THE HARVARD ORIENTAL SERIES
VOLUME TWELVE
HARVARD ORIENTAL SERIES
EDITED
WITH THE COOPERATION OF VARIOUS SCHOLARS
BY
CHARLES ROCKWELL LANMAN
CORRESPONDING MEMBER OF THE INSTITUTE OF FRANCE (ACADEMIE DES INSCRIPTIONS ET BELLES-LETTRES), ETC., PROFESSOR AT HARVARD UNIVERSITY
IDolume twelve
CAMBRIDGE, MASSACHUSETTS
s Tbarvarb THntver0itE
1912
THE
PANCHATANTRA-TEXT OF PURNABHADRA
CRITICAL INTRODUCTION AND LIST OF VARIANTS
BY
DR. JOHANNES HERTEL
PROFESSOR AM KOENIGLICHEN REAL-GYMNASIUM, DOEBELN, SAXONY
CAMBRIDGE, MASSACHUSETTS
bp tmrvarfc
1912
The volumes of this Series may be had, in America, by addressing Messrs. GINN AND COMPANY, at New York or Chicago or San Francisco, or at the home-office, 29 Beacon Street, Boston, Mass. ; in England, by addressing Messrs. GINN & Co., 9 St. Martin's Street, Leicester Square, London, W.C. ; and in Continental Europe, by addressing Mr. Otto Harrassowitz, Leipzig.— For the titles and descriptions and prices, see the List at the end of this volume.
JAN 1 9 1967 'SITY OF T
PRINTED FROM TYPE AT THE
UNIVERSITY PRESS, OXFORD, ENGLAND
BY HORACE HART, M.A. PRINTER TO THE UNIVERSITY
First edition, 1912, One Thousand Copies
TO THE MEMORY
OF
RICHARD PISCHEL
CONTENTS
PAGE
PREFATORY MATTER ix
Preface ix
Delayed appearance of the volume ........ ix
The material intended for volume xii divided between volumes xii and xiii ix
Designations of the manuscripts in the list of variants . . ix
The list of variants refers directly to the manuscripts themselves . . ix
The numbering of the single tales in volume xi ..... ix
Editor's non-acceptance of corrections of real errors . . . . x
Purnabhadra's attitude towards his sources . ... xi
Acknowledgement of obligations . .xii
Key to Tables I and II of volume xi 1
Aksara- forms in the manuscripts * and bh ...... 1
INTRODUCTION TO THE EDITION OP PURNABHADRA'S PANCATANTRA . . 5
Chapter I. Survey of the single recensions and of their MSS. . . 5
Pedigree of the recensions of the Pancatantra, in tabular form ... 5
Key to the pedigree under 25 headings .6
1-6. Kashmirian ........... 6
7. North- Western copy (N-W.), represented by
8. The Pahlavi recensions 7
9. Pseudo-Gunadhya ; 10. Somadeva ; 11. Ksemendra .... 7
12. North-Western epitome (n-w). From this flows
13. The source (n-w1) of 8
14. The archetype of the Southern Pancatantra 8
15-19. Manuscripts of the last, grouped under 5 sub-recensions . . 8
20. A lost manuscript (n-w2) agreeing with n-w1 . . . .10
21. Nepalese recension (v) . .10
22. The HitopadeSa 10
23. Textus simplicior (Jama) : H-class ; cr-class (Kielhorn-Biihler) . . 11
24. Pnrnabhadra's text (Jaina) .14
25. Later contaminated Jaina recensions ....... 15
25*. Textus simplicior, interpolated from Purnabhadra's recension . . 15
25*. Purnabhadra's text, interpolated from textus simplicior, &c. . . 15
25s. Single books of different recensions combined 16
25*. Other recensions (25* a to 254g) contaminated with Jaina recensions . 17
25°. Jaina recensions moulded into other forms ..... 20
25'. Versions from which the frame-stories have been eliminated . . 20
b
viii Contents
PAGE
Chapter II. Purnabhadra, his time, his work, his language . . 21 § 1. Previous statements . . . . . . . . . .21
§ 2. The date of Parnabhadra's recension 22
§ 3. Parnabhadra's work 27
§ 4. Parnabhadra's language . ...... 31
Chapter III. Account of the manuscripts on which this edition is
based 37
§ 1. Description of the manuscripts ........ 37
Manuscripts bh, N (first group) ........ 37
Manuscripts *, PL1, Pr, M, p (second group) ...... 88
Manuscripts A, Bh, 4> .40
§2. Value and mutual relations of these manuscripts . . . .41 § 3. The manuscript N goes back indirectly to bh . . . .42
§ 4. The manuscripts PL1, Pr, p, and M go back to * . . .42
§ 5. Critical discussion of the manuscript A 43
Discussion illustrated by text of Tale III, viii, Self-sacrificing dove . . 44 § 6. Critical discussion of the manuscripts Bh and 3> . . .56
Discussion illustrated by text of Tale V, v, Ass as singer .... 67
§ 7. Books I to III in manuscripts Bh and 3> 67
Discussion illustrated by text of Tale I, xiii, Lion's retainers outwit camel 68
Chapter IV. Principles which guided the editor in constructing
the text 73
§ 1. Basis of the text of our edition 73
Manuscripts bh, N, A ; *, PL1, p, Pr, M ; Bh, * . .76
Manuscripts bh and * differ very little from Parnabhadra's autograph text 77 § 2. Emendation of the text ......... 77
LIST OF VARIANTS 83
The variants are given for each page of text, from page 1 to page 290
INDEX or STANZAS ........... 201
It covers all the stanzas of Parnabhadra's text (HOS. 11) And also those of the textus simplicior (Kielhorn-Btthler)
ADDITIONS AND CORRECTIONS ...... . 230
Additions and corrections to volume xi .
Additions and corrections to volume xii 231
PREFACE
Delayed appearance of the volume. — This volume appears later than the editor of this Series and the author expected. The printer's copy was ready several years ago ; but circumstances for which nobody is to blame prevented its being issued earlier than now.
Distribution of the material originally intended for this volume. — The general plan of this volume may easily be seen from the table of contents. As appears from page xvii of volume xi, it was my original intention to give with this volume parallel specimens of the text of the various recensions. Technical reasons, which have to do with the size and shape of the sheets on which these specimens are printed and with related problems of the bookbinder's art, made it seem more practical to issue the specimens in a little atlas, and to designate the atlas as volume xiii.
Designations of the MSS. in the list of variants. — In the ' List of Variants ', the designations of all the manuscripts (sigla codicum) of which the variant readings are given in full have been printed at the bottom of each page, whereas the designations of manuscripts to which only occasional reference is made are there omitted.
The list of variants refers directly to the manuscripts themselves. — The occasional quotations from the Berlin manuscript K (see below, p. 15, 252) are given from the collations of Benfey and of Professor R. Schmidt. Unless the contrary is expressly stated, all the other references are to the respective manuscripts themselves and not to collations or to secondary copies.
The numbering of the single tales in volume xi. — In volume xi the superscriptions qraTT ^ and so on, at the head of the single tales, are given in the usual Occidental fashion. This is a deviation from the procedure of the manuscripts. Partly by way of justification, and partly for the sake of beginners, it is well to state here that, in the first place, Sanskrit manuscripts never have such story-numbers at the beginning of the single stories, but always (if they number the stories at all) at the end ; and, in the second place, that even the best manuscripts are inconsistent in this matter.
