 54 8a. mostly in the archaic period. 128 7f. and as frequently found in one or both of the later periods. and in other 39 6c. x 126 5c. Such a connexion seems highly improbable for extended Tristubh verses. principally in the cretic period. and (b) of Tristubh stanzas by Jagati. 25 18c. 26 22c. since in the intervening periods very few hymns composed in Jagatl metre at all. however. the ' catalectic Jagatl We seem verse was equally extinct in this period. The use metre of the is for the trimeter verses that occur in (a) frequently in 186) including viii 35.

i 88 lb. 165 lc Id 4d. 75 7a 7b 8a 9b. 164 14a 14c 24c 39c 39d 44a 45b 47d 48c ii 42 v 83 10c. (b) *in the popular Rigveda i 24 15c. 16 3b 3c 8a 8a. 56 5b. 74 Id. vii 4 6d. 85 18b 18d 27a 27b 44d. 14 la 10b lib 12a 12b. 94 la. 37 4a. iv 19 5b. 182 la. 25 ii iii 114 4a 5b. 27 24c 24d. 66 2a 10a 11a. 93 8c 8d. 140 10a. characteristic of the archaic period. 69 12c 12d. 53 5c. 59 2a 3b 4b 5a . 17 2b 4c 12d. vi 13 5d. 16 5a. vii 103 viii 100 12b. 4a. 114 la 2a 2b 3c 4c 4d 5d 6b lOd. 162 4d 6b 8a 8c lib 12c 12d 14a 15c 21a 22a 22b. lc 2b 3b. 18 10c 12a. 44 7b. (c) *in the popular Rigveda: i 164 36a iv 58 lid. They appear to be 224. 163 2a. 26 6d. 122 la 3b 3d 8d. 99 4c. 128 7a 7c 8c. 84 4b 5b 5d 6c 7b 7d. . 149 2c. 94 14a. 130 2c 2d.208 : Extended Tristubh 7a. . Id 2c. 85 lie 12d i* . viii 48 5c. 72 lb 6c. 96 lib. Hypersyllabic verses' containing thirteen syllables (Tristubh) or fourteen syllables (Jagati) are usually formed by an extension of the rhythm of the cadence. [61 11a]. [46 7c]. 110 6a 5c 5d 6d 8a 9b. 36 6c. 1 . ix 82 5c. The following are examples abhi vo wrdm ekah swparndh \ { of hypersyllabic verses : \\ dndhaso mddesu gaya sd samudrdm a vivesa viii 46 14a. Rigveda proper i 35 3b 3d 9d\ 101 5b. 180 4a 9c. 76 3b. 145 la 4b. 53 lie. 115 5c 5d. 62 la lc 2c 3b 3c 4c lie lid. 117 2a. 125 4a. 161 la 4c. vi 47 29b 30c 31c. 92 5a. vi 61 3c. > 142 iii 6. 53 10a. 181 Rigveda proper: ii 3 3c. 104 3a 7a 7c 23d x 18 13d. 138 6b. 98 la 5d 7a 12a. 100 9b. 82 6c. 101 9a 9b. 54 Id. vii 50 41d. 5c. ' 229. 75 7a 14d. 1 142 ii. 95 3d 11a 12d. vi 75 10a. 129 3b. first ' (ii) Extended Tristubh td is found in the verse of the following stanza : avadan \\ prathamd brahmakilhise salilo dkupdrah viMharas dpo devih ^ \\ mdtarisvd || tdpa ugro mayobhuh pratharaajd rtena *x 109 : t. 3 6c. 55 5a. 34 5c. 59 8b. 168 2c. They are iv 1 2a2 . 88 la. 121 3c. lb 2b 3a 3c. 103 2a 2b 3a 4c 6b 8a 109 la 5a 5c. x 10 14c. 86 3a 48d . 114 11a. *x 114 Hypersyllabic verses are numerous in the hymn viii 97 10-15. 59 5a 5c. iii 31 20c. 35 10b. 120 9d. but there are also ' several occurrences in the popular Rigveda. 63 3b 10b lie 14a. ix 70 la. 81 3a 4c 4d 5a 7a. 91 13d. 113 10a. 50 2 lib. 1 also 'hybrid' verse. 1 16d. 10a 10b. 37 6a. 133 6a 6f iii 59 2d also found as follows 1 : . 117 8a 8c. 7c. 51 8a 9a. [11 3b]. x 23 5a 5b. 71 2b 2c 9c 9d. v 33 4d. . 18 2d. v 44 10b. ii 32 5b 5d . vii 41 Id. 47 18c 18d. 125 2a. 52 14a. Extended Tristubh verses are found as follows (a) in the 53 10a. 25 3a. 53 16c. 64 7a. 2 102 6c lOd. 60 2c. 102 2c. where they occur as 10b 1 13a 13b 13c 13d 14b 15b 15d. : x 23 .

These hymns must necessarily belong as a whole to the archaic On the other hand the Pentad and Bhargavi hymns period. and The hymns if metres. 90 5c 103 5a . ing-point ( 94 iii) the view that the teristic of ' We ' have taken as our ' start- rests : generally are charac- an early period of the Rigveda and this view is borne out by the fact that they are comparatively rare in the popular Rigveda. i 61) only differ from proportion of decasyllabic verses. in shewing a a. and the these verses are chiefly found in the decasyllabic Tristubh hymns. 229. A closer examination however shews that the rests fall into three classes. similar type (ii) verses agreeing on the whole with the normal trimeter rhythm. but chiefly found in decasyllabic metres. their lack of distinctive character in decasyllabic Tristubh and those in decasyllabic examined on this basis. 15 14a. *121 7c. 1 otherwise this verse of 14 syllables. 52 2 2 14a. them much higher 14 . 143 iii : 225. the rest being at the fifth place. verse 1 The ' syllabic variations which affect : one part of the ' but the rests and only have already been discussed verses need to be considered separately. partly according to the rhythm of the verses in which they occur. . but to a certain extent are spread over the whole Rigveda. rhythm sixth syllable being long of the break.Varieties of the 1 1 ' ' rest 209 1 1 vi 10 lb. and partly according to their use in the Rigveda. namely (i) verses in which a rest is associated with iambic . but having the first part of the verse defective by one syllable. Irregular rhythm. 2 3 reading apddm. All the hymns in decasyllabic Tristubh are alike in shewing a free admixture of regular Tristubh verses with decasyllabic verses of the Viratsthana and neutral types. . on account of they are found most frequently in the hymns in decasyllabic Tristubh. the types being the Pentad and the Bhargavi verse and (iii) verses : : entirely agreeing with the ordinary trimeter rhythm. 14b 3 14c. Verses of the last class may be said to contain : ' ' neutral rests. viii 46 *x 87 12b *13c *114 4a. *vii 104 15a. The Viratsthana and Gautami hymns (ii 11. are found to fall into two groups which do not exactly correspond to the differences denoted by the titles. and are of the Viratsthana or some . also a very high proportion of other archaic variations. 26 2c 7c . is a hybrid verse. *139 4c . as they involve hybrid difficulties as to the position of the caesura and therefore affect the general structure of the verse.

*95 4c. if by so doing irregular rhythm is introduced' This rule. of which almost 400 are in the decasyllabic metres. 144 4c. and about 50 in the popular Rigveda. x 32 5b. 105 8a. v 41 9a. . however. decasyllabic metres. according as they are found (a) in the hymns in decasyllabic metres . the types the . and . as shewn in brackets in those sections. 112 6b. 143. as the occurrences of indra in the Vasistha hymns ( 149 i). 132 la. about 100 in the hymns in decasyllabic Tristubh alone. the same number in the whole of the rest of the Rigveda proper. 50 9c . 61 26a 27d. ix 110 10a 79 5b. often give a reason for a preference it should at least be a general rule of metrical interpretation that hiatus or resolution should not be postulated in order to bring a verse up to : Even on the lower computation. 132 2b. 55 8d. From the list of verses containing one or more rests we omit the following. ii 19 5a. 173 4a.210 Doubtful decasyllabic verses consist almost exclusively of decasyllabic verses of their respective types. 150 and 151 above are rightly to be suggested in Of these doubtful cases there are explained as decasyllabic verses. Some of the instances are open to question. The caesura may most conall In veniently be reckoned as following the rest. viii 46 16c 20a. verses with a double rest. 149 ii-iv. The interpretation chosen will therefore materially affect our view as to the extent to which the system of 1 rests ' is carried. iv 26 7a . because the rhythm is so exceptional that it becomes very probable that errors have crept into the text: i 120 lc 6a 7b. On the other hand it is probable that some at least of the verses for which resolution has been 142. The rhythm may. vi 17 10c. *162 16c. and have comparatively few irregularities of any other kind : it is therefore unlikely that they are earlier than the strophic period. 226. about 250 in decasyllabic Tristubh verse. decasyllabic verses are about as common in trimeter verse as is hiatus ( and there is therefore no abstract reason why in 131) individual cases we should prefer the one explanation to the other. (c) in other hymns in the Rigveda proper . The ( the break are early caesura (iv) decasyllabic types which have iambic rhythm at (i) the Viratsthana verse (ii) indra following an 149 i) (iii) the GautamI verse (see below). as already observed. (b) in decasyllabic Tristubh or (d) in the hymns . and excluding the : ' number of syllables. seventh syllable is more often short than long. the regular ' ' . In recording variations belonging to these different types it is convenient to distinguish the occurrences. The lists of decasyllabic verses given in the next three sections include over 900 instances. is opposed to the interpretation indara so far as the archaic period is concerned. popular Rigveda.

151 iii. 11 times: also . ii 19 3c 7c. iv 37 4b. 63 3a 6a. 149 3a. : Of the first variety there are 106 occurrences. 45 2b. 26 7b. 18a. 8c. 66 3b . (d) none. 19 10b. 104 2a 1 8a. 63 2c. 153 2a. viii23 1 x 46 5d 6b 7c 2 (b) i 60 lb. ydd dha syd indra srustir dsti i 178 la. v 33 4a 5a. 33 la. (d) *x 59 5c . 120 4a. 29 la. 99 2d. (c) i 58 8d. 33 4c. and a long Its two varieties are illustrated by the following with short seventh syllable rdsi examples (a) ksdyam u r&si mitrdm asme ii 11 14a. 23 3a. (6) with long seventh syllable dvdbhinad ukthair vdvrdhdndh \\ ii 11 2d. . 2 165 14b. 25 lb. and the Gautami verse in i 61 otherwise these verses are : chiefly found in the decasyllabic Tristubh hymns. 1 x 6 lc 6d. 142 . 89 6b. 148 la. 39 occurrences: also i 61 5c lOd lie 14d 15d. 67 10a. as follows (a) ii (a) 11 la 2a 4c 7a 11a 14b 16d 20d. 117 iv 44 5d vii 6 2a. 41 8b. 50 11a. . 174 9b. . 35 2a 3a. 2 x 77 7a. 30 lb 4a. 140 48 1a.The Viratsthana verse 211 The Viratsthana type is very common in the hymn ii 11. 92 4a. 46 7d. . 127 7a. 173 5c. vi 13 lc. 10 4a. 129 la 4a 167 la. But all the types are occasionally found with Jagati cadence. 21 5a 6a. vii 19 2a 6a. x 29 3a. 177 ii 17 8d iii 32 12a. 25 la. 5a. 2 Jagati verses. viii 66 5c 97 ix 88 la. (b) i 63 la to 9a. Of the first variety there are 98 occurrences. 20 2c 4d . (a) in ii 11. (a) in ii 11. 53 11a. 100 17a. 4c 2 . . x 22 2a 11a 1 12a 13a 15a. 33 14a. 155 5b vi 3 7b. (c) i 33 14c. as follows (a) vii 34 17a. vi 21 x 61 16a. 22 6c. v 7d. vii 1 15b. 17 la. 111 10c. 51 4a. 24 10b. 2 135 5b . 154 3a. (b) Of the second variety there are only : half as i occurring as follows 68 2b. 14b. v 41 16d. 61 3d 4a 6a 8a lid. 7a to 10a. (i) sixth syllable. 13c. (ii) of the Decasyllabic verses in which indra follows early caesura are same types as Viratsthana verses. 58 6a. lc. 5a. 21 8a. 2 3c ix 93 5c 3 x 30 9c. 71 4b. 3 but see 170 ii/. 28 3b. (c) i . 49 Id 3b 5c\ 61 2c 3d 17a. x 6 6c. (b) i 63 9c. 78 la 6c. 173 13a. 19 2d. 131 6f. x 2 2 6d. 117 lb. (d) *x 17 13a. 178 2a. 22 la 8c. 45 9a vi 20 2d 6c. 2b 6c. many instances. 169 la to 5a. 44 15a. 77 2c 3b. as illustrated by the following : examples (a) (b) yd udfci n indra devdgopah ta \\ i 53 11a. 23 5a. 148 2b. *120 lb. ii' 19 3a vi 20 1 la 13a. . 174 la to 5a. 80 2c 4c. : The Viratsthana verse has a rest at the fifth place. 50 3a 4a 1 61 22a. iv 16 21a. 23 62 43 56 . 148 2a 4a 1 . 93 11a 1 : 1 . iv 21 6c 7a 7c. 165 3a. 21 8b. vi 44 9a. which are most frequent in the collections i 58-64 and vi. 63 5c. 122 4d. 47 9a. ii 20 6d v 33 4d. 121 6c. 14a. 133 6a. vi 22 3a 10a. 99 7d 8d. 19 la 2b.

Occurrences with long seventh syllable (a) vi 44 7b 7c 3 iv 3 5d 12c. iv 3 5d. The two forms of the Gautaml verse may therefore be illustrated ' : by the following examples bhdrdmi dhgusdm dsiena (a) brdvah had (b) aryamne kdd bhdgaya u : i n 61 3 b. 173 5a 7c 10b 11a. 3 (d) none. 1 . i 61 Id 3b 4b vi 24 20 8b.186 60. 1 (d) *x Jagati verses. Gotama hymn i 61 . . The tinctive metres are the decasyllabic types which are chiefly found in disPentad and the Bhargavi verses. 3 but see 244 iii. 123 b. 33 variety occurs six times. 2 Alternatively as a Pentad verse. the seventh syllable being therefore short but it is probable that some occurrences with long seventh syllable in other hymns are really of the same type. 20 178 la 3a 4a. . except that the rest Such verses are very common in the precedes an early caesura. Jagati verse. (a) ii (d) none. and belong to the archaic period. 1 indra follows the caesura. . 105 5a (c) i 88 2c. 227. 2 vi 26 la . 1 . (a) The (b) viii i these varieties occurs nine times. . namely (a) x 22 9a 61 24c. x 74 3c. 7d . la . in the occurrences in that hymn the rhythm of the break is always n ^. As the great majority of these instances are in hymns in decasyllabic Tristubh. The majority of those trimeter verses in the Rigveda which (iv) have only nine syllables may be explained as Viratsthana verses with an additional rest at the fourth place. x 132 (6) 3c. ' 139 4c. x 74 lb. : . Alternatively with hiatus. follows: (a) none. 24 la. as (b) 1 8b 167 10a. x 22 la. vii 20 2a . vi 20 3b. 23 la 49 11a.212 The Gautaml verse Of the second variety there are only 35 occurrences. 148 2c. ii 1 1 i aydm first of sd hbtd yb dvijdnmd 1 3b 18 15c. 6^130 2^. x 49 2a 2 . 129 11a 12 174 6a. v 36 iv 24 2b. The name Gautaml verse is suggested to describe a metre (iii) which agrees in form with the Viratsthana verse. ii 19 lb. . (c) i 138 2f (b) vi 20 2b. The following verses will illustrate the (a) (b) two varieties \\ : vdyo nd paptatd sumdydh | ( i i 88 149 Id. v 46 8b . x 1 6 3a. 9a 88 Id. 173 3 61 23a. 21 2a. 1 Occurrences with short seventh syllable: (a) 4c 10b lib 13a 13b 13c. (b) i 129 iv 21 10a. verse is The Pentad commonly equivalent to a Tristubh verse . 50 2b 1 148 la. it becomes highly probable that they are varieties of Viratsthana verse. 39 4d. (b) i 77 5b. (6) The second 1 4c 1 . and therefore also to be associated with the Viratsthana verse. 2 Jagati verses. 5a. namely (b) i 149 5a. (c) i 89 1 Hi 49 la. 11313a H 1 . . (c) i 147 lb . 4a 5b 6a 7a. (c) 2 19 33a . viii 96 20a.

120 3b iv26 7d. vii 52 3c. 174 3d v 15 5c vi 25 iv 3 lib 13d. ix 109. The great majority these occurrences are in the Dvipada Yiraj hymns ( 94 iii b) in the : hymns in decasyllabic Tristubh verses of this type play a comparatively The two varieties may be illustrated by the following small part. vi 24 6d. with the result that the third syllable the seventh and ninth syllables are is generally short. we may find examples of Pentad rhythm with and of Bhargavl rhythm with Tristubh cadence Jagati cadence. 99 8b. one having a rest in the The Bhargavl fifth place. 6 ix 109. cf. differs from all other decasyllabic variations in being based upon The Bhargavl Jagati verse variation. . vi 63 2b. of The Pentad verse occurs about 170 times with short seventh and 130 times with long seventh syllable. x 1 6c 6 6 times. vii 34 i times. 62 (a) 26 times. can be found. ' Examples syllable. 4 151 244 ii.The Pentad and Bhargavl with rest at the sixth place : verses 213 but each part of the verse exercises an influence over the other. x 45 8b. . but these examples are isolated. (c) 4 *x 114 10b. but the resemblance appears to be accidental. 124 6c. For the immediate purpose it will be sufficient to is distinguish two short or long . 56 i-n. . (b) i 153 la. according the other characteristics are discussed elsewhere. : it is One therefore strictly speaking a of its forms agrees with one ' ' hendecasyllabic of the forms of Pentad verse. 18 times. 13 times. 68 5c . 149 2c 4a 6 times. 12a. 50 2c . Both in Pentad and in Bhargavl verse the rest appears to follow the caesura: upon this assumption each type in its more Occasionally : common of ' double rests (i) variety agrees with normal trimeter rhythm. 16a. and often long. (b) i 104 lb. . 5 times. x 99 8c. . 1 . *120 Id *4a. transferring mdnma to 3a but see . ix916a. 67 11c. 56 1-11. examples : (a) (b) tdkvd nd bhurnir u vdnd sisakti i i 66 2c. *101 7a. as the seventh syllable verse has two forms. ii 11 vi 44 7a 8a 8b 9b 9d 55 times. . 14 times. x 1 Id 5b2 6b. Occurrences with short seventh syllable (a) i 65-70. 67 10c. 46 6d. 152 iii. 66 3c. Id. (c) i 62 8c . Occurrences with long seventh syllable: (a) i 61 15b. 87 6b . 46 11 times. 24 times. (b) *i 162 16d . (d) *x 83 3b. varieties. . 3 v 41 19a. ii 11 17d vi 44 8c. x 93 v 41 15c . but are rare. 73 3a. except for the difference in the cadence. ii. . vii 4 3b 6 3c. . dddhdra ksemam M bko nd ranvdh : 66 3a. 89 8d. 7d. 1 3 2 with short eighth syllable. and no importance can be attached to them except as suggesting emendation of the text. 66 6d. vi 17 i 130 10c. 65-70 (b) vii 34 1-21. reading yajiid-yajnasya. and the other one in the sixth place.

Several of these verses may also be interpreted as Gautami verses for instance vii 52 3c. of the verse complete and of the ordinary rhythm the rest therefore must be considered to precede the caesura. (b) i 122 3a. (b) i 129 9f. The varieties that follow bear a resemblance to the Pentad (iii) or Bhargavi types. 1 sixth syllable long. 112' 10b. vi 12 6a. 24 5b. 4a. vi 12 . with short eighth syllable. rdjdno n nd citrdh susamdfsah 1 : This example seems to stand alone. (c) i 83 6b. ix 71 7c 1 79 lc. (c) (d) *x 14 5c. coming either at : the fourth or at the fifth place. . Pentad verses with sd tudrh rest before and after the caesura. Any textual emen- dation which will add a syllable to the first part of the verse will have the effect of restoring a regular trimeter verse. vii 96 2c. cadence: (a) i 61 6c. viii 96 21a 1 ii arvan nidayah 20 3a. n The occurrences are: x 55 6a. (c) i 36 8c. Verses with neutral rest are such as have the latter part 228. viii 27 12c. 100 6b. v 51 13a. as in 226. ii 18 la. 36 la. : Occurrences with rest at the fifth place (a) x 77 la lb lc (a) 2b 2d 3a 3c 3d 4d 5b. x 32 4c 2 : . viii 52 4a . 78 lb Id 2c 2d 3b 3c 3d 4b 4c 6d. 117 22c. but are too scattered to be of historical importance. vii 75 (c) i 100 5b. (a) ii (c) . 87 6b. : The varieties of Bhargavi verse have been illustrated in 52 (ii) the form with rest at the sixth place only differs from the second form of Pentad verse in the cadence. Pentad verses with short seventh syllable and Jagati cadence: viii 98 7b. dididno bhdvati druhamtardh 127 3b. 31 7c. Occurrences with rest at the sixth place (a) x 77 Id 2a 2c 3b 78 4d. 130 lOd. Bhargavi verses with rest at the fifth place and Tristubh (b) x 6 2a. (d) Bhargavi verse. v 33 2b. (d) *vi 47 31c. 1 : 1 (b) 4c. 173* 2c. (c) ii as na (b) 2c. with rest before and after an early caesura x 78 lc.214 Neutral rests : For the quantity of the third syllable see below 249 the instances of a long ninth syllable have already been enumerated in 220 iii. 2 vi 15 Id. The rhythm is that of the normal trimeter verse after early caesura. x 61 27a. 70 10a. 48 4a. 2 fifth syllable long. The two (i) (ii) varieties are illustrated by the following examples i i : anjasi u kulisi virdpatni u 104 4c. *x 56 4d.

48 17c. . vi 20 7b. be obtained at the cost of an irregular opening. ^ 5 4 with irregular break. 53 24a. 13c. 37 4a. 3 x 23 5a 4 50 4c. (c) i 39 3c 2 iii 23 4a. 49 4c. 8a. 173 10a. . ^13 3c.. H . *84 4c. . 3 ii 11 lOd Neutral rests at the fifth place: (a) i 61 10a 2 2 8 2 i 127 3b 9a 10a 10b 2 135 lb 78 4a. Alternatively read maniebhih. 40 2c iv 1 19c. (b) i 60 4c. . 92 14b. 122 5a 7d 2 *x 12 *v 40 7a3 7c. viii vi 17 7b. *164 30a # 35b *48c (d) 8d. . . 88 3c. 60 9b. 93 9a. 45 Id. *95 4a *10a (d) *vi 47 31b. 1 . 41 17b. 148 4b. . . *180 3a. 244 a. 63 4c. 17 15b x 31 6b. 39 14b. *x 16 8c. . at the cost 3 2 Refrain Supply srotu nah to begin the verse. 229. 52 2a. . 97 6a. 68 4d 7a. 138 3f '. 1 Jagati verses. 14 5b 5c. 79 lc. (c) i 48 3c ii 2 la. 141 99 6b lie. be alternatively explained by hiatus or resolution. 15c. *94 11a. ii 4 3b 8b. (c) i 35 9d. . *51 9a # 82 4b. 2 64 9a 2 x 23 3c 2 7b3 50 5c. . 159 2a 180 6c. (b) i 167 lb. 4 lc. *128 8a. 20 2a 2d vi 20 2c 5c. (b) ii 4 5d 145 5a 5d. . . . *x 28 lc. *129 6b. 27 6c. 61 lc. 22 14c. 89 4b. *\1Q 3b. 31 158 3b. 79 3a. x 77 4b . *85 44c. 178 lc. 9c. 110 9b. 61 25a. *i 161 14c. 180 3c. . 1 may 3 2 In these verses the full number of syllables 152 i. . Neutral rests at the fourth place: (a) i 61 11a ii 11 3b 6 8 122 6c 6d a 167 lc 7b'. 3 14b. (ii) . . lOl 10b 1 *130 la. 167 6c. 17a *23d *45d. . . . 169 6c. but see verse mddema. (iii) Bhargavi verse the first early caesura gives us the first variety of 227 iia): and a rest following a late caesura gives us variety of Pentad verse ( 227 i a). vi 4 8d 3 10 7b 12 6d 3 13 6d 15 iv 25 la.. . 186 3a 7c. . . 107 9a. 184 5b 5 190 7c. 93 14b. 4 2 v 30 17 lc. and are there doubtless to be explained by tion. 101 12c. 3 3 v 29 13b. 46 19c. 76 Id. 30 la. 3 vii 5 la. 66 v 33 10a3 41 5d lOd 16e 3 iv 21 5c. 2 2 2 x 3 la.. . x ii 4 8d. 61 27c. The instances are (a) none. # 87 14a. 1 x 122 3b. 29 2b. 2a. 6a 57 8c. 24 3a 10d 4 62 3a. . . 9c 10c. . 120 la 3c v 33 lc 2c. 61 8c. But there are also a few examples in the earlier parts of the Rigveda. . . *ii 43 2c . 104 4c. Jagati verses. 169 7d. . ( A rest following an Hybrid verses ( 56) are most common in the popular contaminaRigveda. 4 168 iii. ^53 4c. 74 3a 78 8a 112 2a 7c. . 132 lb 7a. 174 2c 8c. 16 15a 61 4b. 30 5a. vi 65 5a. 2 2d. (d) *i 161 8a *164 *vi 75 v 44 15a # 15b *15c. ix 88 4c3 24 17c 101 12a viii 19 18a 2 2 3 3 2 1 63 15a 66 4b. . 3 vii 100 3a. 1 . 105 lc. and particularly in the decasyllabic Tristubh hymns. ' ' ' Hybrid Jagati verses are very : ' rare. . 15 7a. 1 The verse may i of 6 an irregular opening. viii 46 17b. 113 7c 115 9c. . 112 lc.Hybrid (i) verses 1 216 . . 59 8c. 105 3c 6c. 5a. 18c viii 59 6a *7a *7d 103 7a *lla. 57 lb. 7b. 50 2c. ix 86 43a. 1 1 .

Table of syllabic variations Table shewing the use of the Syllabic Variations. .216 230.

or smaller groups outside them united by their character or position. . (ii) verses in (iii) the strophic period is marked by the frequency of the which 'secondary caesura' is found ( 213). so that the homogeneity of the The division of the hymns of the larger groups can at once be tested. vi. and even of the group x 29-80 shews that the separate parts (as for instance the Agni hymns. in the popular in the cretic period.Metrical criteria of date 217 tive periods Subject to these qualifications. iv. 232. obtain prominence. are The most important groups united by their character the hymns in the different lyric metres. and (c) hybrid verses ( ( (6) catalectic JagatI verses 225). From 226) : and (c) verses with secondary caesura ( 213) must be excepted. break ( in the cretic period the cretic -w -. the Indra hymns. vii. either after the third syllable or in a compound ( 214): (b) the iambic and irregular forms of the break ( 218): (c) irregularities in the rhythm of the cadence ( with iambic rhythm (a). and therefore it is unnecessary to give details here. to other deities) are practically identical in rhythm. the characteristics of the respecmay be summed up as follows : (i) chiefly in the archaic period we find (a) the weak caesura. in the strophic is I. Table in for the larger groups of Chapter III. popular Rigveda into small groups is of an artificial character. *. ' : (a) extended Tristubh verses ( 223 223 ii). It is unnecessary to give here a list of the hymns in each of the 57 small groups examined. as the contents are sufficiently indicated in the To each of the small groups are prefixed the results 235. since only the hymns x 10-18 form a natural group. The smaller groups to which these tests are applied either subdivisions of the large groups.^ (iv) and normal periods the Pentad opening more common than elsewhere. the hymns. and to a less extent the syncopated opening ^ ^ (v) j. Small groups united : by position are found in all parts of the Rigveda outside the homogeneous Mandalas and in Mandalas vi and vii the presumed later additions form groups of some importance. (b) ( 220) (d) decasyllabic verses or with neutral rests ( 228). Subdivision may be of the homogeneous Mandalas iii. ' Rigveda the rhythm is almost the same as but the following syllabic variations occur i).

the general five verses. shew special favour to the 116-119. of the group i 76-92. but this character is least pronounced in Group III. which yet 234. Amongst the smaller groups the Usnih. In the normal period several of the groups diverge from the The Vamadeva hymns (Mandala iv) have as general standard. and (to a lesser extent) ix The groups of the cretic period shew a remarkable agreement in all particulars. On the other hand the hymns of Mandala v. and The hymns interest. tion. : In the strophic period all the groups agree in shewing the archaic variations about once in every eight verses but the use : of secondary caesura is confined to the Vasistha hymns and one or two of the smaller groups in the Kanva hymns there is no : trace at all of this variation.218 233. the hymns of the group i 58-64. the decasyllabic Tristubh hymns of Mandalas ii and vi. results of this investiga- The three principal groups of the archaic period have on the whole the same metrical character. the deca- hymns of the groups i 165-190 and of Mandala x. as Reviev) of the trimeter groups The following are the principal shewn by the Table in 235. and the remainder of the Bharadvaja hymns have archaic variations once in every fifth verse. rhythm being very strict in all the the cretic break being used once in every four or groups. Kakubh-Satobrhati. and the hymns of the Mana family (i 165-190) other than the hymns in decasyllabic Tristubh do not shew any large syllabic proportion of archaic variations and the two latter use the Pentad opening as freely as the hymns of the normal period. variations occurring in every third or fourth (in verse. and the group x 48-50 decasyllabic Tristubh and JagatI) have the most marked archaic character. The hymns Tristubh in Atyasti and Brhati-Satobrhati. Since it . many period of the archaic variations as the : hymns of the preceding the hymns of Mandala Pentad opening: the group i the third Mandala combine secondary caesura with the cretic break. and uneven lyric hymns. of the popular period present no new feature of appears that the secondary caesura is entirely absent from some of the groups of the strophic period.

be assigned to this branch of the strophic period. as a whole.Parallel developements agree in other points. Popular Rigveda. Strophic hymns with secondary caesura (Vasistha (Tristubh hymns) {Group i 55-57 \ j Vasistha (Br. hymns) Kanva VYJimadeva (iv) 4. Cretic ( Mandala v Soma hymns {Mandala hymns with secondary caesura Group i 116-119 (iii). the Brhatlhymns of Mandala vii. 219 we must necessarily conclude that the metre in this period followed two parallel lines of developement. In this group i 116-119 the developement of the ere tic break is also very marked. one only of these being characterized by the secondary caesura.-Sat. v and ix being from secondary caesura. and perhaps It seems also probable that Mandala iv those of Mamateya. i (Vamadeva hymns) should. Normal hymns ii 5. of bks. the Kanva hymns. the hymns in Mandalas ii. Cretic hymns. of trimeter metre appears therefore : The general developement such as may be indicated as follows 1. v and x. is found By this definition the following small groups will properly be assigned to that branch in the strophic period which is free from secondary caesura: the Gotama group i 76-92. The same division into two branches will then be perceived normal period. Archaic hymns Lyric metres generally ( j Bharadvaja Decasyllabic (vi) ( hymns 3. The second branch will consist of the Tristubh hymns of Mandala vii and the group i 55-57. "(Kusika 6. 2. The general characteristic of the period must then be defined as the use of the archaic variations to a smaller extent than in the archaic period. whilst those in Mandala iii and the group i 116-119 retain the principal form of secondary caesura in the free and discard some of its rarer forms (see subsection iii a below). Strophic hymns Gotama (i 76-92) Mana (i 165-190) . (ix 68-97). 7. ii. . the Mana group also Satobrhatl 165-190 (excepting the decasyllabic hymns).

. though the evidence may not always be sufficient to justify us in naming particular hymns as later in date. In the strophic period there is a considerable divergence (ii) (a) in the proportion of archaic variations between the three groups of Kanva hymns but the number of trimeter verses is so small. 201 iii. The great frequency of the secondary caesura in the Vasistha (b) Tristubh hymns stands in striking contrast to its entire absence in the Brhati-Satobrhati hymns of the same Mandala. iii. period. that we should not be justified in separating them. Wherever the bulk is sufficient to enable us to examine the two sets of hymns. chiefly found in Mandalas vi and viii. because the decasyllabic It variations themselves are not included in the case of these hymns. 105 iii as to the real antiquity of the doubt already expressed in It will perhaps be best to refer them to the strophic these hymns. the following points have some probabilities to (i) recommend them : In the archaic period the Atyasti hymns. chiefly found in 127-139. In all the groups into which the large Group III is divided the (c) question arises whether the hymns in decasyllabic Tristubh are really of the same date as the remainder. The Pentad opening is also much more common in the Tristubh hymns than in those in It is perhaps not impossible that a single poet Brhati-Satobrhati. The Mamateya group i 141-158 does not employ the secondary caesura. If the indications of metre are followed out in further detail as a matter of speculative interest. should have introduced these variations into one metre without employing them in the other but it seems more probable that a difference in time should account for the divergence. for the Brhati-Satobrhati hymns of Mandala vii. seems therefore probable that these groups are not entirely homogeneous. but it has very commonly the Pentad opening in both respects it agrees closely with the Vamadeva hymns. and the Brhati-Satobrhati hymns. but agrees closely with the results Cf. are not so early in date as the other lyric metres.220 This genealogy Small trimeter groups is : of course suggested with every reserve but it is quite consistent with the general view of the history of dimeter verse obtained in 201. lyric metres see (a) the group i The group viii 35-37 contains so few trimeter verses that the (b) So far as they go they revive metrical indications cannot be trusted. we find that the decasyllabic Tristubh hymns have a much larger proportion of archaic variations. Upon this supposition we must assign the earlier date to the lyric hymns. which is distinctly lower than that of the Brhati-Satobrhati hymns in Group I. This result is the more remarkable. 201 ii. For the results of the similar investigation of the dimeter verses in the 187. and the : agreement in other matters so close. which we have now brought into the strophic : : period. If we take the three groups together the percentage of archaic variations is 17.

we arrive at the following results for each 100 verses (a) : : .Small trimeter groups (iii) 221 extremely rare In the normal groups as a whole secondary caesura is its frequency in the group i 116-119 and in the third Mandala is the more remarkable. If we compare the four separate forms of secondary caesura in the Vasistha hymns and in these groups.

222 Trimeter variations 235. Trimeter variations in the smaller groups. .

).Trimeter variations 223 Trimeter variations in the smaller groups (contd. .

the iambic and cretic variations no longer appear as contrasted. which characterize its later are in a position to distinguish two parts. resulting in a rhythm which we may call the less rigid trimeter rhythm. But the different variations which together contribute to this general rhythm do not disappear simultaneously. as opposed to the to cretic variation chapter that assumed us to a somewhat different view of and contamination. History of trimeter verse The leads detailed examination of trimeter verse its in this history as a starting-point in 94 above. we find a developement of the iambic rhythm. Just when the rests and most other variations are becoming rare. as in The iambic 116-119 and in the third Mandala. To a i the hymns certain extent these variations even overlap. and which exactly corresponds to the earliest form of dimeter rhythm ( 201). These facts supply us with an explanation of the developement of the cretic break it is the shifting of the rhythm which : follows a late caesura in the Vasisthl verse to a similar position following an early caesura. but as characterizing successive stages in the developement of the metre. 207 late iii) that they are uncommon some comparatively hymns. Variations of almost every Now. kinds of variation in the earlier parts. This we seem bound to recognize as the earliest form of trimeter verse in the Rigveda. The considerations discussed in this chapter have led us to assign to the same period two of the largest family collections. : of view now appears to be of subordinate importance. especially in connexion with the secondary caesura: this rhythm affecting Thus chiefly the break. all It would the family collections on the seem more larger scale as practically contemporaneous.224 236. however. we kind appear side by side in those groups which we can assign most confidently to the archaic period. and to speak of a correct to think of . but to some extent the opening also. variations cease therefore to be regarded as in themselves evidence of the earliest date : and indeed not it has been noticed above in ( 94 v. those of Vamadeva and Vasistha whilst several others are assigned : Earlier enquiries into the history of the have generally taken the form of an attempt to arrange Rigveda the separate family collections in order of time but this point to neighbouring periods. At that time decasyllabic and iambic variations appeared to be characteristic of the earlier parts of the Rigveda.

find no explanation of the comparative frequency of short final vowels in the second tion of families Of all We except by assuming it to be a survival of a metrical freedom originally used in syllables of all kinds. and that the uneven lyric hymns shew Jagati verse in process of developement Tristubh from it through the influence of dimeter verse. whilst in the hymns only rarely exhibit Jagati cadence. owing to the general preference opening ( 35). and Jagati hymns can hardly be found. that pre-Vedic verse used iambic rhythm at the break suppose In that than : : and the rule laid down in 45. the need for variety in the opening caesura is in that position. not specially characteristic 15 . for is more felt when the With regard earliest period to the cadence we notice that. 237. The caesura appears indifferently after the fourth or fifth syllable in every part of the Rigveda yet there is some slight reason for thinking the early caesura to be the normal position : . the lyric hymns admit Tristubh cadence in their trimeter verses with some freedom. is Although the secondary caesura a. seems most probable for the pre-Vedic period also. In this case the rule prohibiting the use of consecutive short syllables in the place second and third places can hardly have existed in the pre-Vedic period though the occurrence of such a combination would not : syllables in the have been very common. can. parts of the verse. on the increase. for long history of trimeter rhythm in the Rigveda we to the consideration of its probable rhythm in the pass naturally pre-Vedic period. the opening lends itself least to historical treatment. of change from the earliest Vedic period earliest period iambic variations are no more common other irregularities they are however. It would seem to be a probable interpretation of these facts that Tristubh was originally the dominant metre. that the caesura must be followed by two short syllables. however. under certain many There is therefore no reason to conditions. and later periods in which the distinc- had disappeared.History of trimeter verse 225 'period of the family collections' as contrasted with an earlier period of small collections. As to this we can only form conjectures by From the observing the trend onwards.

inasmuch as almost every possible rhythm is occasionally found. in successive waves not always producing identical metrical results.226 The primitive trimeter its verse of the earliest period. : 238. yet recognition by possibility seems to point to the the earliest poets of a division of the verse at that point. encroached upon in the opening by long syllables. as there is no trace of syllaba anceps either before or after the caesura. In the history of the Rigveda the two forms gradually diverge. This process never affected the extremities of the verse. far less This conclusion corresponds as far as regards the opening with (i) that reached in $ 201 for dimeter verse. The chief problems that remain are three: (i) what is the origin of the rhythm ^ ^ {ionicus a minor e) in the second member of trimeter verse ? (ii) what is the origin of decasyllabic metre ? and lastly (iii) why is the iambic rhythm. dimeter verse preferring the long third syllable. . both in dimeter and trimeter verse. Analogies drawn from Greek metre might suggest to us that (iii) the two parts of trimeter verse were derived from two separate verses but we must reject this view. corresponding metres into 'feet. The presumed pre-Vedic forms of dimeter and trimeter verse (ii) approximate to some extent to the non-quantitative verse of the Avesta. But in this direction much is still left unexplained.' more or less to the analysis of the modern From these considerations we seem to arrive at the following scheme as representing the dominant form of pre-Vedic trimeter verse. the trimeter verse inclining towards the short third syllable. The natural pause at the caesura for taking breath made it convenient that short from this beginning a preference for syllables should follow prior to : long or short syllables spread in both directions on the principle of alternation. in the cadence by short syllables ? To given (i) : these questions the following conjectural answers may be In the developement of trimeter verse the caesura was any differentiation of quantity. and as the possible starting-point of the developement of the Vedic forms generally : all the symbols being understood to mark preferences marked than those of the Rigveda generally.

however. If this explanation is correct the GautamI verse ($226 iii) has considerable interest. There can be no : influence of the trimeter doubt that a strong desire to differentiate the opening and the cadence in dimeter verse contributed to maintain this distinction. 1886). to must here to refer Dr R. The neutral rests may well be later in date than the Viratsthana ' ' verse. : (iii) in the cadence in For an explanation of the preference for short syllables 35) it is most natural to look to dimeter verse. a direct and independent result of the natural pause at the that is. that the rest is the starting-point of the developement. easy to explain rhythm at the break. Whatever explanations may be adopted as to particular facts. which we can most certainly trace to the archaic period. trimeter verse after the Vedic time. 152 . nor from contamination of two forms in which the caesura is in different positions: nor is it an indirect result of the process of syllabic contraction. Kiihnau's 'Die Tristubh-Jagatl Familie' (Gottingen. and that this causes directly a preference for a long is caesura ' ' : sixth syllable. Against any of these suppositions stands the fact that the Viratsthana verse. because in this type of decasyllabic verse the rest is actually followed by two long syllables. ' and that the long sixth syllable (here as always) causes a secondary preference for a short seventh syllable. whatever it its origin may have been. and may represent an assimilation of the decasyllabic verse to the normal trimeter type assimilation in the opposite direction being represented by the iambic rhythm of the Vasistha hymns.Origin of the 'rest' (ii) 227 of one long syllable to Decasyllabic Tristubh does not arise from the equivalence two short (of which there is no trace in the Rigveda). The decasyllabic metres we take to be a developement from ' decasyllabic Tristubh verses. has a rhythm which is distinct from that which would be reached by any of these It seems therefore to follow that this form of verse processes. is clear that both dimeter and trimeter verse had a long history For the history of suffice it of developement before the Vedic period. ( most marked. syllable is which the contrast between the opening and the cadence is The comparative favour shewn to a short sixth it may be due to the not. quite contrary to the ordinary rhythm of the break.

and the combination of verses into stanzas and of stanzas into strophes. the elements of the ' external structure ' ( 30) This inquiry is of a much require to be carefully considered. With : trimeter verse the discussion of Vedic metre the conclusion of the analysis of dimeter and is in the main the combination of verses into stanzas and strophes being usually so simple that it has been sufficiently explained in the General Introduction. 239. After the consideration of the separate be possible to discuss the few hymns which do not groups fall into the classification. (iii) Brhati-Satobrhati strophes. the verse. or internal structure is on the whole of comparatively rhythm small importance. THE LESS USUAL METRES STANZAS AND STROPHES. which are constructed upon principles which differ In these hymns the from those already discussed in detail. chiefly belonging to the earlier periods of Vedic poetry. Such groups are (i) the 'uneven lyric hymns. (iv) the Atyasti hymns and (v) the the lyric and dimeter triplets. ' (ii) amongst which Usnih and Kakubh-SatobrhatI and prominent. will be found that most of the hymns to be examined belong to groups that are fairly well defined. . It easier kind than those in which we have so far been engaged. There remain however for consideration complete a number of hymns. found in ordinary trimeter verse. whilst the number of syllables in. that is to say. Gayatii are most the decasyllabic it will metres.CHAPTEE IX. and to consider the relationship of the groups to one another and to the bulk of the hymns in the Rigveda and also such traces of strophic arrangement as can be . and which include one or two standard types and a number of variations which only slightly depart from them.

233). Hence in these hymns questions of emendation assume an importance unusual in the Rigveda. dimeter verses are occasionally requite regular placed by trimeter. since in the Rigveda proper we find that strictness of metrical regulation increases in the later periods. (iii) By contamination verse of the cadences of the two kinds of often. yet it will appear that there is little justification for the complete despair with which the metre of some of these hymns has been regarded by Western critics. The antiquity of these hymns may cause irregularities in the Some are due to the lack of strict metrical text in various ways. The following principles are suggested as a clue to the difficulties of (i) the ' uneven ' lyric hymns : The number of verses in the stanza is not fixed : the con- cluding stanza in particular is often made of greater length. hymns as they stand in the Samhita text is a further indication of their antiquity. 241.The 'uneven ' ' lyric'' hi/mm 229 The uneven lyric metres have been defined in 27 as 240. which characterized the composers: others to the perprinciple plexities of the reciters and copyists in preserving the tradition poems of which both the dialect and the metre were becoming strange to them. have seen reason to believe that the dimeter and trimeter verses which they contain represent the earliest Vedic types of these two kinds of verse ( The irregularity shewn in these 196. and it becomes of impossible to stanzas. it will be justifiable to look to the composition of the hymn as a whole as a guide and though it : may not be practicable to restore with any certainty the original form. and we syllables. . reduced to seven syllables: whilst the trimeter verse is with varying regularity extended to twelve syllables. as in the Atri Anustubh. by repeating one or more of the verses in the scheme. feel Where however difficulties confident as to the original form of many of the of metre and of interpretation occur in the same stanza. verse the dimeter is sometimes almost regularly. (ii) The succession of dimeter and trimeter verses : is seldom in particular. those which combine in one stanza verses of eight and eleven A list of such hymns has been given in 186. and thus approximates to that used in lyric verse.

12.8. 11. 231). come nearest to the regular to interpret are of course those which These hymns are v 24. however.8. and 12a 12b may be read as catalectic: of the dimeter verses 2d is certainly catalectic. conveniently included here. it The hymn x 93 approximates to the type 12. and viii 35.8. The regular type of the stanza is 12. x 93 and x 132.3 than 8. occurring as la 2a lOd He. which we have seen reason to regard as not quite of the earliest date (*$ 201. But catalectic verses are somewhat common. i (ii) i 150 is of its three trimeter verses only one (2c) has the probable standard is therefore 8. ' ' . after a careful counting of the whole number of syllables. the catalectic verses being 5f 8a 8e. 12.230 Uneven lyric hymns (iv) Amongst the trimeter verses decasyllabic variations are common. and the stanzas are grouped in triplets. The easier hymns 242. (iii) treated by the Anukranianl as an Usnili hymn. The hymn viii 35. The discussion will be the more readily followed if we begin with the hymns in which the metre causes the least difficulty. but they occur less frequently. but more than 11 syllables: v 24 is 8. are easily explained by the principles stated in The hymn i 127 can hardly be separated from the whole group (i) 127-139.8. The hymn therefore appears to be associated with the uneven lyric metres in the irregularity of the cadence. and is (v) accordingly described by Indian tradition. lyric metres. the last three verses of the stanza being the same throughout each triplet. Such further difficulties as appear in the separate hymns the last section. Of the trimeter verses. la lb 4a 4b 14a 14b are certainly catalectic (14b being also 'hybrid').12. to which should perhaps be added i 127. as of the Prastarapahkti metre. lyric hymns marked by somewhat frequent : catalexis both of the dimeter and the trimeter verses or all but regarded historically some of them mark a transition stage between the uneven lyric ' ' and the lyric metres. 129 and Looked at from the later standpoint. namely 10b = llb=12b and 16b=17b=18b. like i 127. these are 150. namely 6a 13a 14a 15a: and of the 21 refrain verses 2. Of the 21 initial verses 4 have catalectic cadence.11. and 5c 9b 9d 13d 14d may be so read. belongs to a group from which (iv) cannot easily be detached. and there are also several irregularities of rhythm. but its peculiar structure deserves notice.8. though the type is rather Only in lc (varuthiah) do we find four syllables in the short verse. Of the other hymns in this group only i 129 has the same characteristics. and reserve to the last those in which either the metre is very complicated or the text less satisfactory than usual.

and 9 are described as Anustubh.8.8. . Of the opening verses of these three stanzas 5a is Tristubh. metre and meaning would be adequately represented by some such verse as Ijdndm id y vdsund dyauh prnaktu. perhaps to ydd va In stanza 1 only the first verse causes any difficulty pusyati dadvan. The remaining stanzas 2. 7.11) all four verses are catalectic. and stanza 2 can be read in the same way.8. and it is quite likely that each of the original stanzas contained at least one trimeter verse.8: is exactly of this type. termed Purastddbrhatl in the Anukramani.8. but in any case some emendation is required in each. stanza 6 : : 243.8.8: and : stanza is Tristubh. and probably needs correction.11.8. 142 v): 9a is Tristubh with a double rest ( 226): if we read vdatasya ( 7a alone is dimeter. but the last the type were 12.7.8. Of the initial verses only 3a and 11a have Jagati cadence. of an ' uneven of In x 22 we have an isolated but almost perfect example hymn. In this case they may have run somewhat : as follows : yajile-yajfle sd dev&n u saparyati yah sumndm dlrghasruttamd dvivasati agneh. and there can be little doubt that stanzas 3 and 5 are substantially of the same type in this case 3c requires correction.8. the type being 11. as though stanzas 5. The whole number of stanzas is the same as in x 93. yesai'n rdbhir ^ hiranydyl vavdrta yuktd isd hiranydyl uemddhitd vam fawthsici vfthd viprdh stavanta.8.12.8. The lyric ' position this hymn ( in the Vimada collection enables us definitely to assign it to the same period as the earliest forms of Anustubh verse The metre is 201). 3 and 13 are described in the Anukramani as Anustubh this may be correct. of the type 12. 2 visvesaam irajyati vasundm devdnaam mahdh sd \\ visve hi visvdmahasah 3 visve yajnesu yajniyah. The final stanza 7 has an extra dimeter verse.8: these stanzas shew the irregularity of external structure which characterizes the uneven lyric metres.Metres of x 22 and 93 281 With this limitation ten of the fifteen stanzas correspond to the Stanza 9 is of the type 12. and 7c is In stanza 4 (11.8. (vi) 13 The hymn x 1 32 approximates to the same standard 12. catalectic.8. and stanza 15 type described. but many have rests or other archaic variations of the dimeter verses 3c 5b and 6c are probably to be read as catalectic.

have two irregular forms of Tristubh verse. and x 105. and stanzas 3 and 4 diverge by having trimeter verses in the third place. . The starting-point for the interpretation of both to be found in stanza 8. . etdt tidn namely nd n sasvdr ha ydn yojanam aceti maruto gotamo vah | ( pdsyan hiranyacakran dyodamstrdn y vidhdvato vardhun. but the metrical hymn may have been somewhat as follows : the original kd frddhdti n hotard asvind vdm kb vdm josam ^ ubhdyor a vidhdti kathd jddsdd dpracetdh. and for x 105 11 7 11. .7. If this is so. violently disturbed in some of the other stanzas. .7. we can feel confident that the general type in the mind of the composer was for i 88 In 11 11 7 11 for i 120 11. (i) i 88 stanza 5 may be taken as giving the normal type. i 120 1-9. In addition we contains nine syllables must in any case be emended. On the analogy of i 88 we may infer that the third verse will not always be of seven syllables. . (ii) hymn is exactly what is It follows that the text has been to the Asvins. vidvdmsd it \kHbha jprcchdti jsdntd dvidvdn itthd u dpara acetdh nu cin nu mdrta fasyat. and Id having a double rest ( 226). . but may appear as a dimeter or a Tristubh and also indicates that the matter of the usual in hymns verse. But although it may not be possible to determine the original form of each stanza in these hymns. It In stanza 4 we find an additional verse of 7 syllables. 3 prcchdmi vdsatkrtasya pdkid jcid \vratdni fadhvardsya dasrd 4 td pdtam sdhyaso nah pdtdm ca rdbhyaso nah. lb being hypersyllabic ( 224). many is In i 120 1-9 both metre and meaning are very obscure in of the stanzas. would be beyond the scope of this book to discuss possible effect of emendations of the text as such.11.11. 2c which whilst stanza 6 has a dimeter verse in the fourth place. \\ 1 2 f dvidvdmsah h havdmahe mdnma td no vidvdmsd \\ prd vi j riricydd n ^ \i \vaydm vdm \d jjusethdm ddyamdno yuvdkuh. these hymns In : therefore the Tristubh or trochaic cadence is generally found in both the trimeter and the dimeter verses. stanza 1 diverges by having a full dimeter verse. and needs correction.232 Metres of i 88 and 120 The more difficult hymns of this class are i 88. 244. . which gives us a metrical type 11.

ydyd vdcd h ydjati pajriyo va/nt prd fdevayur nd fsddhat.7. the normal period the construction of a hymn of this type would have been a very severe task. The ingenious theory that the peculiarities of this hymn are due to 1 temper of the bard. 8b is a full abbreviated form for ddsyave dimeter verse.7. in which each of the two last In 7a we have verses is doubled.11. 5 fyad n prd surah sdvasdstaut rbhur nd u krdtubhir mdtarisvd. 2 and 10. carkrse 3 sdcd ay oh fydd fenam 4 \u\ianaksat saparydn naddyor vi. and Prof.sa 6 1 ff. \nii jcic chrutam tdkavdnasya f gdthdm ahd/rh cid dhi ^ rirebhd asvind vani dksi s'ubhas pail ddn. 8 duhlydn mitrd -dhitaye yuvaku rdye ca no mimitdm vdjavatyai ise ca dhenumdtyai. Pischel.Metre of x 1 05 233 prd yd jghosi n \bhfgavdnena j.subhro. The most difficult stanzas are those from 3 to 6. \\ \\ 9 Rigveda the ill For previous discussions of this hymn refer to Prof. A. R.11. Ludwig's iv 39-42. and 9a appears to be a dimeter verse. Pischel in Vedische Studien ii 87-109. md md kdsmai dhdtam ^ amitrine nah fkutra no n grhebhyo dhendvo guh standbhuja dsisvlh. 1 rsvebhir rsvd-ojds tatdk. In x 105 the normal type of the stanza is 11.11. \\ 5 6 yuvdm fydd \gd td no vasu ( | \\ dstam d Jtamsayethe 7 sugopdd sidtam pdtdm vfkdd aghdyoh. Vedische Studien. although it is tretini ydd easy to restore a trimeter verse in the shape urdhud bhiiut te. kesavantd n yah fsubhd adhi tasthaii vydcasvantd nd pustyai biprdbhidm n vanoti siprinivdn. which may : perhaps be restored somewhat as follows \\ : indro jmdrtio nd dpa \ayoli sasramdno bibhlvdn subhe ydd M h yuyuje tdvislvdn. and is still more clearly proved by the form of the final stanza. who purposely made it unmetrical. i .M vratayoh sura indrah. this is (iii) found in the text in stanzas 1. unless we substitute ddsyu as an 8a has a double rest.7. R. is not To a bard familiar with the smoother metres of (say) very probable. a trimeter verse of 12 syllables. giving the type 11.

or Usnih or of mixed lyric (i) Anustubh (iii) Gayatri or Anustubh triplets (iv) triplets in which Anustubh and Gayatri are combined in some systematic way. . or of lyric metres combined with Gayatri. Almost the whole of the lyric verse in the Rigveda. whether of the type 8 8 8 4 or the (i) the list of type 8. Triplets of the Atyasti type are found in i 130. 132. 106 13. No. For : : . The : principal other homogeneous lyric triplets (ii) triplets metres. viii 18 22. In trimeter verse triplets are not common. and ix 111.8. 12. The only stanzas not so grouped are found in vi 43 (4 Usnih stanzas). ix 102 (8 stanzas).12 is found in ix 110 1-3. a large proportion of the dimeter hymns. 18 in iii 21 4 as 11 11 11 8 in v 19 5 (unless the text is corrected) of the type 8. Owing to the disconnected character of the Vedic poems.11. Tristubh. of hymns in which the number of stanzas is some multiple and by the fact that changes of metre so commonly take place after each third stanza. Hymns in Usnih metre. discussion of this point (see It is probable that the system of grouping is at times carried further.8. Nos. kinds of triplets in the Rigveda are the following . . A group of three stanzas may be called a triplet. . 134-137.11. are almost invariably grouped in triplets hymns is given in 186.11. For Brhati triplets see 246 iv. 251 i).8: in viii 46 14 of the type 13.8.14. . the hymn consisting of any number of such units. See also triplet of the type 12. stanzas to make a metrical unit. . : similar stanzas in the popular Rigveda see Appendix III ( 253). and in particular that groups of three or five triplets are deliberately combined to form a larger whole in the former case an extra stanza is often added to the last triplet. 14 (2 stanzas): see also subsection ii.234 (iv) Lyric triplets follows Isolated stanzas of the uneven lyric type appear to occur as in viii 46 20 of the type 11. it might not always be easy to trace this grouping by the meaning: but it is easily recognized by the great number of 3. see Appendix III. shew a system of groupThe most common grouping is that of three ing of stanzas. : A It very often appears from the general arrangement of a hymn (ii) which contains a large proportion of lyric stanzas that the stanzas are grouped in triplets. .11. and 245. and we defer the . nor grouped according to any recognizable system. although the stanzas which form the triplet are not homogeneous. . v 87 247 ii. so that the poem as a whole consists of ten stanzas in the one case and fifteen in : the second. 35 and 39.8.

42. in the triplet is Anustubh 82 1-3. . X. and Kanva The beginnings triplets.8 Purausnih) making a triplet as follows .). and in particular of the Sobhaii. of the system which some third lyric may stanza is combined with Kakubh or BrhatI and SatobrhatI and there is an intermediate stage in which the combination is in pairs collections.-B. 7-9. . 16. in be traced back to the : The of stanzas.-Pur. vi 15 16-18 28-30. 32. .-Sat. 102 1-18 ix 6-13 x 118. 81-85. 74 1-12. 35 22-24.8. 26 20-22. 77 1-9.-Sat.Lyric triplets 235 Mixed lyric triplets are found in viii 9 4-6. 80 1-9. iii 37 1-10. 79. 38. *iii 28 vii 14 viii 5 37-39. Groups of ten stanzas are found in i 4-9. 89 5-7. 19. 73. ix 101 1-^. And also frequently where a Kakubh-SatobrhatI or Brhatl-Satobrhatl strophe ia included for the latter class see 246 i. viii 11 1-9. or the number of Anustubh verses a multiple of 12: see the 'Table of Hymns' appended to Ch.-Sat96 1-3 (B.-Prastarapankti) (B. . Vasistha. ii 41 16-18. (v) Groups of nine stanzas are found in i 1. Similar groups each consisting of five stanzas are perhaps to be found in viii 31 5-9. : l\lore complicated groupings are chiefly found in Gayatri verse. Of the Gayatri and Anustubh verse contained in the Rigveda (iii) a very large proportion is grouped in triplets the system extends to the epic Anustubh verse of the popular Rigveda. 92 1-3. 9 13-15. 40-42. iv 15 Groups of iv 31 vi 16 145 fifteen stanzas are found in i 37. . : A The grouping of Anustubh and Gayatri together in triplets is (iv) a very simple and effective arrangement.). viii 17 1-10. 24 1-3. 38 ix 1-4. 43. 12. The combination of a Kakubh or BrhatI stanza with a SatobrhatI stanza so as to form a strophe is a well-marked feature of certain parts of the Rigveda. 1216. 86. . 34 1-15. but many slight variations are permissible. . 50 1-9. 72.) In vi 48 variations from the strict Pur.-Sat. vi 45 31-33. 63 1-3. 10-14. Triplets in which lyric metres are combined with dimeter verse or with Tristubh are found as follows: i 91 16-18.) 16-18 (Kak. : vii . viii 3 21-23. that they are themselves frequently combined in triplets. Two Anustubh stanzas are followed by a Gayatri stanza in viii 63 4-6. 23-25. 69 7-11.-Sat. 46 viii 6 1-45. 10-12. 9 19-21.) 19-21 viii 4 19-21 (B. v 61 9-11 . 26. x 62 8-10. vi 48 13-15 (Pur. Two Gayatri stanzas are followed by an Anustubh stanza in i 43 7-9. Usually the first stanza alone of this we have examples in v 28 4-6. grouping of this kind is usually to be inferred where the number of Gayatri verses in a hymn is a multiple of 9. 44.-Sat. 98 7-9.-B. iii 11. : .-B. 70 13-15. 68 1-12. strophes are so far treated as single stanzas. ix 67 25-27. . x 62 5-7 (An. 17. 41. 14. . Kakubh-Satobrhati and Brhati-Satobrhati strophes are found (i) combined with some third stanza (often 12 . 246.

types 12. .12. 97 1-9 x 150 1-3. . 48 viii 3 1-20.12. and 6 is Uparistajj yotis (12. 7 is of the type 8. r . 15-16 and ix 108 13-16. and there is an appended Brhati stanza after stanza 12. 46 7 53 18. 69 7 18. 81. 12. composed of these strophes. 63. Strophes not arranged in triplets are only found in v 53 5-6. of the respective In viii 101 stanza 3 is Gayatri. with the following irregularities in viii 19 we find an extended strophe (12. . For the position of this metre in the later parts of the Rigveda see 94 i. In ix 107 we have extended Satobrhati stanzas of the same type as above as 2-3 and 142 ii. 32. In x 140 the first two stanzas are Vistdrajmnkti (8. 90. 59 1-6. 151 i.12. stanza 8 is Anustubh.12. Similar strophes not arranged in triplets are regular in the Kanva hymns. and stanza 12 of the type 12. 47.8.8. 87. . Here viii 46 12 is of 25-26. Kakubh-Satobrhati strophes are found in triplets in viii 19-21.8. . and 11 is Kakubh. 42 4 47 19. (ii) 22 7-18. 46 n-12 1 x 33 2-3. . 78 10. 228 i. of which two only are Satobrhati and the rest Brhati. 39. the type being 8. viii 27. irregularities. 10. Another such triplet is probably to be found in viii 103 8-13. 89 1-4. The hymns named contain the following irregularities. 15-16 for stanza 9 see In v 56 we have a hymn of nine stanzas.8. 144. 45 Detached stanzas are found as i 139 5. though by rearrangement of the words Satobrhati may be restored and in viii 22 stanza 7 is Brhati.8. v 53 2 vi 15 18. 99 1 *x 14 15. . 71 10-15.8. *102 1.8. . and 150. 1 viii 30 3. The Kanva hymns contain no Detached strophes occur i 84 19-20. *101 5. and 37 as it stands is Pankti.12. vi 48 1-12. ix 107. 17 14-15. although stanza 10 is in the text only Gayatri and in ix 108 stanzas 1-6 and 7-12 are probably triplets. 77 10-ir.8.8). 22 1-6. 101 1-12. 10 1-6 stanzas 2 and 4 are variations on Satobrhati. 60. (iv) . 19. and 49-54 are so Elsewhere we find the hymns vi 46. 99 composed. namely x 140. .8.12.8.12. Brhati stanzas occur in triplets in vi 59 1-6. 1 These are Dimeter Brhati stanzas.12.12.8. 60. Brhatl-Satobrhatl strophes : type occur for which see the Appendix to this chapter as follows 15 21 No. 1 7-9 . 88. 103 4-7 the type 8. 74.8.236 No. In vi 48 stanzas 6 and 8 are of the type 12.12. 3. 40. Brhati-Satobrhati strophes are found in triplets as follows (iii) 16.8.4. Isolated Kakubh verses do not occur.8.8.12.12. stanzas 3-5 In x 144 the first Satobrhati. ix 98 n. : iii . . 70 1-6. ' ' Satobrhati and the allied metres are very rarely found outside (v) the strophic system but there are three hymns near the end of the tenth Mandala which are so composed.8) as 26-27. .8. viii 1 1-4. 66 10-15. and viii 103 5 of the type 11. 66 1-14.12.12. . 2728. Other Brhati hymns are iii 44.8.8.12. viii 33 1-15. and i 36.12. 61.8). 1 See 248 iii.12. of the type 12.12. . . 44. 60 14 u.8. *170 1 iii viii 1 5-32.12. . 62 70 7-12. viii 4 1-18. In vii 32 we have an extended In viii Satobrhati stanza as 2-3. vii"l6.12 and 8.

. 77.The Atyasti group two 237 of these metres are combined with Bihati and Gayatri. .8J8. 139.8 8. The Atyasti group of the archaic period.8 ( 246 ii) or through the metre of v 87 (12. v 87 consists of triplets of stanzas of the type 12. in standard Atyasti triplets are i 130. iv.8). group : their | .8. 12 12. The Atyasti metre is the most common form of a group of lyric metres.4 12. and that (with rare exceptions) the stanza begins with a trimeter and ends with a dimeter verse.8. (ii) Hymns composed and j 134.See Appendix III ( 253) Nos. and 88. of which the type is 12. 81. All these hymns must necessarily be referred to a very early Vedic period. and (iii) 83. 247. to be similar in In stanzas 2 and 3 the type is 12. 132. 129 133 6. which are characterized by the large number of The whole group may be regarded verses included in one stanza. . ( 253) Nos. 133 6-7. 12.8: and i 135 of triplets which are chiefly Atyasti. Other hymns in Atyasti are i 127-129.8| 12 8 8 is as common as all the rest put together. but partly of an allied 12 12 8). . ix 111 : . 8.12. type (12 12 8 Hymns i 130 and 136 have each also an appended Tristubh stanza. lyric stanza : of metres is amongst the latest productions and presupposes a long developement of the but the view entertained by many Western critics that hymns is A Rigveda metre is given by A.8. . The triplet iv 1 1-3 consists of mixed metres of this group. | 75. 131. of which two at least are trimeter and two at least dimeter. as a developement of the Satobrhati metre in its extended form 12. See Appendix III ( 253) Nos. but the last two dimeter verses (which belong to a refrain) are unrhythmical. There are about 20 varieties of stanza belonging to the Atyasti common features are that they have at least five verses to the stanza. in which the following stanzas differ from the standard: i 127 6. and 87. and is followed by a Tristubh hymn.12. 8: i 137 of a triplet of the type 8. Rigveda (i) in these metres belong to the later additions to the correct appreciation of this entirely untenable. Ludwig. See Appendix III 8. .8. 9. The standard Atyasti metre.4 12. The concluding stanza is in the text of the type 24 1'2 but it is probable that the first 4 12. 74. principle to Atyasti. xxxi-xxxii.12. 80. 138. whilst i 139 s is a Bihati stanza. but on account of its length it must be regarded as almost the latest in developement. 8. 8 4 section admits of some analysis. .12. The first stanza only differs by having an additional ii In 22 we have a metre which seems | | verse of four syllables after e. : | | | 84. 136. In x 150 a BrhatI triplet is followed by two stanzas in Uparistdjjyotis. . 79.8J12.8. 85. 62.

Such hymns are v 53. 7 Dim. 10. 29-33 23 Gayatri.12.8. as requiring considerable restoration.12.12 8. BrhatI. 70 and 71. . 3 BrhatI) is followed by an appended Anustubh stanza. which the lyric metres are used.8) does not indeed come within our definition in subsection i.8. but has not been faithfully preserved. but 17 The type is 12 12 12 8 8. In this hymn stanza 4 may be interpreted as Kakubh (i) by reading vdHau in 4a ( 170 iii).8. and viii 46. Gayatri triplets 5 is in the text Dimeter Kakubh.12. Here a triplet in mixed metres (1 Gayatri. It therefore only remains here to consider the building up of such verses into stanzas and strophes in the hymns that have already been distinguished as composed in decasyllabic metres. arranged in triplets. by stanza 8. Dimeter stanzas. 24 Pankti. 20 is either disordered or is in uneven lyric metre. (iii) contains many groups. viii 30. but probably suvlricwn should be omitted. 15 as 13 omit the concluding words nundm dtha. 14 is a Satobrhati stanza: the first verse is hypersyllabic. 2 viii 30. 25-28 Brhatl-SatobrhatI strophes. and to measure the extent to which these hymns are differentiated from the less regular trimeter hymns from which they have originated. j : . viii 35 (12. . 14). Satobrhati stanzas (7. j stanzas. except that 30a is trimeter: in 31c said should It seems probable that stanzas 1 to 12 are perhaps be omitted. 13. 24 shew no clear indication of grouping. 11) becomes equal to that of the detached As there are also two strophes (5-6. This is the most varied hymn in the Rigvecla. 9. In addition the Astarapankti hymns of Vimada (x 21. and has points of contact with the uneven lyric metres but viii 36 i-6 is of the type 12. The internal structure or rhythm of decasyllabic verses has been examined in the last chapter ( 228-230). 25) v 53.8. and the concluding stanza For the similar metres differs in having one dimeter verse less. 11-12 BrhatI followed 13 Dvipada Jagatl. 9 Satobrhati. 22 Stanza of type 12 8 8 8 8 representing Satobrhati. 21 Parikti. and 13 to 28 in strophes each containing two 16 : : . in v 87 the third verse requires restoration. 18 Uparistadbrhati. and only 4 stanzas of quite different type. only found in viii 37 see Appendix III ( 253) Nos. and thus the number of the detached Kakubh stanzas (1.12 in place of Satobrhati. : | The grouping of stanzas is so regular in those hymns in 248. and the fourth is represented only by the initial word ydthti. (ii) Purausnih. . 4. but viii 46. 8 Anustubh. in the hymn.238 (iv) Disordered grouping In viii 35-37 the metres appear to be allied. that some disorder in the text may be suspected where an arrangement in triplets or strophes does not come to light. 249. . 19-20 BrhatI. . 10 Gayatri. 1-3. 16 probably the same as 14. it seems 15-16) probable that the original hymn was composed entirely in KakubhSatobrhatl.

as for instance dgne tdm adyd dsvarii nd stomaih krdtum nd bhadrdm hrdispfsam rdhidma ta ohaih. is in these hymns long seventh of the instances (220 iii) these two changes appear to be : : : due to mutual assimilation. always short in trimeter verse. so that the total number is In ix 109 there is also an additional stanza in another 10. and becomes two verses and these two exercise a mutual influence which tends towards their This point seems to be reached in the complete assimilation. almost in'rest' and of assimilation coincide. different in quantity in trimeter verse. 56 i-it. All the Pentad hymns are arranged in strophes consisting of five double stanzas.8) which hardly seems to belong to the hymn. It is remarkable that the the fifth syllable is long in three-quarters of instances. in which three pentads are followed by a Tristubh Pentad hymns are usually arranged in verse in each stanza. but some add a final stanza. metre (8. the second and third of seven stanzas each. vii 34 1-21. these 44 7-9. the pure Pentad hymns. (i) The mixed Pentad hymns are two-thirds of the vi hymns about : : (ii) and ix 109. The metre found in iv 10 consists of three Pentads followed (iii) by a Tristubh verse. trimeter verses. of ten stanzas. The 'pure Pentad' hymns are i 65-70. Tristubh verses are combined verses are so rare that they these. and of the remainder three-quarters are Pentad verses and oneThe stanzas consist of four quarter belong to other decasyllabic types.The Penta/l metre The most important 1 230 ' ' of these metres is the Pentad ' metre. and only the last of ten stanzas. In verses are ordinal Tristubh verses. in the other hymns in one-half but here the effects of the The third syllable. we observe that : the original trimeter verse is split.6. of which In vii 34 and 56 the seventh syllable only nine are Tristubh verses. These hymns contain together over 200 verses. x 1. : | ( . or 21. which we may call There are a few hymns in but in the greater number Tristubh : Dvipada Viraj or which Pentad and In may practically be disregarded. and 4G. so that (in spite of the tendency to assimilate) this syllable does not become syllaba anceps. metre of iv 10.4. each of which consists of two trimeter strophes verses or four pentads. (the initial syllable of the second Pentad) is long in one-third of the instances. is in these hymns short in almost three-quarters of the instances whilst the corresponding ninth in onesyllable. and the two parts of each verse retain their original character thus the quantity of the third syllable is indifferent. The number number of stanzas in the strophe thus corresponds exactly to the of syllables in the verse. but the The strophe-formation varies the ninth is almost invariably short. 11. first hymn named consists of a triplet.

Although none of these metres become established in the same way as the Pentad metre. 5c (dytim 142 ii). The hymn ii 1 1 : consists of 20 Viratsthana stanzas followed by one Tristubh stanza it therefore corresponds in the number of stanzas with the Pentad hymns vii 34 and ix 109. 15a (yyantu 140 iii). we verse certainly has eleven syllables. they appear to be clearly conceived by the poets in the special hymns in which they appear. It is natural therefore to regard these hymns as composite of a Bhargavi strophe of five stanzas followed by a : triplet. and that of the last pentad (the last two in stanza 5) As to the middle the same as in the second pentad in those hymns. The Bhargavi hymns x 77. : : The remaining hymns in decasyllabic metre are ii 11 250. ugresvid nu. An (i) arrangement in strophes of five stanzas is probable. supposition by a Jagati On this stanza is Tristubh. The strophe-formation is very uncertain perhaps we have a strophe of five stanzas followed by a triplet. 4d (ddslh 151 i). There are 4 regular Tristubh verses only. 78 each consist of eight stanzas. In the Viratsthana verses the seventh syllable is usually short. the last . and 5 which shew various irregularities other than those alread}^ described. seem to have contamination of the Here. namely 3b 7d lOd 15c 17d. if the text is correct. (ii) and in each case only five stanzas out of the eight shew the special metre. 13a (sydma 147). of a similar strophe followed and x 78 however. ugresu nti seem more probable than ukthesvid nu. 12a (dpy abhuma 125). 19d : (sdkhydsya 137). Of the 80 verses in the Viratsthana stanzas 43 must necessarily be read as Viratsthana verses . becomes necessary to transpose stanzas 5 and 6 in x 78. the quantity both in the third and in the fifth place is indifferent. In 9d 10b 12d 13c 17b 19a 20a the : to have five syllables in the opening. Bub in 4c 6a it is possible to omit indra as a gloss.240 In the fifth The Viratsthana metre stanza there are four Pentads besides the Tristubh verse. (Viratsthana) and x 77. Of the remaining verses 12 have the Viratsthana ending. namely Id 2b 8d 12b. pentad it is hardly possible to make a safe inference from 8 occurrences only so far as can be observed. but appear two metres. x 77 consisting Tristubh triplet. and yet 8 more if we abstain from resolutions which are permissible but not necessary so that we have 59 verses or three-quarters of the whole number which are of the Viratsthana type. The rhythm of the first pentad in each stanza is the same as in the Pentad hymns. it of which. 78 (Bhargavi) to which may perhaps be added i 61 as shewing at least the beginnings of a new decasyllabic : stanza (Gautami). The verses in which we abstain from possible resolutions are 2d {ukthaih 151 iii). 8 more if indra is read (as seems certain in this hymn) as a disyllable. and in 6b it seems desirable to read stand for stavdma in 3a 17a ukthesu nu. 5a (guhyam 136 ii). the last of these being evidently a repetition of the third.

iv 27-29. ix 110 412. In viii 9 10-12 a triplet is composed of a Tristubh. attach to the Gautami verses. seven. Between the two types of Bhargavl verse it is impossible to distinguish satisfactorily. as in vii 22. Where the stanza consists of three verses only ( 94 iv). They are parts probable in vi 15 1-15. as a large proportion of the verses may be classed with either. the triplet grouping seems to be more regular. except for their extreme rarity in other parts of the Rigveda as it is. In i 61 the regular Tristubh verses number 30 out of 63. and an appended Tristubh stanza. 68. but the traces of them are few. in iii 17-23. Tristubh and Jagati triplets might be expected in the earlier the Rigveda. and ten (or eleven) stanzas. Jagati hymns often have a concluding Tristubh stanza. of (ii) Strophes of five stanzas are certainly found in the group Indra hymns extending from vi 30 to vi 41. Gautami No special importance would verses 9. There can therefore be no doubt that we have to deal with a variation of Jagati verse. The hymn consists of three strophes each of five stanzas. 16 . but the varieties are so numerous that it is quite possible to explain them as due to chance. Occasionally there appear indications of a grouping in three. 251. 44 13-24. and some of the longer hymns have one or two such stanzas towards the end. It is open to question whether this habit has any connexion with grouping in strophes. the Viratsthana verses are 10 in number. and may perhaps be associated a. or (iii) rather less than half. to the later (iii) Strophes of seven stanzas appear to belong rather with the periods of the Rigveda. a Viraj. and a Jagati (i) of stanza. 44-50. ix 75-84. and are further probable See also subsection v. there appears to be an attempt to create a new type similar to Viratsthana and in stanza 1 3 no less than three verses out of four are of this type. and in the Soma hymns ix 86 and 97.The Bhdrgavl metre : 241 In the Bhargavl strophes there are 14 verses out of 30 which do not conform to either of the Bhargavl types ( 52) of these verses 5 have eleven syllables and 9 have twelve. but all agree in shewing Jagati cadence. according as we place the caesura before or after nd. and other irregular verses 14. though the Anukramani by means of counting the syllables discovers most of them to be Tristubh. 31 10-12. In the great body of Tristubh and Jagati verse in the Rigveda we look in vain for any strophic arrangement. five. as follows : : : asyed u turdsya n prd bruhi purvidni ukthaih n kdrmdni ndvya u \\ yudhe ydd rghaydmdno isndnd ayudhani ni rinati sdtrun.