By way of illustration, a few details may be given. Our MSS. of the bh-class have no story-numbers ; the numbering contained in the MSS.
x Preface
of the *-class and in the fifth tantra of Bh are given in the ' variants '. See, for instance, variants on p. 6, 1. 27, p. 14, 1. 5, p. 21, 1. 11, &c. Since the interpolated stories are in most cases numbered before the frame-stories, the numbering of the MSS. disagrees of course with that of our printed text. The frame-story I, xv of our edition, e. g., is numbered as xxi in the MSS. of the *-class, whereas the intercalated stories xvi to xx are thus numbered both in the print and in the MSS. In tantra v, the frame-story is numbered as i in the *-class, and all the following stories are numbered down to xii. In this case the frame-stories are throughout numbered in the MSS. before the intercalated stories, whereas in Bh, which in this tantra is a MS. of the textus simplicior, the frame-story is numbered as i, our first story as ii, our third story, in which all the following tales are contained, as xii (see variants on p. 289, n), our stories iv to ix in like manner as iv to ix, and our x as xi ; whereas our xi, which is inserted in x, is numbered in Bh as x.
Editor's non-acceptance of corrections of real errors. — At the end of
this volume, some additional corrections to vol. xi are appended. Several learned friends of mine have sent me emendations which they will not find among these corrections. I need not say that — thankfully and carefully — I took all their proposals into consideration ; but in the course of my critical work I have become extremely cautious in correcting the readings of good MSS. Nothing indeed could at first sight be more convincing than the emendation f*}U!!9J)tmtft, which one of these scholars proposed instead of f^BJnprffo, as my text reads p. 2, 1. 12. But as the best MSS. of both the recensions of the textus simplicior, from which Purnabhadra took this passage, confirm the reading of the MSS. of his own recension, this con- jecture is inadmissible; see variants on p. 2,12.
The same scholar proposed to read with the editions of Kosegarten (V, 49), Blihler (V, 60), Jivananda VidySsagara (V, 60), and Kasinath Pandurang Parab (V, 58) f*nft?if instead of f*HI«!li, as our text has in its stanza V, 46 b. But again our variants (on p. 273, l) show that Purna- bhadra took the wording of this stanza into his text exactly as he found it in his sources, and we have no right to alter what he approved. There can be no doubt that in this as in other cases the later printed editions simply follow that of Kosegarten ; cp. below, p. 53, and Indogermanische Forschungen xxix, 215 ff.
I now regret that I followed Parab in correcting the chandobhanga in stanza II, 155 a. This stanza is absent from all the other recensions of the Pancatantra including the textus simplicior. I found it in a metrically correct form in Parab's Subhashita-ratna-bhandagara and in his edition of
Preface xi
Ballala's Bhojaprabandha, and as the correction seemed to be an unavoidable one, I adopted it. But later on, I found the same stanza with exactly the same chandobhanga in Jivananda Vidyasagara's edition of the Bhoja- prabandha and in the metrical version of the Campakasresthikathanaka ; see below, Variants on p. 163, 13. And when my ' Variants ' were already printed, I found again this stanza with its chandobhanga in the printed edition of Devavijaya Gani's Pandavacharitra (see Yashovijaya Jaina Granthamala, 26, p. 152), and in three old MSS., the one containing Dharmacandra's Malayasundarikatha, the other two Hemavijaya's Katha- ratnakara (story 211). Hence it is evident that this stanza was current in its faulty form, and that this form should be restored in our text. As here, Purnabhadra in several places took over into his text anomalies of his sources ; see below, p. 30 f. and p. 36. All these cases should be carefully observed, as they afford us one of the best means for constructing a pedigree of the different recensions, and for finding out their most trustworthy MSS.
In my emendations I always carefully examined the best MSS. of att the old recensions, including both classes of the textus simplicior, and I beg my critics not to venture conjectures of their own, without comparing the same sources and without taking into due consideration their genetic relations. The text of Kielhorn and Blihler cannot replace the MSS. of the textus simplicior, as will appear from our parallel specimens, from pages 58 ff. of this volume, and from the occasional quotations strewn over my notes.
Purnabhadra's attitude towards Ms sources. — Purnabhadra no doubt knew Sanskrit well, and if he had not been renowned for his pdndityam, no minister would have entrusted him with the revision of so celebrated and widely-known a nltitastra as the Pancatantra already was in Purna- bhadra's time. Moreover, his work would not have been so widely circulated and copied again and again to even recent times, by Jainas as well as by Brahmanas, if it had not been approved by the most cultivated people of his own time as well as of later times. Hence it seems to me now quite possible that he was well aware of such anomalies as he took over into his text, but that he intentionally refrained from altering them. In stanza 5 of his Prasasti he says :
xii Preface
This shows at all events that he had a great consideration for his sources, which, as appears from our parallel specimens, he followed pretty faithfully.
Acknowledgement of obligations. — Once more I have the pleasant duty of making public acknowledgement of invaluable help and kindness received from very many scholars. First of all I must thank Geheimrat Professor Boysen, Director of the Leipziger Universitatsbibliothek, Professor Miinzel, Director of the Hamburger Stadtbibliothek, and Mr. F. W. Thomas, Chief Librarian of the India Office Library, London. These gentlemen sent me the Pancatantra MSS. preserved in their respective libraries and permitted me to use them under the most liberal conditions. To Mr. Thomas I owe the possibility of collating again the London MS. A, and the Poona MSS. bhPBh. Moreover, this scholar procured for me copies of the most valuable MS. h (see below, p. 12 f.), of the Ulwar MS., and of the MS. preserved in the Raghunath Temple Library (cp. p. 231 of this volume). Both these latter MSS. are inferior fragmentary copies of Purnabhadra's recension. But to know this is a great relief for an editor, for whom nothing can be more painful than the thought that there may still exist some MSS. of very great value which he is not allowed to use.
To Professor A. A. Macdonell of Oxford I am deeply indebted for sending me the Pancatantra MSS. of the Max Miiller Memorial. Besides, he as well as Mr. Thomas collated for me a passage of the originals of Kosegarten's MSS. BCDEF which were not in my hands, when I needed them for this single passage ; see below, p. 44 f . To Mr. Premchand Keshavlal Mody, M.A., LL.B., of Ahmedabad, I owe the use of the MSS. pr (see p. 12) and Pr (see p. 14). Sastravisaradajainacarya Muniraj Shrl Dharmavijaya Suri, the founder of Shri Yasovijayajainapathas'ala in Benares, and his head disciple, Muni Indravijaya, who unite in their persons the truly Indian pandityam with a keen sense for philological criticism and with a far-sighted benevolence to all the scholars interested in Jaina literature, have sent me many valuable Jaina MSS. and books necessary for my further work, and have given me many items of informa- tion of the utmost value which it would have been difficult or impossible to get in Europe.
In 1910 the Munich Academy awarded from the income of the Edmund Hardy Foundation a prize of one thousand marks for my Contributions to the History and Criticism of the Pancatantra Literature (Arbeiten zur Geschichte und Kritik des Pancatantra). Since Theodor Benfey was a member of that distinguished corporation, and dedicated to it, over half a century ago, his celebrated pioneer work in Comparative Literature, entitled ' Pantschatantra ', and since so eminent an authority in that same
Preface xiii
field as Geheimrat Ernst Kuhn is closely and no doubt authoritatively concerned with the administration of the Hardy Fund, it is a matter of deep satisfaction to me that I am here able to record the Academy's approval of the way in which I am continuing the work of Benfey.
If I am able to continue this work, I owe the happy privilege in largest measure to Professor Charles R. Lanman; for at his instance an inter- national memorial signed by seventy subscribers (more than a third of the signatures came from India), was addressed to the Koniglich Sdchsische Gesellschaft der Wissenschaften with a petition that this corporation request the Koniglich Sachsisches Ministerium des Kuitus und offentlichen Unter- richts to allow me the leisure necessary for bringing my work to a satisfactory conclusion. Rektor Professor Dr. Curt Schmidt of our Real- gymnasium in Doebeln most kindly supported this petition, and the Royal Ministry granted it. I may now hope to complete the literary-historical part of my undertaking, as with the present volumes (xii. and xiii. of this Series) I am completing the philological part thereof.
I need scarcely add that for these volumes, as for volume xi, Professor Lanman has laid me under deep obligation by revising my Introduction and other preliminary matter in respect of its English style, and by arranging the contents of the volumes with his well-known editorial skill.