. 121. 2. 185. 4. 186 : : . their original form :. about 40 have a concluding Tristubh verse. There are about 170 Jagati hymns in the Rigveda: of these (vi) nearly 100 are composed in JagatT alone. x Other probable examples. *81-84. The Apriya hymns in must have consisted of seven stanzas only. and *87 1-21V vi 79. 41. are the hymns 1-7. 52. they may be traced in i 71-73 . and leads to greater irregularities in the two last periods. 49. 34-36. 54 vi 49. In the archaic period irregular combinations of Jagati and Tristubh are proportionately common. (iv) Groups of ten stanzas are probably due to the influence of the Pentad hymns ( 249) iii 5-7. The hymns i 116. though additional hymns verses have been added at a later period to two of these three hymns. 33. 79. 30. and the 17. but in the latter the use of final Tristubh stanzas begins to assert itself. 54. suggest strophes of five stanzas each. 122 J *vii 35 x 91. v 29. ii 33-35 . Hymns of eleven stanzas may be explained in the same way. : : Table shewing the varieties of Jagati verse. 5 . The relation of these varieties is shewn in the annexed Table. ix 68-70. 43. possible examples are i 53. iv 1. Period . by the addition of a concluding stanza x 48. all in Tristubh metre. 51. 117 each contain 25 50. 80. The history of contamination within the stanza is exactly similar see 223. 92. corresponding to the uneven lyric metres in the strophic and normal periods pure Jagati is most common. (v) Hymns of 15 and 20 stanzas are more common in the Rigveda than those which contain an intermediate number. stanzas. and about 35 have some further admixture of Tristubh. and ix 5 1-7 are of this type. . iv 19-23. and thus indirectly Such hymns are i 32.242 conduct of the Possible trimeter strophes sacrifice by the saptd hotdrah.

162 .Decay of the strophe 243 have their counterparts in modern metres. though they were diverted into directions which : appeal less readily to modern sympathies. It is therefore perhaps not too much to claim a more human interest but on the other hand it may be said that for the earlier poetry the metrical powers of the Vedic poets were not lessened in the later periods. whilst the developement of the internal structure appears artificial and even pedantic.

Type 12. *x 157 25 also according to the text in vi 10 7. 6 and 8. Type 8. 5. 69. A variation of Gayatri. The following list which occur in the Rigveda. 1 . abbreviated quotation of x 25 1. The three stanzas are together equivalent to a Mahapaiikti stanza. see No. Stanzas apparently of this type seem always to be capable of resolution into three verses see Nos. : see No. but these verses may more appropriately be considered as parts of the 3.12. .8: [Dvipadd Satobrhati]. 253. : This stanza occurs vi 47-25. v 41 20. instances the verses seem to belong to a preceding Satobrhati stanza In *x 157 1 the corresponding type 11 8 is found. The number of verses in the stanza varies from two to eight. used in This metre is regularly Type 12. In the Samhita text the single verses iv 17 15. and vi 63 11 are treated as stanzas. shews the various forms of the stanza For convenience of reference they are and arranged in the ascending order of the number of verses in the stanza. and also appears in the text in viii 46 30. Type 11. Anukramam are given when they are practically convenient. but an to the preceding stanza.APPENDIX III. vii 17. 42 17. 12 viii : [Dvipada Jagati]. in the A. but omitted when they are misleading. 15. THE VARIOUS FORMS OF THE STANZA. occurType 8. ring in ix 67 16-18. Stanzas of two verses. 43 16. 2.8: Dvipadd Gayatrl. 72. The titles given of the number of syllables in each verse in order.11: Dvipadd Tristubh. : . New titles are given within square brackets. In other viii 29. but in each case the verse really belongs The verse x 20 1 is not a stanza. 17 15. A possible interpretation of the stanzas 46 13. : stanzas preceding 4.

For the use in lyric triplets see in the Kakubh-SatobrhatI strophe 246 ii. in all periods .5.8. 14. : 21. vii 1 1-20. 19. 4. 7-9. This uneven lyric stanza occurs in x 93 2. 46 5 : in each case the short verse is perhaps a and the stanza really Gayatri. ix 109 27. 7.8.4. Type 8. A variation of No. Kakubh in the usual strophe Type 8. 20.8. iii 25.8. see last in a lyric triplet. see 249 iii. and is Type 12. Type 8. . occuris ring in iv 10 22.8. 20. in the text as *i 23 Occurrences in the popular Rigveda are doubtful. : Type 11. 8. but appear 19. 242 Type : 8.3: of the type x 172 see 3. 12. : C. 15.12: Type 8. Kakubh.11: Padapahkti.8: [Dimeter Purausnih]. See 249. Type 11. TJsnih.4. 192.4. This stanza is found in triplets in ix 110 4-6. perhaps the basis of viii 46 14 (13. see 244 iii.8.4: i. 14).11: probably aimed at in i 150.5. Stanzas of four verses. 4: x 172 2. 11. 25. 9.8: Anustubh. 12 . This stanza found in the text in viii 28 gloss. Regularly used in triplets. *vii 66 16.8 ii.8: Gdyatrl. 4 . 30 2.8.5. for which see 243 v an occurrence in the popular Rigveda *iv 57 5 may be due to some error in the text. Type 8.8: Purausnih.8. : 17. for the .Forms of B. 8.8. 18. 244 ii. 245 13. Type 11. 246 i . and For the occurrences see arrangement in strophes 245 i. 16. 12 \ Tripadd Jagatl]. Type 5. and the stanzas are usually grouped in triplets 251 i. Type 12 12. i This uneven lyric metre is imperfectly pre- served in 120 1-9 . 245 i. vn throughout 10.8: [Dimeter Kakubh]. 15.8. Type 8.7. This metre is established in the strophic period. vn throughout. of which the first is 8. Type 5. viii 245 ii : This stanza usually occurs as the other occurrences are vi 48 13. This stanza occurs in ix 110 1-3. of four verses. Type 11 11 11: Virdj [Tripadd Tristubh]. Type 8.8. variation of 24. See Ch.11.8. as a 246 ii. for the strophes see Common 245 iii. Type 8.4: [Dimeter Usnih]. 6. 12. 23.8. where each of the last two verses has nine : syllables. see 242 ii.8. Type 8.5: Dvipadd Virdj [Pentad]. the stanza 246 Stanzas of three verses. 8. It is possible that this is also the metre aimed at in ix 67 30. see Ch. This stanza occurs in i 120 3 in place of No.8. Used in x 105.11. Type 12. : see Occurs viii 22 n. : see . see Hymns not so arranged are i 149.8: Type 8.5.7. occurs v 24 in a hymn iii.

8.11. 10. lyric metre.12. i 144 6: see 32. Nos. 39. Cf. 13 39.11. in connexion with it: it occurs x 140 l. Viratstltana. 10 are rare. For the occurrence in *i 164 42 40. 11 . 8 8. : A stanza of this type Skandhogrlvl. and 21 in the text in 41.246 26. It is perhaps only au complete stanzas accidental variation of Anustubh. is . Viparita. and 46.8: Purastadbrhatl.8: Visamdpada.11. clusively found in strophes (occasionally in triplets) combined either 246. homogeneous hymns and in detached verses. with Kakubh or with Brhati see : : . 14 ( cf.8 occurs : 187 11. accidental cf. 35. Type 8. The type 11. Type 11 11 . for which it has suggested a only found in x 93 . to Satobrhati Type S. 246 iv. See Ch. 4. : See In ii 11 250 i. 246 : iii. 10. i Type 11. is found in the text v 19 b.11. is . 50. Type 11. In *x 170 4 it is probably the accidental result brhati see cf. i This ( is i). Type 11. : 37. 46. This stanza is almost ex44. important uneven 35. 15. 12 8.8. by omitting vayund in i 29. Stanzas of four verses Type 8. 4. 8 is : . . 11.8. 88 . This is found in the text in iii 4 : see 244 .). 13 ( 242 v. see Most regularly in the Brhall-tiato246 iv in occasionally in triplets. Type 8. No. 5 ( 242 vi).12. 5 either it or it is to be read as a Pankti verse 244 iv.11.11.8. metre : 36. and is a variation of 38. . found in i 88 6 ( 244 i) and in x 93 46.8. lyric Type 11. There See 250 iii. Type viii 8 . Type 12 8 12 8 Satobrhati. Nos.8. iii : brhatl strophe.11. 12 12 8 and is used . is This regularly used in x 22 ( 243) and may probably be The occurrence in *x 17 13 is perhaps restored in x 93 3. title. important uneven lyric metre is 242 v). but only occurs in viii i even No. it is found in .7 occurs Type : 120 4.8. 28. 246 v. iv.8. Type 10 10 10 10 . 46 see .11. Type 11.9. similar to and in x 132 3. Type 12. 8. 38 and r. 27. 39.11.8.8. 42.7. 5 244 This an uneven lyric metre. 27. This variation of No.8.8: 246 Brlwtl. 35. but i 61 13 is nearly complete. 43.11. 33. 2. Nos. 12 12 . 122 5. i This occurs in 88 1 ( 244 i). see parallel to No. 30. no complete stanza.11: Tristubh.7. 15. The type 9. In viii 10 4 this stanza represents SatoType 246 iii. 27.12. of contamination of Anustubh and Jagati verse : .11. [Gautami\. 8 : Occurs 175 1 in an Anustubh hymn. but one is quoted in 34. 35 and 46. This appears to be an un46 20 see 244 iv. Nos. . 12 : : This stanza represents Sato- brhati in 31. Type 10.11. viii throughout. This stanza is equivalent Vixtarapankti.8.

46 7. The type 12 12 14 12 appears . 80. iv 10 5. 13. 31 4 . Tristubh stanza is fairly common in the archaic period.7.8 [Dimeter viii Brhati]. Type 12.5. 75. This variation of No. 21. : Type 5. 12 is found in x 105 see 244 iii. 97 n. j popular Rigveda in *x 86. 49 15. and x 93 9 No. 46 ai.12.5. This stanza is closely OOO- x 93 and 132.8. This metre was originally an 55. 25 is (5 irregular). vi 2 11.8. This is a variation of Satobrhati. i>47 of E Type 12. vi 48 7 .7. vi. Type 8. 50. is found as an independent metre in i 29. and 150 4.8. Type 8. 24 v 6. See Chapter viii throughout.8. 62 7-9. and only used in concluding stanzas of Anustubh hymns 84 : 10-12. See 246 v. and invites emendation. : 4 6. and x 140 6.11.5. 11 .8: Mahdbrhatl. extension of No. Type 8. and in the viii 31 15-18. 19-20. 13 Atijayatl. 11 . : 59. 48. Stanzas of five verses. Seeg 94 iv. There is an occurrence in the popular Rigveda 35 and 39. 22 12 see 246 ii iii. (cf. Type 12 12. 79 .12. 52. 82 1-5. 13. .11: Sakvarl. Type 12. Type 20. 27 3 . 8 | 8. : to occur in viii 97 10 : cf.8. 46 22. 12.8: Uparistdjjyotis. This extension of the 58. and the two following stanzas appear to aim at the same : result. No. D. In x 140 6 it is described in the Anukramani as Tristubh. 246 i. : Jagatl. 26. Type 12 . 144 5.11 Mahdpadapankti.11. and is See 246 iii.Stanzas of Jive verses 45. 81. This is found in *vii 50 4 .11. This variation Madhyejyotis. 15 15. : Common only in the normal and cretic periods.8. the regular metre of viii 35.8: Pahkti. and is found frequently in concluding verse of a triplet in vii 96 3. 251 .4. and as the See 242 v vi.8: Prastdrapaiikti nected with No.12. Type 11.8. 25 occurs 54. 25. it occurs iv 17 14-15. viii 46 18. 12 is : brhati found in viii 10 2. This variation of No. and also found in viii 103 5.4: Uparistddbrhatl. This stanza occurs in 53. 5. 56. 49). Nos 27 47. 11 : This extension of No. later 43 15-16. . 41 16-17.8. and is an extension of No. see 114 Note 4 even in these the metre is almost unknown to Kusika (Mandala iii). x 115 9. 8. 39. 57. 12 . in *x 18 u : cf.8. and occasional v 2 12. the only lyric metre which is found mostly in the later parts of x 126 1-7 the Rigveda: it occurs *vii 55 2-4. This occurs i 105 a variation of No. 39. 12.8. but it 105 . Type 11. viii 97 13 A complete stanza is found Type 13. 42 16-17. 12.11. Type 8.8:8: viii 35 23 (cf. the next No. 63io-n . 51.8.8. For the relation to Tristubh see 49. 11. 48). 12.11.

Type 12. in *i 191 | 13. but found in the stanzas vii 32 2-3.4.12. Type 8. 12 | .12 . | 8. Type 11. found in vi 15 65. 10. This stanza is found in viii 36 1-6 : see 247 74. | A : see 245 64. perhaps by some as 68. | Stanzas of six verses Type 12.12. No. Cf. . the verses varies. It occupies an intermediate position between Satobrhati (No.8. ix 107 2-3.8. 66. even in the same hymn.12. 71.8.4. 2i. Type 12 12 . Type 12. with a still longer stanza it appears not to be earlier than the normal The grouping of period. 8 1. and 25 see 248. 54. 8. being found in x 21. | error in the text. 61.11. and found in iv 1 2.8: Mahdsatobrhati.8. Cf.8 12.11. viii 46 17. is found. Type 12.8 12. Type 12. Type 12. group. 44. See 247 ii.. This exten- E. This stanza No. This stanza and x 132 7.8. 17 14-15. | 73. 8. and occurs in viii 39-41. 75. This stanza is an exand takes its place in the strophe ( 246 iii) in vi 48 is tension of No.8 8. brhati. This metre first appears an extension of an Anustubh stanza in v 86 6 it is used in associaAs an independent metre or associated tion with Paiikti in *x 59 9.12.8: Mahdpahkti. Type 12 12 . 24 1-3. | 70. 134 1-6. x 133 4-6. 3.8. Type 12. A variation of the last No. 12. Type 8. 15-16.8 it is 12. = 37 7.248 60. 44) and the Atyasti group .8. 8.8.12.11 11. 12 8.8 8. Type 8.8 8. found in 37 This is an extension of Sato72. 12 | sion of the Jagati stanza 12.8 12.4: Astdrapankti.8. See 242 vi. This extension of the Anustubh stanza is peculiar to the Vimada group. This is an extension of the Jagati stanza.8. 63. but is a probable interpretation of the pairs of stanzas which appear as vi 10 6-7. 12 [Pancapadd Jagati]. | : 67. 12 iv. 6. not recognized in the text.4.8.12.8. : This stanza also belongs to the Atyasti group it is found in iv 1 1.8. ii. 12. 4. viii 12. and is a The verse of four syllables always consists Type 12. of an unaccented word. 12 . | only found in viii 36 7 stanza belonging to the Atyasti 62.8. and is also found in vi 48 15. This stanza is regularly employed in v 87.8. viii 19 26-27. and see 246.8 | developement of Satobrhati. Type 12. 8. is found in *ii 43 2 and vi 15 6. See 247 ii. Type 12. 8| 12.8. | .8 8.11.4. This stanza is not recognized in the Samhita text. Stanzas of six verses.8.8. 69. | | 12. This occurs in viii 37 2-6. | This stanza belongs to the Atyasti group : found in i 135 7.

8|8. .8. | 247 . This elaborate stanza is found generally in the hymns i 127-139. Type 12.Long F. See 247 iii. This stanza of 76 It is an extension syllables is the longest recognized in the Rigveda. Stanzas of eight verses. 8. and it is 8 8 14 8. strophic period. iii. found in 40 2. Type 12 aimed at in ii 22 85. 81.8. Type 12 4 . . 8 4.8: 12 | Asti. . 8 | 12 . 83. Type 12. i : ii.8. . 247 ii.8. j 79. This is the metre of i 137. . 8 . . : Dhrti. This combination of a BrhatI and a SatobrhatI stanza was very possibly regarded as forming a single stanza 87. 12 8 . 12. 12 . | | Atyasti group. 76. | 86. x 133 13. Type 13 . Type 8 12 8 12 8 12 8. This stanza is found in i 133 doubtful whether the text requires correction cf. Found in in i in ii 22 see see 2.4 Type 12 8 . Type 8 8 8 . . | 8. Type 8. and belongs to the end of the archaic period or to the See 247. Found Found : 129 129 9 8 : 247 247 ii. See 12 . 80.4 | J 12. 82. No.8.8. : for the occurrences see 246 .8. though with many slight variations. see 8 8 the metres that now remain. 3. Found in ii 22 1: see 247 iii. .8 .8 12. . 8 | 8 8 | | . 82. 4 | 12 8 . . 12 . . | Type 12 12 8 8 8 Atyasti. it belongs to the . 4 | | 12 4 | 8 . . This combination of a Kakubh and a SatobrhatI stanza was perhaps regarded as a single stanza in the For the occurrences see 246 i. strophic period. : | | : . 50. 88. 12 .12. .12. . This extension of the Mahapankti The grouping of the verses 77. and occurs in i 127 6 247 ii. 12 | . This appears to be the type 6. ii iii. 4. 12. 8. is stanza varies.8 8 . 78. see of No. .8. .8. Type 12. Like all 12 8.12. Type 12 12 8 8 8 12 8 8 Atidhrti. . viii Type 8. Found in iv 1 3. 84. 8. stanzas 249 Stanzas of seven verses. | : G.

CHAPTER 254.' and ' ' But popular periods. and subject-matter in such a way as may seem most helpful to those who in the future may be concerned with the investigation of any one of them. . and the division into ten Mandalas was replaced by a division into ten groups. and which does not in every particular basis. language. and subject-matter of the hymns first of the Rigveda. with the object of finding a basis for the more exact metrical investigations which were to follow. GENERAL CONCLUSIONS. in accordance with their general correspondence to the respective types of the bardic. language. The historical study of the hymns by means of internal evidence is only possible by the process of successive approximations. and lastly to trace in general outline the developeinent of the metre.' normal. X. and three with the ascertainment of the laws of metre. In the was made three chapters of this book a general survey of the metre. in the six chapters which have followed.' ' ' ' cretic.' each group being based upon one or more of the collections found in the Sariihita text. support the rearrangement which was adopted as a Thus a revision of the grouping becomes necessary. then to restate the principal elements of the evidence in agreement with the revised grouping. and in this concluding chapter the endeavour will be made first to revise the grouping in harmony with all the evidence now before us. we have met with a mass of evidence which has a considerable bearing upon the grouping of the hymns. As a result of hymns were rearranged. of which three have been occupied with the establishment of a metrically correct text. but being modified by the detachment and attachment of individual hymns and the ten groups were arranged in this survey the ' : a rough chronological order. each of which The earliest reacts upon the evidence by which it was obtained.

Brunnhofer in KZ. but many addressed. therefore. Trochaic Gayatri. pp. the popular Rigveda being denoted as before by the asterisk. Poems which are not of the nature of hymns have no entry in the column for the deity.The Table of hymns 251 attempts made in this direction broke down because their authors took as their units complete Mandalas. and C. and Epic Anustubh are recognised as distinct metres. 255. the bardic period of Chapters i-in having been replaced since by the archaic and strophic periods. but it is very clear in ( every part of our subject. examination the plainest of all. is. the Rigveda proper and the popular Rigveda is recognised.' except ' that the words 'cosmogony' or 'charm' appear where the poem has . extra verses in Gayatri or trifling variations (such as occasional Tristubh hymns) are not noticed. though not p. The period to This difference. in in dimeter rhythm ( 195) and in 180). to this chapter. states the Following the practice of the native Anukramani's. x The latter writer has however correctly stated. in opinion. the student will be prepared for the further distinctions made at the opening of this book. whilst the further corrections made in this chapter. in syllabic restoration ( 153). 576-581 (1878). JAOS. These difficulties attend the first investigations of any intricate problem but as soon as a beginning has been made. and in the present work and the papers which have preceded it I have done no more than attempt to carry out Lanman's programme. the method by which progress is possible. xxv Wl 3 t\\ (1879). that quantitative change ( trimeter rhythm ( 235). Contaminated Tristubh. as stated on ( my 581 of the work referred to. ' ' ' The ' ' ' The difference in character most marked in the use of hiatus between these two periods is perhaps 131). pass unnoticed in a first review of the Rigveda. as H. : unimportant in themselves. R. though it may becomes on closer which a particular the left-hand hymn is assigned is shewn by a capital letter in column. the Table metre of each hymn and the deity to which it is ' ' ' ' Uneven lyric. The is revised proposed indicated in the grouping of the hymns which is now Table of hymns which is appended ' ' periods to which the respective hymns are referred are five in number. scarcely affect in any important particulars the arguments of the preceding chapters. It is obvious that the process of repeated correction is not only laborious for the investigator but also for the reader. Lanman in his Noun-Inflection in the Veda. who may further be inclined to dismiss as artificial arguments which involve a mass of detail which is constantly changing and presents no clear picture to the mind. they diminish Thus as soon as the general difference between rapidly in seriousness.

three state the number that is. hymns of this period may generally be recognised by the small total number of variations (on the average about one in every four trimeter verses) and by an approach towards equality in the number of early and late variations. In the case of each hymn or small group explained in the statistics as a whole combine to form what we may hymns call a metrical 'picture. a numerical formula by which the Thus since general tenour of the metrical evidence is indicated. . of those variations which are chiefly found in the archaic and strophic periods. number of Lastly. about one in every two trimeter verses. because it is entirely independent of the metrical evidence. has only an indirect bearing upon the different periods of the former. as the variations is much smaller. this indirect evidence of great value. But when applied is to large groups of hymns. that is. The general harmony between the metrical and linguistic evidence appears clearly from the Table in 257. according to the lists given in Appendix I. Hymns in dimeter verse. half of these being in the first column alone.' that is. 258-260. all variations are comparatively rare in the normal period. that is. of which of ' early metrical notes. however. cannot be so readily distinguished by their metrical pictures. being mainly directed to ' the difference between the Rigveda proper and the popular Rigveda. and a hymn or group of hymns will be judged to be either earlier or later according to the extent to which the metrical picture deviates from the standard thus laid down: the archaic hymns being most plainly marked by having a very large number of variations. and such as occur are evenly distributed between those that mark the earlier and the later periods. in the two right-hand columns is given the number of occurrences of early and late linguistic variations/ including both T grammatical forms and w ords. poems which are not hymns are found almost exclusively in the popular period.252 Metrical pictures It thus appears at a glance that either of these characters. This linguistic evidence. The normal period thus becomes the central point in our investigations. and popular of of those variations which are chiefly found in the cretic The nature of these variations is more fully periods. and the last three the number of 'late' notes. The metrical the first statistics are arranged in six columns.

in the normal period. which the metrical and linguistic evidence may be combined with regard to the larger groups of hymns which appear there. first * three are early variations. the linguistic . B. quite common in all periods. verse being the more significant. and the corrected statistics shew that they are naturally divided into six groups. also seen to the archaic period. the first due to the frequency of latter to the frequency of the cretic break. The Table in the following section shews the manner in 256.' A detailed list of the variations is given in 260.Final arrangement of of ' the groups ' 253 in the Table the preoecBng chapters to possess historical importance. Lastly. whilst 127-135 and i 165-190 belong in the main i 74-92 cannot well be earlier than The greater part of the fourth Mandala is i hymns The trimeter to be probably of the strophic period. there can be now little reason to doubt that the groups the strophic period. but much more frequently employed after the end of the strophic period than before. Variations distinctive of the same periods. D. and the whole amount of evidence is much smaller than in trimeter verse. Variations distinctive of the archaic period. Of these the ' late variations. the We must further notice that the average metrical pictures of the last three periods differ but slightly. Variations distinctive of the archaic and strophic periods. but also in use. E. some forms of secondary caesura. Variations distinctive of the cretic and popular periods. C. and that therefore the metre becomes here an uncertain guide. in all the trimeter groups. due to the frequency of secondary caesura: and the group i 116-119 shews a high proportion both under C and under D. The figures in thick type apply to the whole mass of hymns assigned to each of the five periods. and give the average number of variations under each heading in every 100 trimeter or 150 dimeter verses whilst the figures in ordinary type give the : same information with The pictures as given for trimeter regard to the separate groups. as follows : The metrical variations referred to in 257. 260 and hymns include all those which have been shewn in A.' and the last three are In dimeter verse only the first two and the last two classes are represented. The cretic break. F. Variations distinctive of the popular period only. though to a smaller extent. of variations under of Mandala vii shew a high proportion C.

directed to two points. the metrical arrangement previously adopted has been left unaltered. followed if it is well marked but if picture has generally been or if the group or hymn is very short. and for this the metrical evidence is of With regard to dimeter verse. In the earlier periods the grouping is most successful little value. with regard to the lyric metres. archaic period. Except as regards the groups here discussed. the grouping found in the Samhita text. In these large groups the metre has the regularity of the normal period. contain almost twice as many early variations as those of the Kanva hymns.254 Dimeter groins and single hymns evidence very closely approaches the standard of the period to which each group is assigned. list of hymns previously assigned to the popular Rigveda no the alteration has been made. period. important of the two. and the ' direction. attention has been mainly gathered.' indications that may have been noticed any divergent in Ch. : now usually disregarded. ill are accordingly. whilst the language has the variety of the This contradiction is with our present knowledge but the metrical evidence seems to be far the more insuperable. . whilst lyric hymns of the Kanva groups correspond precisely to the dimeter verses of the lyric hymns of the same groups. It seems also probable that the section i 24-30 belongs to the strophic In the other groups the results seem fairly consistent. verses of the group i 127-135. and those of the group x 20-26. From this assured starting-point we go on to notice that the dimeter hymns of the early part of Mandala v. an italic letter in the left-hand column being used to indicate that the metrical picture by itself would point to a different conclusion. the this is not the case. as very little new evidence has been In the Rigveda proper. the revision of In the grouping only affects single hymns or very small groups. less successful. Where these both point in the same metrical picture. and the grouping is simplified Where this agreement is wanting. just as its trimeter verses. and therefore only a very small amount of dimeter verse appears in the Table as belonging to the cretic and popular periods. 61-67 30. in which it is guided by the Here for instance we see that the dimeter trimeter variations. it is easy to see that our work is Anustubh is happily for our purpose sufficiently Epic distinguished by its external form. are undoubtedly of the archaic period. the dimeter with the striking exception of the Soma Pavamana groups ix 1-60.

column represents the 'archaic' variations has by far the most distinctive which gives the occurrences of the cretic character. In tracing the history of single variations we have often (ii) needed to collect the occurrences from 1000 verses at one time. But the different which columns are by no means of equal value as evidence . Where (as in the third and ninth Mandalas) there are considerable series of hymns which are identical in their metrical character. we can trace this with the same accuracy either by observing a single variation in 1 000 verses. it is essential metrical picture for each part. however small it to record a separate ' ' ' may be. A D ' ' hymns It must be noticed that the numbers given in the 'Table of whilst those in the Table are absolute. space has been economized by On the other hand. The variations of which the occurrences are counted up in (iii) the six columns are those which are shewn in the Table in 260. But when the most important variations are combined. and cannot be very clearly followed unless we can first group on some other ground hymns containing together from 100 to 200 verses. or all together in 100 verses. single hymn consists of parts that are not homogeneous. In fact the archaic variations are so numerous that we have little difficulty in recognising a hymn of archaic type by its metre. (iv) ' : . even if it only includes from 25 to 50 verses but in the other periods the evidence is less direct. and column break is the least important. and include all those which have been shewn in the previous chapters to The evidential value of the separate possess historical importance. The extreme shortness and fewness of the dimeter hymns in the family books greatly hamper the application of the metrical evidence. if a treating each such series as a single group. the history can be traced with equal certainty from comparatively small bodies of verse. must therefore be reduced to the same proportions before they are compared with the standard pictures for each period. variations included in one column is approximately the same. and sometimes from as many as 5000. Supposing that 10 variations have exactly the same history. as appears from the corrected statistics as given in that Table. and not proportionate in the next section are proportionate to each 100 verses. and we can therefore do little more than assume that in each Mandala the dimeter hymns are probably of the same : date as those in trimeter verse.Detailed evidence of date 255 The following notes deal with points of detail with regard to the use and value of the evidence of date furnished in the 'Table of ' hymns : The amount of matter contained in the 'small groups of the (i) 'Table of hymns' varies considerably. three dimeter The pictures in the 'Table of hymns' verses being counted as two.

256 257. No. . Combined pictures Combined Metrical and Linguistic Pictures.

and also the Anustubh hymns of the Kanvas and of the cretic period.Metrical developement 258. according to the rhythm ( 225). developement of the Rigveda runs parallel with that of the metre. decasyllabic hymns. that is. can only be followed systematically so far as we find general types consistently followed. both for trimeter and dimeter verse the only all : : exception being the increased use of the cretic break. we put aside all hymns of special types. such as the and those in Trochaic Gayatri and in Epic Anustubh. which results in the use of Sandhi combination. whilst Pentad hymns are found as late as the normal period. and 7 are connected ' with the restoration of the text. one of the most important criteria of the popular Rigveda. therefore. one has to do with the external form. and only 4 which retain some importance in the normal period. either to 257 traced with regard The developement of metre may be the ' ' form ( 30. ( ' ' ' ' to be a mark of the archaic or of the strophic period. but with regard to it there is little to be amended in the statements of the earlier chapters. the consonantization of semi-vowels. : although there are 17 . This is to be seen first of all in the use of Sandhi and the linguistic forms which are reached by metrical 259. Of the later variations almost have to do either with the text or with the external form so that we may say that at the end of the normal period the internal form is fixed. date of the Trochaic Gayatri metre is not satisfactorily determined. The developement of the internal form. 31). Lyric metres generally characterise the archaic The period: but the regular Brhati-Satobrhati strophe is later. history of trimeter verse and (less clearly) of normal dimeter 260. The The general explanation of these changes is the extended increasing rapidity of pronunciation. We notice internal external or the ' ' External metrical form is however that contamination is also a mark of the archaic period In the Rigveda proper decasyllabic Tristubh is found 223). linguistic restoration. of the less striking variations. verse can then be followed by the aid of the Table in The The of ' which early variations included in the Table are 26 in number. and the shortening of vowels originally long a. which approach the type of Epic Anustubh ( 200). Of the remaining 18 variations there are 7 which are almost restricted to the archaic period. For this purpose. 7 which are common to the archaic and strophic periods. and are discussed in the next section.

-i. Notes to the Table on the page opposite. -u. though not to the same extent. It is sufficient here to note that the linguistic features of the archaic period differ from those of the normal period somewhat in the same way. 3 3 143 ii. 129. -i. archaic period than elsewhere. etc. accompanied by caesura after the third syllable. -anam when restored on 5 6 metrical grounds. but they are used in the periods which precede it.e. -vanam. so far as grammatical forms are concerned. 151 i. also 220 iii-vi. Catalectic and heptasyllabic dimeter 7 8 verses. when shortened |[ : . of the whole number of these words and forms there are some that go out of use at the end of each period in succession. and contrast them with those that mark the two earlier periods. as the dialect of the Homeric poems differs from that of Herodotus. This line of evidence has been pursued. and are more common in : the strophic and normal periods than in the cretic hymns that is. same The same progress can be traced in the use of the linguistic features which characterise the Rigveda proper as contrasted with As appears from the Table the late Rigveda and Atharvaveda. 174 i). and after duals in -a. -manam. hi. which are explained by the influence of analogy. 128. also i. and generally reach their minimum in the normal period : but the linguistic developement can be traced in the later periods by the rise of new variations which are due to the general causes as the decay of the others. All the figures in the body of the table are proportional to each 1000 trimeter verses. This variation is not taken into account except in the treat14 ment of dimeter verse.for the n also u with consonantal value of y other forms see below. 170 i.' 1 is in accordance with the arrangement adopted in the ' Table including hiatus after -a when shortened ( 172 i). -ii with 15 dissimilar vowels. The Table in 260 shews that those variations which are becoming rare in the archaic period die out rapidly. none of them are in regular use in any part of the Rigveda proper. when 166 iv. in the older forms and words are much more common in the 257. Vi in each case. Combination by Sandhi of final -t. and verses with double rest. in the author's Historical Vedic Grammar. cretic period twice as often as in those Very much fuller evidence would be available if we were to take as a starting point the linguistic features of the normal period. also except -nam in ( 4 dimeter cadence. as in -saham. 9 10 indra verses. or to each 1500 dimeter verses and the assignment of the occurrences to : the respective groups of hymns. As to the later forms.258 Linguistic developement a few instances of change in the opposite direction. or by the break ~ . with the final vowel in each case shortened.. Viratsthana and except when accompanied by secondary caesura.

.Table of variations 260. Refer to 259 Table of Variations of historical importance.

perhaps of later date. heroes.2G0 Origin of the ritual As the parallel developement of language and metre 261. or a means of warning off thieves and beasts of prey. The preparation of the sacred drink. originally mead and in a later form Soma. affords adequate proof of the general chronological sequence of the Vedic hymns. was an even more direct means of strengthening the clan by raising the spirits of its warriors. In giving here a general sketch of the conceptions developement of the conceptions which form the subject-matter of the hymns. and so that the one becomes to a great extent out of harmony with the other. It is perhaps peculiarly difficult to trace such a developement in a ritual literature. in which antiquated may easily linger for long periods by the mere force of daily repetition. The people in heaven is variously described as consisting of the we gods. necessary is accompaniments of either of the primeval ceremonies. and that there is reason to suppose that investigation may lead to more definite results. it is unnecessary to trace the developement of ideas with any purpose of obtaining from this study a corroboration of our theory as a whole.' to which list ' ' ' ' . or spirits which may be adduced from time to time in explanation of them. 262. the poetic inspiration which finds expression in chant and recitation. Throughout the Rigveda the 'priests' are primarily expert craftsmen. The sacred fire is now kindled to act as 'messenger' between two peoples. The kindling of the sacred fire before dawn may have been originally an act of sympathetic magic calculated to ensure the The appear to return of daylight. skilled in the kindling of the fire or the preparation of the nectar in accordance with ancient rules: a third attainment. might some day fight on his and making welcome the visitor who hosts' side. but But even in the earliest parts of the Rigveda we find each of the two great ceremonies interpreted in a different way. Mitra. closer ritual practices which are fundamental to the Rigveda be essentially older than any beliefs in gods. Aryaman. and therefore as practical in its aim as the kindling of fire to be a source of light or warmth in the house. long after they have ceased to express any genuine feeling or belief.' or by name of Varuna. reaches almost equal importance. we can only attempt to point out that here and there such developement plainly runs parallel with that of metre and language. the one on the earth and the other in heaven.

' since so far as our knowledge goes the study and admiration of the heavens is in the first instance associated with the shepherdThis ' peoples of the Chaldaean plains. that we must ascribe to the enthusiasm of conquest embodied in the praise of this warrior-god the chief impulse towards the creation of the literature. and fighter.' ' A great feeder. he stands out as the type of the Aryan adventurer-prince. the two worships seem to be a gulf both of theory and of sentiment. swiller. We have therefore in the archaic period a clear conception of a company of deities living in the sky. taking This conception seems so natural possession of lands and herds. and Pusan the rustic deity of the religious conceptions there is field path-ways. clouds. Zet>9 or Jove. cut off by the fact that he dwells on earth. rushing to the invasion of the lands of the 'seven rivers. that we are hardly surprised to find that there is practically no trace of the worship of Indra amongst other Indo-European peoples: and it is so predominant in the earliest Vedic hymns. but his name Dyaus is still held in .' storming forts. chief deity of the Indo-Europeans. a product of the times. In these a striking lack of coordination. releasing captives. is honoured in the Rigveda by no hymn. not in the sky less sharply by the lower moral tone which he . and an interpretation of the sacrifice as a festival to which these gods are invited by the fire-messenger to descend. to which we are hardly likely to find the clue The unless it be in the history of other Indo-European peoples. Both appear capable of naturalistic interpretations. On the other hand the drink-ceremony or hero Indra. circle of deities we may provisionally name the Chaldaean gods. well described is associated with a god by H. separated by Two other groups of deities are prominent in the earlier parts At of the Rigveda. the Asvina as representing some phenomenon the ' Chaldaean of the sky. and we can only think of them as fragments of some earlier system or systems of deities.Gods of heaven air and earl It -jin may mentally add 'and so forth. From the ' Chaldaean gods ' Indra is and again not represents. this period then. deities ' . In a subordinate position we may notice especially Usas the Dawn. the Asvina and the Marutah. Oldenberg as a barbarian-god. by which they come to be associated with the Marutah as representing the storm- which as they sweep furiously over the earth may be compared to or associated with the devastating onset of Indra.' since the names given are plainly no complete catalogue of the gods meant.

making (ii) the Indo-European nature-worships. the daily course of the sun occupies a position of very subordinate importance. be disentangled. who with a blow parts heaven and earth. and can therefore only with difficulty . and makes the light appear. and we may largely interpret this Indo-European system as a deification of natural phenomena. ceptions cross one another and the primary conception of the warriorgod in all imaginable directions. . name stands always prominent in the catalogue of the gods. which may have been originally independent of him and much earlier in date. sents the Wind this later interpretation of his character may be In no case is Vayu to be identified either a reminiscence or a guess. : by Brhaspati. with Vata. the invasion of India. the thunder-bolt (d) the conqueror of some miserly These confoe. (c) the storm-god. who rides at the head of his host. such as the Panayah or Vala. for they appear side by side in many hymns to the Visve In the later parts of the Rigveda the place of Vayu is taken Devah. . we may reasonably speak of all the deities mentioned in this section as Indo-European. To the primary conception of Indra as the warrior-god several (i) others are attached. the begetter of heaven and earth. Failing fuller knowledge. as indeed it does throughout the Rigveda in the hymns addressed to the Visve Devah. : (i) the primitive ceremonies of and drink. and then admitted to their company. the slayer of the dragon who guards in his rock-cavern the seven sacred streams (b) the dawn-maker. however. Our general conception therefore of the subject-matter of the Vedic hymns is that it is composed of heterogeneous elements amongst which the following stand out prominently. Indra and Vayu and at times he seems to exercise a shadowy sovereignty. Such are the conceptions of (a) the Vi-traghna. arranged in an order of time based upon their relative clearness in earliest the minds of the fire(iii) hymn. the universe. the protector of the He is chosen by the gods as their champion in time of ceremonies. the Marutali. We may consider here in more detail those deities who are most prominent in the archaic period. In the archaic hymns Vayu appears to be the charioteer of (ii) There is no hint in the hymns themselves that this god repreIndra. In the latest poems his danger. and the savage traits of his character are He assumes the position of the creator and supporter of toned down.262 respect. Amongst these phenomena. the Chaldaean deities of the heaven (iv) the warrior-god of . In the later periods of the Rigveda Indra is brought into closer relation with the other gods. whose cows he seizes.writers . and who strikes down his foes with his mace.

known as the seven Adityah. are variously given find with some regularity Savitar and Bhaga. if indeed that was their original besides the three principal figures we number.g. or those of the male deities Daksa and Amsa. In the periods next following either Aryaman disappears from the group. but not so frequently in hymns addressed to this triple deity as in Agni hymns. This number seven we may perhaps trace back to the archaic period by the help of viii 28 5. The Marutah are frequently brought into relation with Indra. a door is open for H. In later times (e.' The names of the seven Adityah. 193). In the earlier parts of the Rigveda they are also entitled Rudras. who did not practise even the fundamental ceremonies who kindled no fire. In truth Indra appears to have stolen his title of Vrtraghna from some Later the Marutah become the fighting-men of earlier god or gods. and the Vedic statement that they are sons of Dyaus (viii 20 17) may be interpreted by us as meaning that they are IndoEuropean deities.The Adityah (iii) 268 is The group Jlitra-Varuua-Aryamav \crv with in the archaic period. RV. It seems clear that the Marutah are personifications of the (v) storm-clouds.' In the pantheon thus formed the war-god naturally took . ix 114 3c) these gods form part of a group of seven. and the list may be completed by adding the feminine names of Puramdhi and Aramati. or other names are added. Vedic pantheon appears to have been formed under the pressure of war. as twin-charioteers. In the later parts of the Rigveda the Asvina are rescuers and physicians. Indra's host. and it would seem that through them Indra first became associated with the region of the clouds. p. on the ground that he has conquered the Vitra. and if so. The first different Aryan clans: but any such difference became small ' in the face of the dark-skinned natives. : . the sons of their mother Aditi. not they. and pressed no . Oldenberg's of this group as ultimately based upon an older interpretation worship of sun. moon and five planets (Die Religion des Veda. who travel to the ceremony and bring with them mead by their beauty and their courage they win the favour of the Sun's daughter she mounts their car and becomes their bride. Soma. who becomes their father. In the Mana hymns (i 165-190) Indra robs them of their share in the sacrifice. The association of Mitra with the sun is frequently based upon the parallel with the Iranian Mithra. but later a deity Rudra is evolved. The various cults noted in the last section may have been maintained with various degrees of zeal by 263. ' : The Asvina are twin deities who are usually interpreted as (iv) In the Rigveda they appear denoting some phenomenon of the sky. and play a part in numerous myths. But it finds little support in the Rigveda. It is however quite impossible to connect the Vedic deities commonly 111H directly with the separate heavenly bodies. in which Agni_ is often said to be like Mitra. unless it be in the early Agni hymns.