Mr. J. C. Pembrey, Hon. M.A. (Oxon.), the Oriental Reader at the Clarendon Press, has not only done his work with his unfailing care and pains, but has also given me many valuable suggestions which I was glad to follow.
To all the above-mentioned gentlemen and corporations, and to the Royal Ministry, I here record my heartfelt thanks. The great and sympathetic interest which they have shown in this large and laborious undertaking, gives me courage to address myself to its second and perhaps more difficult part; and this, when complete, will, very likely, and as I hope, turn out to be a History of the Indian Narrative Literature.
JOHANNES HEBTEX. OrossbaucUitz bei Doebeln, Saxony, December 10, 1911.
KEY TO TABLES I AND II OF VOLUME XI
Specimens of the MSS. * and bh in facsimile. — In volume XI, between the end of the introductory matter (p. xlviii) and the first page of Purnabhadra's text, are inserted two Tables, containing collotype re- productions of fifteen specimen-parts of the MSS. 41 and bh. These facsimiles are designated as ' No. 1 ' and so on, and are thus referred to in this Key. They are especially useful as showing the old aksara-forms used in our MSS., and as making it very clear how some of them were easily liable to be misread by later copyists.
Facsimile No. 1 shows a verso-page of *, corresponding to the passage beginning svayutMntikam (169, l of our text) and ending with bhadra (170, 12). Our reproduction is a trifle smaller in size than the original. The original has three red spots : one in the centre of the middle square beneath the (correct) leaf-number 60, and encircling the small hole ; one in the right margin, covering the figure 5 of the (wrong) leaf-number 56, of which the figure 6 and part of figure 5 have been lost in the course of time with part of the margin ; and one just opposite to it in the left margin.
Facsimile No. 2 shows a recto-page of *, corresponding to the passage gatam vyddham to krtah inclusive, 173, 2 to 174, 23 of our text. The original has only one red spot, namely, in the middle of the blank square.
Facsimile No. 7 shows a verso-page of MS. bh, corresponding to our text 125,11 rtkdni to 125,29 samdptam ce inclusive. The original is a trifle larger than the facsimile, and has neither the blank square in the middle of the page nor the red spots in the middle and in the margins.
The Jain diagram for the sacred word arham appears in ¥, see No. 2, 7 e. — Aunsvara appears at the end of the line in *, see No. 2, 12 g. Cp. Variants 183, 11. The same character is used merely to fill out the blank space at the end of the line in *, No. 8, 6 g and 7 g. Variants thereof in *, No. 1, 1 z, 4 z, 9 z, 14 z ; No. 2, 7 g. — Red markings. In the originals of * and bh, the words uktam ca, api ca, the ends of the single padas, and other important places are coloured with red.
Aksara-forma of the MSS. * and bh. — Under the headings of such forms as are for one reason or another of interest, are now given references to the facsimiles and to the places thereon where such forms may be found. References for ¥ are on the left ; those for bh are on the right.
Key to Tables I and II of Volume XI
Aksara-forms.
Initial i : Post-consonantal « :
Post>consonantal ai: Post-consonantal o :
Post-consonantal au : Modern forms : gga:
gha, old form : {/ha and ppa :
gha, modern form : gha and tha :
jha: jjha :
tit and nu :
In facsimiles of MS. ^.
No.l,2ik; 7e; 8n; 14r. No. l,2b dhe; 2m, 5r ye; 2u se ; 3 y le ; 5 d Me ; 5m sthe.
No. 1, 4bg rai; 6u, 12 1 dai; 14 c tat ; 10 a smai ; 14 e thai
2g yo; 3f 3 n <o ; 3 s go ;
No. 1, 1 m tro ; 8 1 sto ; 3 v 6/10.
No. 1, 1 q mau; 3q, 8b dau;
6 z saw.
No. 1, 11 n ro; 11 z yo ; No. 2,
7 b fyo.
No. 2, 15 f rggam (distinctly two ga's). The usual form No. 15, 2 in durggam.
No. 8, 4 b c rgha ; 7 a pfea ; No.2,9bg*u; 10f,Ucgha; No. 9, 1 a gha.
The old form of gha is distinct from that of ppa, but easily confused with it : cp. No. 1,
1 c rppi ; 4 h rppa ; No. 8,
2 a, 6 d rppa.
No. 1, 5t gha; No. 2, 1 c, 2a ghu ; 4 d ghra.
The modern form of gha is distinct from that of tha, but easily confused with it : cp. No. 1, 1 v tham ; 6 y, 13 s tha ; No. 8, Igtha; Qftham; lOe rtham.
No. 1, 15 c jhum (quite distinct from leu 15 e and g) ; cp. jjhi, No 12, 2 a.
No. 9, 3 b jjha (cp. Buhler, Palaogr., Table V, v-18 from inscr. ca. 807 A. D.) ; No. 12, 2 a jjhi (cp. Btihler, Palaogr., TableV, xiy.xviii, xix, xxi-18).
Very nearly alike. For tu, see No. 1, Id, 3k, 7x; No. 8, 8b ; for nu, see No. 1, 15 b.
In facsimiles of MS. bh.
No instance.
No. 7, 1 middle tye ; 1 first quar- ter ye ; 1 third quarter ste ; 1 fourth quarter te, ke ; 2 first quarter sre, &c.
No. 7, 5 and 6 middle, and 9 first quarter thai ; 6 first quarter kyai ; 10 third quarter dai.
No. 7, 1 fourth quarter no ; 2 first quarter yo ; 3 second quarter vyo, dyo, to ; fourth quarter vyo, dyo.
No instance.
No. 7, 7 second quarter ptai. No instance.
No instance in the facsimiles (and none in the entire MS.).
No instance.
In bh, the form of tha is like that of our printed texts : No. 7, 1 a rtha ; 2 third quarter tha ; 5 middle thai.
No instance.
No instance.
For nu, see No. 7, 7 first quarter.
Key to Tables I and II of Volume XI
Aksara-forms.
t and n in conjuncts : tra:
In facsimiles of MS. \^.
Very similar. Thus :
No. 1, 2g, 5k; troBf; <H9b.
In facsimiles of MS. bh.
Cp. nnr, No. 7, 1 first quarter ;
tra, No. 7, 6 first quarter. No. 5, 1. tsa, No. 7, 2 second quarter and
9 first quarter ; No. 5, 3 end. No. 7, 2 last quarter, exactly
like sru. Identical in bh. For tya, tya,
see No. 7, 4. first half, four
examples ; tye, No. 7, 1 middle ;
tya, 2 end. For nya, No. 7,
2 third quarter ; 3 beg. ; 6
middle. tha after s, cha, b before dha, and s before ta or tha or i\a, are very similar. Thus :
tvd: tsa:
stu: tya and nya :
No. 1,2 x, 8 h, 13d. No. 1, 8c; fcw No. 1, 12 a; tsne No. 1, 14 v.
Quite distinct in \fr. For tya, see No. 8, 3bc; tye, No. 1, 6 be, 12 d; tyu, 2 q. For nya, see No. 1, 9 c, 10 v.
stha: No. 13, 3a; sthe, No. 1, 5m;
stha No. 3, 2 middle ; 4 end.
ecfco : No. 1, 4 u v ; No. 10, 2 and 4.
bdha: No. 1, 11 p; cp.No.10,1 second
half.
ita: No. 1, 10 q, &c.
stha: No instance.
sna : No instance.
ddha and dtw : Identical in i/c. For both, see
No. 1, 14 q, niryativasad vad-
dha.
stha, No. 7, 2 third quarter;
sthi, 6 second quarter. cchi, No. 6, 2 middle. No instance.
No. 7, 1 third quarter (twice).
No. 6, 3a.
No. 5, 2 middle.