The worship of the fire-god charm and and of the Soma was already in existence in the archaic period. and under the names sometimes of the pair Mitra and Varuna. senes perdenda fateri. Every stage in the ritual is described with insistent minuteness in hymn upon hymn. Savitar or Bhaga. quod iuvenes didicere. Beginnings of philosophy But when this pressure was removed new forces and aesthetic began to assert themselves. in mountain and wood. and in the variety of opinions the uniformity. were unwilling priests. repeating on earth the everlasting harmony of the kingdom of heaven. as we have attracted the more attention. and the existence of a beyond known to the initiated only. Rigveda proper they are increasingly prominent in their ceremonial meaning only. and even In the later parts of associated with Indra in his warlike feats. Under such influences the half-forgotten lore of Chaldaea was in part revived. were originally celebrated for severely practical purposes: but when the fear of days without dawns and sons without spirit died out. New deities are the . possessing a moral grandeur not altogether unworthy to be compared with that of Ahura In the exuberant life of a tropical country the of philosophy arose in questionings as to the source beginnings and the unity of the manifold vital powers of the universe. and their sovereigns and and chancellors prided themselves in establishing an order of peace justice. political. as it seemed. Thus the fire-god took ' ' again a new character as the source of cloud and rock. and the third mysterious region and from these in turn heaven. and. the beauty of the ceremonial The ceremonies indeed. but the deities were conceived in human shape. philosophical and to influence the forms of religious worship. sometimes of a single deity as Varuna. Great kingdoms were established in the New Land. and even a three- a giant who fold order of the universe. Mazda setting-out of its parts in the order of space. life in man and old-world beast. religion became more and more divorced from practical affairs. the or Yahveh.204 the first place. three heavens. the fancy builds up three earths. they were continued for their inherent : as exercises of priestly skill perhaps too because the like highly educated people in all ages. in Some : myth of takes three steps was turned to a picture of earth. seen reason to think. and embracing the light-skinned and the dark-skinned as members of one family. a sovereign power is established in the pantheon itself. But as the speculative spirit grew.

an Agni : Raksohan. Of the worship of Varuna alone as a supreme deity there is hardly any trace in the Rigveda.. dawn is the hour of the ritual. These deities are therefore no longer nature deities in the strict sense. and one to Bhaga. in those of the singers (Agni Narasacrificial post. But this developement is most marked in connection with the ritual. and apparently from the penalty of dropsy. Agni is therefore at once one and manifold by the side of Agni stand Thus in the later parts of the Rigveda proper Agni the Agnayah. in the doorways. Several hymns are addressed to Savitar singly. is That Visnu represents the marking out of the bounds of space (iii) the view of H. in the straw-carpet. and mrnsa). Vac and above all the deities of the successive verses of the Apriya hymns. Gauh. but in the persons of the worshippers (Agni Tanu-napat). an explanation of the new importance attached in the normal and cretic periods to such deities as Usas and DyavaThe Prthivi. (i) but are part of the equipment of the ritual. and at the same time to hold them apart. and also as a warrior-deity: in the latter capacity he is Even in these hymns he bears the title frequently joined with Indra. that it is in connection with the ritual of the fire-worship. Perhaps to all previous writers on Vedic chronology.Agni Visiiu and Soma 21. the hymns to Soma to . Macdonell suggests. according to the set type of the Apriya hymns. an abstraction from the qualities of the group. in the in the solemn concluding cry of svahd. which seem to be Indo-European in their origin. two to Mitra. Mitra and Varuna reaches its highest developeto these deities. and the twin deities of darkness we may find and light nurse the infant flame. indicating that all living things are his household. p. Again. In the earliest hymns Agni is worshipped as the messenger of (ii) the gods. Aditi. with the result that Agni is discerned not merely in the sacrificial fire (Agni Samiddha). 228). certainly (iv) the present author in his earlier papers. Havis. Jatavedas. The fire is set up as a pillar to join heaven and earth. Oldenberg (Die Religion des Veda. but there are hymns addressed to him in which he is appealed to to release his worshipper from the burden of his sins. introduced which are without meaning except for their part in the ritual. the mother of the Adityah. Jatavedas and Agni Vaisvanara are deities distinct from Agni himself. the soul of the universe.. as Ghrta. In the Vasistha hymns the daily course of the sun is the chief wonderdeed of Mitra and Varuna and from this time on the worship of of The worship ment in the Vasistha hymns : Surya has a growing importance. which attain an elevation of moral conception which has often recalled to their readers the Hebrew psalms. and we can trace the beginnings of an Agni Dravinodas. is probably. as Professor A. or in other words that he is the source of all life. and so forth.

cosmology. altogether deficient in myths. in Indo-European fashion as parents of the gods. however. goes to shew that only a few of these hymns can belong to the two earliest periods of the Rigveda and it therefore seems likely that the number of these hymns has been artificially increased so that they may equal those The hymns are purely of a ritual addressed to Agni and Indra. Indra and the clever rescues of the Asvina. and takes the place left vacant by the disappearance of Vayu. turn by Vac. few concluding words may be devoted to those myths of which a considerable number which appear to constitute are to be found in the popular Rigveda. as in Greece with that of Dionysus. and its earliest part since in the mythical : hymns the linguistic forms of the Rigveda proper and the popular The Rigveda proper is not Rigveda are used side by side. The deity of the sacred song is properly Brhaspati or Brah(v) ma/mspati. (vii) of the skill Amongst the deities which are prominent in the later parts Rigveda are the Rbhavah they are craftsmen who by their have attained the rank of gods. For this view there is the obvious argument that both in language and sentiment these hymns are entirely cut off from the later parts of the Rigveda and all later literature. and in hymns of the two earlier periods this character is In later hymns the god becomes associated with strictly preserved. Earth ritual can readily be traced in the Apriya hymns. Indra. The metre. which opens a world of thought entirely different to that of the Rigveda proper. Rigveda myths appear in a dramatic form. and In separate hymns Usas alone takes the place which is filled by the pair Usasa-nakta in the hymns to the Visve Devah. but they are baldly related and generally in single stanzas as for instance the warlike deeds of . : 264. and obedient subjects of the law of Mitra and Varuna. and magic.266 Lesser deities Paramenia have appeared to be amongst the oldest in the Rigveda. in It hardly falls within the scope of this book to discuss any detail the subject-matter of the popular Rigveda. and records for us the first distinctively Indian efforts to But a lay the foundations of philosophy. But in the popular and the theme is not . The hymns to Usas in their feeling for brightness and beauty recall to us the Greek Eos and the Latin Aurora. and perhaps are our best representations of the So Dyava-Prthivl appear occasionally Indo-European type of hymn. As god of song he appears to be replaced in as Indra's charioteer. : character: only in the popular Rigveda (ix 113) do we find the doctrine of immortality associated with the worship of Soma. but more usually they are merely guardians of the sacred rite. (vi) The Heaven with the close association of the pairs Night and Day.

The poem is of special interest to the folklorist. 120-126). At the first glance it becomes probable that we have a variant of the tale of Adam and Eve. The best known of the dramatic myths of the Rigveda is the (i) dialogue of Pururavas and Urvasi (RV. and his subsequent desertion by her. One It seems probable suggestion may be made here by the way.Vedic myths historical : 207 but social here too the poets for the first time wrestle with the moral and aesthetic problems associated with the relationIn these myths the woman ship of man to woman. the parents of the human race (x 10). The divine Soma may well forgive the sin. To her So far the poet has only blame passionate appeals Agastya gives way. x 95) on the interpretation of which much light has been thrown by Professor Karl Geldner ( Vedische Studien i. to the advantage of both the light and the dark races (Lopamudra. pp. without any regard even to the claims of We find in fact the germ of the ascetic theories which marriage. 243-295). and which have spread thence over the Western world. though it may not be easy to guess from what common source or by what channels the tale has reached us in such . afterwards became so firmly rooted both in Brahmanism and in Buddhism. being based upon the marriage of a prince to a fairy maiden. we may conclude. However dramatically the sneer may they have the hearts of hyaenas. it reveals a bitter grudge in some poet's mind against feminine influence. pp. Agastya as a sage is bound by a vow of chastity: Lopamudra's youth is passing. that Aryan princes may often have been led by native enough wives away from the practices of the Brahmanic religion. Special attention may be called to the cynical reply of Urvasi in stanza 15. and the question of their origin seems to call for investigation. is a native woman) and the high purposes of the gods are after all fulfilled. for desire is human besides Agastya has won offspring. Such theories stand in striking contrast to the patriarchal sentiments of the Rigveda proper and of all other national religions. not perhaps so much from a sensual impulse as from the desire of children whilst the duty of the man is always abstinence. and in particular from liberality to the Brahmans themselves and thus that an opposition of interests between priests and queens may : have become traditional. : In the light thrown by these two hymns we may consider (iii) the tale of Yama and Yami. when Pururavas threatens There is no truth in the friendship of women : to die for love of her. for the woman. But in the concluding stanzas excuses are found for the pair. : generally appears as the temptress. yet she remains childless. The tale of Agastya and his neglected wife Lopamudra (i 179) (ii) has been lately treated by Dr Emil Sieg (Die Sagenstqffe des Rgveda. suit the occasion in the mouth of Urvasi.

as well as the penalty inflicted on the latter. irritating.268 different forms. Yama does not yield yet it is obvious that in the original form of the tale he must have given way. . and the rival claims and the divine law to man's unquestioning obedience. however have in reality the same meaning as the Vedic: at any rate " well represents the late the phrase " the knowledge of good and evil Vedic conception of the marriage relation. The solution suggested in the Veda is perhaps near akin to that of Genesis of reason : had the first human beings had more faith. they are conceived (and it cannot easily be otherwise). and yet would have found a way to perpetuate the human race. Yarn! is full of a longing to Yama on fulfil her destiny. So far as our poem reaches. such as the instinct of shame in regard to sexual relations. yet replies her brother. and bids his sister look This suggestion is under the circumstances elsewhere for a husband. and to become the mother of mankind. and Yarn! fairly loses her temper and speaks out her mind. knows their secret intent. the laws of the gods the other hand is overwhelmed by scruples permit no exceptions. Our first parents The Vedic myth has the advantage of logical consistency. both as The Hebrew version may brother and sister and as husband and wife. or will venture to be their spokesman?' Heaven and Earth were brother and sister. In the Rigveda however the position is plainly stated. but the gods designed us for wedlock. and the shame which overcomes Adam and Eve after their sin. Who Nay. for the temptation deals not with the apparently meaningless prohibition of eating from a particular tree. the prohibition of marriage between near relatives. both point to the real nature of the sin itself. but with the real moral problem necessarily implied in the relationship of the first human pair. Much more light from other literatures is needed before any certainty can be felt in this case as to the details of the interpretation but it seems sufficiently clear that we have before us the sincere attempt of a theological school to grapple with fundamental problems if : 1 ' ' : : : of morality.' Yami suggests. the heavenly powers would have spared them the sin and pain of an animal mode of procreation. why not you and I then of they became the parents of the gods mankind?' But Yama is still unconvinced. and their eyes are open wide to mark offence.

Piisan. J at. with rhythm of Epic Anustubh). As. . Visnu. J or Cont. Agni (A. Jt. DP Dyaus-Prthivi. Raksohan . Dvipada Tristubh . Mar. Padaparikti. Sar. Mahapankti P Pahkti (E. Explanatory Notes. or Mahap. . Tr.APPENDIX IV. Satobihati. Marutah. 265. Tristubh stanza. Asvina. Devah. S Strophic. Deity. An. 87. or Ep. Brhati. V Forms and words 86. Normal. C Cretic. Tr. S. or Ag. Mitra-Varuna. (E. or Cont. A N Metre. AV. B G KS A. Sarasvatl. Usas. Danastuti. Ct. TABLE OF HYMNS. Prast. J Contaminated Jagati). Kakubh-Satobrhatl. Vais. Tristubh (Dvip. Soma. Padap. 84. or Usn. BS Brhati-Satobrhati. Us. Linguistic notes. lyric Uneven lyric. Agni A. period is indicated by the metrical variations alone. Ma. Forms and words characteristic of the Rigveda characteristic of the popular Rigveda and Atharvaveda. Purausnih. 85. with rhythm of Epic Anustubh). Usnih. proper. Brh. Yaruna-Aryaman. Sav. Tr. Trochaic Gayatri with extra verse). or Brh. or So. Jagati with final (Ct. PrastaraPur. VD Visve Us. See 255. Tr. Bihaspati or Brahmanaspati. J Jagati Gayatri. An. Pus. or Ind. of hymns For a general explanation 255-260 above. Archaic. or Ep. MitraI. Contaminated Tristubh). Dan. or Brhasp. Agni Jatavedas A. MV MVA Metrical notes. Raks. Anustubh (Tr. 260. see the arrangement of the Table of The following are the abbreviations used in the respective columns : Period. Indra. Vis. pankti. Un. Yaruna. # PopuIf these letters are in small italic characters the corresponding lar. Sat. Yar. Savitar. Agni Vaisvanara).

TABLE OF HYMNS.266. Mandala I. .

I 44-50: 44. 47. 49 j Kanva BS An.. Agni. etc. 43 7-9 Soma. 41 7-9 MVA. 39. G 90 144 28 2 2 54 5 52 12 12 3 2 39 Mar.. 40 Brhasp. etc. I 51-57 51-54 55-57 j : Indra collection I J1 Indra N ... I 36-43: 36. Epic 1 An 47 As. S F. 43 i-6 Eudra. 48 45. 38 i-6 MVA.). 10-15 Mar. etc. (2 37. 48 Usas G. E. 41 1-6 38 7-9 Mar. G. Dim.i 1-93 I. etc.. Tr. 40 2 37.. I ABC 12 13 4 20 1 D E F Lingu. Various 2 3 2 1 -w 38 1 3 7-9. 271 Mandala continued. 1 ! 46 50 1-9 10-13 G . 42 Piisan. Agni. Tr. Kanva BS G Tr.

.272 Table of hymns Mandala I. continued.

continued.i 94-191 278 M and ALA I. .

.274 Table of hymns Mandala II.

ii l-iii 62 275 Mandala III. .

276 Table of hymns Mandala IV. .

277 Mandala .iv 1-v 87 V.

278 Table of hymns Mandala VI. .

continued. .vi 1-vii 33 279 Mandala VI.

continued. .280 Table of hymns M and ALA VII.

vii 34:-viii 31 2M Mandala VIII. .

282 Table of hymns Mandala VIII. continued. .

mil 32-103 2*3 Mandala VIII. continued. .

.284 Table of hymns Mandala IX. 2 5 6 Soma P. Apriya Soma j> P. N Soma Pavamana 1-4 collection Soma P.

ix l-x 26 2 so Mandala X. .

.286 Table of hymns Mandala X. continued.

.x 27-114 287 M and ALA X. continued.

. x 115-191 Mandala X. continued.288 Table of hymns.

the verse (iii) : : : : (iv) of aa for a in the gen. even though restoration is required by the ordinary value after a heavy rules. in final (ii) evd. in the words guhya. in the optatives asydm and sydm . and only notices those metrical irregularities which are very exceptional. -I. fern. tvdm. -u. as Hiatus. -a. restorations The Metrical Commentary does not form a complete record of the required in the Samhita text. mfl (mflaya-. matroh. in -vd nouns in -an. su-candrd: and vocative singular. ndrya. dadhydnc. pratydnc. syd and some others and in the suffixes -bhyah. In particular. nfndm.CHAPTER XI. sumfUkd) vowels in positions which are metrically unimportant. in the suffixes -ya -yd after heavy syllables . the following restorations are very in addition to those mentioned in the Commentary : commonly required or exceptional. sakhyd. mfllkd. -u. -dm at the end of the sing. divyd. as when the suffix -ya has monosyllabic syllable. svasroh. -man. tvd. -yd and the dual form -yoh after (ii) heavy syllables . sing. whether regular. -van verse (v) of -an for n in the declension of : : : ar for r in the dual forms pitrbh. Quantitative restoration: (i) in the words correctly written dcchd. 267. Combination of final -a. Emendations suggested in the Commentary are not intended as definitive proposals. -i. Introductory Notes. -bhyam after heavy syllables at the end of of u for v in tvdm. 19 . pavdkd. and at the end of the verse: (iii) of cch for ch in Aufrecht's text: (iv) of c for and (vi) of : sc in locative compounds as puru-candrd. but only as indications of the general requirements of the metre in accordance with the period to which the hymn is assigned. pi. mddya. Syllabic restoration (i) of i u for y v followed by the grave accent of i for y in the instr. -a with initial r (text -a r-). (v) of n for nn in the Where the text gives a correct reading this is not usually noted in the Commentary. tva. as in the third and fifth syllables of trimeter verse. as after after -a. METEICAL COMMENTAEY. tyd. before consonant groups. yiijya. jdnya. svd and in the instr.

*21c jiok. is 3b and elsewhere indra vdruna. 18d sumati. 3a indrd-. B.] For the date see 88. 86 35c mddhvd. 15c jiok. Bonn 1877) ix 47 2b dasyutdrhand. . 13a dhuat. 9b see 152 ii. . : 33. [Gayatri hymns of a very regular type.290 Metrical Commentary : The following slight corrections are needed in Aufrecht's text of the Rigveda (2nd edition. somavrfham su-dvasam. 17. 6a dmte. 26. perhaps aaditya.g. e. 3c pdhi. 14a ghd 30. 150: extended Tristubh verse. Later additions appear both at the beginning and towards the end of the series. 14c Viratsthana verses. 14a perhaps ydsmi. la vdsisva. 27. 51 lc bahud/td. to sbmdnaam suarnaram or 3a probably ntisatidbhidm. I. Sandhi is (130 iv) initra vdruria 3a indrd. 18. 19c *18c kdrtuam. or mitrfigiti as a compound. 12c dhuat. non-metrical. Metrical Commentary. perhaps pardkdat. tutivdn. needed. la correction . x 4 5d pra?idyanta." 8. or read te for tdva. vii 66 16a. C. 5b smd. 17a dsva valid. 151 i. 21b a dntdd : its but 33 in I 31-35. and the metre corresponds fairly with that of the dimeter hymns of irregularly arranged. I 12-23. . see 174 ii. 8 3c dsvabudhnd. 174 ii 6c dsathe . 4a u. I 24-30. 12a visvdha. except as regards some of the more common restorations. normal period. [Trimeter hymns of the cretic period metrical character is of the strophic or normal period. 25. to be deleted.3b either miirdm aynim with irregular as in Latin. 6 2a bhdniibhir. Mandala III. 9 perhaps in Trochaic Gayatri 8d asmdbhya. Kanvdh 14 2a. 106 2d mahishevdvaptindt. 9a a 'double rest' ( 226 iv) is hardly probable in this hymn perhaps add tvdm at the end of the verse. 139 iii. [Dimeter hymns of the normal or strophic period. 28. : 31. 22. References are given throughout the Commentary to the sections book in which the points raised have already been discussed. 7c hdrlva with Sandhi. 12d sd. 15a utd yujia. for which the references are given in the Index. I 1-11.] 12. 6b 15. 15c *24. 20. 12a santia: the statement in 137 14. [Dimeter hymns of the strophic period.] D. cf. of this 268. 2b. 10 is addressed to Indra of the Kusika family. *19a probably 23. 6a smd. Mandala A. not earlier than the. 14a. 143 3a daiisishtJidv.

ib. : . the metre being in disorder and the verb wanting probably brhativa with Sandhi . lb Viratsthana verse. and connects viii 80 and ix 93 see 106 i. 13c sambhu is probable. 4c sd 8c Bhargavl verse perhaps read vfsd kdnva abhuvat . 1 2c 36. 10a 6/f for bhdh: the repetition of the particle w is 174 ic. 3a ucchdt. 7a 178. srutia . . 170 iii. 3c kua. I 51-57. 139 iii. 14c tdtra su\ 15c dyu. 5c 53. 147. . 166 vie. 173 iii. . 4b ardbhyd. 6a upa . 35. : 15a pdhi. . 5d cdratah. 17c. es-d 2c perhaps vasu166 iv 6c rdsathdm. impossible. 5c prd . 56. Id dadsvatl. 8b dhdnud. . 10a prdmaM. lOd srudhi . . 4c trsu. 2c kd. 14b duria. 9 c asmdbhya. 9d the hybrid Tristubh verse invites correction. 59. see 178. For the peculiar metre of i 61 see 250 iii. 7b probably -asya with hiatus. 12c gdvah. llb^awa^a^. 11a vdmsud 12a vahd 16b mimiksvd. 8c bhava . 1-61 . mahind. 151 iii.] . lid Viratsthana verse. I 58-64. series i [Of these : Kutsa 51. 4a requires correction. see 152 ii. vocema. 9a ndsatid. 8d Viratsthana verse. 10c. 4b mdrtidsah. . 142 ii 15b Gautami verse rather than Pentad perhaps imnah 15c probably surye (Viratsthana verse) . 37. . 7c ddsudn. cf.i 34. ?/> . 2a nemanisah 4a yatR 57. 6a correction. I 36-50. i 146 lc. 7c sakhid. H. 4a Bhargavl verse. 175 i. 10c. . 2 b the verse is disordered. : . : . 46. as by reading d for . i 62 13b. 2c pradivo 'kdma52. 6c ndribhiah. EF. triui dhdnvd would be more regular. 48. not mdhya for the rhythm cf. 45. 6c perhaps asruvat. metrically more convenient. 39. 43. [Two typical Kanva collections but both from the form and the contents a still earlier date is suggested for the : first hymn. 6b vdsi-. . . . 291 lc himid. 167 iv 6d evd. 160 i. 47. lid su-apatyai. i .] as : : 58. which is rare. 2b. 8d nd 'not' with hiatus. 2c smd . though the reading -driamndh is 42. 13a ydanti. 3c perhaps needs 40. 6a tud. [The Gotama Nodhas hymns appear to be as early any in the Rigveda the refrain marks out i 59 as a later addition.] lc perhaps read vdcobhir Imahe su-uktaih. 44. 4b trayidhd 7c rat hid 9d naasatyd. ib. 7c yujia. 54. as a Vasisthi verse. cf. . 6c dyu. 192 . 1 0c 1 0a evd bddhe cannot stand 8a probably Viratsthana verse 142 ii 11a Jagati cadence perhaps dpah should be restored gdah. la. lb dditidn 5d <wa 10b ydksvd. 168 ii . *50. 13c mdhyam. 94-115 : hymns 51-54 are metrically akin to the 55-57 to the Gotama series i 74-93. abhi. 6b. 9a duih . : G. 4a gdthd-. . 166 ii. 6a md u su. 11a medhidtithih or medhya-atithih 1 1 2a perhaps needs correction . 41. 60. 2b avrta d. though against the metre 3c asidh is doubtful. the true type being found in 3b: 61. see 168 ii.may be the negative particle. : . perhaps SCT for utibhih. 38. 5d dsvavatyd. . lb tud . requiring correction being against the rhythm perhaps prd ye tdva 4d ndmd plural. 16a hariyojana.

84. . 4b Rest at the fourth place diauh . 6a suasti . 76. 9b vanuyama. lib srinanti . 16b dadhidn. and needs correction where it conflicts with the standard metre. cf. 2c requires emendation. 4a nfndam. 226 iv a. correction. 5a ksdam. 2d Viratsthana verse or read devdndm devdh dmartam. 5b Gautami verse. evdm v 6 10a . cf. . cf. 5c asmdbhya. 7b vajri. probably of the normal period. 4a irregular cadence. 4b sdmid . 83. 10b sammayd. 8c asthaat.. K. su . verse. by the use of the verb d pyd in each stanza 23b 4d avar. srnuhi as in text needing correction. I 65-73. 7c grbhdyd. 67. more probably Viratsthana verse see 151 i. See 106 ii. cf. 51 2b . 6b Bhargavi verse. 80.] 9a perhaps svdslti. 10b 65. 9b kartd. . lOd rdnia . cf. . iii 1 10a. 2a dcidhuam. . 12a sdrmd . perhaps mdrtidndm. 142 v. : 9a nu 90. 80-82. 9a cardthd requires developement of this metre. 6a. I 74-93. 6c tud. 151 iii. . lb hypersyllabic verse. vi 31 3d. 12a evd. v. 8a probably read ydd dhdn 8b viver 70. 158 iii: prthvim. : . . 169 vi. 66. perhaps read vedo vi babhruh. 5a 3a mdria. [The Pentad hymns 65-70 shew a very regular rhythm. 178. tuotdh rather than tvd-utdh. 4d 'mrtdh. 178. 142 i 7c perhaps sudadse. 178. 33 6b. 4c sdhyase. 12d siauti. 71. 175 evd. 151 iii 8a mdrtidn. the hymns appear to be [Gotama collection generally of the strophic period. 88. 73. 10b the long 9th syllable marks the 249 ii. Hymns 71-73 are in very regular Tristubh. 6b Viratsthana verses . i 58 5d 10a read purutrd ndrah. 3c Rest at the fifth place. . 3a tud. . . 5b Gautami. . . 7a bhadsvatl. 79. 244 i. but see Id verse with double Rest. and cannot be earlier than the strophic period. 226 iv b . 5b dydam 9b Tristubh verse. 91-93 being perhaps later. 5c kasthaah. For the metre. 5c 88 . 2a vilu plural. 11a tud 35. uktaih 7b cdratah. cf. cf. lc Rest at the fifth . 63. ii. 3b cdratdm. 89. . 178. 85. prindndh. . giving Viratsthana verse 2d 151 ii 3b Viratsthana verse. . 2b read havirvdt. 9b perhaps suura or su?-ia. 6c probably tarksio Wista. as indicated sahdvan. 72. 20b cand. see on i 87.] 2b. 92. 3a prehi. 15a 81. cf. *16-18. *9 cf.292 62. 5b sudpdihsi. 64. full . at least there is no other example of a gerund in -ya in the Rigveda 4b perhaps proper. *vii . 151 iii . see 224. 6a tudt . place. L. 3c vahd. but 220 v. : . *18b yajdtai is a very late form . i 72 7d. 3a Rest at the fourth place 5c bhumydh 8c Pentad verse. fifth place: dasema. The text is rather imperfect. 4b avrtah. 173 i. Metrical Commentary . 3d dhd . see 175 ii. 16-18 form one 91. 7b d nakam . 2c mdrtidya. 3a su ca. 9b ydtra 9d dyu. possibly . 9c ndri' dpamsi 151 ii. 86. 178 . 6b sumddudhnih. The collection is characterized by the comparative frequency of Jagati : and Pankti 74. 6c asmdbhya 82. la upa iiahi nu. 8d yena nu. hymn. 69. 8c asmdbhya 8d visvddhd. 9a Rest at the fourth place . 124. Id Rest at the 77. 2d visvdhd. 15a nu it. see 168 ii. 4b Viratsthana verse. 151 iii. lc siona-. dliatta. 68. ib. 6c vdsimantah.

109. 22c catalectic Bhargavi verse: correction is 214 iii. 2c netari 13b avah . 6b. combined by Sandhi with the initial vowel of a word following. 151 i ddasasya. lb jlrdsua. see The first two hymns. ni/iid. 2c following vowel . 111. The remaining hymns correspond generally to the types of the periods to which they are respectively assigned in the 'Table of hymns. 7a Sandhi at caesura. 16a Epic 106. . correction. and to a slighter extent 118 and 119 stand alone in the Rigveda in the regular use of iambic and cretic variations at the break side by side see 110 iv. lc avasdya final -d shortened before lb Pentad verse 104. aghdsua Viratsthana verses 2d asmdbhya 4c long ninth syllable. The most modern the gift of a passage in the Rigveda. 6a Tristubh cadence. 18c rjrdsua. 7a sd. 10b Bhargavi verse. the final vowel being often 116. ' . or read raydye. 108.'] : 2c naasatyd here and frequently. 145 ii b 10c dyu. probably required 23d srutia. 118. rjardsuasya. 158 i. 5d asmdbhya. 8c as 4a. I 116-126. 114. 2a svetid. recording the poet's gratitude for 151 ii. : . it is possible that some of the dramatic setting fragments are of : earlier date. N. 2c ydtra.] 94. 120. 171 v (to be added). *10-12. . and presupposes a peculiarity. cf. 5c. 95. 8a sindhum 'va. . . 4a bhadsvati. 5b catalectic Bhargavi 227 iii b. 6b seems verse. 149 ii. : : 12c sd. [These hymns are very slightly connected by 105 v. 18c nicdyid. 5a rudard. . 16b rjrdsua. 103. lb prthividh.i netari. 16b Rest at the fifth 139 i 151 ii. 19b ghd. probably to be corrected see also 169 iii . syllabic verse. 115. . *93 1-3 for the metre see note 167 : for M. 62-120 : 298 on p. . 9b tdtra . . 2a vd u. 16a irdhuam. references to the Pajra family. *97. . unless we read indrarbhumdn. 19d janayd. 7c vianti. 2b bhutd. 119. hiatus. 19a vis'vdhd. 15d ndvia . 14c utd. 6b bhajd 8a as 2a 9a tud 9d srnuhi. : . 3b mldhicah 4a rudardm. to require correction. as to djd d. 5b nd 'as combined. 2d pariddyd . 8c sd: irregular cadence. 2d dgacchatd. cf. 17b . 17c uccha. id a with semi-cadence. I 94-115. . 4a 100. 2b sidldt. lib irregular break. 4b u. 5d sd. 112. 178. 102. : 145 ii&. . 3d tvdyddbhiah 7d ddha. 7a 9b hybrid verse. place. 14c vi uccha the date see 88. 5a hepta105. 18b sdrud. 145 ii b. . rected by interchanging rinlthah and ddmsobhih . Both the text and the metre of stanzas 1-9 require emendation on a large scale see 244 ii. dyaje. 122 6a sd . easily cor14d rjarebhih. 149 : ii . [This collection stands alone in the Rigveda for the frequency of the cretic break: but 111 does not share this 105 is of the same type as i 24. probably needing 3a taksatd. 6b lb lc 117. 5b dhattand . 17c rjrdsuah. and differ greatly in metre . 2a probably Viratsthana verse (indra type) 4c Rest at the fourth place. 113. 110. . . 5b susrava. 6c drtvijyd rtviyd is a probable correction 151 ii.

10a as 3b. 12c irregular cadence. 8b -cdksia . 3b probably apdaiu . 13c istd-asva or istdma. 6c d art-. 13c adyd. 151 i. . 1 0c ucchd . 126. 242 i la ddsuantam. 8f a trimeter verse should be restored probably catalectic verses 9f Bhargavi yerse. 3a adyd . . 4b ardnl as 219. 227 i b. 5c u 124. 6c Rest at the fourth place . 4c the same 4d Viratsthana verse 5c the verse may be completed by interpreting dm as an abbreviation for d voceya .] authorship is not known. verse. 151 i. . lOd perhaps suurah. srdyinim. 166 iv. cf. 151 i. hymns are also marked by metres in which the dimeter verses greatly The preponderate. outwardly similar. . lOd catalectic verse. 2g bhadr. lb sasahydma. 8a prd-pra. but see hdrlhd. and by occasional Brhati and Tristubh verses. . 7a su-idhmd . mrina plural. 149 iii. 128 ii. [These hymns. 5f. . 142 iii a. 127. 7a Viratsthana verse rather than verses. 7djahi. 125. 152 i. 227 iii b . 5d ndyisi. . 8d vaatdpyam. 8a verse. . 4b. 8a vdsudhiti probably. . 12a Viratsthana 145 ii&. 10b nardam: irregular cadence. . 3b ydvia. 5g catalectic 151 ii. .294 Metrical : Commentary : motor-car (rdtha anasvd) the extreme regularity of the metre is in striking contrast to the preceding stanzas see also ^88. 8d. : . : . 132. 151 ii 6a utd . . 226 iv a llg read tud and omit jijanat. . . different in their internal structure. 127-139. He probably catalectic verse. catalectic verse. See . 6f Viratsthana verse. 4c siat 4d vividnd 6a vasuydntah. dual with Sandhi. probably 9a perhaps rayina. cf. . 131. la Viratsthana verse (indra type) 2a s'rudhi 3d tilbhya anavadya and read as Viratsthana verse 4b visvd-dyum . 217 perhaps bhrajdm for 2b jydyisthani caesura after third syllable viblirdstim. catalectic verse If hypersyllabic which requires correction. . 3a catalectic Bhargavl verse. . : . . . 13c prdsyd (gerund). 10c Pentad verse. 6a bhavia. 8e 175 ii hotarayd. 8c pajrebhiah probably. 7e d %m. ix 98 3d 2d dydam 3b Rest at the fifth place 3e vdnam 'va . opening. *6a hypersyllabic dimeter verse. 121. 6d. 151 i. 11a irregular cadence. . . . . % 140 iii 130. also as the feminine 4b 13a dstodhuam. 3c pdyayd 3d vardhayd. *i 50 13c. 5e vdsunaam 128. . 12d ndmud . 178. I O. . 7f. 142 iii b. . 15d suurah. with double Rest. 6d sd. 15a probably ddsit. 168 iii. 4a upa. . 15c bhajd. 123. verse requiring emendation. 8c irregular cadence. 169 iii. lOd probably dhabhih: Rest at fifth place. 8d mdhya 9d hotardbhih is possible. 6d supply srotu nah at the beginning of the verse. 5e vllji. 149 iii 5c probably Viratsthana 149 iv 6c the same. 7d pasu-ise'. la pdtaram is possible. See 242 i. 6h pdnthaam . 139 i. 15c irregular cadence. 5a nama. 8a adyd. 7a extended Viratsthana verse 9a. 3b vaydh. 6e catalectic verses. 8a jydyasyai has probably replaced an older form jydyase which served cf durvdsase vii 1 1 9a . 122. If omit 129. see 8b sdrvdsaam 6g. 6c mdhyam. 7g catalectic verses. plural is probable. . : . Id tud 2a Viratsthana verse (indra type): svandm. are very 127 and 129 are of very archaic The later character: on the other hand 136-139 are very regular. 2b two syllables are wanting. 6g srudhl. la pdantam. See further 201 ii. cf. 9b ydhi 227 ii b 10a perhaps rayina 11a pdhi. 151 i.

verse.vdrund. la pdantam. la Pentad verse .. 136 i. 8d madayddlami probably) with irregular Sandhi idbdakdm. cf. . 7c devdnaam. lb visva-pusam. 3c anapdvrjyd. . 142 iii. 144. 9a dadhathe. ii. 6d asate. 148. 155.. : . cf. 3a rodasoh. with Sandhi 142. 143. 151 iii. 13c perhaps to be corrected.). read as 16c. 8b slrsanya (y cons. 9a yayw. unless irudki is a -loss: Jmlrhi 6f hypersyllabic verse. ji6k. 8d dsathe. i 162 22b 2c Gautami verse. la simid. 167 i. . 178. 135. vol. lb Gautami verse. 4d 7c didiat. 5a. 146. 139. 5d Rests at the fifth place. 8a Rest at the fourth place yaatdh. lc aucathid probably. 3f eakfayd 135. 2f vdhd. verse. 158. 4b Rest at the fifth place. See 107 140. 3c suurah. . 142 iii b. lc indravdyd. 16c contains no trace of rhythm. For the metre see double Rest. 15a tud . 5a sahasia. . 151. 4a havisydm (y cons. 2a the semi-cadence here and in 7a indicates 168 ii. 226 iii a. 5d probably Viratsthana. or (more 130 iv. 2f Gautami verse. 5a perhaps add yam at end of the verse. xlvi. rather than candard 5b Viratsthana verse 5f indrdvdyu 8a vaJtai/ir. 8a hybrid verse with irregular opening. yufijathe. 150. 157. c utd. p. 4d Sandhi at caesura. 3d nd ha. 3b Rest at the fourth place. 154. 2c wrvidk 4b yujia. 175 i. j 170 i. 151 iii. : . H. . purd adverb. 7a probably snuht ma place. lb I. 2a Viratsthana verse 153. 295 *l-5 see 133. lie rasmimr va 141. 156. 5b abhikhydyd: Viratsthana la vedisdde. dsathe. lb. cf. 5c ddslt. 14a bhumi' dgnih 1 2a sammi lyd *162.). 174 ii 4d gdam. 6b somdnaam 6d at the fifth place. 142 ii.lrrf. lb adyd. 7e . 2d perhaps irddhiai. 5g asurydt. 3a Viratsthana verse. 6b mitrd. is I 140-164 [The collection 166 ii similar in character to Mandala vii. satd-atmd 4a Pentad verse . time. 6d utd. 3a Viratsthana verse. are missing. 8a mcf sd lib the insertion of . . ye at the beginning of the verse P. . lc ddha. : . 6a hypersyllabic. Oldenberg in tiacred Books of the East. but see 151 i. is \ . to j ujuvah or dhljuvah. 4c apdam is doubtful here 5a verse with 242 ii la dctfudn. 170 ii/ 136. 134. 167 iv. 3b jlrdsuah probably. . 149. 166 i. la jagdma . 3d see 178. 160 i 145. 226 iv b.] 4b ujuvah requires correction. 2a" Rest at the fifth place. w. 4a asurd. 159. 4f Gautami verse. . 88 *lc ahhivldgyd *2c chindhf. M ! . 3f hybrid verse. is metrically desirable. perhaps 10c probably avdsia. gdvyam ydnto ydviam dlrghd aha.i 121-*163 . 170 i. njdsaa. lc nindima. : . 5a dpia. . . 3d ft&fe 4e adyd. 14b probably four syllables adhdyi. as to : } . Id trayidhd. 14c Rest at the fifth 12c asapatd. a transition to Epic Anustubh . . 144. 6a esaam . 175 ii 4c sd'. 3a nu . 7b evd. lb sd. 13a abhi. and needs correction: read ydt samd&nam ydc'ca pddtnfam <irvatah. 224. 8d adyd. *161. : . . 2c Pentad verse. 13a simivadbhih. 10c nu 151 ii jlrd-asvah 12a sudiotmd. 174. 4a sii 138. 2c tnah the second 147. He i syllable wanting. 146 ii 3d catalectic verse. lc perhaps restore sdnavi: delete the comment on p.ih.