Distinct in bh. For ddha, see No. 7, 6 first quarter ; ddhya, 1 third quarter; ddhi, 8 first andfourth,andlOfirst quarter. For dvi, see No. 7, 7 second quarter.
INTRODUCTION TO THE EDITION OF PURNABHADRA'S PANCATANTRA
Chapter I. Survey of the Single Recensions, and of their Manuscripts, as used by the Author of this Volume.
Pedigree of the recensions of the Fancatantra. — Below is given a statement of the various Indian recensions, and in such a tabular form as to make clear the genetic relations. The Brahmanical recensions are marked with a star (*) ; the Jaina recensions with a dagger (f).
1 'Original work (Kashmir, about 200 B. C. ?) 2 *t (Kashmir)
6 *K (Kashmir)- before 570
9 'Pseudo-Gunadhya
7*N-W North- Western India
i
10 *Somadeva,
Kathasaritsa-
gara, 1063-1082
(Kashmir)
11 *Ksemendra, Brhatkatha-
manjarl,
about 1040,
Kashmir
12 *n-w 8 Pahlavi recen- North-Western sions, Persia, epitome after Western Asia, Kalidasa (fifth North Africa,
century) Europe,
from about 570
13 'n-w1
14*SP Dekkan
20* n-w2
3 *£ (Kashmir)
I
4 *gar. a,
Kashmir,
before 1000
5 *£b. /3
Kashmir,
before 1000
Tamil recensions
23 ttextus simplicior, North- Western India,
after 850. H-class and a-class
I i I V
15*0 16*0 17 *y 18*8 19 *£
I
21*K 22 'Hitopade^a Nepal Bengal ?
before 1373
24 tPOrnabhadra, North-Western India, 1199
25 Later mixed recension*
6 Chap. I. Survey of the Single Recensions
Key to the pedigree. — There follows now, under twenty-five headings corresponding with those of the table (1-25), a brief statement as to each of the inferrible or extant recensions of the Paiicatantra, so far as known to the editor of Purnabhadra's recension, and as to the MSS. used by him as editor, and in his studies of the history and sources of that text.
1. The original work.
The author's MS. of this work and all exact copies of it are lost.
2. t.
Some copy, inferrible but no longer extant, of the original work, which copy already contained certain mistakes and interpolations.
3. 8'.
The lost Sarada archetype of the Kashmir recension or Tantrakhyayika. S contained many corruptions and gaps, and some more interpolations.
4. S'ar. a.
The more original text of the Tantrakhyayika. Known from the MSS. :
P = Puna, Deccan College viii. 145.
P1 = one leaf, containing most of the kathamukha, Decc. Coll. viii. 145.
p1 = the greater part of MS. p, belonging to Dr. M. A. Stein. The Sanskrit text of MS. P was printed in Abh. der Kgl. Sachs. Ges. der Wissenschaften, vol. xxii, No. v, p. 1 ff.
5. S'ar. /3.
The slightly revised and enlarged text of the Tantrakhyayika. Used by
Ksemendra. Part of the interpolations contained in Sar. /3 go back to some
MS. of the K-class (No. 6). MSS. :
p2 = the smaller (last) part of MS. p, belonging to Dr. M. A. Stein, z, and its derivatives p r R, MSS. belonging to Dr. Stein.
Critical edition of these recensions : Tantrakhyayika, die alteste Fassung des Pancatantra. Nach den Handschriften beider Rezensionen zum eraten Male herausgegeben von Johannes Hertel. . . . (= Abh. der Kgl. Ges. d. Wissensch. zu Gottingen. Phil.-hist. Kl. N.F. Band xii. 2).— Translation: Tantrakhyayika. Die alteste Fassung des Pancatantra. Aus dem Sanskrit iibersetzt mit Einleitung und Anmerkungen von Johannes Hertel. 1909. Leipzig und Berlin. Druck und Verlag von B. G. Teubner. 2 vols.
and of their Manuscripts
6. K.
A lost Sarada MS. which was the source of all the other recensions of the Pancatantra.
7. JS-W.
A North- Western copy flowing from K, not extant, but represented by
8. The Palilavi Recensions.
The Pahlavi version itself is lost, but very numerous offshoots of it are preserved. See V. Chauvin, Bibliographic des ouvrages arabes ou relatifs aux Arabes public's dans 1' Europe chre"tienne de 1810 a 1885. II. Kalilah. Liege. H. Vaillant-Carmanne, Imprimeur. Leipzig, en commission chez 0. Harrassowitz, Querstrasse 14. A new edition and translation of the Old Syriac version has been given by Prof. Friedrich Schulthess of Konigs- berg. Title : Kalila und Dimna, Syrisch und Deutsch. Berlin. Verlag von Georg Reimer. 1911.
9. Fseudo-Gtmadliya.
The lost metrical extract from an old text of the Pancatantra, interpolated in a North- Western recension of the Brhatkatha.
10. Somadeva.
The abbreviated Sanskrit translation of No. 9, contained in Somadeva's Kathasaritsagara. Editions used by the author of this volume :
Br = Katba Sarit Sagara. Die Marchensammlung des Somadeva. Buch vi. vii. viii. Herausg. von Hermann Brockhaus. Leipzig 1862 in Com- mission bei F. A. Brockhaus (= Abh. fur die Kunde d. MorgenL, herausg. v. d. Deutschen Morgenl. Gesellschaft, ii, No. 5).
Du = The Kathasaritsagara of Somadevabhatta. Ed. by Pandit Durgaprasad and Kasinath Pandurang Parab. Printed and published by the proprietor of the " Nirnaya-Sagara " Press. Bombay. 1889.
MSS. used by the author of this volume : A = I. 0. 1881, E. 3957. B = I. 0. 2165, E. 3949. C = I. 0. 1102, E. 3955. K = Sanskrit College, Calcutta, No. 1796. P = Deccan College, 1887-1892, No. 660.
11. Kaemendra.
The abbreviated Sanskrit translation of No. 9, contained in Ksemendra's Brhatkatha-manjaii, xvi. 286 to 567. Editions :
8 Chap. I. Survey of the Single Recensions
v. M = Der Auszug aus dem Pancatantra in Kshemendras Brihatkatha- manjari. Einleitung, Text, Uebersetzung und Anmerkungen von Leo von Mankowski, dr. jur. & phil. Leipzig, Otto Harrassowitz 1892.
£ = The Brihatkathamanjari of Kshemendra. Ed. by mahamahopadyaya (!) pandit Sivadatta, Head Pandit and Superintendent, Sanskrit Depart- ment, Oriental College, Lahore, and KasMnath Pandurang Parab. Printed and published by Tukaram Javaji, proprietor of Javaji Dadaji's " Nirnaya-Sagara " Press. Bombay. 1901.
12. n-w.
A north-western epitome, in which all the stories and nearly all the verses of N-W were given. It must have been composed after Kalidasa's Kumarasambhava : see vol. I of my translation of the Tantrakhyayika, p. 158, middle.
13. n-w1.
This is a derivative of n-w, and the immediate source of
14. SF.
The archetype of the so-called Southern Pancatantra. Of this archetype no quite faithful copy has been handed down to us. The MSS. known to us belong to the following five sub-recensions :
15. SFa.
A | A B palm-leaf MSS., C a paper MS., all of them belonging to the B I late Prof. Leo von Mankowski, and kindly lent to the author C ) of this volume by him. K, a copy of the MS. of the Madras Government Oriental MSS. Library,
Alph. Index, p. 46, No. 7-1-7. L, ditto, 7-1-6. N, ditto, 7-1-8.
Q, a copy of the Tanjore MS., Burnell, Class. Index, p. 165b, No. 5,110. P, a copy of the beginning of the Tanjore MS., Burnell, p. 165b, No. 5,109.
R, ), !, » » !> 5,111.
S, „ „ „ ,, „ 5,113.
TJ, „ „ „ „ „ 5,116.
V, „ „ „ „ „ 10,^40.
W, „ „ „ „ „ 10,241.