. very doubtful. Id dhattd) 6a sdhyasah is probable. 175. 10b. . 166. 147 isdam. 178. . . 14c varttd.) is lc Rest at the fifth place. . and more probably we should read aid and then svadhd ah 7a bhuri 158 i. 4c cdkdnanta. 38c perhaps 166 iv 40c addhi 40d pibd 45d hybrid verse 48c Rest visu-ancd. 35b Rest at the fifth place. i 181 8d 6c Rest at the fifth place 5b perhaps gdvah. Benfey . la sudh 4a catalectic verse. (plur. 186 have the archaic Rigveda character more strongly marked than the rest of the hymns. 4c vdhd: catalectic verse 4d vdtasya dsuaih. 12d drddhuam . 5a vdhd. prdyistha. 178. lc marutaam. 8d yddl 9b nardam. 4a extended Tristubh verse 180. 10c perhaps Pentad verses: otherwise nardam. 178. 5c u. . 4a aydasah 173 ii are doubtful. ii 473. 7b ksadm. 4a nu perhaps sdsmi. 3b dydani or didm . 13d eva. . la Viratsthana verse {indra type). 10a tvdm. . 3c Rest at the fourth place . . : . 151 i and yavlyd 178. 165. probable. \ . 3a sddmd 4a requires correction. 23d hybrid verses: in 23d perhaps omit it. Viratsthana verse lb omit u. lie perhaps tatrsdnd. 2c Rest at the fifth place: anavadya drndh . 4a devayd 'yam with Sandhi. la catalectic 172. 5a sustutah (nom. and O. . 167. . . . . 225 4c double Rest. 13d bhutd. 8c bhuut. at the fifth place. 2c Rest at the fourth 169 iii. . 8d cakard. 174. 130 ii 2c. 8c Rest at the fifth place. 12c perhaps yavlyd. 14b Viratsthana verse. . 226 iv b 4d Viratsthana verse. ix 22 5a 168. 32a sd. : . I 165-190. lc aidhe 'va. 2c apdam 3a trptd-arhsavah 5c dhanua-cyutah. . 169. 151 iii after Th. . 142 iii b 5c su. 16c as 7a. 168 iii. or read <Iii ma 5b vdatasya dsvd a. : . Id yudhe 'va ib. 13b purusdrhsa appears to la Viratsthana verse {indra type) be a proper name. 151 ii. The late hymns 170. 9b Viratsthana verse. 6c cyavayatha.296 *164. 6c ddhd: hybrid verse. 3a Viratsthana verse {indra type) 6a svadhd (final -a shortened before following vowel): but dsit is a very doubtful form in this hymn. 226 iv a mitrd-lrun Grassmann but emendation is rather required. 17a. 5b probably Rest at the fourth place. . 176. 15c a dative vaytiyai seems required. Mana collection. 1 Od nardam. . . . 171. . 9d caesura after the third syllable. : : . lc Rest at the fourth place: raydyah. collections in the : of its ascription to Agastya. . Bollensen O. 160 i. cf. 169 iii 12b avaydjah. sprdhdam. 151 i 7 c irregular cadence the resolution of o in ksonih is but see 151 i. 5b Viratsthana verse: but see place. cf. Metrical Commentary 7 a i. 3a tisthd 4d nisddyd: hdrlhd with Sandhi. 177. lb hybrid verse. . 139 i. 2c ddhd ydd . 3a nU 3b vriiktd. [One of the most homogeneous but 173. 3b ksitlndam 5c catalectic verse. See 106 iii. 6a double Rest. but see 11a Viratsthana verse {indra type). or perhaps vdatasydsuaih with long fifth syllable 6b mayo 'va. 13b ydtand. . giving a catalectic verse lc rodasoh. 142 ii. 10b nu lib snitia: cakrd 12c -cdksid. 173 ii. 5d gdtuyanti. 2b brhddivaih. Qhdyayd. 227 iii b . . . . 173. .). *170. . 6c Rest at the fifth place. verse 2c catalectic Bhargavi verse. 179 have found their way into this collection in consequence Q. 7b Rest at the fourth place. 174.] . 128 ii. 3d Pentad verse. . 8a padnti 8b aryamd 9a nahi nu 10a adyd: 10b sudh.

11a. syllabic verse. cf. : . 7a sudh sudya. 2c Pentad verse. 160 i. such very regular. 7b trayidhd. 2d probably devdn as gen. 13a aaditydsah. 9a nti 9c ddhd . ksesidntah. cf. suydvasah 7c Rest 'raks-. 10c vdatah. with Sandhi at the caesura at the fourth place. 6a tvdsta d. 8c pipdya . lb adbhidh. 11. 5b mdniebhih or Rest at the fourth place. 182. Tristubh verse. 5. 6c krsnd-adlivd 6d diauh . 6a supra-etuh. 22. 19. giving an extended Tristubh verse. 5b anarvdnam. 4. 6a perhaps tatysand^ 169 iii. In 13c there is a further disturbance owing to the words are asya being 14a heptasyllabic verse. 9b pipdya: brhddivesu. 2. vahathe. 3a isukfte 'va. lc s<mta with -a shortened. 8c caesura combined with Sandhi . 14c suadanta. and jlrd-asvah. 6b vayie 'va 6c catalectic verse la -diotmdna?n. 11a sahasia.] 151 iii ta dvidhat. 5c catalectic verses 6a prthvi. 2b tud. 7d nardam. see In 10c the 194. lie valid. lib it. 7b suayuh 8b Rest at the fourth place 8d su-apatydm hybrid verse 9c Rest at the fifth place. : . 2d drum (dual) with hiatus 5a prati-dncam. 9d perhaps jrrusaydnta. . 42. 11a wo We. 20. 6a sadhu. .cadence *8. 7d probably ddma d as in 8a. 6. . 3c valid'. 178. 3a. . 4a suydvaseva. . 7c jdnyo 'va. la may be 187. 151 iii. extended Tristubh verse. 151 i. [The collection as a whole is of the normal period. 6b probably svanista. 16d hyperplaced earlier in the verse. 3b Rest at the fourth place: (cf. 181. 51 2b). 4a tie. *191. la. 7c dtra . la Rest at the fourth place. 5c rjuyd. : . 2a apdam. as 4. cf. 2d didiat. 5b caesura after the third syllable. la adyd. 6c catalectic verse. 8d as 180 5b. lib probably need 5a we should probably read vibhuh prabhfih. 5c apdam. Maiidala II. vardhatd. 8d no caesura. and in lOf the unaccented tud is clearly out of place at the beginning. lb ucchdntidm. 3a Rest at the fifth place. 1. 3a yuyodhi 4a pahi . Bollensen . 151 iii. 4a mddhui. see p. 188. 170 ii c. and the metre is The few archaic hymns. 6a ghd tudvdn . . 168 iii. $ 168 iii. 7d an 190. 3a dsathe . la prdyistha isdam. 151 iii. both in the opening and the break. vi 11 3b. lc u sU. 41 16-18 seem to represent the original collection of the Grtsamada family. 151 ii. Jagati being prominent. 7d catalectic verse. pi. 4a. 166 iv. 173 iii. . 4 : . 2c 170 i. 2c Rest at the fifth place puuh ( 151 i) 189. 214. is more than doubtful . 3. 3d bhlmd irregular verse. 142 v. Id vdsudhitl. 9c probably sudsviam. 10a prd u . 135. 1 42 v. 9c bhdrl cid perhaps as adverb sudads-. 5d Rest at the fifth place.i # 164-iY 10 2!)7 vdatah. la ndyd . 2a u. . 8c adyd. The metre here is disturbed. 10-13. 10. 185. restored thus: pitum nu u astavisam. 186. 6a Epic Anustubh semi. 184. 2a sS. . emendation. 6b devdnaam. cf. See 109 ii. 169 145 vi. 150. 9. Id sudads-. 183. yet strange in this hymn 142 iii a 7b srutiai. . 5a vasavydm. sense requires that no should belong to the following verse. 2d bhaasi. vi. 5d krdhi: su-apatydsya. 166 iv 5a extended . 174 ia. : .

4c asurydm. 27. 217. 2d Rest at the fourth asi with Sandhi. . 19a. . cf. 20a as 9d . App. 151 ii 20. 13a vasu. 5b perhaps insert vi before durah. 9d asurydm. 14b kdrmd. 7b asvaristdm. 4c brhadivd cons). . 175 i. 2b utd tritiyam ndva ayoji 2d sd. 6a perhaps saptati. 8c jydyisthe 8d jigivdn. but not certainly. . . Nos. 7c vihdvia 140 iii: Viratsthana verse. : : . *7c vispdtniai. 17a omit it . 25. 5d. 29. dmartam 3a it . Id here and often rudard. 13c dtia only in 17. 213 ii: dydam. 5c sdtvd. 5d secondary caesura. 5c Rests at the fourth place. 2c perhaps should probably be omitted nU 3b rudriyesu ca requires emendation. 2c bharatd. bhydsam. stanza properly belongs to 14. 9b vdsu 2a asuriya. . 3d perhaps dhnaam. but see 79 and 84. lib ksdmia. verse (indra type). 5a mimaya. 4c jigivdn. For the metre see Ch. . cf.298 11. 19. 135. with -u of dual shortened dditidh 10c rdsvd 174 i c. 3d urnutd 5b. 6c perhaps sayddhiai. 9d. la perhaps read 18. 6d Viratsthana verse. 17d ydhi Pentad verse . 8d hatvi is suspicious is probably required. 6a as 4c . 6d u. 4c arpaya. . lc yaati. 5d dhattd. 2b Tristubh verses. 8d krdhi. la tdsidh. 34. 7c usdsam. 170 ii/. 130 ii. 8d adyd. The cadence is probably. Commentary : 250 i. 7a Rest at the fourth place. 10b Tristubh openings. : . 13c dvr/iita. Id juhotd. 5a 13a catalectic verse. 151 iii 7d Rest at the fourth place. 15. . 4c rdksati. 21 concluding verse borrowed from ii 19. 146 i. 5c utsndya 16. 151 iii. 13. 175 i. 4a sd. 14c krsvd. 5b apijti. la dadityah 151 iii. . as nrdv dristdh sdrman d siama. . 178. 6c probably 13a suydvasa. 7b bhrindnti. lOd Rest at the fifth 15a vydntu in nu . navati. 5d d sastid. a form probably invented by the 149 ii. cf. 169 vi. pro5b sdrma. 7b perhaps dravasid. requiring emendation. Id vasuyu. 178. 9a d 170 ii a. : . 23. 31. the omission of ca being the first step. 26. 15d sadhu. 145 ii c. 151 iii. *5d catalectic 32. 145 iii. 5a requires emendation 5b probably sdtvd. 6b stavd. 8c for the caesura see poet. . ix. lc didhiatah 214. 178. i 1 40 2c 3c md sd 4a dudbliydm 5b hypersyllabic opening. 4b rayim va. 8a tanunaam . 10c jiok. 87. . 2b asdnl W. (plur. 219. 170 ii. . 5c utd tastambhudmsam.) bMvltud. 12b Tristubh verse. 7d abhi nu . 5c usdsah. *5b. 5a catalectic verse. that of Jagati. 142 iii. : } . be sd Wydh. see 167 v . 139 iii 14. 4a dsmdsiam 24. 227 iii c 3b nardam 3a double Rest of an irregular type. 3a simivan bably plural. . . place 5a sd. 6d on dvrnak see 169 vi. 3a Viratsthana lb svdndsya. 2a Rest at the fourth place. . lid srutia. 4e probably omit indra. $ 139 iii. . 7b ddaslh ( 151 i) is here unlikely 8a tavasydm. in so early a hymn by restoring ghndn the metre becomes correct 22. Metrical For the metre see . Id hhivant. 135 a. 6a dpa su ment. 8d Tristubh verse. la sti Id no caesura. 2a vihi 4a dvidhat. . 5c prthivl 7c irregular double Rest ( 227 iii c) 168 iii. 3a. . . 135. 33. 5a ksipa 9b abhikhydyd 10b vlrya (y 30. . 16d the verse seems to require rearrangebefore a vowel. 146 ii . 12. 13c as 9d Rest . . see 213 ii. 13b vamvydm. 130 i. . verses . if we 226 iv 2b -pah read tudbhih but more probably a double Rest. 7 b mdrtiah. . 15c double place. 2c Viratsthana verse. 28. . a change of accent 135 8b Gautami verse. . Id. The last 'diah. 17b as 9d . 17c prindnd . . : . 178.

. 5d dditydn. 6b 39. 35. 9c nu . . 38. 5a sdrmd 16. 14. 11. See 1101] lc didiat 2a cakrmd. Sandhi. . : Mandala [This collection is III. . 16a dmbitame is probable. 15a sudhii. . 8c rathio 'va . lb nayatd 3a diauh 6c m^d probably Viratsthana verse 128 ii but prdk or prdcd as an adverb is more 10c prdcl 'dhv-. 5a avyath\ ix. lid sd . 7b devebhiah. 5a ddria. 7d tiibhya. and so frequently in this hymn. 34. 15a pnrutdnua. 94 viii 5b perhaps 6a plpayO. 3c vdstav usrdh. 7. 5c stokdh scotanti te vaso is a probable arrangeJagati cadence. 'va. 2a aviratdya (neuter). 227 iii a verse. . la nicdyid. unless we correct by reading 1 30 iii but indra tvd?h nah. 4c vahd. 151 ii 3d ydthd. 9a bhard. 20c extended Tristubh verse. 2b tiibhya 167 iii. : hymn and Mandala 135. . 6a tudt. Id mahdbhih. 6a probably abhlyujah. lid msi^ (plur. 2. 42 . 16b nd 'not' with hiatus. 13c sd pdm with Sandhi. 145 vi. 30. . 5d gdnia 6a. nd not with hiatus . 6. 142 i. tdpd. 6c srjd. . 9d prd 'nam. : : . lb catalectic verse 7a rdsvd.). ment. id u. 38. 13d karma (plur. and see 19. 10c i^d. . Id sasanvdn. : 9b itand. 167 iii sd. . 178. 8a sidd p. la usdsam. 4d nimite . 17. 33. 6a vavrdja 15d ddmia 16b dhdnid 1. 4. ' ' . but characterized by a fairly frequent use of the VasisthI verse. . *43. . dhitdvanam. 17c matsud. . 2c d. but perhaps rather extended Tristubh verse. bhdrl cit probably. 9b siasva. lc srnuantu. la sd and perhaps ydsmi. ' : . % 178. 21c divdksd 'si for the Sandhi see 130 ii 2 Id sd. 3b indra-vdyu 4a perhaps mitra varuna . 2a /?r 148 vi. : . 15. 10. 9d. : singularly homogeneous it is also very regular. . 5a jidydn . 20c suarydvah. la grdvdnd 'va with 170 ii e. .ii this 11 lit 39 298 6a svdr. 3. 36. 10a pibd. giving irregular break (218 vi). *13a Epic Anustubh semicadence. 10b abhi. 13a yddl 13b rodaslyoh. 167 i . 2c hybrid verse apitid. 173 ii . 3c scf '(/ne with Sandhi. 15a dfhyd. . 147. 170 ii 14a vfsd 15c didiatam *28 see p. *42. 3b prthividh. 171 v 8a kdrmd . 4a Rest at 27. sdrtava u is probably the right reading. probable. 18. . su-apatyd. 7c asuriyam. . . 3d bhavd. . 8b sasanvamsam . 2b brahmaputro 'va . 6b kdrmd. 13. 42 *29 see 4b prthividh 6c perhaps dnlvrta. % 170 iia. 7a dpia. 20. 18a brdhmd (plur. 3b siksd. 35. 2c perhaps 26. 7c wanto 8. candrdm . 3a upa. 5. 'not' with hiatus. it calls for emendation. lb dpdsid. 9a srayiiiisah. the fifth place. 7h 36. 10b gndaspdti. 9c nd 37. 5c dsate 7a 41. : . followed by short eighth and tenth this seems more probable than reading aaditydn and then syllables 21. la extended Pentad 7c. 7a su-apatye: dyavi is probable. 22. . 23. pipdya. 3a krauda. 24. : . 2d nd . . 4a sdtvd. . . 166 ii 16b devUame as 16a 17b devidm. 2a tdbhya. lc bharatd. 3a adyd 7b dsvavat 8a. 7c s^. For the metre see 213 ii. like 24 5b above. 8b ddmia-.) 31. 23c sicf. naasatyd 12a dsabhiah .) 18c mdnmd. 39. 2a ardnioh . 6b for the possible Sandhi see 32. 8a namasydtd. 3c abhigurid.

la usdsah. 20a. 8d ddsudmsam 10c asvi are interchanged. . : perhaps Mandala IV. 2. 59. u. 6a 41. 3b didie 17b s<f 'nyasmin 56. 151 ii. . 53. 5c me 'rsim for the Sandhi see 1 30 i 7a pibd. 2d %%. 4a sdmiai. 175 i. . 6d rdsvd. 75. 2d 224 Sandhi combination of all the words in the hypersyllabic verse. le janatd 224. 19d caesura after third syllable. 145 vi. . 135 a 15b gdah. 2c jyay istham 3a vavrtsud 5c yaksvd omitted. 2d no caesura. 4c rdsvd 7a pusan. 14c siidayd. . 62. except so far as the metrical standard varies between those of the strophic and normal See 109 i. 48.] periods. 151 ii 5d mfllkdm. 4b probably usd d ydti. . 5d 6d probably a Gautami verse 6c ndasatydya yakse. Lyric verse is rare. . 49. See p. 7d probably aridh. . . lOd diauh. 19c sucandrd. lb tujia. . 6b apparently jahndvidm.300 . 62 2a hypersyllabic verse. 4d dhanad. on account of the rhythm 13d nd ha. rather than with rudardya but see 149 ii lib Pentad verse. 1. 40. Id vidhya is probable 4. .). 5a -didvd. . . . 178. [This Mandala is also of a homogeneous character. . . 18c mdrtdnaam. 5. . : . 45. . 5a vidntah 5d ajanata. 20b visvesaam. . 6c tud. 19c Rest at the fifth place. ca. 47. lie su-apathotard. 4a utd 54. : . 3b tri-udhd is not 18a. lc mahind. . 142 ii. first part of the verse is unlikely. 170 ii. a heptasyllabic verse being improbable in this collection mitra varuna. 2c hdrioh. . . : . Ludwig Gautami verse. 175 ii 58. 3a srjd 147 i 175 ii 14b as'idma. : . 43. 3b sumflikdya in spite of the metre. 2a mdhya. . 3a upasthdyd. 135 a. . *52. . . App. 12d paantu is doubtful. 5d matsud 6b sdcid. . 7d dadsvdn. 43. 10b rddhdnaam. Metrical Commentary . 170 ii. . lb yaM . 44. *22c Epic Anustubh semi-cadence. . 12d for the Sandhi see 127 b 13a manusyah (y cons. 9c dilria. 46. but see 177 iii 7c tubhya 15a dyu 16a abhi should probably be read for d. as 5d. . . . but seems probable. : . 12c 9d probably Rest at the fifth place: pipdya Gautami verse 13b ma for md before vowel 13d Pentad or Gautami verse 14b Rest at the fifth place prindndh 14c rujd 16a tubhya 16b ni?iid. Nos.). 61. unless ague be 152 i: vavHsud. : : . 150. lc ddmia 4d prthivl 17b bhdvathd 20c aadityaih. . 3c tud. 2a usdsam. 51. 14c ihd. 12b duria 12c dfsia 13c bhard 149 iii. H. mandasva. . . Oldenberg 2d. 17b devidh 17d rju (plur. 151 ii. with irregular cadence. 3. 147 11a pretd lie extended Tristubh verse 16c catalectic verse *17c pdtalye (y cons. . . . la perhaps ydsmi. . and 83 . ix. lc Rest at the fifth place: drundndh. 60. 5a dadhisvd. 18b f4& 5d sddayd 57. For the metres of stanzas 13 see Ch. 55. 178. 2c jahi. 214. . required by the metre. 8 b sdm devdtdtd. . . 170 ii c. 2a ma. 178 2a raz. : . A. 8a bharatd. 2b haryd 3a dhitdvanam. valid 5a the cadence becomes regular if yajndh and 142 v. 2d the verse as it stands has a Rest at the fifth place and a long ninth syllable but sukrdn should probably be corrected 4d u. 2a.).

39. 301 pasupdah urdhudm 3c navajdah. 5b ha. lc perhaps viruructih. 4c nahi nu . 5a sdcia 'kartd. 142 i. 5c sdcyd '/carta. 171 iv: ddtard. 3a. 151 ii 3d sdrud. 149 iii. 8b ndsatia. 7a catalectic verse dasema. 4a probably abhi. 5d ca dechd 6b srdyinibhih. 16. 159 iii. 2c perhaps ib. 5b sfma 4c pdah This hymn contains several Viratsthana verses. improbable durdt and indrah may well be interchanged 4a u sU . 2d vicakrd. 29. . tuavatah. 2c. 3d didm. 6c tve 7a tvdm. 6b evd. lid caesura after the third syllable. . The original may have run somewhat as follows pardvdtah sakuno mddhumantam somam bharad devdvdn dddrhdno diva addya uttardd amUsmdt dddyd somam abharac 168 ii. 130 ii . 2d perhaps tdrutram. ksdah. iavavrtsud. 9. . cf. . 5b. 4b caesura aiter the third syllable The 26. 4b 23. 35. Sandhi. : . 20b probably Rest at the fifth place. . . 6a bhuydma su 32. for siendh 145 vi for devdvdn see chiendh 7d Pentad 3c jydm with y consonant stands here alone in 27. 36. 140 iii 4a indravatah. 17. . 5a yddl. la the resolution duurdt (151 i) is 20. 14b matsud. probably plural. 5b extended Tristubh. 4c. l&'adyd: Rest at the fourth place. 3a devd 178. 2a veda perhaps vdsudhiti. 151 i. 14a irregular cadence. 15. For the metre see 12. . 10a deva with -a shortened before vowel. 167 iii. 6. strange in this hymn. 169 iii. \\ . 3b mahdbhih. diaiih. 178. 4b vdar. la tubhya. 19. 11a ihd. . . lc Rest at the fifth place. and probably isdnah. 7b citana. 15a extra verse belonging *18. to 14. 142 i: tdtrddnd. dual with beginning of this verse 6a yddl . unusual in this Mandala namely 6c. 178. 147 i 4c perhaps 10. extra verse. 169 iii. 6d pdtd. 6b sasanvdmsah. long fifth syllable. . namely in Id. 160 i 7. 10c abhitid . lib hdrioh 9a. 3c krdtud the Rigveda proper. . 8b apdam -tamah . 5c Rest at the fifth place. 7b irregular break. 8d extra 30. 4a su violent disturbance of the metre which extends from 6b to 7a must be due to some corruption of the text. fifth place: vasuydntah. Viratsthana verse. 24d verse. 174 \a. 139 v. 2c visvesaam 3a sddmd (plur. Id agmata. 9d ahuanta. . 2b vdstav. 2d rdsvd. 142 i. There are many irregularities in the break. 10a caesura after the third syllable: but. 9b catalectic verse. : | \ \\ : . 6d tanul. 20b dsiat 31. 22. vrnlsud. 2a stavathd . 168 ii. but very rare. as this 175 ii. 28. 6c evd. 38. 3b yddl verse. 4c pibatd . gddhia. we may perhaps read ye medhdyd hdri ulilui. Id rdhidma. 21a nu u. 249 iii. 10a.). 2a durgdham etdd with Sandhi. . 3d agriydh utd combined. 4b utd. 175 i . 37. 7c. 3a tud. la mf[d 166 iv. unless we transfer d to the 9a sukrtd. which are 21. 6b bhrdtardm. 142 iii. : . . 4c 158 iv 9b rju-dncah. 8a dhisdndbhiah. 4b 150 ii 8d Sastd. 7a. lie gddhia. la su. 9d jahi. 4a 149 iii lc ksetrdsdam.' 33. 158 i. . . 3b perhaps yuyudhuh. . 8. 130 iv. 25. lib tud. 175 ii 3a long seventh syllable 4b tud . . 158 i. 23c sobhate. 5d perhaps is probable. 7a duddasa. 3a abhi su. 5a avadydm 'va. : . Id diauk . 4a cakra. 4d see 178 . 2d tuena twice. 175 ii . is 34. 11. 3d.iii 10-iv 39 . la utd. 170 i. 12c mrnd 15a Rest at the 168 iii. 169 ii 11a trs4. 23a Jcanlnake 'va. . . : .

3c veda. 51. 4a tud. lie devebhiah. [The hymns in Anustubh with occasional extra verse stand alone in the Rigveda. . 169 iii .302 Metrical Commentary 169 v. 4a bhuut or bhuvat 4d probably we should read mddhul and omit nah 5c mddhul 7d naasatyd. 44. 145 vi. 7. verse. 13. only a few bearing indications of early date. lb perhaps 151 ii. . *6c surabhi (plur. . lOe cata8b svdditir 'va. 5. bhdsmand. 2a u . 5c Pentad verse. 6c extended Tristubh verse . 7b perhaps verse ucdtliaih is an attractive correction. 167 iii. la perhaps rtaydvah. cf. 2c vdsiulhitl. . though against the metre: see 178. 169 iv. : . 8c pdat la: the latter part reads prthivlhd jydyisthe for the 56. 11. : Mandala V. *57. 135. 147 i. 5a hfdia 9a kanyah (y cons. 1. 7c 10. 2a sd. 170 iie. asmdbhya. lOe sasahidt. . 6. 159 iv. la 2c Rest at the fifth place: 4c -tdvituat. : . 1. 147 i. 3a. 6a prd u. la hotara. lOd asu-dsviam. 7c hiri-smasdruh. . p. 15. Id ydhi . 5c 6b dpia . 8a smd 151 iii 7d hypersyllabic verse. 11a indara. 2a kutra. Id pdnthaam 173 v. . 10c siondt. 10a sasanvdmsah seems necessary. 4d indrdvdyu. 3d sdcid . 43. . 151 ii. 109 iii. 2c Pentad verse. . 5a. 55. nu u. 3b etc. 135 a or kanidh 'va . cf. 151 ii . la rdstardm. 166 iv. 46. lib indrd . 9. 149 iii 2a mdhya 2b asurydni. 166 i. perhaps read asya 'sail vd u.). : . lid catalectic verse. 9a vaam. . 7a Pentad 4.). The trimeter collection. 2b adyd 166 iv. 12d uru-dncam. 147 i. . *58. 5d tubhya . 22. 3c tud. Id -ebhiah. 2d mahdbhih 3a dhdyistha 3b sasamanebhiah 5b vrsabhd 'va 10a rdthia. 3a va 'sd u 17.. lc perhaps candrd (instr. 8a vasavydsya . 19.2 Trochaic Gayatri 5b vayUnd appears to be a gloss on 152 i. lectic verse. 12e extra 11a adyti. 149 iii. 40. 2c tue. perhaps maghdvanah 130 iii. lib tistha. 5a jusetlwim is probably a gloss 6b tud. 171 v. 45. 10a bhdri ndmd. 4a perhaps brhdbhih. : . . 149 i . 3a aridh. 3b yuiijathdm 147 5b nti.varund. 2a dadsvatah. and appear in the main to constitute the original Atri In these hymns hiatus is extremely common. 5e -sthdah. 4c it. 48. 3a maghdvanah or 16. This hymn 47. . 2. 41. 129 ii lOd. 5d catalectic dimeter verses. See 105 i. 5a smd. See 3b due . 5a ml. 5c rasvd. 50. 2a citra with -a shortened . 5d Viratsthana verse. . . vibhfimham. 5b mdam. . lc long ninth syllable. 5b 166 ii. 43 10c nay aid. 151 iii. 42. 191 iii. hymns agree closely with the second Mandala in character. 7a abhi. 10b irregular cadence: perhaps read ndma guhid 174 ii hotdra. 12. 10a evd with hiatus. unusual. : . 9c utd pupuridh. . 2d tanudntah. contains several catalectic verses. 3. 4b pdanti nfndam. .). 4a smd-. 168 iii. indrdvdyu. 178 6c uhidthe. 8. . lb trdsithdm. . 5a. the Sandhi is 5a nu u. 3a tud . 4d yesaam . 5c. 6a nu u. 84 1 . 3c probably vdrenyasya with y cons. . . dudrd. 6b sudyai. 3c yesaam . . 5b probably avari. vdstav. 21.] 178. 5e extra verse 5b dsvdnaam 18. 52. 4d perhaps indaram. vedlsdd. . 170 i. 6b as 2d. A . Sandhi see 128 ii.

3b dsan. 3b catalectic verse. 5b probably caesura after third syllable su-uktd-. 16b perhaps mahaah. 4c yddi . . 7d 2d ydcchd. *7a. : . Viratsthana verse. tiibhya 151 ii. This hymn contains numerous Rests. 7c Rests at the fifth place. 29. . 10b catalectic verse. rdthasas pdtih. 32. 6a evd. *14b . . 3c pathesthdam. 23. 40.) read evd na indra dkavdbhir uti 10a Rest at the fifth place is probable. 42. 5b cdru (plur. . 3b mahtnaam 5a eta 8c asaam 9a yaatu. 35. . *15a. . . 147 7c dtra 9c uMt with final shortened 151 ii.iv 137 iii. . verse. 166 vi 6 7a perhaps (y cons. 2b mdrutam utd with Sandhi. lc etc. 13c gndah 14b $ 174 ic. 9a requires correction. 4a utd. . 25. 5c vidmd. 3c vidnii. 1G7 v. . . final -u being shortened before the vowel. S 24. 174 id. 149 ii *14b. 4d apparently daasdsya extended Tristubh verse 5a as 4a 5b yaatdh 6a -enyam tue or Viratsthana verse. 6d vahd. 148 vi Id Rest at the fourth place . 5d sugopdah. 139 iii. 8c utd. . 151 i. 242 iii. see Rest at the fourth place 16b. cf. 15b yuvanyun is suspicious on account of the 4d sumati. 142 iii 6. 142 i. 139 i. 43. . *15c hybrid verses. 128 ii. 34. 8c ahuanta. 3b vdstav usrrfh. 36. 2b ca with hiatus. lid diaiih. : : . 4a ydsmin. lc trasithdm 3a ydyistha (-a dual shortened) 4d dsu-asva. and 4 should probably be read throughout. 4d probably Viratsthana verse 5b srutia. 48. . ix 70 Id 49. la nayituh. 3d mi tubhyam id. rhythm: perhaps yuvayuh . 3a probably sdktivah. 13a kathd nit 14d perhaps tdviseh. viantu 8b Gautami verse. This hymn has many archaic variations. lOd ganta. . : la didhie. *14d suasti gives a better rhythm. : 12b suasti.. . 7b area 7c. 2a 142 iii b 50. lc sunotana. . Anustubh semi-cadence. . . words with resolved vowels. la visidnn. 3b suniatid. 5a 170 ii a: perhaps uruyd. 2b Viratsthana verse 2d. 130 iv 2c rudardh. vfsann. . 6b sunudnti. 3b bharatd 4a bdhu. . 6a sataddvani. . 6a perhaps 30. 41. *15b. 2c as lc. 5c. . 44. 168 ii rodasl 'bhe. . lb usdsam. hdrlnaam. Oldenberg) 3a sardhd. 6c before vowel. 9b su-aitavah Pentad verse lOd Rest at the fifth place. lib perhaps raydye. 149 ii . 37. : . 13a mahatah or 178 14c candrd-agrdh . 2a dravd . . 14d see 170 iie. . 4a Viratsthana verse (indra type). 12c probably Pentad or Gautami verse. 31. 2b hfdia 3&iraya. 13a Bhargavi verse. . 15b ndvia. 16c decha-uktau 16d Viratsthana verse 16e Rest at the fifth place. } . 51. 27. 4c caesura after the third syllable the verse becomes regular if tatakse is placed at the end . . 5b heptasyllabic verse. 18 borrowed from v 76. and examples of hiatus . 2d prd Widh. . lc Rest at the fourth place 2b catalectic Bhargavi 227 iii b. 5b irregular cadence. .sthuna iva . 17a extra verse to 16. la Viratsthana verse (indra type) 5a tud. . 8d sd. 7d ha 175 ii 8b Viratsthana 10b apdam or verse. . 225. tudt 28.). 1 9a Pentad verse. . 227 ii b the resolution adid can hardly be justified 13d rudardh. 13b md. 6b gndam. saptd-asvah. 38. 4b sahasia. *9a Epic 177 iii. . 170 iie. 145 vi. *15a gdvia. 45. but see 33. 6d tuvl-. . 2d su 8a indra ehi. 178 (H. 170 iii. . . 46. 4d tvdstd with hiatus 8a gndah. 151 iii. For the metre see 40-v 51 303 lb perhaps prasdhd. 142 i.

7b sudm. 142 ii . v 74 la 151 iii. 81. 5d deva with -a shortened. Similarly in la ku is a probable correction for ke. 129 ii. 3b yunjathe. 170 i. vadd . . 2c Epic Anustubh 85. 2a su 3b mitrd rdjdnd. . 145 ii b . 43 . . 3c yesaam . 80. lc 2a. 14c vrstui. i 5864. 2c ucchd. 2d paanti 3c marutaam 4d pdanti 9a smd. 169 iii . 3b maghdvanoh or maghdvanoh. 6b kapanti 'va. 6b pinvatd . . 3c simivdn 5c Rest at the fourth place.. 9e sydtd. 67. 6a mitrd . 5b yuyudhuh. 59. nstisam. 69. 75. 2c extended Tristubh verse 61. 4b bhujema. 55. but of larger extent. sthd. 151 i. 4b asidm. 17d gdvia. with long fifth syllable: but see svasti with v cons. 7a srdyiiiih. . 14b avathd . la kua sthah the old form ku is curiously preserved in a verse where it cannot stand.304 Metrical Commentary 135 *14c with Sandhi. 7c krnutd. 5c tudtayah 135 aiate . 3d pdanti. See p. 58. 20. 3a 166 i. 2c dsathe. 7d 170 ii c . 6b mitrd. 70. 6a nU. No. see 248 i. 166 iv. la sydvdsua. 14d dhatthd. deva with -d shortened. la prajdvat saubhagam suvd dusvdpniam pdrd suvd. tisrndm. 2b ahdm sand requires correction. For the metre see Ch. 2a vtisimantah. S. lc srdyistham .puru. [The original Bharadvaja collection was of the same metrical type Its general as that of Gotama Nodhas. The archaic variations are 106 iv. 7b sm. 9d probably prthvidm. 76. 9d jivantidh. 10c extended Tristubh verse.. 158 i. 125 Hi a. 4c w&t 56. perhaps rdjanam. : . . $ 160 i. 13c dhattand. 87. 2d nadinaam . nu u 151 ii. but are The important also found to a smaller extent in many other hymns. 10a dhenundam. 79. 7a ucchdntidm. Additions lyric . 9a wf w. . la indrd-agni. 3b yayidm. 4d stoifndam or spuurdhdse. 151 i. 139 iii. semi-cadence. 15a. 82. hymns 46 and 48 also belong to the original collection. 74. 171 v. 174 ii . 7a adyd 8b ydyistha. 53. . 73. 4a vdsisu 8a ydtd 12a adyd. . 2c 57. 6c nil u. 14c dhrsnava ojasd with Sandhi seems unavoidable: see 16a catalectic 130 i. 63. 9a catalectic verse . 4c probably require correction as follows *83. 5d sdkhlnaam. 168 ii. 2b s'ekd. 4c sudt . heptasyllabic dimeter verse or read indraya agndye. 60. 4b. puru. 5ab mitra . 4c dew. 174 ia. 16b gdam. 135: but perhaps the true reading is yajyavah. 2a the probable reading is ku v6 'svdh ku abhisavah. 2b asurydm. iv. 68. and as 7d. lc marutaam is doubtful. 10-13. Id 166 i. 6d met 66. .bhujd. 54. 5b. cf. 2c vaam. *15a svasti pdnthdm dnu : 130 8c. 84. characteristics are described in most marked m the hymns 4. 166 i. 63 . 151 iii. 9d jigivdmsah. . 174 ia. 3a probably rathir iva (Lanman) . 8b pibd. 151 i. 6a evd indrd-agnibhydm. . 10a nayatd. 65. . 174 ii. lOd vaam. 16c yajiiydsah. probably svasti pathie. lc rdthid. 64. 62. 5d sydvdsuah. 52. verse. 5c sydvdsua-. 5b susyantidh. 72. . App. 142 iii a 4a perhaps mitra varuna. 170 ii a. 6c vasatd. and 63-68 8. dhayidm ib.varuna. 4a 5d ma^f is probable. 24-26. . lc *78. For the metres 142 ii . 4c djathd. 86. ix. la deva is doubtful. : . 10c mdhya. 3a irregular cadence. 2b probably ndvdh. . 10c. 6d cdksur 'va ib. : \ : Mandala VI. 2b ndasatyd. 9a pdrusnidm.