Y, „ „ „ „ ,, 10,242.
Z, a copy of a not numbered MS. of the Palace Library, Tanjore. In this copy, the text of the SP is wrongly ascribed to Ksemendra.
and of their Manuscripts
After my edition of the SP was printed, I got, through the kind help of Prof. E. Hultzsch and Govt. Epigraphist V. Venkayya, the MS. b, i. e. a copy of the beginning of the MS. Hultzsch, Reports on Sanskrit MSS. in Southern India, No. II, p. 45, 1219. This MS. goes with B.
16. SP ft.
F, a collation of the MS. of the Madras Govt. Or. MSS. Library^ Alph. Index, p. 46, No. 3-2-20.
H, a copy of the MS. of the Madras Govt. Or. MSS. Library, Alph. Index, p. 46, No. 6 B-3-15.
. first group.
0, a copy of the MS. of the Madras Govt. Or. MSS. Library, Alph.
Index, p. 46, No. 3-4-19. E, a copy of the MS. of the Madras Govt. Or. MSS. Library, Alph.N
Index, p. 46, No. 7-1-5. I second
1, a copy of the MS. of the Madras Govt. Or. MSS. Library, Alph. [group.
Index, p. 46, No. 7-1-10. j
M, a copy of the MS. of the Madras Govt. Or. MSS. Library, Alph.j third Index, p. 46, No. 5-3-13. J group.
The text of the MSS. F HO, El, with the variants of the best MSS. of
SP a has been published in the following edition :
Das siidliche Pancatantra. Sanskrittext der Rezension /3 mit den Lesarten der besten Hss. der Rezension o, herausg. von Johannes Hertel. Des xxiv. Bandes der Abhandlungen der phil.-hist. Kl. der Kgl. Sachs. Ges. d. Wissenschaften No. V. Leipzig bei B. G. Teubner 1906.
17. SPy.
D = India Office, Biihler MSS. April 24, 1888, No. 320. G = India Office, Burnell 211.
A useless attempt towards publishing these two MSS. has been made
by Dr. Michael Haberlandt, Zur Geschichte des Pancatantra, Sitzungsber.
d. phil.-hist. Cl. der Wiener Ak. d. Wissensch. 1884, p. 397 ff. Cp. ZDMG.
Iviii, p. 3 ff.
18. SP 8.
T, a copy of the MS. of the Palace Library, Tanjore, Burnell's Class. Index, p. 165b, No. 5,112. Cp. Introduction to my ed. of SP, p. xxxiv f.
19. 8F ;'.
X, an old palm-leaf MS., presented to the author by the late Prof. v. Mankowski. An analysis of this southern ' textus arnplior ' has been given ZDMG. Ix. 769 ff. and Ixi. 18 ff.
c
10 Chap. I. Swvey of the Single Recensions
There are, furthermore, two Paris MSS. of the SP, and one MS. belonging to Prof. Teza, which were not available for me. Cp. ed. of the SP, Intr., pp. xxix and xxxiii.
20. n-w1.
A lost MS. agreeing on the whole with n-w1, but having numerous more original readings. The first and second tantras were transposed in this recension.
21. v.
A. Nepalese recension, containing only the verses and one prose sentence
which the copyist evidently took for a stanza. First and second tantras
transposed. MSS. :
n1, a copy presented to the author by the Durbar of Nepal, and containing books I to III incl. Cp. ed. of SP, p. Ixxxviiiff. The complete variants of this MS. are given in the ed. of SP.
n2, a copy of the beginning and of the end, transcribed from the same original as n1, and procured for me by Prof. Sylvain Le"vi. This MS. contains the stanzas of tantras iv and v, which are missing in n1. They are printed in the appendix to my Introduction to the edition of the Sanskrit text of the Tantrakhyayika, p. xxvii.
22. The Hitopadesa.
This is based on n-w2 and some other story book. Its author was a Saiva
called Narayana, who wrote for some king Dhavalacandra, probably in
Bengal. Books I and II transposed as in v.
Editions with critical notes: 1. Hitopadesas id est Institutio salutaris. Textum codd. MSS. collatis recensuerunt interpretationem latinam et annotationes criticas adiecerunt Augustus Guilelmus a Schlegel et Christianus Lassen. Pars I. textum sanscritum tenens. Bonnae ad Rhenum MDCCCXXIX. . . . Pars II. commentarium criticum tenens. . . . MDCCCXXXI. (The translation promised on the title has not appeared). 2. Hitopadesa by Narayana. Ed. by Peter Peterson. Bombay, 1887 (= Bombay Sanskrit Series, No. xxxiii).
As to these and other editions compare : tJber Text und Verfasser des Hitopadesa. Inaugural-Dissertation . . . bei der hohen philoso- phischen Facultat der Universitat Leipzig eingereicht von Johannes Hertel . . . Leipzig. Druck von Breitkopf & Hartel, 1897.
A truly critical edition of this work is still a desideratum. The Nepalese MS. N, mentioned by Peterson in the preface of his edition, p. i, did not belong to the British Museum, as Prof. Peterson thought, but to
and of their Manuscripts 11
the late Prof. Cecil Bendall, and now belongs to the Cambridge University Library. A MS. which Prof. Zachariae presented to the Library of the German Oriental Society has been described by him ZDMG. Ixi. 342.
After the publication of SP and v it will now be an easy task to determine the best MSS. of the Hitopadesa and to give a truly critical edition of this work.
23. The textus simplicior.
Called in the MSS. Pamcakhyanaka. This text is the work of some Jaina author who seems to have lived after the middle of the ninth century A. D., since he quotes a stanza of Rudrata 1, and before 1199 A. D., as Purnabhadra used this text as one of his main sources. The author added new tales and new stanzas, especially from Kamandaki 2, transposed the stories, especially in books III and IV, and greatly amplified the bulk of the Pancatantra, especially in the fifth book. As for the single stories, he not only altered their wording throughout, but also their purport. The stories of the textus simplicior have many features in common with Buddhistic forms of these tales, which deviate from the old Pancatantra texts. The MSS. of this recension disagree very considerably, and in most of them the text is in no good state. All of them are revised copies. MSS. of the textus simplicior : H, No. 281 of the Hamburger Stadtbibliothek. Not dated, but older
than I. I, No. 280 of the Hamburger Stadtbibliothek, dated sarn. 1701. As to
H and I see my paper ' Kritische Bemerkungen zu Kosegartens
Pancatantra ', ZDMG. Ivi. 293 ff. O, MS. of the Bodleian, Oxford, Aufrecht's Cat., p. 157% No. 335, ' ex
eodem codice atque Hamburgenses H. I. videtur transcripta esse.'
Dated sam. 1709. This MS. I have not seen. Bh, fifth book, contains a text very closely agreeing with that of the
Hamb. MSS., but without two interpolations of H I. See below,
p. 56 ff., and cp. No. 25, Later Mixed Recensions. <r = Decc. College, Peterson's Fifth Report, No. 356. Not dated. Copied
by gani Caritraklrtti, disciple of gani Tejastilaka. Corrected by pandit
Sukirtti and pandit Amarasimha. s = Decc. Coll. i. 17. First leaf replaced. Copied from an old original
(many small gaps). Not dated.
1 See Pischel's edition of Rudrata (Rudrata's (Jffigaratilaka and Ruyyaka's Sahrdaya- lila. Kiel, Haeseler 1886), p. 26.
* Benfey, Pantschatantra I, p. xv, note 2.
12 Chap. I. Survey of the Single Recensions
S = Decc. Coll. xv. 147. First two leaves and last leaf missing. The conclusion of IV, vii (Mouse-maiden), with IV, viii (Saints' clothes) is an unintelligibly short abstract, after which the fourth tantra is concluded. Dated samv. 1534 caitramase suklapakse 5 pamcamyam tithau somavasare atreha Harsapure
sutena savakena pustika likhita punyasyarthe tena punyena bhagavan sri Mahat'WiM piitostu, &c.
a = Decc. Coll. xii. 252. A fragmentary MS. of the text contained in S. The conclusion of IV, vii (with the emboxed story IV, viii) is literally the same in both MSS. The following leaves are missing : 1-55 incl., 59-61, 63-74, 77, 80, 85, 86, 89, 93 to the end. At the end of tantra iv the copyist gives his name : likhitam idam pro(!)hita-Ramacamdrabhi- dhena nijapathanartham paropakrtaye castuh I
pr = a MS. belonging to the Bhandar of Ahmedabad, and lent to me through the good offices of Mr. Keshavlal Premchand Mody of the same town. It bears the marks da° 7, pra° 25 ml, and da 13 pra 15. Colophon: sarnvat 1592 varse vaisasasiditraravau lisitam. This MS. agrees very closely with the edition of Kielhorn and Buhler.
Bii2 = India Office, Buhler MS. 86. Fragment, leaves 1-39 incl. of 88 leaves missing. Dated Samvat 1804, Saltern 1669 praihamMe pausavadya 2 dvitlydydm ludhe Bhisagupandmnd Sn-Ndrdyanapamta(i. e. pamclitaftutena suhrdvarenedam pamcopdkhydndkliyam pustakam likhitam svdrtham par&rtfiam ca, &c.
Mil* = Max Miiller Memorial e 11, Bodleian, Oxford, 50 leaves. Begins kd sotkamtMs tutati iva (corresp. to Purn. 230, 3). The text of this MS. belongs to the o--class. Dated sake 772 Sdrvarindmasamvatsare vaisdkfia- $uddhanava»<ydm.
h = a copy of the MS. mentioned in Sh. R. Bhandarkar's Report, Bombay, 1907, p. 55, § 46. The original lies in a dilapidated fort in Hanumangad or Bhatner (Bikaner). Bhandarkar says : ' The place in the fort where I saw the box of manuscripts is also dilapidated and deserted. The heir to the manuscripts is a young boy who, I believe, is studying at Patiala.' . . . Bhandarkar calls the original ' a copy of Panchatantra made in Samvat 1429, while Firuz Shah Taghlak was on the throne.' Mr. F. W. Thomas kindly procured me the copy, which was ordered by the Durbar of Bikaner. It has been made by two copyists, neither of whom knew Sanskrit, and both of whom, especially the second one, very often misread the old-fashioned characters of the original. The colophon does not mention the date given by Bhandarkar, as the copyist of this part of the MS. evidently altered it to give the date
and of their Manuscripts 13
of his own copy. As the colophon gives an idea of the knowledge which this copyist — the better one of the two — has of the Sanskrit language, I give it here : sarapum [for samaptam !] vedam aparlksita- karanam nama pamcamam tamtramm iti i vrhatpamcatamtram samattamh (corr. to samaptamh)!! samvat 1965 ramitimigasaravadl 12 ne lisamtamm atmacatarabhuja FiMneran&g&r&m&dhye Saratare gacchai II yadrsam pustamka drstva tadrsam lisitam maya I yadi I suddhamm asuddham. va mama doso na diyateh II srir astuh II srikalpamnam astuh II s"rl subham bhavayat h II h. This copy is very faulty. Moreover, very many corrections and glosses are entered in it in some places. In spite of all this the copy is valuable. But a future editor of the textus simplicior must try to get the original of our copy for his work. This original must contain a good old text of this work. The story I, v is concluded in it as in H I.1 In the Introduction to my transla- tion of the Tantrakhyayika, p. 158 (Kap. II, § 2, 1, S. 31), I have shown that the stanza which contains the argument of story II, iii of our text has been altered in all the descendants of K, and has been well pre- served only in Sar. Our Specimen III, 1. 133, footnotes, gives the reading of h, which, though corrupt, proves in an evident manner that originally the textus simplicior also had the reading of Sar. c = Deccan Coll., Bhand. Cat., xvii. 637. Fragment. Leaves still extant : 2-10 incl., 14-21, 23-33, 37-41, 43-46. Goes down to acimtaya, Kielh. p. 89, *. Rather faulty ; modern.
The textus simplicior has not been handed down to us in its original form. All our MSS. show interpolations,2 and the original wording has not been preserved in any one of the MSS. that I have seen. Our parallel Specimens and, above all, the text printed below, p. 58 ff., show that the MSS. of the textus simplicior may be roughly divided into two groups :
(1) The H-class, to which belong HIO and book V of Bh; see below,
p. 58 ff.
(2) The <T-class, to which belong o-sprh (and book V of 4> ; see Later
Mixed Recensions).
As to the Vaisnava MS. S and to the MS. a of the purohita Ramacandra, I cannot say to which class it has to be assigned, as unfortunately I failed to copy the greater part of their text. At any rate these two MSS. are worthless. Of the two classes, each at times excels the other in the greater originality of an occasional passage. Our parallel Specimens I-III and the text
1 Cp. Uber die Jaina-Rezeneionen [see below, p. 15], p. 97 ff.
1 Cp. Her. kgl. Bachs. Ges. d. Wissenschaften, ph.-h. Kl. 1902, p. 68 f.
14 Chap. I. Survey of the Single Recensions
printed below, p. 58 ff., show that Purnabhadra used copies of both these classes. Wherever he follows the textus simplicior, nearly his whole wording can be reconstructed from MSS. of these two classes. It is scarcely possible that he had before him a MS. from which both the H- and the o--class are derived, as in some places either the H-class or the cr-class is more original than Purnabhadra's text.
The text of the H-class seems to me, on the whole, to be the more original one. It has not yet been edited.
One single MS. of the tr-class has been edited by Kielhorn and Buhler in their well-known edition of the Paficatantra in the Bombay Sanskrit Series, Nos. IV, III, L Cp. ZDMG. Ivi, p. 298 f. This edition agrees very closely in its wording, and completely in the arrangement and number of its tales, with the above-mentioned MS. pr.
As to Kosegarten's edition, see below, p. 15, ' Later Mixed Recensions', and p. 44 ff.
24. Purnabhadra's text.
Called in the MSS. (like No. 23) Pancakhyanaka. Purnabhadra's text is a compilation of Sar. 0, of the textus simplicior, and of sources unknown to us, amongst which there was a source composed in Prakrt ; see below, p. 27 ff. He seems to have known Ksemendra ; cp. WZKM. xvii. 347. According to his prasasti, he completed his work in A. D. 1199 by the order of a minister named Srt-Soma.
No other recension of the Paficatantra has been handed down to us in BO authentic a wording as Purnabhadra's work. The MSS. which contain it are the following :
bh = Deccan College, Bhand. Cat. x. 190. N = „ „ „ „ x. 189.
A = India Office 2643, E. 4084 (a revised copy). * = Deccan College, Bhand. Cat. iv. 55. P= „ „ „ Report 1897, 419;
L1 = Leipzig University Library, A. 404. M = Deccan College, Bhand. Cat. iv. 54. p = „ „ „ „ ii. 46 (a revised copy).
Pr = a MS. of the Jaina Bhandar of Ahmedabad, da. 28, pra 10. Not
dated, but old.
B = Oxford, Aufr. Cat., p. 157*, No. 337. It contains only the first two tantras and the greater part of III (down to 227,5 krtaghnd incl.). Written after A. D. 1810. As I know this MS. only from Tullberg's collation, I neglected it. As for the other MSS., see below, p. 37 ff. The text of this recension has been published in vol. xi of the HOS.
and of their Manuscripts 15
As to the textus simplicior and Purnabhadra's recension, cp. my papers : ' Kritische Bemerkungen zu Kosegartens Pancatantra ', ZDMG. Ivi. 293 ff., and ' Uber die Jaina-Kezensionen des Pancatantra ', Berichte der phil.-hist. Kl. der Kgl. Sachs. Gesellschaft der Wissenschaften zu Leipzig, 1902, 23 ff.