5b see 151 iii.v 52-vi 21 305 have been made at various times. . 177 iii. vidmd. lib rodasiyoh. . *19c asthuri. 6a bhara . 7a jdnvmd (plur. 20.) 12b u 6e extra verse 8c mdrtidsah. la dyauly. 4b bhaasd. 151 iii. . 22c area. lie viht 3. 3c cdrmani 'va . 149 iii. 18a jdnisva. 6a vdstav 8a yiljia 8b as 4d suebhih 8c marutaam. cf. lb hypersyllabic verse. 3a vasavyaih. . 167 iv. 13d tuotah. and cannot always be clearly disbut a considerable number. 4d caesura after the third syllable 6c as 2d . 4a irregular cadence . apdam 4d vasavyaih 5d probably extended Tristubh verse. . Id vardhd. lie -vdstuam 13a as 11a 13c tubhya. 152 i 3a 224. cf.) 7b perhaps dmartena. 12a Rest at the fourth 9a sd. Id jiok Bhargavi verse. . : . . 15c Rest at the fourth devdnaam. cf. verse. 9a sd 13d vdsu. 9c dhisva. iv 51 lb. 6c as 2b j 6d evd. 2a. : : . 227 iii c. : . 10b long ninth syllable: the occurrence is not sufficient evidence of an adverb end'. tuotah. . 7'd jigivdrhsah. shew the metrical type of the cretic period. .sthdvirasya. 7d Pentad verse . lOd dhdah. 46a mdrtiah. 14. 151 i. . yacchd. 12. 2. 6b caesura after third syllable. puru-anlka 6a krdhl. 42c si one but see 130 iv . 226 iii b 2d Viratsthana verse . la su. 12d caesura after the third syllable. 227 ii6. . : . 27a tud-utdh. 6d gddhia 7a. at the fourth place. 8. 6a double Rest. . regular Rest. 3d the position of the caesura is 129. la tudt . but see 18. uncertain perhaps indrabhi dvavHsu' djau . 7b extra verses to stanza 6. krsnd-adhva ddhd . 135. 6b srutia. 9a tid . 7c probably Viratsthana verse. . 20 . 224. 15e extra verse. 3c caesura after the third lc ndasatyd. lOd probably as 7a but see 151 i . 2b heptasyllabic verse. 12a read 19. : jj . tua vayunavat. 16d nayd. or omit agnini. 227 iii 6. a. 2b asurydm: Gautami verse. 3d kutra. 5d vrscd. . 11a Viratsthana verse (indra type) . 14a hypersyllabic verse. 177 iii. 20 8c 3e extra (dimeter) verse yacchd 151 iii . lc vrsann. 4 b ksdam. . 3a dhdnid irregular cadence 120. but cf. 3c nu. 6. 151 ii syllable. 3d purdam. gdyd. nemdnaam . : : . place. 14a dadhid?in. cf. 5d dvrtah. 6b sdcid. 2d Rest at the tifth 4b hiatus after -a and -. 2b. 173 ii. 147 . 26a srdyisthah. 10. 5c. 2d Viratsthana 152 ii. 18b 16. 10c requires correction. 6c 13. 151 ii. 159 iv. prd tuvidyumnd. 9a ubhdyd (neut. 175 i 151 i: 13b ajurid ( 151 ii) seems unlikely: hence irplace. 8c probably sasvadibham in one word. 7b Viratsthana verse. 3d yaasi: irregular cadence. pipdya . 3b catalectic verses less probably aridh in 3b. 4a perhaps diotanam. 15a. 3c yiithdm 'va. 4d. 2b Viratsthana verse 7r cf. 3a bhuuh. 5d dhdnud. especially tinguished amongst the Agni hymns. 12c perhaps prthvidm. 7d probably Viratsthana verse. : . lb visvesaam. bhuma with hiatus. . 4a sd with hiatus . 2a Viratsthana verse (indra type) . : . 10b as 2d. 7b hybrid verse 225 . . 5. 15b extra verses to 14. 4a Viratsthana verse . 4. 33b 151 iii. 4c krdtud probably. 2a evd. 3d 15. 4c 160 i. . 7a Rest at the fourth but place 7b Rest at the fifth place .] : 1. lc Viratsthana verse. 11. 149 iv. pi. 3b 21. 17. . 9d tuotah . . bahu cit. 13c sahasdvann. 4c perhaps tudm 5a nitiktl 8d Rest place 2a tve.

7c gdam. dvisam should probably be 21c pipdya. 38. 2c gdam. 12d for dvesah. 224. giving a Pentad verse. 7d probably bhavat. 7c. 8a druhydv. 44.). . 8c. lb dadsvdn. 8b. 22. . 5d avd. 32. after the third syllable 34. 7-9 249 i 7a. 4b dviver dpdmsi . 14a puru plur. 2b cdyamdndya. . u 8a as 2a 8b as 2b 8d d-isiau. 5b diauh. 3c tudm esaain. 173 i. 2b gddhia. 7c Gautami verses. . 178. 159 iv. 5b pinvd. 3d nfndam . fjiantah. . 170 ii e. 37. 170 ii 6. 3b dbhiah.. cf. 9d Pentad verses metre see 7b. lid jahi. 173 i. 3a extended Tristubh verse. unrhythmical perhaps read ksatra. 3a nahi nU 151 iii. 26. . 2c irregular cadence 5b tu 6b uti 29. 12b sdrmd. lid perhaps -miirdhanah. 8d. 17 cmrlayd. Viratsthana verse {indra type) the fifth place. 5c probably Viratsthana verse: for possible ddasa see 151 iii. . . 8c the 7c tudyd and as 2c: but see on stu. 7a probably ndvyasid. . 9c vrhd 5b dudyoh. 139 i. 142 ii . . mahdbhih. 9a Viratsthana verse 7c sasanvdn. or rather visvasdham. lib ava. 9c uti. 4a vrsabho 'va. la Rest at the fourth place. 2c hypersyllabic verse. 4c Viratsthana verse. 41. . 7b sU. 39. 4b perhaps brdhmd (plur. . 5d d ubhe prdah 46. 13b bhard. Viratsthana verse (indra type). 16b krstlndam. but see verse (indra type). 4c vatsdnaam. 2c nardam. 16a tidt for the . 8c: see also 149 ii. 8d venydh (y cons. Id rodasl 'bhe. 2b Rest at the fourth place with long final vowel. ibjidydn. 25. la GautamI verse. 3d 31. last half of the verse is entirely -srdyisthah with divided compound. 2b caesura 2d tuotdh . 226 iii a: see also 151 iii 8b 167 iv 10b 173: dris-. 19a raylndam 22a gdyd 29b stotfndam: catalectic verse . 40. . 129 ii . 178. 5d srdvayd. Id ukiha-arkd. lc su.306 yujia . . . : restored. Id as 24 6d 2a perhaps 167 iv. la Viratsthana verse ((indra type). . 175 ii . . Id asurydm. 12c yacchd: the latter part of the verse is very irregular tanve is not without parallels 135 6). 2d possibly uruyd. . 2b.). 2a prd u with hiatus. 15a Viratsthana 149 iii. 9b. . 24. 128 ii 2a 30. sdtva. 9c dhisvti. -d being shortened before a vowel 6d Pentad or Gautami verse. 5a Rest at asurydm'. 10b vdjdnaam. 3a Rest at the fourth place 3cntt. 35. cf. drisanyann. 36. where a represents &. 18a dhisvd. *28. giving a Pentad verse. 6c Viratsthana verse. 3c 151 ii. 8b prdyisthdh. 4a tidm u. : . 2a pibd. 5a adyd 5d pari-etd. . : . . 170 ii c. 3a Viratsthana verses (indra type). 31b asthaat. : 22c sudsya. : . 2a tudt 4e extra verse . but on the whole a hybrid verse with long ninth syllable is ( more probable: cchadih. 3 c veda. 42. 8a. . sudm . 178. .astu 152 ii. 3d jibk 6c kpiuthd. 9c krnuhl or more probably krdlii. 7a suydvasam 7d rudardsya. 2a. cf. Metrical Commentary . 151 i. 5a vavdna gives a hypersyllabic break. 4a tidt 5b probably a compound abhydvarti27. 9b jdndnaam 45. 4 being shortened. . 4c visvdsdham. . 5a s'rutia. 33. 8b mdhya. Viratsthana verse. 7d GautamI verse. 219: perhaps read vdvna or vena Gd kriydsma: this form is suspicious in this hymn perhaps kardma 8a probably mandasva. 142 ii . lOd tud. 2a 3d bJiard. 8c ndmd (plur. 9c yacchd: mdhya. 5a probably Viratsthana verse.). 4c sdcid : . 23. la tubhya.

174 ib. : : . . 2d nd 'not' with hiatus. 4c Rest at the cf. . 54. 3d dhdtd 4d smd 5a Rest at the fifth place. cf. 6d Pentad verse. 6b updvasu is probable. 58.. 151 i. 15e extra verses.. 2c perhapt 3d ydbhiah 6a pibd . 13c isdam. ace. 135.with -a shortened before vowel. 17c requires see also 151 i. drier 'va 49. 151 ii 10b dsvanaam. 51. 14b.. aW . 6c. 23c oasa.). 4b cinuhi 5c.-. perhaps requiring correction. 4a heptasyllabic verse. 15c gndah. : . 15a mf . 18a correction. 2b sanu (plur. 130 iv rdsvd. cf. 142 ib 15d bhuuta. 9d dpia 174 ia. . pi. 4a Viratsthana verse. 9a spuurdhdn. 151 ii lc vasavia exceptionally. 2b Pentad verse 2c Viratsthana verse with irregular cadence. 9c. . 12c tanui. 15e extra verse. 14a see 135: but perhaps yajniydsah. wn2.): perhaps 50. 6b dpia 7a citrd-ayuh. 8a visvasaam. . : . . 202 . 2a perhaps uruyd . 174 i a. . u 168 iii. Id prdyisthdh. 9a Viratsthana verse {indra 9b s'atavan needs correction. 4d vdjam utd with Sandhi. 7b dhdtd. . 2b duih. 6b icchd. cf. 6d d atan. See p. : . ajdsua. lie Pentad verse.). and a-arksd. vii 79 4d. 53. 16d manusyhn (y cons. 160 i. . 6e. 24b gdah 142 ii irregular cadence. 145 ii 6 18c. lib uru is unlikely. 10a srdyistha-. 158 iv. 5a diauh. 174ii. See 149 iv j . 10b coda yd 14c probably 160 i. la satdant place. 4-4 . . 142 iii 6 . 8b dvathd . 6d extra verse. dydam 10a 151 i. 3c nd. . type). 7e extra verses. 9c daat see 11a extra verse to stanza 10. and 64. 2a vaam.). 8c ghrtaanna. 160 i . 3a u.. iii. 18d extended Tristubh verses. *31c extended Pentad verse ($ 227 iii a). 44. Id dud jdnd dual with hiatus 67. 52. 62. 151 i. i 91 23b. perhaps to sahdvan for saha*<< . probably to be corrected. 56. 16b nU . 60. 9b tYmt (plur. y : . and for the metre i 60 4c. 4a 166 via. 3b. 6a adbhidh. 174 i a: also agni. 13b anydh 3c. 10a mrla mdhya. 151 iii. fourth place. vasuyuh. .vi 22-67 47. 8d extra verse. 55. maghdvanlh 66. 2a ajdsua. . 9c aratnau (BR. . : . 63. 178. 5c naasatyd with Sandhi. . 135a. *28e sd imdm . 13a indra. la -huta. 2b devdn. Viratsthana p. . 9b 151 i. 5b nu. % 121 c 8a puru. ii 28 2a. $ 10c Pentad verse. craia d. 4c possibly mahdna or mahdd. 7a rikhd. 2c rju (plur. ..). 17d vaydh. 7a perhaps dsuasah. 307 cyautand. . *29a svasayd *29b. 14b vasavyaih. 9c isdam. 151 i. . 3c maghdvanlh or la sid 65. and so as the next two verses. lc naasatyd. 13d diaiih. day imam. vamsud. 11a chadisah. gen. 10c vira. 14c Irypersyllabic verses. 4c valid 5a sd d vahd yd probably. 9b dhdmd (plur. vii 42 4c.).puru. yd having the vowels shortened. 8e. 7c asmdbhy a. 10c Rests at the fifth lc rasmd 'va. *i 24 15c. 2d Viratsthana verse. 48. 8c mddhui. a-urva-. 61. pi. 3d sd it 151 i. . lc iitd nd. *30c extended Tristubh verses *31b Rest at the fourth place . 21e extra verse jydyistham. lc pipdya. sd. 10a pardstaat. 23a dsudn. 14b apdam. . 2b ydsya with hiatus. 4d dhudma. catalectic verse. Sbcauiiit: Viratsthana verse 3c Pentad or Gautami verse 5c probably read aydaso mahimnd. less probably suurah.. verse. 8c candrd-agrdh lib gantd. perhaps evdtha 21a dydam. 17c -dhunudndh.bhujd. 59. 224. 7c ksdyathd 12d perhaps caesura after the third syllable. 3a Viratsthana verse 3d a-aiijan.. 10b catalectic verses.

2d vdsu (plur. 149 i a. 168 iii. 6c ksdyasi (H. 130 iv. . 1 3b smd . . 7a tue. 17. see 139 vi. 3b are not necessarily indications See on of late date they may be connected with uneven lyric metre. sahasia. Sandhi. perhaps savistha. 2a see on vii 1 15b: dhaasim is 173: asurydm.] : 1. 18c hybrid verse. 167 iii. 20. 9a nu. 8c utd. valid. 135a. perhaps puuh.308 68. 2c dydam 5b srutia. . 3a tudt. 14a dhir 'va . The name of Vasistha and the characteristic refrain verse are however found even in the hymns which have not this metrical type. 2a probably u sid. 4a gndah. la 8. 5c satatamdm a. 142 i. verse . 213 ii. 16. The 71. 4a tad . 3b su-urmia . just possible. 6a adyd. lc. 8c sahasia. 8d sumdt. 18c viantu. Mandala VII. 151 iii Pentad verse 6d md with hiatus extended Tristubh verse 8d abhisdt. 149 iii visdam. 15b Viratsthana verse: but such verses are so rare in this collection that they invite emendation 16a 8(1. di&m . 11a -ddah. 142 i. 19b for the caesura see 205 ic. 151 i. 7c perhaps dmarta. 160 i. 9b sdktivantah. la bhdva. : . . 2. 6a tue. 151 iii. .). 2b rodaslyoh 3c Pentad verse 7d d with hiatus. 4. 2b. 7b dayisndm. 168 ii *75. 6b asmdbhya. 14c 15. . *74. lc prinandh. : . 173. 9c mdtroh: this is a very unusual measurement. 4b nardam. 2d smd 3c . . aridh. . 2c mandard is : 11. 4a suydvase.). 18. maghonaam : or 2b surdnaam. 4d rdsvd. 2a indara. 9b drcd. 7c for the caesura see or dydam 4a perhaps prthvi&m. 242 iv viii 35. lie as 5c. 4d diauh or dydva Rest at the fourth place prthivl . 20d pdtd. Id prati-dncam . 1 4b jidyah 16b sdravye. 7a pro1 49 iii 7c yesaam or sasahvtin j bably Gautami verse. 5. 6. extended Tristubh verses lb. 7a dahd . as described in 213. 8a nU u 7d perhaps tdrutrah. 9. 12. 3a prd-iddhah: perhaps dldihi. . 9b srutia. la asurydm. Id a with hiatus. . [The great majority of the hymns are characterized by an approximation of the trimeter rhythm to that of dimeter verse. but see . Gautami verse. : . . 5a tve. pdhi. 2d bhurl (plur. . 13a. 18a irtid.). 72. 4b trsii. Id vdsu (plur. 3d jid with hiatus. . 3. 173. 5b varuna 5c dadsvan Pentad verse . 3b mdrtidsah. 3a asthaat . 6c perhaps barhih-sddd. 2a indarah. 22a due 22b vadhumantd is 218 iv. lb ndrte tudt 5a vdmsva. 2c perhaps sucidantah. 178. 5 a. 151 iii. possible 151 ii: the interpretation is doubtful. 19. 3c vdsti (plur. 17e visvdhd: extra verse 18b for 205 ic. 7a dasema. 19a sudh heptasyllabic the caesura see 19c semi-cadence of Epic Anustubh.with possible. and are therefore probably of a different period see 107. 5b su-apatydm. tua . : . 6a for the caesura see 214 ii. 3d satruydntam. 5c catalectic verse . 13. lc sprsd. but the rhythm is irregular. 3c iria. Id ghrtd-anna. Oldenberg). .). la Rest at the fifth place. Metrical Commentary 2a srdyisthd: irregular cadence 2c 174 ii. 169 v. : 4c probably prd ebhyah with hiatus .

6c rod with hiatus: 24. see 8a avata. . la brdhma. 10b Tristubh verse. 1 7a Viratsthana verse 21b Tristubh verse. tistJid . . 48. or (with Lanman) Qdjigrtd. . . lb probably devaya ksiprd. 170 ii e.vi 68-vii 59 7d 309 213 ii. 151 iii. . 3c sd puru-anika 42. 32. 166 iv 151 iii.). utd: the double hiatus in this hymn is suspicious. stanzas 2-4 is without parallel in the popular Eigveda. 2c krnuhl. 5c for the caesura see 205 iv a . 6a. 159 iv. 2a ksdmia 2c sd. 3a dayisndm . 3a utd. 175 i. . jiok. 3<. 3c jahi. 1 20. 139 v. . . 171 ii unexplained. 6a indara. 15a smd 21c tubhya. . 5a d ii . Tristubh verses must be restored at the beginning of the stanza. 16a abjdam ( 142 i) gives an irregular *35. 2a u. 24b jdnanaam. fifth place 2d sadatd . 53. 173 v. 4b bhutd. 13c perhaps apdam. 4d ydtd . la mdam (151 i) or Rest at the fifth place. is : . . 5c vidntah. 17a mflantu with long ninth syllable. 6d vdatah. 52. requiring correction. 140 iii 6a jdaspdtih. i 165 15c. 6b bhtiri. 9a sredhatd . . 8c nd 'niena. 167 i 4c pipdya. cf. as Sandhi combination 39. 3b emendation seems required. *33. 10c jdnma. 14b prdyisthah. prdyistha. 217 . 5a d with hiatus 149 ii: . 6b astaut. lc bahvoh (v cons. 14c ddmia 23a 18a Viratsthana verse. . 26. : . 12b krnotd. 37. 2d avriiita. brdkmd. . 21. probably requiring correction la indara. 151 iii. see 175 i 3b Viratsthana 4d the metre requires mayinah.). 149 ii. nadydh. . sdkhayah: for the opening cf. 6b su-uktdm. *55. 4d gantana 151 i. tubhya 2c s4 3a su-uktaih dbdasema. 5a evd. 20 8d. 7b prthvi. 4c viddhi td. 142 i long ninth syllable. 22d bhutd lb Rest at the 57. . 11a sua-. 25a catalectic verse perhaps read amitridn 26b putrebhiah. . 167 iii. 158 i.^ isave 45.6a evd with hiatus. 3d vasavyd . . 20d dhaitd. 5a probably vaydyai. 12c hypersyllabic opening. verse. 3c tudt. 41. 4a sud-yasah. avdci 56. 149 ii. 1 30 iii. bhdrl. 9a irregular cadence 220 v) ( improbable in this collection perhaps read sydma ta indra vi. 2A& probably abhi mtdh. 29. 162. 219 . 4e extra verse 4b read sdnti for ydh 4a perhaps ends udvdto va 54. : : . . . : . is but the form : . 6a Viratsthana verse. 8c catalectic verse. 142 v. 2d rudara. 2c rdyiknah ( 151 i) is certain. 4c bhajd. 3c janaya . 2d pdrlvrta probably. 4b siona. 22. 44 opening. The four-syllable verse in each of the extended Tristubh verse. 5a. - .sd 59. cakra. cf of a locative from an -u stem seems unknown 5b vaha. See p. 46. .). 135 b. . 4c ddsudmsah 8c read payur dividh. 10a sudadsah. 28. . 4d ndasatya . 145 vi. Id catalectic verse. 162 iv 31. . 4 two *50. 5c no 'vata 6b for the caesura see 205 iv a 58. . 4a utd glid. 2c strnitd 43. 14d gdvajatah ( 142 ii) 6c rudarebhih. 27. 3d sdsmin . 152 i. . see 192. or ye gojdtdh. Id visudriak. 135 a 7c sahasyena (y cons. 2d suvandh. Id bhava: 3c Pentad verse. 6c pinvd. 5b for the caesura see 205 iv a. . 4b perhaps brdhma (plur. la Viratsthana verse. 4b dradhuam . 5d rudardya. 129 iii. . 5b dsakrah. 8c as la. la d u with hiatus. la indara.). 40. 34. probably an adverb with plural form. . . . 6d either rdtham 'va. 4d dayisndm. as also in 7c. . 8b probably munir 'va 6a ydyisthdh 3b vdtasvandsah 9b perhaps ma durmatir vah 10a ndma (plur. 25. . 36. or omit dsvdh as a gloss. 21c vasavye. 2d bhard 2b mdrtidsah. 5c tue. 38.

asmdbhya. 6a u: 6c vdrdha. 2iajahi. jydyistham 2a td for td dual. . 10c sdptivanta. 168 ii. 19a vartayd. 6b sudm. 25a caksvd. etc. 5d dsvavat. 6a naasatyd. lc vdha 2b tubhya 90. 2d rju (plur. 151 iii. 177 i. 4c vrsabhdsya. . 68. lb jdnimd 62. 2d sdrud. 102. 2a catalectic verse. 3a pathdam. 128 iii. 2a u. indra-agni. 63. la su-uktaih. Id 214 ii. is a monosyllable. etc. 4b adityd. *104. 8c Rest at the fifth place 129 iii. 45. requiring correction. Metrical Commentary . 13a nd. 2c purusinaam. 4a usdsah. 6a pasavydm (y cons. 5d 3c for the Sandhi see 74. 3a catalectic verse. 89. 151 iii. indrdvdyu. 8b. 2 2d mrnd. 142 i. 13c succhadis-. . 3d prthvim. 3b upa . 2d dhama 1 5 1 iii . 130 i. 101. 93. upa. 100. 2c for the Sandhi see 128 ii. . asmdbhya 80. . 15a hypersyllabic verse. la. lc upa . 5d rasvd. 3d for the caesura see 88. 91. 99. 2a madia. la varund. lb juhotana. 168 ii. 151 iii. 71. esaam. 170 ii e 70. 9b probably 67. 170 ii a . 3d su-apatydni 4a indara. 12a devd. 96. vi 63 3d. 2b upa 135 6. 149 iii. 19d jahl . 2a rdstardm. 8d starydm. 3c sdnubhih. perhaps 2b sasanvdn. 5b as 60 la. Bollensen) . 20c -ebhiah. 60. y . 205 iv a. 18b grbhdydtd. 149 iii. 95. trdsithdm.. asmdbhya o& adyd. . 6c probably requires correction. la for the caesura see 205 iv a. 84. tareyam from tar. 101. 83. 7c su-dvesdh. 4d emendation is required. Id 78. 7a mfldyati. 77. la md sU. 2c asurydya . .). 3a tanupdd. 8b extended Tristubh verse . 92. 2a requires correction somewhat as follows samrdd vdm anyd anyd ucyate svardt 169 iii. 3a as 87 2c.). lb dsuriye (vocative) or (as 2c Bhargavi verse. tie. 3a 75. tur but see 6b Pentad verse. 2c. 142 i. 4a 79. 98. 3b usdsah. 5a vdstav. etc. 6c nU d -a^Aa for -dthe. 3c kdsthaam. . but cf. 3a upa. 6d nrpdtl va 175 ii 158 i. 12b su-uktaih. 9b dsvavat. law. 139. 85.310 6d -adhuai . 170 i. . 170 ii 227 iii b . 87. 174 i e. 5c. 73. lc asurydm. . 4a trdsithdm. 4d dayisnaih. . perhaps requiring corasuriydm 3a jydyistham 205 iv a 97. lb. caksua. 5b caksathe. 170 i. 8a extended Tristubh verse. rayind. 224. 4d coddyd . 174 i a. *103. 3c as 60 la. 81. 66. . 19a perhaps mitra varuna. 12d extra 94. See p. 72. la perhaps ntt u mdrtiah 3a Rest at the fourth place. 6a nu u.). 3c upa. : verse. 4a jydyistham . 61. 15b tatdpa. : c : cf. 3a perhaps & suasvd yasdsd 69. 7b perhaps yuyudhuh. . 65. 3b for the caesura see rection. 6d brdhmd (plur. 3a ddsudmsam Tristubh verse. . lc vidmd. 6a u su. 4a sia: catalectic Bhargavi la 8b dsvavat. 151 iii. 2d mddhui.). 6a Rest at the fifth place. 86. indrdvdyu . *12c-to'w. 9b Rest at the fifth place.% 170 iic. ib. 9a as 3b. Id dhdyistha. a. . : : . 6b asmdbhya. verse. 135 a. 3c see (plur. 5d anarvdnam. 3d yatb y -nimisam. 5a 102. 11a sd. 82. 8c iydm . 4b trayidhd. .5a extended 5d prinand. la i7. 2a sicl 121 175 i 4d a-urnoh. 9a hotardbhih.

37d ustrdnaam. or catalectic lc prthu cchadlh. . 142 ii.^ 3a sustutid or sustuti d. 32a mdhya. variety of metre than 1. 31b utd.. 166 iv. 152 158 iii a. . 19c ddha . . 11a tud . 151 ii. 29a vaam. 22a sifted 22b vidmd. 33a brahmanid. VIII 12-31. 37b perhaps vididtam. 145 ii b 20a kdnud 21b gdam. 6d svdndsah. . 142 iii a. 2c perhaps visvddevdn. 14. . 16br<ft>va. 19d hinudndm. . 24a Epic semi-cadence. 158 i. 147. mdndasva 46b pdrsav. 142 i. 147 i). 28d extra verse. 31a perhaps parakdat. la u. . 31c utd. 13c md. 20a tud. 10c su-apatydni . 160 i 35c vdsu plural 38b purutmdnam. 142 i. . 4a catalectic verse. 30b tubhya catalectic verse. . 21a perhaps yddl. 19c vipanyu is possible. 6a purubhuja. 3b srindntah. metres. 5d dthd with long fifth syllable. 31b asndnta. 4b perhaps 4a heptasyllabic verse 3b catalectic verse prasdsiah 151 i. 147 i. 0*2 2b gdam. 7b perhaps sadhdsthaat. 18a dvd.] 142 ii . 12d probably anusatam. 7a catalectic verse. 5b bhurl ndmd (plur.% 162 iv. VIII 1-11. 28b heptasyllabic verse. 28c. cadence. 15a ndsatid. 142 v. 10a adyd . 21c visvesaam. or dhtaih. utd. 178. 25a vdrdhasva 15. 24c trochaic semi-cadence. 151 i. 16c perhaps tvd with hiatus: otherwise long ninth syllable. also 148 i. cf. 10. 15c as 13c.hutdh. . 5. dniani if the word is correct. or 170 iii 33a d u sd. 10c sudm. 12a tud. 21a sma. 170 ii c 40b kdnudm. 10c . i. 2. . 3. 30c trimeter verses. 13b sitit tvdvatah . 17b probably 8. 29c. 24d read pdkasthdmdnam abravam. 2b probably dcidhuam. 23a mdtsud 26c either rasinah iydm with Sandhi ( 130 i). . 39a gaat. as nah is unaccented. 13c u. . 174 ii purubhuja. 19d sd. A. . 11. 6a 16. 30a irregular cadence. catalectic verse. 2 Id extra verse*: upeva requires correction. : . . . 5b dayiyam. or hybrid 151 iii. 32c -vdsibhih. 142 iii a. 28c bhadh. verse. 170 ii e. 30d viedksid. 151 i. 26a pdatd. 4c sd. 30c -jidh. This is probably a Pankti verse 22b -praam. but see 151 iii 32b purfi. 151 ii. 28c utd. though with dimeter rhythm. 26c diauh. . 170 i. lib srinihi. B. 2a 158 v. 32c ndasatyd 35b probably catalectic verse: 151 ii 35c as 32c. 32a sa/m lie tu. 20a md sti adyd. 32b perhaps liiranydyl. 14a vdvatd requires correction. 15b daah. . 139 iii. 10c sd. [These collections are almost exclusively in lyric They have every indication of antiquity but the few dimeter : hymns which rhythm are included in the collection appear by their smoother to be later additions. 13c is a trimeter verse. icyenanu. 19a mtdsuam.). 3a tid.puru. 23d jivebhiah. verse. 151 ii 39a 6.] 13. 6b devd: 9a naasatya 9. cf. 4. found elsewhere. . mdrtiasya 10b ndvia. 33c vavrtidm (cf. 18a. 12c probably sagdhi. .vii 60-viii 16 311 Mandala VIII. *34c abhicdksid. 108 : but with more or catalectic verse. gha ( 158 v) being unexampled. 7. 2a irregular sU. 174 ia. 13c nahi nu *24. 14b as 2b. 9a 12. . 15a ndsatid. [Kanva is collections.

7c irregular rhythm. heptasyllabic verse 23b hdrlnaam 23c wtfa ?iu for the caesura see 205 iv a. 15a catalectic verse. 160 i. is unrhythmical 142 See iii b . 30a perhaps ydjamdnah . 4c sd. but a trimeter is not easily restored. 16b ndsatia . Metrical Commentary . 27 extra cf. 27. 104 23. 14c midhudmsah. 142 iii a 142 v. 16b aprindt . 13b gopaydtiam ( 136 i) or catalectic verse sindhu. 13c cakrdta. : : : : : . 19-22. 33a two additional syllables are required 226 4a perhaps repeat anye . 7c suvira. 16c . lib bhadh. easily corrected by interchanging drvatd is also possible. 3b marutaam. otherwise a hybrid verse. 30a sd . verse is required . : . vdjind and drvantd 26. 15b tudvatah. . dsiknidm. 25a 151 ii . 8c utd. 19c gdyd 24a dvathd 18d probably vavrdhuam. 10b ndasatyd 170 ii/*. 19a eta . 22. 2b naasatyd. 3b asurydya 4a perhaps mitra varuna 25. 23-26. : . 17. 7b uti d (for uti d) is the more probable reading. % 142 iii b . 23c the rhythm is irregular. 14a upa: hdrlnaam. 20. lib svasti. 24c irregular break. . anticipating . . 1 8b prthividh 1 8c catalectic verse perhaps 136 i. . 4c probably dhdmd hdndma is probable. . iii. bhuutam. 17a tud . 1 4c purdam 1 4d muninaam. 14b tesaam . This stanza must certainly have been Satobrhati but the original form originally. 24a area 24c ddmia. 22b perhaps hdraydyane. and emendation is desirable. 13c suaih . : : . 23a mahitvand 22a uksaniayane. 5d ndsatia. . 12c Bhargavi verse 14c adyd 4a 28. 19c -ydvand. 22c probably dyu. almost without parallel. 30 probably in its parlvfjam.puru. 142 i. . 21c dhatthd. 8b ndsatia . 178. 169 vi 20b yem^ . . 28a susdmane. 16b apdam. verse. 22c 21. 5a adyd : 15d dvidhat. . 12a tudt nd 'not' with hiatus. . 5c grbhayti 11a catalectic verse . 18. 14c catalectic verse. 6a diauh. 37e diydnaam. . la gopddj 12c 4b deva. 5c. 168 ii. 30c esd original form an Usnih verse perhaps omit dpasritah. 25 23b vdhasva . 16b sincd. 220 iv. 158 i. 14d udno 'va. .312 drianti. verses to 26 . 17c Rest at the fifth place. : . 24a airayata. catalectic verse 8a ydtd lid dnidm . 174 ia. lib 170 iic. 18a Rest at the fifth place. 17a dsvavat. dyu. 24b perhaps 167 i. 174 ie. 20. 174 ia. . 15c 24. lib. . 14c ardnaam. sd. 21b dvidhat. 18a vdjdnaam. perhaps 14a apdam. : . 14c hybrid verse. lb 145 iii 7b purU. 12c mdhya. 6c adyd. 2b dditidnaam . 12c 21. 18c jigiuh.see 151 ii. . 17c long eleventh syllable. the text for once gives correctly pibd with hiatus. 34b ndyathd maghfrnaam belongs to the second verse. 24c for possible dsua. 7c the rhythm is very 4a asthaat. : . 24c as 23c. 37d praimyitd. 3a adyd. See 19. 175 i. 20d vane/na. 10c 104 ii. . . 7b urjaam. . . 169 vi 23b jydyistha . 9b irregular cadence.hutdh. irregular. 18b 178. la d u 2a purvawith internal hiatus. 4b srdyistha129 ii . the words tdthed asat are probably a gloss. 9c upa 166 iv 12a urjaam 12b rdsvd 12c avd 16a probably vasuvidam. lie krnutd. 37. 17c the first part of the verse seems to be defective. 3c extended Viratsthana 5a tisthd . la fljsva. 151 ii . perhaps with an extra dimeter verse For the first verse see 152 ii. %& it is probable that all these were originally Usnih stanzas.