The lack of critical spirit, which is so characteristic of the old style pandits, was the reason why the more complete, i. e. the interpolated and contaminated MSS. of celebrated works, were always copied, whereas the old genuine texts disappeared. Hence the textus simplicior and Purnabhadra's recension completely ousted the old Pancatantra from North-western India. But Purnabhadra's compilation was not the last stage of this literary development. Numerous new recensions arose, and these have been copied and enlarged even to our days. These mixed recensions may be classed under six heads.
1. The textus simplicior was interpolated from Purnabhadra's text.
2. Purnabhadra's text was interpolated from the textus simplicior and
other sources.
3. Single books or tantras of different recensions were combined.
4. Other recensions were contaminated with the Jain recensions.
5. The Jain recensions were moulded into other forms.
6. Versions from which the frame-stories have been eliminated.
25. Later Mixed Recensions.
251. Textus simplicior, interpolated from Purnabhadra's recension. MSS. : D = I. O. 2790. Dated samv. 1796 asadhavadi 3.
b = Deccan College, Bhandarkar's Cat. xii. 253. Fragmentary MS. The following leaves are still extant : 49-79 inch, 81-5, 87-9, and one leaf, the pagination-number of which is ninety-something (the unit of the number is not to be made out).
d = Deccan College, Bhand. Cat. ii. 44. Has 54 leaves, incomplete. The text goes to sahanam aha, Kielhorn, p. 93, 14.
Mii1 = MS. Max Muller Memorial e 10 of the Bodleian. 100 leaves. Fragment. Begins with ca vilokya Purn. 130, 10. The rest complete. Dated tsambat 1776 varse I Sake 1641 pravarttamdne \ jyestamdse I krsne pak*e I caturlhydm 4 tithau I ddiiyavdre\ Written Sitjdnasimhajlvijayardjye I Sri Vlkdnera-madkye by one dcdrya Rdmakrsna.
Mii3 = MS. Max Muller Memorial d 40. 71 leaves. Contains books III to V inclusive. The pagination begins with 1. Not dated.
252. Purnabhadra's text, interpolated from the textus simplicior and other sources. MSS. :
K = Berlin, Chambers 176. This MS. is known to me from the collations of Benfey and R. Schmidt.
16 Chap. I. Survey of the Single Recensions
L2 = Leipzig University Library, A. 403. 84 leaves. From the beginning to trdsaydm dsa 219, 2. Title HitopadeSa (only in mg.). Very faulty Saiva MS. After I, i the story Hitopadesa II, iii, ed. Schl. (' Dog and Ass ').
Bii1 = 1. 0., Buhler MS. 85. Dated sake 1788 ksayanamavatsare. This is the recension translated by Galanos, and used by Meghavijaya (see below, p. 19). Cp. WZKM. xix. 62 ff.
Mii2 = MS. Max Miiller Memorial f 1. Complete. The pagination (lost on some leaves) goes from 244 to 395. Leaf 279 wrongly inserted after 379, leaf 337 after the first leaf of Sivadasa's Vetalapancavimsatika, which follows in the MS. Down to about p. 22 of our text Mil2 contains a mixture of Simpl. and Purn., the textus simplicior prevailing. But also in some other places the textus simplicior has been compared. After truyate ca, 94, 4, for instance, Mii2 continues : tat pranastam knlam paksikulam samprati I any an api svechayd vydpddayixyati I yatah ( = Kielhorn, i. 72, is). Thereupon follow, with variants and corruptions, the stanzas Kielh. 342 to 346 incl., and then the MS. continues with our stanza 344. Nearly all the mistakes common to *bh appear in our MS., and many other corruptions besides. Of the prasasti, Mii2 has the two stanzas of PPrM. — Mii2 is dated tunyaSdstramuniScam . . . [supply dra\ 1760 Yikramasya gaidvdayah II &ucih Sukle trayodayydm by one Sukadeva in a village of Gujarat. It is useless for critical purposes.
253. Single books or tantras of different recensions were combined.
Bh = Deccan College, Bhand. Cat. xiii. 68. The text of books I, II, and III is a Purnabhadra text, mangled and interpolated from the textus simplicior. Book IV (incomplete) is a Purnabhadra text. Book V is an old text of the H-class of the textus simplicior. See below, p. 56 ff.
<£ = Deccan College, Peterson, Report IV, 719. The text of books I to III inclusive is the text of Bh ; book IV is a Purnabhadra text different from Bh ; book V is a textus simplicior of the <r-class. See below, p. 56 ff.
C = Bodleian, Aufrecht's Cat., No. 336. Dated sam. 1856. This MS. I know only from Tullberg's collation. The beginning of the first tantra contains the textus simplicior, the rest of that tantra and the remaining ones are Pumabhadra's text.
F = I. 0., No. 2319. Books I and II contain the textus simplicior, the others are copied from the same original as C. To Mr. F. W. Thomas I owe the statement that codices F and L (cp. Kosegarten, p. vi) are identical.
Bii5 = I. 0., Buhler MS. 89, a fragment containing leaf 1 and leaves 53 to 119 incl. Nearly all of book I is missing. The conclusion of book I
and of their Manuscripts 17
and books IV and V contain the text of Purnabhadra, whereas books II
and III contain a textus simplicior with stories interpolated from
Purnabhadra. Cp. WZKM. xix. 75. Bii3 — I. O., Biihler MS. 87. Fragment ; 47 leaves. Begins with the
description of the hunter at the beginning of tantra ii, and goes to
Bubler's stanza iii, 163. Purnabhadra's stories inserted in the frame
of the textus simplicior. Cp. WZKM. xix. 73. IT = Deccan College, Peterson, Report III, Appendix iii, No. 313. Not
dated. Modern. Books I and II, Purnabhadra ; the other books,
textus simplicior.
The following MSS. contain in their books I, II, V, Purnabhadra's text ; in their books III and IV an interpolated textus simplicior : — n1 = Deccan College, Bhandarkar, Report 1894, No. 371. Quite modem. II2 = Deccan College, Peterson, Report V, No. 355. Dated mitau (?) kartika-
krsnacaturthl bhuputravasare samvat 1811.
n3 = Deccan College, Bhandarkar, Report 1897, No. 418. Modern. n4 = British Museum, No. 277. This MS. I have not seen. Q = Deccan College, Bhand. Cat. viii. 144. Last leaf wanting. q = Deccan College, Bhand. Cat. xii. 251. Missing leaves : 1-70 incl., 77,
78, 122, 140 to the end. Not old. In the parts extant, q has the
same stories as Q.
Kosegarten's edition of the textus simplicior belongs to this class, and his edition of the textus ornatior as well. Both are mixtures from MSS. of various classes. The edition of Jivananda Vidyasagara is based on Kosegarten's, and so is that of Parab. See below, p. 51 ff.
254. Other recensions contaminated with the Jain recensions.
2S4a. The MS. E = I. 0. 1812, E. 4086. Kosegarten, p. iv of his edition of the textus simplicior, says of this MS. : ' Textus ad codicem A. prope accedens passim verba sensum supplentia adiicere videtur.' On the contrary, this MS. deviates from Purnabhadra in the most remarkable manner. It is based on the textus simplicior and contaminated with Purnabhadra's text. But the author of this text has used still other sources, e. g. the Mahabharata, the Vikramacaritra (or some work quoting a coherent passage of it ; see below, p. 44 ff.), nay, even an older recension of the Pancatantra, from which the author inserts his story III, i (' Ass in panther's skin'), which seems to be based on the Tantrakhyayika. The text of this story, printed from Tullberg's collation, is given ZDMG. Ivi. 317. The order of the stories has been altered throughout; see ZDMG. Ivi. 326.