For the metre see 35-38. 140 iii. 174 i a. 8a 44. caritraat. 6b caksayd. 142 iii 6. has a distinctively archaic character. 32a heptasyllabic verse. 16a 129 ii. 4c pdanti. 152 i. . 10c sddhu. 6b asate. 6a pipdya. 6a perhaps vasu. 142 i. 13a omit -vantd the first time. 52. 152 i mdrtdnaam . 28a add 142 i. lOd svdnasah. 167 iv 32. yuvaso 'va. 2a catalectic verse. 15c etc. 2a -ydvand. 17b Rest at 152 i. . 38a as 22a. 32e 130 i. 22c sydvlnaam. 170 iic. 38. 15b dhisvti 177 iii. . 3a 146 ii 109 iv. g 170 i. 152 i. 50 7a. 9a indra may be a 6d vdsu may be a gloss. 8a dud 9a cakrata dual. VIII 32-59. abhlyujah. 142 i. 7d drddhuam. 139 iii. 2a diotanah. 4a probably prthividh. Id vilu (plur. 45. 25c long fifth syllable. 22e gdvaam. 151 ii. 2c nfndam. dsi: see also 152 i. 32d. 16b prthividh.). Id etc. 3a mahimdnam. 5b probably isdnakrt. 45. 170 i. . 19c pibd. 7a irregular cadence. 14c srutia. 6a as 51 6a. 35. 142 i. 151 iii. 175 i. See 247 iv. yayd. 170 ii/. -gopaah. 4a area. hymns and others of the same period. 47. . 28c rdjo-'sitam. 4a bhavd .). 8c tud . 12b -dsuah. 19b etc. [This part of the Mandala consists of Kanva C. . cf. 4e -hutiai . : . 8c indrdgnioh. 13d brdhmd (plur. 28b ghrtasndah. 10c heptasyllabic verse. lb pranayitah. 6a perhaps yddl. 151 iii. 49. 140 iii. lb perhaps yajddhiai. 12b vfsann. 7b prthividm. 5c suddanti. devesu. 6b apiciam is perhaps a gloss. . 39-42. devdnaam. 145 vi Qdpipetha. See 152 ii . 25b ydhi. tubhya . . 2d tuotdh. or restore -bdhave 34b dudyoh. 2a perhaps pfsadraye. 26a vdstav and as 20c. 168 ii . fourth place: or omit milhuse. 170 ii c. 16b requires correction. 17b. 2b pitfndam. See p. and mahd should be omitted there: cf. . 24b deva. 13b gopdah. 145 ii b . 14a rathesthdam. 7b prthividm. 19d jydyistliam 19c probably with yiijia ( 136 i) as hybrid verse 20a defective verse. 18b dvrtah. 29. 225. 167 iii. 40. 2c 46. 248 ii 3b u. See 39. 51. : . 234 i&. g 15] omit sucih. 26b -bdhue. For the metres see Id *^//v* 31. 36. lib dsvavantah. 9d extra verse: dsvdya. 173 v. 152 i. lib smd. 23b sidh. lib catalectic verse. easily corrected by reading pakvdm gosu.] 160 i. perhaps brdhmd (plur.) 158 i. 6c -ddah. 22a vrsabha. 24a tu. 50. 2b svdne. 242 iv . 3c u. 175 i. 3e ndvyasi. 10a gavyd srf vidmd. ydsya stomesu tudm indra. 5c perhaps *48. however. . 31c omit ddha. 3d naitfd 174 ie. 14d extra verse. 142 i. lb area. 129 iii. 4d ksudrdm 'va. 17c omit visvdmanusdm. 10c. 145 ii 6. 2b. 4a probably requires emendation: ydsya tudm gives an . 224: but perhaps gdyd belongs to 14b. 4c hdrioh. Hymn 46. 5b hydndh. 4c Rest at the fourth place. 24d abhuut. For the metres see 158 i. 178. 34. 28a tue. The verse may be easily corrected thus irregular opening and break. 5a 105 iii'. 30. 33. *17c utd. 15a rdyiknah. 16a visvesaam. 14a hypersyllabic verse. 1 3c pdnthaah. -dsuasya. 151 i.viii 17-52 i 313 the sense of the next verse 5c saptd. 20c catalectic verse. lie utd nu. gloss. 6a as 32 18b. lib suasti. 41. . 142 ii. 248 iii.

but see la dravd 8a ii nti 9a catalectic verse. 4b 6AifrI (plur. . . No. . . D. lb *91. 3c devdnaam 17b indra-ute. 9a naasatya. 12a yuyujmd 158 i. . 2d matsud. . 151 ii 8a heptasyllabic verse. 55. 4a 5c vasuvidam 8c hybrid verse dpraml-satya by the derivation 9c tadyd . 160 iv. 9c kena nU . . 5c hypersyllabic 90. Id. 66. 11a Epic Anustubh semi-cadence. varan a. 166 via. 5b siuh. 12c tanvdm. lOd piirdhi. % 142 i. 13a catalectic verse. 6b irregular cadence. 13a tue. 7a. . 13c. . le tua. 151 iii catalectic verse. suggesting emendation. lc diauh. 177 iii. 173 v u at the end of 7a hidh. 3a Epic Anustubh 60. 5b isdnakrt 5c d ihl 65. 169 iv. 9a dvidJiat cakrma. 6a srudhi. 166 iv. . 68. 10a -praam. 8a the verse requires correction. 170 iic. : . 6c tvdm. 120. 17a perhaps mitra 10b pdrijmanam. 7-9 for the metre see Ch. 151 iii. 12d probably vasuvidah 13b ddvidhuat. 9b. . 84. 75. . or martidsah. 16d tisthd 20d sedhd. . . cf. 6a bhard 4a eta la prdyistham 6b ddhvan. 10b heptasyllabic verse. 17b trdsvd. 8a devdnaam. 17a sudh-suah. *59. la ^?ra w. 7d perhaps sdkhiuh. 142 i. bhard 2c probably hiraiiydyi. 8b as 2a. 170 ii/". Id. 4b jydyistham 166 i. 54. 4c heptasyllabic verse. . 7d hybrid verses. 6c naya. 9d probably pdhi . cf. 6c avd. . . 4b perhaps raydyah. 77. 10b smd . 5c extended Viratsthana verse (indra type). 9a sne 9c rdjann. . 140 iii. 13d taniipdam. . 9c urjaam. 4b pHhvydh. : ) . 4a bhard 89. 20a vesti requires correction. 82. 4d. . 7b dlrghdyu. 4a tU. 3a tudm. 5b dntebhiah. App. 5c sthd. [Short collections of the archaic and strophic 6a probably dldihi 7c dahd 9a. 1 8a utd nfi. 63. 2b isdnakrt. 4b area. 14a dud-dud. 13a isdam. . . lc visvdsaam. la tidn. 53. 6a. . 85. : . 81. 78. 14a perhaps pratldhfse. 4a ca. kavlvrdhdh. 88. 83. 2a 76. 7d krstindam this appears to be the original form 1 6a catalectic verse 1 6b a-arkse 151 i: catalectic verse. 19c asmdbhya. 6c catalectic verse. 5d extra verses. 135 b. 7c mdrtidya.314 53. 7a catalectic verse. 2c probably 171 ii 8a bhaag. 8a u. lc naasatya. 4a M^ra. 151 i. : . 3b srinanti 69. . ddsudmsam. 62. l&midhuah. 151 i. 15a dpa the verse is only found here. 2c brdhmd (plur. 16a tua.] VIII 60-103. 4a mahimdnam. heptaT 6b asmdbhya. 56. 140 iii. 224 but it may easily be corrected to tvdm vrtrd hamsi verse. 167 v. 71. ix. 139 i 4c ddsudmsam. : . 2a catalectic verse. 10b yddl. 17a catalectic verse. 6d ddyisthah. la. lc y<5 'niicdndh. sit. 151 iii. 64. 4d 57. 7d sasanvdmsah 8c evti.) is probable. lie perhaps iirukrt. 5a tidni. : . . 6b dsvavat 6a utd nU 2a catalectic verse 7a as 2a 72. . s} llabic verse 5d dydam 6d jdntuam. 178 2c yujia. 12c dhdndnaam. 9a tuotdsah. 13b rodaslyoh. . 70. 2d dlienundam. 14a ?ndam. . 6d sasanvdmsah . 4a perhaps la catalectic verse. 74. probably to visat 61. 13c irregular break. lie rdnia. 171 ii. 170 ii c lc dUria. 79. of the patronymic. 6a tua. 149 iii 3b the same. 4a sthd. 171 ii. 3a visvesaam. 7a mfhiyakuh. . 139 iii. *58.). 13b pdnthaam. Metrical Commentary . periods. 80. 67. 5a hdnta u nd. .

4a arsd 9b srindnti. 9c. cadence. 142 iii a. 13c citra iva. 11. 5a bdhubh. 1 2a irregular cadence 1 2b catalectic . . 12. 4. A. also the word dive-dive probably belongs to this 166 iv. 4b dasat. 178. 10b hotfnaam. . 9d vapd 10b prd waya. 140. 5a hypersyllabic 152 i. 3b caesura after the third syllable. then irregular break and Jagati 3a perhaps mitra varuna. 3. [This large collection of Gayatrl hymns does not correspond to any collection or set of collections addressed to other A few of the hymns may have belonged originally to such deities. 5c catalectic verse. la hidh. 3d srutiai.] : 9b devebhiah. cf. . :. 3b carkftia. 6. 7a tidm u: 8a anarvdnam. or read devdn. verse. 4a etc. 6d upa. 103. bhydsam. 21. and from the regularity of their structure cannot be earlier than the normal period. *100. or omit vdjam. 8a vdar. *8a. probably requiring correction. verse. . 173 ii: hybrid verse. . IX 1-67. 151 iii. 151 i. 3b vdjdnaam. Gbjiok. 13b ro/u/jia satrdsdham. 5c extended Pentad verse. 224. lb vajri. 170 ii b. 10a 10b four syllables are wanting at the end of this verse. 13. 13a caesura after the third syllable. 7a eta nu for the metre see 94 viii. perhaps diauh. 6b 19.). 6a utd nu 8a adyd 8b dev&naam 10a etc. la giristhdah. tation is doubtful. aiti. 4b heptasyllabic verse. . . . cf. . cf. this stands first see 224. verse.i:> 3c sdnair 'va. 53-ix 21 129 ii. lib jydyisthah 14a dvrtam 18a tua. lc upa su 2a catalectic verse. 9b probably purd 3b dsvavat. 167 iii. 7b extended Pentad verse. 168 ii 93. but some correction i required . 9a. 3b svandh. 20a Viratsthana verse (indra type) 20b sustutiti) 21a emendation seems required. semi-cadence. 8b prthividh. 1. 10b for the numerous hypersyllabic verses of which 97. . tanul 13c sdcia 12c bhusd 15d. 31a indrdbhi. 151 iii. 94. : . 4a pdvitdrah. 6a hydndh. 170 ii c . 7 c as 3a. prdyistham as yasdstamam 1 0c rdthanaam . 151 i. 12c asuriyah. lb perhaps aksanh. . lc prindn. : . 166 i. 178. . 3c suaih. tud. . . 7b kdrtud 7c naasatyd. 8. . 101. 4b tdnua. 2b sdnU (plur. 10a catalectic verses. . 12c rdsvd. 178. 177 iii 33b somdnaam. . la perhaps read dhiy* 15. 9a tud. 10a vwvesaam 102. 14c tudt. 3a bharatd 12b the interpre140. 2a catalectic verse. 6a rdthia. 30b vdjdnaam. t . Mandala IX. 6c catalectic verse. 9. 227 iii a. 7b heptasyllabic verse: perhaps read pdvamdnah. . lie Epic Anustubh 5. collections as those of Kanva or Kusika the majority must be imitations of these. or read updstha d. 7a catalectic verse. 6b srinltana. 4b nadio 'jinvat. 142 i. 99. 5c sindhunaam. 96. 16a irregular breaks. 3b catalectic verse. . 4c sd. 151 iii. 92. 7a arsd. 9c srlndndh . 95. 2a perhaps abhlyujah. Inplural. 18. verse may readily be corrected to gomad ydvamad d&vavat 5c utd 21a abhi stt 31b ydhi maddnaam 33a read tvdrii hi wtrahan esaam. 9b perhaps uruyuge. lOd dhehi lib spr^a. 14. tidn n4. la pdantam. 4a a-iirva-. . . 151 i. 5a m. 98.vlii semi-cadence.

3a probably omit asisyadat. 151 iii: catalectic verse. . . 2b. 42c dud.9. 2a bhurisdt. 4c srinlta. kdrtua. verse. 2b dtha . 2a 43. 58. 82. 7b catalectic verse. 7a svandh. 8b dsvavat. lb maghdvanah or maghdvanah. 3a perhaps nu na indo rayim lc perhaps vipram sumbhauti 40. 151 i. 34. 10a ajdsua. la drava. IX 68-97. . devebhiah. la rdthia. 3a -a 'hian. 28c. 4b giristhdah. la etc. 44. 31. or restore mddhunah. . 151 ii.] 1 78 lc perhaps barhih-sddah. 2b asuriyam is required by Vedic usage. 66. verses. 56. cf. lc srindndh.9. 5d codaya. 28a irregular cadence. probably lc caesura after third syllable. 6c vfsann. 3c asmdbhya. prthividh. . 3a tud. 2c vdjdnaam. 23b midhuiih. omitting jdrdm: but see 3c sldd. 4b utd. 41. 87. 45a dpia. 77. 25c mdam. 59. 6c devavih probably. 60. 83. 26a na . 29. 63. 18d extra verse. unless we read mddhuah. 62. . 79. 41d asmdbhya. 55. 30 the metre is of the type 11. la 30. 145 ii b 9a smd. 171 iv . 42a sd. 46. 48. 64. 19c probably devavih. 80. 3b dratiah. 16a prd 43a Rest at the fourth place. 48b dhdva . lOb&amnaara. 5b srinanti: suarvidam. 1 2d extended Tristubh . 8b sasanvdn74. lie. . 142 i. 71. 57. 145 vi. 34d dhdnia. 2a upa.316 22. B. 84. 39. 2b devebhiah. 9c mdsvd. to be corrected. 2a long fifth syllable: 13a upa su. 8a sd. and requires correction 5c extended Tristubh verse. 85. 3b dhanuantu. 147. 178 75. 81. 10b giristhdam. 70. 1 a catalectic verse . 88. 6b dsvavat. 4a perhaps rnahd. but recurs ix 107 26b. 47d svandh. 5a dhanua. g 142 i. la catalectic verse: but SV. mandi is probable. lc tdnud 4b suarjit. 78. 5c iAa sit 26c srindndh. 2 c irregular cadence. 2b rdthia. 42. lc dtiah. 5b for the caesura see la Viratsthana 142 ii 8b urue. 29c. . 166 vii. 24b arsd. [Only a few of these hymns have the metrical marks characteristic of the archaic and strophic Otherwise the collection both in the smoothness of its rhythm and in the equal use of Tristubh and JagatI metre is closely akin to the second Mandala and the later part of the fifth. 15a sd. 177 iii. lc extended Pentad verse. 3b bhdva. irregular. lb svandsah brhddivesu. 18b dsvavat. 2a the reading is perhaps prd dh&rd asya. 9b diauh. 4d 179. verse (indra type): tubhya 178. 5d tuotdyah. 5b dtha. 24. 5c dtia. 28a perhaps aksaah. . : . variations or other periods. 30 la. 3b requires correction. 3a hyandh. 6a d indo. na. . 69. 168 ii. Metrical Commentary 7b vdsu (plur. 3b 142 iii a. 135 a. 3a vfsanam. see 135 duduhrire. but is perhaps not so intended. 28. 2a prthividh. 178. 6c d %m . 36. 166 vii. ^ 151 i 27d extra 67. 20c devebhiah . Addenda 4d srindnti . . 68. 4c dsvavat. mldhuah. 65.). 76. 214 iii. . 30c mfl_a. doubtful. 145 ii 6. 61. 2b devebhiah] 6a adyd. 6c rdsva. 72. 7d gdah. 53. 32. 6b mdhya. lc the rhythm is la dhanuantu. 179 3a. 26. 47. 48d catalectic verse. 139 i. 147. 151 ii. lOd dhattd. . 28a ddvidyutatid. 3c rujd. maham . 86. 170 ii c. 3c pinvd. 4c prinanti 2d perhaps tritiyam. perhaps kanid 4b abhitid. 2c pavitdram. 26c hdricandrah. 3d dtia. la arsd . 30c pdantam. lib devavih.

6a dpia. 27d hrdhi. 21a devebhiah.5 ii prd u. la abhi. . 10c svdnd. 142 iii b. lid vasiivid. 49c.) . 21c camvdh ( 135 6) is improbable either camti (sing. i 17. la 10a requires correction. 2c 103. 6b perhaps vdcam probably chandasidin. 2a iidt. 5a madanaam. Pentad. 9c catalectic verse. 151 ii. [The Agni hymns 1. and Yiraj hymns. 6d jiok. . 40d 43c perhaps read jocfo/o 'bhisri?idn 48b sram . . 8b pdantah. the rhythm of 3a 4c Rest at the fifth 89. 225.). dhanud. 7b irregular cadence. 173 ii . 3a mddhuah. 8a heptasyllabic 2d. X : . 52c vdata. 26a perhaps devaviyah. 5b sid. 167 iii. 6 shew frequent and other variations characteristic of the strophic period 249 L] x 46. 145 ii b. : . [Lyric hymns of the archaic and strophic periods. 4d asmdbhya 6d pdtd. 2d hyand. 99. 161 ii 6b vasuruc. wm . 7b devebhiah. perhaps devaviyam. 105. 151 iii. 17b srinand 18b irregular cadence. hdri?iid. 98. la catalectic verse. 167 iii. 18a syd . 5a 'dsuam. 142 ii aksaah 9b aksaah.] lb arm. 168 i. but the hymn is one of Gotama Nodhas Id dtia 3d srinanti 5a indsvd 5b caesura after the third . but cf. 4b urnuhl. . 10c apdam 13b d-nayitd. 110. 3b arm. la dcchd. la vd U) *112. 9a a7i ?/ac?. and 8d xrayd perhaps suasti should be read for the caesura see 2 3 ii 9b rdnia. 4c utd verse. la rdthia. 106. 21b sravd. . 171 v. 5b heptasyllabic verse.). 22b srindnn. *113. 3. . . : Mandala X. or Pentad verse . 33c. : . la rodaslyoh. 129 ii. . 9a Rest at the fourth place: gdvamdn. 169 iii. and see A. 3a perhaps aksaah. 2a svdnd. 4b svastdye with v cons is doubtful in the Rigveda proper. 2b svdnd. visa . 104. 16d. 95. . 20c dhanuanti. Id sddmd (plur. 151 i. . . 90. lc the irregular rhythm suggests emendation. devavih. 2b midhuah 3b hypersyllabic verse: omit tdm. 11a mddhud (fern. lb vibhusdham. 93. 148 vi 96. 142 i. 3a abhi. 15d urur 'va.ix 22-114 . 167 iii. . 101. 107. : . 7c sagdfit Id vdsu (plur. 5a nU. 142 iii a. pra-dcah. : 1 . 166 iv. 12c sasahvdn. 3c mdrtidya. 4a dadhidhn 108. 152 i. Pentad 1-9. la place. 4b dhanud 5a harlnaam. : : . 170 ii f. 19d sravd . 111. 4b hinotd 97.).). 151 i. 16b arm . 5b 91. 36c diviydjah read diviah. . 139 i catalectic verse. la svdnd. 9c giristhdah. le the refrain is borrowed from ix 106 4b. 16a svdnd. 21c syllable: vaatrfpyam. 317 159 iv. 6a Pentad 92. cf. 51a a6A. IX 98-114. 32c matsardvdn. 94. 2c 173 ii. lib rbmd (plur.) or as a hybrid verse visa. as a long fifth syllable is very unusual in the full cadence the Sandhi combination of of Epic Anustubh 7c hypersyllabic verse two verses ($ 119) is not probable the omission of mtim is perhaps on the whole the best course. 3a svdnd. 6b devebhiah. 17c dhanud 17d indu. 100. 109. lc suurah. 4a vasuvidam. 166 iv. . with which the editor has included with good reason the Anustubh. 2d srinantah. 5c Viratsthana verse. panindam. verse. 2b ndnid (plur. 26d see 219: for asmdbhya. C. 3c svard. 3d paid. 166 iv. verses cf.

2d dtya more probably . 8. 4c sudsah. 9a perhaps devdtd. as is *X 10-19. lc. . 4a devebhiah. . . 6b sdptivautah. 178. 16. i 164 42. . 8a irsva 18. lb 2a catalectic 2b. 7c pitvy&nam pravidvdn is probable here. but cf. . 6b. 11. 6d Viratsthana verses. 5c irregular 7c jiok. 6c su-apas. .318 Metrical Commentary 1. and in doubtful . . 6a.] X 1 20-26. 152 i. . 2c ydtra. is very doubtful. 3d svdha 'nie. 12a requires emendation 1356. 8b perhaps siuh. . both on metrical and on : is linguistic grounds. On the other hand the 10-13 is open to question. . 2a rodasiyoh. cf. 3d. see 2a emendation is an imperfect quotation from x 25 1 but see perhaps omit agnim ( 152 i) and read bhujdam required: cf. agnim sit dimeter verse perverted by the recollection of 12a: omit skanndh. 12. . but see 136 ii. Thus the decasyllabic verses in this hymn are more than half the whole number. . 3d sd 7a tud . la suasti *9. 7b hdvya more probably. 7 b Pentad verses Bhargavi verse. 13a tudt. For the metre see 249 i. . . 2d extra verse. . lie tdbhiam. rhythm. 3c srdyinidantah 130 iv 21 etc. 2b simivdn. . 3c gacchd 14d long fifth syllable exceptionally: perhaps dydam. 8d vrha. 5. sucidan. 173 i. 2a tubhya. 4a barhih-sddah. 3b didm or 3b. lb sudh 2d bhavatd this is therefore the third 19. 13b vidmd 14c svardjdh. 145vi. 3b tritiyam. . 151 iii. 168 ii . 5d prthivi. lc bhaasd. 6c.) 6c vdsimantam. la Rest at the fifth place. 10c barbrhi . la. more probably 7. or svdhd anye 5c catalectic 9a ita twice 8d Rest at the fifth place Bhargavi verse. 14. 6b semi-cadence of Epic Anustubh verse of an Anustubh stanza. 20. . Id. sing. See p. 11a sudnau. 7c putrebhiah. 201. 13. 3c. . 13a and 13b also require perhaps nd te tanvam n tanud. 5c sddma (plur. 152 ii 6a tn 7d vahd: sahasia. . . emendation perhaps read batb bata n asi yama nd evd tdva mdnah 14c for tdva read te. 9c sahasra-arghdm 17. 7a. . : C. 45. 15. the first two being similar to those of stanza 8. 3a Gautami verse. . 6c Pentad verses. 151 ii. see 160 i 4c dpia.8d perhaps interchange dtra and vltdt 9c vahd. which position of hymns [The hymns 14-18 are clearly of the latest date unconnected by subject. . 6. 3. 4a. Metre and language are closely allied to those of x 95 lb purti as ace. 5c prdti-ardhirii devd-devasya. . . 2a Rest at the fifth place 2d irregular break 4a as 2d.] 10. also 19. 5a srja 13a appears to be a 5b sd. 142 v. Id pratidh . : | . lid svasti 15c iddm should perhaps be omitted. . 5b caesura after the third syllable. 2. 145 vi. 168 ii. . For the metre see 249 i. . 178 3c. [The Vimada hymns are of a very early type. 3c Rest at the fifth place . 2a dpia 3a sd 11-13. : \\ . 174 ii. 2c. 168 ii. 151 iii. : . cases the preference maybe given to this interpretation. 5a. 7b dhatta. . 1 lb sadatd. lc jyestha 2c sakhyd more probably . 5d sasanvdn 6a Iraya . . 7d yacchata. 2d. B. 11 cf. 7d rathia. lc nhiioh. ^ 227 iii b 10a drava 10a.

7d jidk 158 i. 12a the text gives the hiatus correctly. . 44. 13c ydsaam. 175 i. 5b dthd 39. cf. *28. . 2d svdndh) 7b suvd 35. 142 v 5a Rest at the fourth place with irregular break emendation 6d as la gopdah. For the metre see 22. Id rdmayd 42. 45. 66. 4b sdhasdvann. 12a. 8b diauh. 177 iii. 170 ii e 142 iii a. 9a ksdam. 7b as 3c rather than is probably required . 45. 168 ii 6d bharatd. 4c k/-*ra 2a tuadrik. 7c semi-cadence of Epic Anustubh : perhaps read prd dildhod. 142 v. la ydantnm. 21 7c. cf. . 11a adyd'.] *27.. : carseh. 9e is that of trimeter verse. 3c for 139 iv. cf. lib perhaps yuvateh. 5c the rhythm here. \ . *27. 10b catalectic verse. <:/ . 31. 12d duria. 24 3c. *5b samicyoh. 13a read dyatir ddrsram 14c dhattand. : \\ : . : . . lc hybrid verse. 5a %rayd 8d sumati. lc smd 23. : ca indra 24. 29. 32. 174 ii. .) is quite uncertain 3c Rest at the fifth place. 21 7c. 10a bhavd verse. 4c as la. 13a as la. 25. 4d vdatah. 9c. . . 2b dhdmd (plur. as also in 6e. D. 2a catalectic verse. . 10c carkftiam 5d dsvavate. 4a sd. 173 v. . X See p. 40. 7a visvdhd. 5d su-dstrdn. the cadence cf. 175 ii lid sd. 18a vxsu-ancah..) . la extended Viratsthana verse. . dnasmd. . 14b Rest at the fifth place. 2d visvdhd twice. 139 iii. 3c 12c sthd: su-apat158 i. ix. lOd samidm. For the 2a tidt. 14a read pascdt 36. 45. lc ddha. : . 11a extended Viratsthana verse. 148 vi 9c a-usdndm as a-arksd. . No. . lid read i for *28. viii 68 16b. sud. 12a cakpnd. or Rest at the fifth place. 177 iii. . 2a vdstav. 26. [The small groups of hymns here included are generally of a type which stands in contrast with that of the archaic and strophic periods. 151 ii. lie siadhvam. 7c bhfl rhythm is influenced by that of trimeter verse. im the second time. . lb hdrlnaam. 151 i ddasasya (ib. rdjann. 4c adbhidh. . 4c matlndam is uncertain in this hymn. 6a catalectic 37. 243 la Viratsthana verse (indra type) 5b vdhadhiai ( 146 ii) is very doubtful lc fslnaam. . *34. 4c Bhargavi 142 i. unless we read ucdthaih. 2a trayidhd. 8c nd. 27-84. It is very probable that many of the hymns assigned to the cretic period really belong to the popular Rigveda. 8b bhadsvantam. 13d dhenundam. 151 iii bhuut 2d perhaps kpiavd/ui. but maintain artificially the linguistic features of the Rigveda proper. fifth place. . . verse solid 5b two syllables are wanting after the caesura perhaps supply 5d caesura after the third syllable: 4mebhiah\ *6-9 see p. lb smd. App. ix 70 la 4c for htm restore hi. grbhdyd: tue. 4a tudvatah. 6c Rest at the 178. 170 iii. 7b catalectic verse: perhaps yuyujrire. 30. % 158 i. 4a vdatasya as d no valid indra pfksa adyd 9a double Rest. . 14a avathd. 2b bodhaya 3d vasilvidam. 43. . lc taratd. 4d sd 6b Rest at the fourth place. 145 ii b 6b mdhya 10a -ds'ua?n. . 11a vidmd. 2a dia/tih) 8b devaviyam. 7a requires correction. . 170 ii 1 4b read uttardt. 7b kulidh. . 191 metre see . *7b prd eti *9c sd. 7e isala. 142 i. 33. . . 226 iv a perhaps dvd tvdm nah . 2d sasanvdn .X 1-45 319 For the metre see Ch. 166 iv. 151 iii. without having the specific character of the Kutsa hymns. .. . 7c sedhd. 8c.

. 9c sdnitd 142 iii b . 135 6. 4c Rest at the fourth place then perhaps asuriydya 149 iv 4d jydyistha extended Viratsthana verse. . Alone in this part of the Rigveda it is companion hymn. 5c. and yet a parallel hymn can hardly be found amongst the archaic collections. 8d diauh. 8b pdrd ait . 5b gdvajdtdh. : . 54. 8a tud. 46. 8e perhaps arista-. . 5b extended GautamT verse. 2b tanvah. 3a abhi su . cf. 149 iii. 5d vdlm. *4d extended Pentad verse . 78. 3c srutdrsim 5a dsvavantam. 140 vivie. 2c Viratsthana verse. 4a mdam. frequent Rests with Jagati cadence but cf.] origin la raudaram. lc hybrid vert>e . *5b catalectic verse. 5c bdhuoh. but see 149 iii see 149 iii. : 130 iv. rather than as in 5c the cadence requires correction. 3b. 6d rocandvi. 140 iii (Addenda). 129 iii: probably as Id rather than ddasam. 7c for the caesura see dissociated in structure from its : . 6c. 5a probably extended Viratsthana verse. *59. : . *51-53. pdtaram. . but more probably Viratsthana verse. *60. : 9b prthividm. 142 iii b. 4c jidk. 45. 2b asmdbhya. 8-10 may be Tristubh stanzas appended later. *52. . . 2c irregular double Rest. 2c ku 'aha. lb mdhya 151 i double Rest. 142 ii. lc brutd. 168 ii. 5d hidh. 10a. 148 i. 227 iii c 3a u.). lb. These hymns are almost alone in the Rigveda in combining lc mdam 48. 61. 9a 2b. 55. .320 8b Pentad verse. la daam. 3b sdcid . *57. 2a mdam. 140 iii. 151 i) is doubtful: perhaps Rest at the fifth place. 47. 178. . 6d jidydh. . 7b abhi. (plur. Pentad hymn seems to have consisted of 5 double stanzas. Id extended Viratsthana or Rest at the fourth place. For the metre see 249 i: it suggests for this hymn an earlier date than its neighbours can lay claim to. . 2b. ( : . 6d rdjann. 2a. 3c pitrndam . An Iranian has been suggested by the proper name Ndbhdnedistha in 18b. 2a ddbhia. [This extraordinary hymn has a metrical unity which stands : it has all the marks in striking contrast to its discontinuity of subject of the archaic period. . 6a dud-dud. See p. 6d bhavd 142 ii. 2b apdam. *53 4c the opening requires emendation 6c vayatd. 3d dsrinlta cadence 205 i c . The Pentad verses are la. 2a sd 2c hybrid verse extended Viratsthana verse {indra type) 3a 214 iii: Viratsthana verse {indra type). hybrid verse: but see $ 130 ii. 6b vrtrdm 'va. . but cyautandh. 167 iv. cf. thus corresponding in length to the hymns in i 65-70: whilst 3. 7c. . 48-50. 2d sudm . x 77. 142 iii a. 3c for the caesura see 4a as 2b . . 6a. 9a adyti. tvdsta d- . : 9a bhuut. 6d Sandhi at the caesura . 5c verse. 10b bhajd lib vdsu 12d dhattd. 11a Viratsthana verse {indra type) or Rest at the fifth place. 5b. 167 iii 49. 3b diauh 4b nu . 10b iii. 7c the interpretation is doubtful JagatI cadence. 4. lie visvd it. 10a Irayd: gdam. 226 iv 6. 8d requires correction. 1 b 2b carkftiah 50. . Metrical 12a nardam Commentary . 5a. 10c -rdnidh. lc. : . with irregular Viratsthana verse 4d ndyisam. 151 i. 11c visvdhd. perhaps visva-dbhuve (Grassmann) . perhaps vardhitti ca nah. 5c perhaps drhhah. 4c as *51. 2c carkftiam. 6c ji6k 177 iii. as Id. 5a ehl. whilst 5d. 8c pfirusam. la sdstdnd. 56. 6d as 4d 7b vdsunaam. 6d. 6b. lb sdcidm. 6c probably 149 iii . vi 3 Id 8b tisthatd. 5c extended Viratsthana verse. Hence the 7a. 7d are Viratsthana verses.

13a mdrtiah. 3b catalectic 76. 6c stiro 'va. omitting u. . 8 Tristubh stanza 8a probably omit devdh ( 152 i) and read with Rest at the fifth place. 78. 62. 7 JagatI stanzas. dsiat. 4b hybrid 2a add at the end kathdslt. 9b utd for the ninth syllable see 177 i. 6-8 For the metre see 77. 12b perhaps pilruih. 137 iv a syllable is lb srutia. lc see 227 iii d. 129 ii. 20c urdhud. 4d perhaps devaviyam. 8d perhaps madhuvfdham. 173 ii 5d area. . 142 i. : . 169 vi. 5a siksd. 16a srdyisthd 16c sd. 63. 2d suaih. . 64. 4c havydm. : : : . Viratsthana verse 7c prd 'va. Id ksitlndam. 8b sunuthd. 6d mehatnud. 3a Rest at the fourth place 5b djyaih. 6a. la brhdc-chravdn.x 45-83 verse. 79. 139 v: tid. 9c prthivi&m. 77. 70. 226 iii b in the cadence probably sddhdyantah (151 iii). 2d devebhiah. 3d pitfndam. wanting at the end of the verse. 14b rtajiidah. 10c su-aritrdm. 321 with hiatus the first time. la su. Id bhavd. 4a rdthdnaam ye Wdh with Sandhi. la vdsunaam. the others being 24c. 6c extended Viratsthana verses. 12c sd. . lb Gautami verse. 3c slmivatitah so too 3a. 12c pipayata. 10c rdthasas pdtih. 78. 151 iii.) . 15d pdtd. 15b caesura after the third syllable. 2b JagatI verses with irregular cadence. 7c siond. 168 ii. 2d saptarsin verse. 69. or read visvani ye. 5a jydyisthdsah. % 166 iv and via. . 3c tvdstare. . 4b Rest at the fifth place 66. 145 ii d or Rest at the fourth place . or Rest at the fifth place. 2b puru adv. 65. 26a. 3d vasaviam 151 ii 4b the rhythm is easily made normal by reading exceptionally gomantam urvdm abhi ye titrtsdn 4c irregular cadence 6b probably ndma. 17 djydyistha. 13a tan nti asya is perhaps most probable. 75. Tristubh stanzas. 52. lie Viratsthana but see 149 iv. . *72. 3a Rest at the fifth 74. 12d brdhmd (plur. la. 4a hatd . 8c satdsuam. 14d perhaps read srudhi hotar n rtdsya hotar adhruk 15a as la16a Viratsthana verse. 227 ii. 16c sd. : . 5. 4a dudbhyam. 129 ii. : . 9b sd. 11a vahd. cf. 12d tistha. 9b indrd-vdyd. : . 68. 73. 3b agnindam 2a. : . . 9b divo 'va. 151 iii. 80. 5c. 27a. . 151 iii. 5d pratidnn. 3a sadhu-arydh 10b for the caesura see 205 ic. 8d 145 iii. . 17a as 2c. . 12b ddaddtd . 21 .with Sandhi. 4a daat. 2c sthti: adbhidh. 149 iii. 4b ndsatid. 12d catalectic verses. 67. 7a catalectic verse. 130 ii 6b susdrtud. 12c. rather than esaam amftdnaam 3c Gautami verse. 178. giving a Viratsthana (indra) verse. a. 7a ayaje. ii 3 8a. 21a ddha (indra type) 23a a series of defective verses are found towards the end of the hymn. 3b requires correction: perhaps . 3a bhdmidh u. 136: hence 142 iii a. 2c Viratsthana verse. verse. *71. lie devdhutl. lc trayidhd. . 27d these require emendation 25a yddl Rest at the fifth place 26b su-uktaih 27b bhutti 27c Rest at the fourth place. 22a ddhd tudm: Viratsthana verse srdyinih. 6b . la ilrjdam. 9a. 14a avathd 15a suasti: Rest at the fifth place. . *83. 7a Viratsthana verse 7b marudbhiah. 160 i. *81. . 4c utd tuasmai . place. 3a Pentad verse. 139 iv. 14c etid. . cf. 2a caesura after the third syllable. . 2a indra-agni . 7d probably sndtud. la tividhat. % 145 ii 6. 5d nu. 2a srdyisthant . 6d hfdid. *82. and 250 ii. 8b sidatd: siondm.

6c. 4c Rest at the fifth 5d sua 7b ddha. for the caesura see 205 i c. 12b vidndh. *84. 174 i a. 174 i b . with several archaic variations. 217. 6a dbhutid. 10c trayidhd. 129 ii. 144. 191 iii. 8c sma . 9b Rest at the fifth place. see 160: perhaps the adverb. 20a. rather than ksavanibhih. 14b hybrid verse . [Companion hymn to x 10. 15a. 198 iii. 12a Pentad verse. requiring correction 2a agnir 'va. the fourth place. verse.] 2a ptirusah. 4c sua. 44c Rest at the fourth place: siond. 151 i. place: tudyd. 130 i. 17b Rest at the . la pdantam. sapdtndn for sdtrun jahi . 12b. la catalectic verse.) . *90. . The first three hymns are of very : unusual length.] 2a sprsd. . 145 vi . 5b smd. 151 iii. 15b tdnuah. 4a hyper130 i. kdtukam Sandhi. . 12d dpa.] *85. or combination of urdhvdh and ut. 224. 10a snid .] 3a A. Ludwig completes the verse by reading asakta 4a perhaps sd a vdsu. 23b smd. 170 ii c . : 173 89. 31a vadhvdh. 9b require correction see 217. 9a hdri. 14a. ucyate. 130 i. 5c irregular cadence. 14b Rest at 160 i. *94. 5b simivdn 6d vlhi (plur. 8c vdyavyan. Sd nd 'nidm. lOd tredhd exceptionally. : 2b rdthia 2d tvisid . 46b svasrudm. 45a midhuah . rhythm. 10b Rest at the fifth place. 8a sd. 6a srdyinih . lOd rdydh utd with Sandhi. trsti. 14a hybrid verse. 9a rudardya. [Although of late date. lie perhaps srotaram. 11a as 6a lid pdda. 146 iii. 145 vi. 8b pdrvd 8d Pentad verse 9d vfsatiam. 15a. 4c two syllables are wanting: perhaps ydsmi. lOd read hdvio yoge. 178. 149 iii . requiring correction. 7a caesura after the third syllable . . 34a perhaps omit etdd . 12c dpia. 38c pdtibhlah is doubtful. 5a apdam . 9c sahd. 20c siondm. 6a purusena. 10a dhaatam. 30c. 14a extended Tristubh verse. 40b -e u. tudt 13d catalectic 4d dthd. 7b rdthasas pdtih. 1356. sahasya. 16b dsvyena. 178. 15c sdrava Wchantu with Sandhi. see 140 iii. 11a Rest at the fourth place. *87. . 9a Rest at the fourth place. (plur. *86. 13d probably dnu dpah. 7c jahl. 13c hypersyllabic verses. the Tristubh verses of this hymn are free from contamination. 135 a. as in the semicadence in 10a. *88. 142 iii a. cf.322 : Metrical Commentary . *95. 93. 7b 92. 19c dahd . . 5d puru. . 9c tanvdh. fourth place. 12a evd. 22b tud . 2d dhatsud. 142 iii a. [Very late rhythm. and the Anustubh verses have the earlier 198 ii. 9c sn-dvdn. 170 i. 46c hypersyllabic opening. 22b 135 a . *X 85114. 145 iv. there are only slight indications of the Epic Anustubh rhythm. 5b dviatyai . E. 91. [In the single hymns which conclude the tenth Mandala those which belong to the popular Rigveda predominate but they only occur in groups on account of incidental agreement in the metre and number of stanzas. 149 ii.with 130 i. [Although the fifth verse in each stanza is a refrain. 17b perhaps yajiianiyoh. 23c] 2c nd u: nd 'not' with hiatus. 5a. For the metre see 242 v . 16a due. 7c diauh. syllabic verse.) ii. 16a probably paurusyena .

but see 151 i. 19c catalectic verses. 10c evd . : : lOd suyavasdd.). cf. la drhyd. *108. 6d mflat. 10a liarsay d. 5d for the caesura see 205 ic. 205 i c . lc Rest at the fourth place. 9b catalectic verse. 7d sd extended Tristubh verse. cf. 12b apue. . 151 ii . verse. bharatd . 11a hybrid verse. 7a cakrmd. 105 ii 99. khuddtd. 7a evd with hiatus . 158 i. 135 b.] 5a irregular 115. 10b vdsibhih. 149 iii. 11a itd 11a Rest at the fifth place. rather than lc represents a stotardm. trimeter verse: Rest at the fifth place. $ 129 ii . verse. 10b stentih with z'm (Grassinann). Numerous duals combined with pdtare. 13b dthd. : . 125 iii b or read karad asmai suastim. or hybrid verse. 10c tanvdh. 168 ii 10a pariffMdafa 142 i. 4a vedd (1 sing. 10c 151 i. 13c Rest at the fourth place. 6a ntt . 3a yacchd. *102. 10b dpia lOcjdnista. 7c as 2a 9a sldd 9c ndrte tudt. 4b smd. 9a bhdrl 10a extended Tristubh verse (adv. . 8b Pentad or Gautami verse . lc vahd. la Viratsthana verse. *102. [Very late hymn. cf. stanzas 3 to 6 see 244 iii . *114. lb nfndm. then vaatdpidya. . siondm 8d su-dpasah. 173 ii 158 i: ddbhia 145 ii b. 4d suapas. 10a nd. .] 7a asvavatim. with irregular break 7d Viratsthana verse. rather than short eighth syllable. 10c Viratsthana the text. 2a su. 8b vdrmd (plur. 8a sd. . lie Rest at the fourth place.). 6c vimdyd. 8c Pentad verse. 129 i b. 2a smd. : include many charms. rather than duurdm. 106. . 12c dahd. ydd. 10a smd . . : 323 135 6. 46. or tdva for te. 3a srutiai. 5d catalectic verses . 6b requires correction. 2d caesura after the third syllable cadence 5c. . 12d coddyatd. . 10b angdhue. lOd tiratd. 6b Rest at the fourth place. 96. [These shorter hymns belonging to the popular Rigveda. 12d ddhd pithidya. 10b Pentad verse. . 4d devebhiah. iva. lOd. *97. 10a 'dsuam probably. See lb Viratsthana verse. 8a two syllables are wanting perhaps supply indra before the caesura . 12c rdjid. 5a d-icchah. 112. lib 3a. 122 *107. See p. 12c svastim exceptionally. 2a Rest at the fifth place. 2c jayatd .x 83-115 . or read yuyotand . 8d as 7d. 4a hypersyllabic verse. 46. lOd vidd fifth place. 7c nd 'as' combined. 6b perhaps yajnaniyam. 7a hybrid *103. lc yddl.) 224 see also 130 iv. all F. lib pipidnd. 3b Rest at the fifth place.% 149 iii: the verse ends haryata d. 7a. 8a nayitd. . . 7d pddid. 10c sue Viratsthana verse rather than 149 iii. 105. For the metre see 244 iii . 5d tesaam .7c Rest at the 113. as end somam piba satakrato tvdm. see p. 212 . probably brdhmanaspdtih. 121 c. 142 iii b verse. 14c. *X 115-191. rather than vdsardm. 4c irregular cadence. 7d for the caesura see 213 ii. 170 iii: caesura after the third syllable. 5b divo 'ntdn. lie by Sandhi pradisi 'rtdsya. . . 7a Pentad verse. 151 ii. . 11a gava142 ii extended Tristubh verse. 20a as 13a. 11a asuriya. not as in 146 ii. 13a heptasyllabic *98. lc. bhavatd 110. 2b tndt . 15c. 100. s4. 2a for the caesura see 151 iii. 5b rdiiia. . 10a tmdnid. *101. : 104. lib 120. 3d s'dcid. ill. *101. 5d devebhiah. .

5d. 3b usdmm. . . See 246 v. but cf. . For the metre see 242 vi la see 6a. 13a probably grhdm. 4b ddah. 5b tujia. la Viratsthana verse (indra type). etc. 2b rdtridh. . lg. 224. 131. 178. 144. 148. *117. p. 4a indrdvdyd. *142. 6e extra verse. 173 v. 8c perhaps rdjanam. 6c dyu. 3b hybrid and catalectic verse . 6a tid: srutia. . For the metre see Ch. . *2d probably su&t sakhydt. 151 iii . 151 i. cf. 5a catalectic verse Ictud. *ld jyok exceptionally. la Rest at the fourth place. abharat 5c. lc yacchd . 2a heptasyllabic verse. *139. . *3b dhdmd (plur. ib. 7d djiam. 7c (the second time) tubhya. 7c hypersyllabic verse. restoration as a trimeter verse. 4b yacchd. 3d ratha2b Viratsthana verse. 46 . lc. 124. 5b krdhl. 7d ndmd (plur. . 4a dsiat. .] 171 v. 145 iii. 224. 178.): brhddiva. See 140 iii. iii. or iydm monosyllabic. 6b catalectic 138. Viratsthana verse . 177 iii. 156. 129 iii. : . 2c hinvd. 8b osd. *157. 145 vi. Metrical Commentary 151 lc. *128. 2c il sddah *130. 6c tud. . la tve. *155. See 246 v 3c vahd 4a omit devdh. 151 i . 170 ii a. 4a as Id. 170 ii c. 129 ii . 9c Rest at the fifth Id pibd .). jidkdh tdm should perhaps begin the verse 7c dcchidra-udhni. 4c perhaps vivartanih. 46. 123. : . rather than ddasih: but see 142 iii a 4a brdhmd (plur. 143. saddnue. 6a manusyah. 4a sd-. *136. lb verse. 150. 142 v 4a gdvo 'va. . 152 i. 2d pdthd. 134. 133. 4c tudvrdhah *146. . 142 iii 6. *129. 177 i. . See p. 8c evd sU. 3c tue\ 173 v. lc dsvam 'va. see 225. 135 a. 5a siendh. See 134. lb dsti is probably to be omitted. : . 9d extended (plur. 152 i 4b manusyah. 147. : . 2a 152 ii. 135 a. 135 a. . 7a sd 2a smd. 5a Rest at the fifth place 5b matsud 5d perhaps rurucuh. *119. App. 242 vi . See 246 v verse. Id needs 135 b. 5b d 151 i. 4c bhavd. 9c omit dditydh. 2 may perhaps be restored as follows sdvitar ydsya perhaps. 8a brdhmd 213 ii . 5c catalectic verse. Id Pentad verse . 4c pdtir 'va. 7a perhaps apna-rdjand. 4b i. [Very late hymn. 4c hypersyllabic verse. *135. : . No. *125. 174 i a 2b for the 149 iv 3c Gautami verse caesura see 213 ii 3b rekaimh possibly. 132. . 6b diauh. cf.): tubhya. 142 v.). 133. 6c Pentad verse. 8a hybrid verse. 169 ii. 11a heptala jydyistham *120. 118. 2d utd. . 6c catalectic verse. . 151 iii. . 7c sxv-dvan. . lb vdatah. . 8a evd. syllabic . 122. ix. lb gacchd. 142 iii b. 5b juhue. 151 i. verse. 129 ii. lc devd. *151. 151 iii 6b hybrid verse 7b add dadhe. 145 ii b 2c arayyam. or read nahi anydd dpyam ah 5a srdyinayah . lc bhard. 9c for the caesura see Tristubh verse. 8a sahasdvann. *152. 152 i. 7a see requires correction. 5c rdthia. 140. *149. 116. 8a ahuanta. place. *145. 3a ddmsistha. lb sasanvdmsah. 2b dadsvate. *121. *127. la /wa sii 109 iv. 6a mdrtidh. lb Rest at the fourth place . . 149 ii catalectic trimeter semi-cadence. 6c namd. rdyiknah. 3d u. 3b duih . 7a. . 8d pdtd.324 mdrtiaih. % 178. 142 v: -ksaat *158. or as 5a. 6a tudydvah. ulha nfndam. 5c rudardm. 4c hrdayydyd. 4b vij?rebhiah 7c iydm perhaps monosyllabic. 4a adyd 6b ydvdyd. 2c bhuuh perhaps: cf. 54 2c Epic Anustubh 126. 2c -bhir 'va. 174 i a. *141. *137. 6c mdam.

2b nitah. *159. 4c cata*174. 2a catalectic verse. 4a heptasyllabic . 173 iii. cf. lc *164. 172. .x 115-191 te 325 hdrah 151 ii. 129 ii. 4c catalectic verse. \ 4a tena no pdhi didyutah. *162. metre is out of harmony with the period perhaps read ydd dsastyd : 158 i. 2c dosanydm ib. *169. nihsdstid 'bhisastid. 2b pdrvato 'va. cf. 4d irregular cadence. 5c hypersyllabic verse. *161. la correction. 2b for the caesura see 2c indro 'va. em. stanza 5 requires gobhiah purusebhiah. *191. See p. 2b nd 'not' with *184.). 2b indro 'va. *163. *173. 151 i. lc hep tasyllabic verse.. 4& jvud. 3c verse. : rhythm than tdpasa ddhy. perhaps sdm na astu *166. 4a probably diauh: then prthvi. 3b updrimd. 4b -dbhidm. *180. catalectic verse. la tidm. 176. 178. Id ava-sdyd. 2b *183. 2c samsdyd\ 3a Rest at the fifth place. 3a the 135a. lb tdpaso 'dhi gives a better *189. but cf. *167. *170. 3b Rest at the fourth place. ib. 124. 5b tud. lectic verse. Id ihd. hiatus. 167 i catalectic verse. 129 ii. $ 151 iii. 2a. 4b m. 2a ahhlvftyd is probable. 46 . savtinam satdm arhati 4c catalectic verse. 3a dhdmd (plur. sirsanydm. 213 ii. 175 ii. tanu. . 3a heptasyllabic verse. 185. 160. *165. 5a tud the second time. 4c vaam. 5d perhaps tud. lc bhard . 3c requires correction 5e extra verse. *168. : nu. 5a tud. *190.

: p. Omit the words except in sdntya good '. . 21. p. Sept. saptati ii 18 5d. 160. Add to the occurrences: (i) Long fifth syllable.' to '. p.. p. pi. is 18. 142 : see also the p. 1. 19. 143 1. 14 p. for ' Bhargavd' read i * BhdryavV 30 5 for i 22 8d read i 122 8d. . p.' 4 from bottom of page. 1. 1. 16 for i 120 15a read i 121 15a. .. Also dual of a verb in -e in i 2 9c (5). Omit the sentence from The fact. p. p. 131 p. 67 19c. The name Pajra Omit 130 iii. 119 lines 1. ' p. 1. For add 105 2a. x *141 3c. (ii) Heptasyllabic verses. and 7 for long read ' ' short. etc.' p. (v) Irregular endings. p.CORRIGENDA AND ADDENDA p. p. ix 26 3a. 84 p. 134 . 65 79 1. omit the exx. 1. more Add vivie x 49 11a.-Oct.' ' p. 1. i p. 84 p. 5. Add u sddah *x 130 2c. 15. 11 2c . 10 add 11 for 11. for Hi read Ik. p.. in i 120. vii 94 8b. . 11 svdh tomorrow And after line 17 hydh 'yesterday. 137 1. Add after 1. 1. 23 read apuvrktd. For dev&n gen. add v 19 5c 5d. 135 . 25. and add v 19 5b. 25 for heavy' read ' ' light. 221. 159. 17 14a 1 55 3b 2 56 5a. 289325). third Mandala has been carefully analyzed by A. 141 line 12 from bottom of page. 85 1. 9. 29 for 44 1-3 read 44 16. Add navati ii 18 6a. 1.). 6.29. i 2 8a 3 9a 3 9c. (iii) Trochaic ending 5 37a. 87 last line. p. ' ' : : usually 20. p. Meillet (Journal Asiatique. (i) Catalectic verses ' ' : add *x 152 5c . 102 125 Add ahian 5. 79 5c. add v 19 3c. 1. 1. The suffix -iya is also found in asuriya (adj. 209 1. ' 170 vti u ' ii read i 168 ii. p. ' 234 apddm read apdam.reading. 1. 126 128 1..' and add vtisisu v 53 4a. : (ii) Short p. 6. 143 6a. 144 lines 20-22 are to be deleted. 101 p. (iv) Syncopated ending. Aufrecht justifies his transliteration in his Preface. 3. 1.' 15 from bottom of page. But indra-agni occurs vi 60 13a (9). 1. In addition to the passages here referred to there are many others as to which the views expressed in the body of the work are modified in the Metrical Comiii a. 3 for p. sudh. p..' The metre of the ' mentary (pp. 1-13. 1. and p. which required in occurrences (except ix 71 2b) for asurya of the text. 25 for 16. 1. add ' ii 4 2d. *x 141 3a. not the ninth. 1897). 1. 89 1. p. viii add p. 1. 161. 61 1. sixth syllable.. also occurs in viii 63 12c.. viii 3 22a 2 .' ' ' ' 6 from bottom of page. tdmmi being locative singular Metrical Commentary. 1. 96 9d has -a in the eighth place...' always hidh. for 172 ' i read ' 173 ii. 136 all 1. 8. viii 67 19a read viii p. We viii find -u long with hiatus before a similar vowel in 39 2b (2d).

237 in triplets 234d Aufrecht. 62b Bhargavl 14c. the seven Agni's. secondary 189191 weak 191. . 140a. Cadence 10a. Th. 66b in triplets 236d Brhati-Satobrhati 220a. Agastya 61d. Names of metres are in italics. 30d. 199 irregular forms 200. a-aorist 32a Aramati 263b Archaic period 104a. 264b Chance. Bardic period 17b Benfey. 220a. 51b. of 75. 201 Brunnhofer. -u. 179-182. J. 58c Atidhrti 249c Atijagati 247b Atri 17c. 265b Bharadvaja 17a. of duals 77d. 296b. Th. 78ab. M. 56d Ahura Mazda 264b Amsa 263b Anukramanl's 2a Anustubh 7d the Atri hymns earliest 170d. 187b Arnold. 267d Agnayah 265c Agni 265c . -i. x b. 174a. Agni Vaisvanara 55d. . 56d. dimeter 161c . 57a. 60bd. Caesura lie. in the order of the Sanskrit alphabet. Koot-aorist 31bcd.INDEX OF SUBJECTS The references are to the pages. 12d. 251a . the older rhythm in the popular Kigveda 168d. 297a Brhaspati 56c. 183. . 264b. 214a. 0. ib. V. xii. P. 192 Catalectie Jagatl 207c Catalectic verses 7c. 296c Bhaga 263b. The arrangement follows the English alphabet but in the case of Sanskrit words the special symbols used are arranged.76. cretic form 53d. 171b. in relation to the nearest English symbols. 268 Aditi 265b Adityah 263ab Apah 57a Apriya hymns 41b. 61b. . his text of the Kigveda 142a. xi. laws of 176c Charms 26d Combination of final -/. 290a Augment with long quantity 129b Aurora 266b Avery. irregular 79ab Composite hymns 8c . iambic forms 198. See also Epic Anustubh Aorist. see Brhaspati Break lid. 266b Brhatl 8b. in the cretic period 169b. 226b . with iva 78cd. denoted respectively by the letters abed. 266d Asti 219a Asceticism 267a Atharvaveda 25d. in the Kanva hymns 168d. strophes 236 Bradke. -u final -a. Anustubh triplets 235b. Ceremonial 264c Chaldaean deities 261a. of trimeter verse 185 . . 265a originally of seven stanzas _ only 242a Astarapankti 248b . Abhinihita Sandhi 77c Adam and Eve 267d. 37a. 263b. H. 67 note. 64b AU/asti 8b. Agni Jatavedas 55d. ix C Bollensen. E. 169a. Thus the whole alphabet employed in this index is as follows a a b c d d e e f g h h i i j k 11 m m n ii n n o 6 p q r r : : m sssftuwvwxyz. lid. of dimeter verse 152a of Trochaic Gayatrl 152d . -a 72-74. 44c. 240d BLOOMFIELD. 37d Avesta 19d. Each page is for this purpose divided into four equal parts. von 172d Brahmanaspati. 258c in the ritual Art in Vedic verse 19b 264a Aryaman 261a Aryans 263d Asvina 261c.

Kanva 17a. H. in dimeter verse 159b. 161d Heroic dialect 24d. 11. 166. 227b Decasyllabic Tristubh 178c. 123-128 Consecutive short syllables in dimeter verse 157. 66. 106c Hiatus 70-80 . 207-8 External form 8d Jagatl 7d. 53c Trochaic Gayatri Geldner. 108-128 Fire. 67b Earth. prevented by use of id. after -d denoted by m 72c of duals 74. 37a. 51. catalectic 207c. 61d. its Lankan. of locatives in -a 73a. in trimeter verse triplets 235a. 149-174 Dionysus 266a Disease 39b Distich 71d Divodasa 17b. see Prthivl and Dyava-Prthivl See Consecutive Eos 266b Epic Anustubh lib. 205. 241b Gayatri 8a. after -a 72-74 after nd as 74b after sd he 74c . 253c Dlrghatamas 17c Duals 72d. 51a. 14b. 67c Kakubh 8b. 20c. . See Epic Anustubh Late caesura lid . 64ab. 291c Grtsamada 17c. triplets 241d Jamadagni 66b Family collections 16a Feminine influences 267b Fifth syllable. short syllables . See also Dyaus and 210c Dhrti 249b Dimeter Brhati 247c Dimeter Kakubh 245d Dimeter Puraumih 245d Dimeter Usnih 162c Dyava-Prthivl Heptasyllabic verses 7c. B. 51c. 323a Havis 265a Dadhikra 56d Daksa 263b Daksina 57a Dawn. 10. 121. with Tristubh stanzas 242c . as magic 260b. Hybrid verse 15d. dimeter trimeter 198a. trimeter : Heaven 260d. 236a 57c. 163b. See also 195-197 Contamination 15a. 262c. indra verses 186c. 266b . 206 Immortality 266a Indo-European deities 262a Indra 261bc. 211c lndravajrd 184b Inspiration 260d Internal Contamination 15c. 173c Ksetrapati 56d KijHNAu. with extra verse 50b. 53c Internal form 9a Irregular cadence. see Usas Danastuti 56b Decasyllabic metres 178d. 64d. 254b heaven 260d Foot 9c. Kakl 267c Genesis 268c Ghrta 265a Glosses 102b Gotama 17a. 296d. Early Caesura lid. 158. 253c Gotama Nodhas 61c. Dimeter verse 7b. as messenger to Jove 261d Kakslvat 64d. 151b. before hiatus 137a Dvandva duals 78b with caesura 180d derivatives 102cd Dvipadd Gayatri 244c Dvipada Jagatl 244d Dvipada Satobrhati 244d Dvipada Tristubh 51b. x d. 209-215 Defective verses dimeter 162a. . 239 Dyaus 261d Dyava-Prthivl 56d. 265a. 257a Decasyllabic verses 7d. 205a Irregular openings. 261a. ' ' ' ' . 227a. 202a. 167 Extended 'Tristubh 208b External Contamination 15b. 220c. 167d Iambic rhythm in trimeter verse 9c. 257a Cosmogonies 27a Cretic break 53d Cretic period 18a Cretic Tristubh 12c. 184c. . C. of final vowels in the prior element 120d. treatment Index of Subjects history 171cd . 75 72d. 227d Kusika 17d. 169a. 172b Kutsa 64d. 182c Gauh 265a Gauiami 212ab. E. in dimeter verse 160c in trimeter verse 204d. 266b . 61a Grassmann. 309d Late Anustubh. Iambic reopening 52d. 65a. 215d Hymn 8c Hypersyllabic verses. 244d Dvipadd Virdj 8a. 53b. 77a. in trimeter verse 182d Final Tristubh verses 242d Final vowels (variant) xi.328 Compounds. lm 73b Homogeneous groups 48 . 53b. 175-227. 235d. 208d 16 Id . 50. 251a. 67c. not sought in dimeter verse 174a Iambic Tristubh 12c Ictus 151c borne by short svllables 156c.

57b. 55a. 142 Positions of Protraction 119a Pragatha 8b Praskanva 64b Prastdrapahkti 247a Pravargya ceremony 43d Pratisakhya's 2a Pre-Vedic verse 174a. 225b Priests. 215. 265b compared with Mithra 263b Mitra. 221a Semicadence of Epic Anustubh 11a Short eighth syllable 185c. 137-141 Sandhi 70-80 SarasvatI 56b Satobrhati 8b. in the second place of trimeter verse 197a. viu Perfect forms 31ab Philosophical poems 27a. 237b. 257a opening 182d strophes 239d verses 213cd Samhita text xi. 261d Quantitative evidence 3d. 253c. 263c Max Muller. 172c Pahcapada Jagatl 248b Panayah 262c Par j any a 56d Pentad Pentad Pentad Pentad Parts (of the verse) 10a. lie hymns 239. of trimeter verse 182.' 64b Protraction 119a. 106d Restoration of text 5b. 264b. 233b. 81148 Rests 13d. See Gotama Nodhas Normal dimeter rhythm 153c Normal period 17d Rigveda proper 4b Ritual. 236d Savitar 263b. its origin 260bc Ritual hymns 57a Rudra 263c Sakvari 247d Sloka lib Oldenberg. 5a. 263a. 322d Lyric metres 8a . B. 258 Linguistic evidence. ix d. 18c Linguistic developemeut 257d. 71a. A. 26c. 265b Madhyejyotis 247a Magic 39b. 60bd Sadasaspati 56d Padapankti 239d Pada-patha 5a Pajra 6ia (also Addenda) Paiikti 8a. 246c Purausnih 245c Purumilha 60d Puriiravas 267c Pilsan 44c. 260b Mahabfhati 247d Mahdpadapankti 247c Popular dialect 24d Popular Kigveda 18d. A.Varuna 56a. 73b. originally craftsmen 260c Pnyamedha 60d. 303a. 308a Opening 10a. F. 192-194 Order of hymns 26b. 256 Mitra 261a. 151c Metrical Commentary 289-325 Metrical families 9a Metrical pictures 252b. 227c Nodhas. 160a Short syllables bearing the ictus 205. 264a Pischel. 265b Mitra-Varuna-Aryaman 263a Mixed lyric hymns 8b Monosyllabic stems 30b Motor-cars 293d Mythological hymns 26d. 265b Secondary caesura 180b. 12a. 68cd within the Kigveda proper 25 2d Long eleventh syllable 204d Long ninth syllable 204cd. 295b. in the MaMpankti 8a. 190. lid. 209-215 Riddles 27a Reopening Nabhaka 65c Natives of India 263d Nabhanedistha 320d Neutral rests 214. 206 Short tenth syllable 185d. lid. 143a. eighth place 203d Proverbs 27a Purarhdhi 263b Purastudbrhatl 231d. 109 Quantitative restoration 108-148 Quantity not determined by position 157b Rbhavah 266c Reduplication 128b Refrain verses 57d of Epic Anustubh 11a Resolution of root-stems 104c. 221a (Addenda) Members 10a. 191. 888b Macdonell. 220d Mana or Manya 17a. 58b. 263c. 22-46 Position before A h ch nn m Ih 141. 28-41. H.Index of Subjects Late Rigveda 4a. 58b &unahotra 61a Syavasva 17a. A. 300d. of dimeter verse 150d. 266-268 . 146a. 248c Muliuxatobrhati 248a Marutah 261d. 296a Medhyatithi 64b Meillet. of Epic Anustubh 163b. 61d. 169b. 204 Shortening of final long vowels 145b . A. in Pentad hymns 239c Lopamudra 267d Ludwig. 300d. 2 note Mamateya 63d.' 202-204 Short sixth syllable 159cd.

106d. W. Vasistha 17a. 263c Vyasva 17a. 268 261d ZuBATf. 262d Vastospati 56c Vdsisthl 181b. 51b. Index of Subjects Emil 267d UrvaSl 267c Sindhu 67b Skandhogrivi 246a Sobhari 17a. 245b Viratsthfuid 14a. Gayatri 10c. 175-227 Tripada Jagati 245c Tripada Trutubh 245b Triplets 234b Tristubh 7d. 146ab Weak caesura 191b Unaugmented forms 30d Uneven lyric 8c.Varuna Vatsa 64b Vac 265a 265b. 51b. Zetfs "jHtrixtaribrhati 247c xii. in Trochaic Gayatri 163a. 240c Virupa 60d. 164d. 64b ViSvamitra 66b VifamSpadd 246c n'starapankti 236d. 8b. etc. 264b. 192a Verses of four syllables 162 Vimada 17a. 265a. 254c. in trimeter verse 182d Syncopation 10b Table of hymns 269-288 Trimeter verse 7c. 229-234 / Yahveh 264b Yama and YamI 267d. J. Vamadeva 64d. 165 Vrtraghna 262c. 170d Viparita 246b Viraj 8a. 108-128 See also Varuna 261a. 59c. 266b Usasa-nakta 266b Umih 16b. 211. in triplets 234d Soma Pavamana hymns 266a Stanza 7d Strophes 235-237 Strophic period 104a Surya 56d. 64b Soma 260c USana 64d Usas 261d. 241d.'62d. 53a. 64b. 246b Visnu 56b. 189 Verrall. 265bc Syllabic measurement 19d Syllabic restoration 81-107 Sympathetic magic 260b Syncopated cadence in dimeter verse 160c. 220cd. 64b 140a . in triplets. Mitra. Uparistajjyotis 236d. A. 265c Syllaba anceps 181a Vala 262c Variant final vowels xi-xiii. 60c. 59c Wackernagel. 20c. 218c Vayu 56b. 11-13. 63. 64b. 265d Vocalis ante vocalem 134d 242ab Trochaic Trochaic Trochaic 172a Trochaic Anw<tubh 53a Brhatl 53a cadence in dimeter verse 160b. 247a Upastiita 59c.330 Sieg. 67c. 65d. J. in trimeter verse 205a Syncopated opening in dimeter verse 150d.

SANSKRIT INDEX The transliteration follows the forms which appear in the Samhita forms being added where this seems required within brackets. lOlcd -a locative 29c. impv.) 37c pi. -du 135b. dthd 112c. 112c -at (-aat) 99d -dtama 29b -dtara 29b -utdm (-atdm) 130a -ate (-ate) 130a -dtliam (-athdm) 130a -a*7i (-athe) 129d -a nam (-an am) 130c <7a7tyd 99c. 78a dccha. -i. a/u7 114c dhobhih (dhabhih) 294d indra (Indara) 98bc.) 32b avar (dvar) 129b dvidhyat 129b awm 129b -<7*' (-aas) 92ab a's^ 38a dxlh 38a a'sAra (dsakra) -*. 78cd . dual 137c -a. 92. 106d -dya (gerund) 112c. 145d a u 131d. 113b -I instr. 140b dtho (dthd) 132d adds 38b adyd. -ami 142b pi. 113c -a. 112c -a 2 plur. sing. 142a dtha. -a 2 sing. perf. adyd 117a. 134ab. 75 . 73a -a dual. 29b. 113d. asdnyd (-i) 87d asydm (4dm) 96c diva (diua) 94d dsvavant 127d dsat (dsat) 127a dsan (dsan) 303a -asi 30c -ase (inf. 113a. and Addenda a (-aa) noun-suftix 90 -a instrum. 140d ddlui. pi. 3 sing. text. -a neut. developed to -o. 93. 136b (Addenda) asmdbhyam (-bhya) 30a asme 136d asmai 136d r//j. 142b -anta 37b dntja (dnia) 84b dpd 126c dpo (dpa) 133d dpya (dpia) 84b abhiyuj 126c abhi (abhi) 126d arista (dr-) 127a arydh (aridh) lOOd av locative 131a av vocative 132b dvidhat 129b dvrnak (a-) 129b dv'ria (d-) 126d (1 p. xiii. 106a -a. 130c -a (-aa) verb-suffix 91 -a 2 sing. 140c dyuh (dyu) 142d dyeji (dyaje) 144a -ayya 30c -dr (-aar) 99d -dv 131d -am (-aam) gen. 28d -a -aam gen. 120a final with hiatus 74. 113c -inn 142b (d 73b asuryd (asuryd) 83c. sing. dccha 112c. -I neut. impv. 106a Indravant 127d -inn (-<) locative and vocative 142b iydm (ydm) 101b 'irya (Ma) 84b iva ('ua) 68b. dcihd 117c. 140c -a 1 sing. quantity 134c. subj. 140d ddha ydd 140d -an stems 88c. combined 77d. pf. perf. 101a 112c.ti -ann locative and vocative 142b ann 2. pi. -a 3 sing.initial 77bc -a final with hiatus 72-74 -a voc. restored a. 113 -a.

112c. 94d krdhl. 140b (-ate) aih adv. i) 73b. sing. 113c. 141d. krdht 118b. -e dual. ustisam. cchardih (cchadih) 143a 113b utd (utd) 133c lipo (upa) 133d uruvydc 143a jdnya (jdnia) 84b j'a/u. 29b. fr. 112c. ofta 112c -Iman 30c cardthd 143a Udna w. 141a krdtuna. 112c. ccfc 141d u lokd 115c -ui. pi. (kan'ui) 83d 114c tredhd (trayidhd) 89a tva (tua) 97a trt kild 112c -taa (-fwa) 88d. 136c -u neut. 105d tva (tud) 85d tad (fad) 95b. 105d to<fm (tudm) 85d. 29d. ending 38b -o -o (-) vocative 131bc Ui 132c dsadhi 40a -au dual 37a. 140b tu. 113c. 95b (gdam) 90c tvd. 140b. 29d. etc.332 -isya fut. 37d -i Sanskrit Index ao/j (gdvah) 90d gdjata (gdvajdta) 90d gopithya. tu 115a. 100c -tva {-tud) gerund 29d. (isdna) 142d carkftya (-tia) 84c cyautnd (cyautand) 99a 3 115c. 305d end 112c. tubhyam 30a tuvimaghd 125b fr%a (tritfya) 143b fWM 142a fr. -trd 120c. 136ab instr. 141a -tai 3 sing. 141a -tvana 30c tadf (tudt) 97a. 130b 115c. 112c. 86c tvdyd (tudya) 97b. -u.(tud-) 86c -tvl giihya (giihia) 84c (-tui) gerund 95b . 38a. 114b -ui loc. gavapithia 90d before hiatus 134c. (krdtud) 88a kriyasma 306a (/ctfa. 113b. -tana 113c 130a -effce (-af/ie) 130a -ena. -fi instrum.N//.su 114c fe'wa. combined 78ab -ithdm (ithdm) 130a \m (t. pi. 113c. subj. combined 78b etdd 38b -<?te -fa 2 plur. 140d tubhya. 142c 37c tyd -ft/a (tid) 97c m (-tia) after light syllable 84 -tyd (-tid) gerund 95b. 113c ii jya' ( jia') bow ' 89a j?/a (j/a) 'strength' 97a jydkd (jidkd) 'bow '89a jyestha 91d jj/oA(jirffc) 89a jananta 134c a juvah 134c fl fdda$ 134c (Addenda) -urn Ui 132c fi ? 36d nd 36d -fa. 105b Una. 140d before hiatus 134c. 141d. -end 120b ?/u7 112c. Una /cvo 120d. 142c denoting hiatus 72d 6d. xiii. fi?za 120d. 140c -tra. 140b evd. 114a -tana. ja/if jig'tvdifis 118b 143a urviyd (uruyd) 142d usdsam. 132b tanvah (tanuah) 83d tanve (tanue) 83d -Mwe 32a tort. -tori 112c tacasya (-yd) 83d tava w 134b tdtisydh (tdviseh) 131c tdsmai 136d -f/. 142c kavyd klla. taam 101b fat) gddhya (gddhia) 84b aa"/> jiaffl (gdah) 90c (torfm) 85d. m ih 141d. 141a h 6d. 142d i gdmdn (gdvamdn) 90d o/ta. 113b -I dual. evd 73b. 140d -u neut.

130c manusya (manusia) 83d waai 37b mdrta (mdrtia) 101c mahdnt: forms mahdnah. mahimnd 88c. -thana 113c -nn 141d -nya ddbhya (ddbhia) 84b (-nia) after light syllables ddmya (ddmia) 84b das (das) 143b das (das) 143b dasvdihs (ddsudihs) 95a ddsa (ddasa) 100a ddsvat (dadsvat) 92a -p causatives 38a pdnthdm (pdnthaam) 90d pavitdr (pavitdr) 143b pasavya 83d pdntam (pdantam) 91b pdvakd (pavdkd) 143b pitrdh (pitardh) 88b pipdya (pipdya) 128d dasvdms (ddsudihs) 95a dive-dive 29d (divid) 84c df'dyaZ (didiat) 88d divyd didyana (dtdiana) 88d d?dae (dldie) 91c didhyat (didhiat) 88d didhydna (dldhiana) 88d didhye (didhie) 91c durya (duria) 84b <ta?-a (daard) 100a dWM 142a deydm (da-idm) 91b devdvant 127d devavt. 101c ma find 88c tiuihya. mil trnh na ci 112c -md nrndm (nrndm) 143b -ne (-ane) 88c . perf.i). -thd 2 plur.<. 140d nu (nuu or nu u) 100a : (ana$) 88c astu 40b ndrya (ndria) 84b ndvyasyd (ndvyasl) 87d niaininn. 101b (and Addenda) destha (ddyistha) 91d dsnd (dayimd) 92a dwai (didm. 85d. dydam) 90cd. 134a barhimd 143c brhaddivd. devdvi 125c. 97b d?/n< (dia) 100c dyaiih (diauh) 97b dt. 99 maghavann 142b maghonah 90d. 136b devdhutya (-ti) 87d d<?van gen. plur. 100c. 140c -tha. 120a liaiih (nmii. (mddhud) 88a -wan stems 88c. 388 91c 136d netdr (nayitdr) 91c n^aw -tha. 92a (|wd w) 132c. 141c dheydm (dha-iam) 91b bMsvat (bhadsvat) 92b onuwa 139b -fenna/j (-bhiah) 93d. 100c -dnt>e (-rf/iue) 94c -dhvai (-dhuai) 94c nd 'as' combined 74b na 'not' with hiatus 73cd. 127c maghdnoh 90d mddya (madia) 84c mddhva fern. pi. 113. mdhyam 30a 112c (ma tat oh) 88b mddhvt (mddhui) 95a 1 plur. 94 -bhydm (-bhidm) 93d. 86c. -ma 1 p. -thd 2 sing.Sanskrit Index tve (tue) 17&. -dM imperv. 106a na. 143d -na/i ndma nani.-ni stems preceding vowel short 131c ndsatya 99b. n-n 115a.a (dna) 89a dvi/i (da?/?) 97b ptpydna (pipiana) 88d pipy and (pipiana) 88d purutdma 124b purubhuj 124b purma (purusa) 143c i>aru 114d purudhd. hdbhih 143d mahdm 29b mahind 88c. brhdddiva 143c brhddbhih (brhdbhih) 143c 6Ms (6nads) 92b dhdnya (dhdnia) 84b dhanv (dhanu) 96d dhdnvan (dhdnuan) 96d -rfM. 118d. naM 139c -na. perf. 94 -ma. 114.i) nd (nd) 133a {iiaiuih) 90d 84 -thana. 74ab -fia (-ana) 97. purudhd 112c par 100b pusan 142b prihivf (prthvl) 101b prthivydm 136b jprajtf 40c pratlvydm 136b prdmatya (prdmati) 87d _prri prestha (prdyistha) 88d. 119d -maife vocative 29d maksu 114c dhestha (dhdyistha) 91d -d%ai 32b -dhvam (-dhuam) 94c.

srudhi 118b. 140d : . 96a yayam (yayidm) 91c //.srdn (foonj 89a (Addenda) 95c rajfi- 87b sd 'he' combined 74c sakhyd (sakhid) 84c 37a 1$ 141d -ra (-a) 94d.95b sd. 106a reknah (rdyiknah) 100b reknah (rekanah) 99a rai: cases formed from the stem rddasoh 101c rddasyoh (rddasiyoh) 136b Z srestha (srdyntha) 92a ii'dA (sud/?) . a}uf 32a sonw (s/id) 101c sano (sdnav) 131a sasdha (sasdha) 129a si 'bind': stem sia.suffix 95a. 141a sre'rc? yimne 136d yena. visvdhd 112c visvdha (visvdhd) 112c.tr. srnudht 118b . vdstav 131a -rd (-ad) ins.srim/if 118b.vanam 130c -vane 32b vavdna 306 a : sakhydh (sakhidh) lOOd sdptivant (sdptivant) 127d sriVra 38b sasavdihs (sasanvdihs) 144b sdsmin (sdsmi) 144b -sd/j 125c ace. 112c. 95d yatdh (yaatdh) 91c yati (yaati) 91c -yam (-iam) locative 37a. after light mna -y (-mand) 88c stems 30c -ya (-ia) noun-suffix 84. yend 120d. 87 -yd (-id) optative 96c -yd (-id) gerund 95b. 141a c initial 144b s//a? (sj'flf) 91c smdsru (smd. 130d -rl (-z) 95a rr/' 126b -rrta 126b -iTdft 125b vrsann 142b vfsdnam (vr*anam) 130d -w>$ (-ho/?) dual 88a -rna/i (-rana/f) 88c -tme (-vane) 88c rydnti (vidnti) 89a ra'si ydsmai 136d -yd (-to) noun-suffix 84. 95d yoh (-ioh) 87 -ra (-ara) 97-99 rdnya (rdnia) 84c rdf/ta (rdthas) 101c -ratha 125c (srdyini) 89a rdthya (rdthia) 84c rdjann 142b rudrd (rudard) 98d. yddi. 140c -yah (-idh) 37a. -sdham 130c sahasdvann lOld. 85cd. sing. 141a yestha 92a -r/at (-ia/) dative 37a. sing. 142b sddlw 114d sddhu. 140d yavyti (y arty ti) 84b. visvddha 112c visvdha. 136b ydsmin (ydsmi) 144a rata (vdata) 92b vdtdpya (vaatdpya) 100b ram (uaam) 100b -ram (-warn) 96a *> (rdar) 100a 131c vdsimant (vdsi-) 128a vibhvdsdh 124d visvddha. loc. 117d. lOOd -ya (-7) noun-suffix 83d -yd (-id) instrum. lOOd -yd (-la) noun -suffix 83d -ya verb. su 116cd. 201d. lOOd -vaj[i (-ua$) 100d -rah vocative 29d -rat 32c -van stems ace. 305d shnictutt (ximivant) 127d Itrto' 114c sura (sdura) 100b snnidhi.334 mtim (mdam) 100b thu 114c mllhvtfms (milhudms) 95a 7)i> Sanskrit Index vasavya (vasavyd) 83c -rasu 125b vdsudhiti 124c vasurdc 124c vasuvid 124d vdstu. 100c -yah (-iah) lOOd yajhanyam (-nfyam) 136b yajhlya 83d yddi./> (U"ji) 84b yuyudhdh 129a yitnitydh (yuvateh) 131c fcbya (sdsia) 84b iaivddibham 199b.Sdru) 101c srdtya (srtitia) 84c sntdhi. syllables 101a -ra?Hs (-wawts) 95a tnrl (mfl) _143d mrlikd (mrlikd) 143d memyat (memiat) 88d mo (md u) 133a -mnah (-manah) 88c 88a.

-*va -*va. . mmnayu 128b suvdnd (svdnd) 89b sfnjavasa (su-) 127b (sua. M. 97c -sya future 37d sydm (sid)n) 96c sydma (sidma) 96c syond (siond) 89a svd (sud) 97c CAMBRIDGE : PRINTED BY JOHN CLAY. AT THE UNIVERSITY PRESS. 141a hiydnd (hydnd) 102a hiranydyd (-ydyi) 87d hrdya (hfdia) 84b %aft (Atrffc) 89b (Addenda) stavdn (sdtvd) 144c sma. 120b. -sua) -*va 112c. sing.Sanskrit Index sudds (sudads) 92a -sva. svad (mud) 95c 8var (suar) 83b svardj (svardjd) 102a svarga (xitaryd) 89a svasrdh (svasarnh) 88b sura (sriura or stfruz) 100b st1n/a (suri'a) 85b -se 1 pers. 113a. smd 112c. 885 Mc Mi).' -M 118cd. sumatyd (sumatt) 87d sumnayu. 31a -se infinitive 32b -sai subjunctive 37c svdhd 265d -fcan 125d hdvya (hdvia) 84c havyd 84a -to.A. 140c syd (sid) 32c.

%. .

.

.

& H 0) > CM O 1 SCO -P * o H T> T5 * m 0) > en v*v aJ CO h3 I < .

Sign up to vote on this title
UsefulNot useful