D
18 Chap. I. Survey of the Single Recensions
254b. Ananta's Kathamrtanidhi.1 This is an epitome of an old textus simplicior, interpolated in some places from Purnabhadra's text, and even altered by the redactor in some features of the stories related. The single books are not called tantra, but urmi. Cp. ZDMG. Ivi. 296 f. ; Saxon Berichte, p. 117, note 1. MSS. : G = I. O. 2146 = E. 4088. A modern Nagari transcript from a no doubt
Southern MS. (/ frequently appears instead of / between vowels). G1, Aufrecht mentions a second MS., Hall, A Contribution towards an
Index (Calc., 1859), p. 183.
The two following Nagari copies, derived from one and the same original, belong to Prof. E. Hultzsch (Halle) : G2, 93 leaves, and G3, 69 leaves. In the first pada of the concluding stanza
of book V (see Saxon Berichte, p. 117, note 1) both of them read vijf
for kanva*.
25* c. NP, the recension mentioned by Aufrecht, C. C., p. 314 : mptr? kavya, by Dharmapandita. MS. mentioned in 'A Catalogue of Sanskrit Manuscripts in Private Libraries of the North- Western Provinces. Parts I-X. Allahabad, 1877-86, ix. 14.' I got a copy of this MS. by the good services of Mr. F. W. Thomas, and of the Principal of the Sanskrit College, Benares. In this copy, the name of the author (Dharmapandita) does not appear. The original, as the librarian of the Sanskrit College informs me, is written in Tailanga characters. The librarian says that the modern pandits designate nlti works which contain stories as kavya. The original belongs to Pandit Nrsimhasastrin, and the Nagari copy sent to me was made by order of his son, Pandit Gangadhara Sastrin, C.I.E.2
The author of this version has used several sources, the textus simplicior, the recension of Purnabhadra, the Southern Pancatantra, the Hitopadesa, and in some places even Sar., or some MS. which contained passages that are known to us only from this source. Only the first two tantras are complete ; of tantras iii to v there is only a very short abridgment. Books IV and V are transposed. After the fifth book there are several story- stanzas ; no doubt the author intended to use them for the composition of books III to V. Cp. ZDMG. Ixiv. 61.
As to this recension, see Journal Asiatique, Nov.-De'c. 1908, p. 400 ff., where also the stories I, xvii, xviii, xix are given in Sanskrit and French.
MS. : np, new copy in Nagari, 51 leaves, 12 to 13 lines on a page.
1 As the author's prasasti tells us, Ananta was a worshipper of Visnu. He belonged to the family of the Kanva's, and his father's name was Nagadeva. According to Aufrecht's C. C., i. 13, 771, and ii. 186, Ananta Bhatta is the author of many works.
2 See Journal Asiatique, Nov.-Dec. 1908, p. 400, where °pattrena (1. 8 of the Sanskrit passage) is a misprint for °puttrena.
and of their Manuscripts 19
25* d. The recension of the Jaina monk Meghavijaya, compiled from an interpolated Purnabhadra text (Bii1, above, 252), from the textus simplicior, from a metrical version of the Jaina Pancatantra, from the Jaina work Dharmakalpadruma, and from one or more other sources. The prose has been rewritten, and new verses and stories have been added. Meghavijaya wrote in sam. 1716 in the town Navaranga. He belonged to the Tapagaccha. MS. of the I. 0. : Buhler, ZDMG. xlii. 54, No. 6 ; fols. 35, 11. 17, samvat 1747, Puna (No. 90).
An analysis of this version, with the Sanskrit texts of the new stories or interesting variants of old stories contained in it, has been given in my paper ' Eine vierte Jaina-Recension des Pancatantra ' ; for a German translation of these stories, see my paper ' Meghavijayas Auszug aus dem Pancatantra', Zeitschr. des Vereins fur Volkskunde in Berlin, 1906, p. 249 ff.
25* e. The MS. Bhandarkar, Report 1897, 417 (Deccan College, coll. of 1887-91, 153 leaves; col.: sam. 1728 sVavanakrsna [!] caturddasyam somadine -P^te^puramadhye divan iri-Aliphasdmr&^yQ Porz^anvaye MiSra- &TiRdmena]eb.hi II subham II su ll) contains another Jaina recension. The text of this MS. is compiled from the textus simplicior, Purnabhadra (bh-class), Hitopadesa, the metrical source used by Meghavijaya, Sar. ft (with the mistakes of our MSS. of this recension), and other sources.
Tantra I contains the same stories, and these stories in the same order, as Purnabhadra. Only story xxiv and part of xxiii have been lost by a gap (not marked in the MS.). Tantra II: i = Hit. Schl. I. ii (Pet., p. 7, 4) ; ii = Purn. II. i ; iii = Sparrow's allies and elephant, with iv, Lion and woodpecker (from the same metrical sources as Meghavijaya1); v = Hit. L iii (Pet. I. 41 and following story) ; vi = Hit. I. iv (Pet. I. 42 and follow- ing story) ; vii = Purn. II. ii ; viii = Purn. II. iii (but the text of Sar. /3) ; ix, corresponds to Purn. II. iv (stanza and first sentences from Hit., the rest of the text from Sar. and Purn.) ; x = Hit. I. vi (Pet. I. 80 and following story) ; xi = Purn. II. v ; xii = Purn. II. vi ; xiii = Purn. II. vii ; xiv = Purn. II. viii ; xv = Purn. II. ix. Tantra III: i = Sar. III. i ; ii = Purn. III. i ; iii = Purn. III. ii ; iv = Purn. III. iii ; v (intercalated into iv : the lizards, elephant, and water-animals 2) ; vi, corresponds to Purn. III. iv (from 6ar.) ; vii = Purn. III. v ; viii to xv = Purn. III. vi to III. xiii ; xvi, corresponds to Purn. IH. xvi (from Sar., and again from Purn. ; two foil. 166 ; the text in disorder). Tantra IF: i to x = Purn. IV. i to x ; then xi = Simpl. Buhler
1 Cp. Zeitschr. d. Vereins f. Volkskunde in Berlin, 1906, 256 f. (with German transla- tion). Sanskrit text and French translation : Journal As., Nov.-De'c. 1908, p. 425 ff. * Sanskrit text and French translation: Journal As., 1908, p. 432 ff.
20 Chap. I. Survey of the Single Recensions
IV. vii, and xii = Simpl. IV. xvi. Tantra V = Simpl. V, with all the stories given by Buhler, except V. v.
In the wording of the frame-stories, the texts of Hit., Purn., Simpl., Sar. are equally contaminated.
254 f. The Buhler MS. 88 of the India Office (ZDMG. xlii. 541), though complete in itself, contains only tantras i, iv, and v. It was copied in sam. 1830 sake 1695 by Vasudeva, son of Ramacandra, son of Ramakrsna, of a Maharastra family. The faulty colophon seems to imply that Rdma- candra (a Vaisnava) was the author of this recension. Like the recensions recorded under 254 c and g, the text of Ramacandra's version represents the copy of a rough draught not finished. The first tantra is based on the textus simplicior, but interpolated from Purnabhadra ; the fourth tantra contains a text of SP/3. The fifth tantra is contaminated from SP and the textus simplicior. It begins with the frame-story and the first and second tales of SP ; then follow all the stories of Biihler's edition from V. iii onward, except Buhler V. ix and V. xiv. In general, cp. WZKM. xix. 74 f.
25* g. The MS. Deccan College xvi. 105 (30 leaves, not old) contains the Kathamukha and book I of an incomplete new recension. But the first book, numbered as such, corresponds to tantra ii of Puraabhadra's text. It contains all the stories of Purnabhadra's text in the same order. The wording of this MS. has been contaminated from Purnabhadra, from the textus simplicior, and from the Hitopades"a, and many new stanzas have been inserted.
255. Jaina recensions moulded into other forms. A metrical version of the Jaina recensions must have existed before the time of Meghavijaya, who has very largely availed himself of it (see above, 254 d). The same version was used by the compiler of the text 25* e.
256. Versions from which the frame-stories have been eliminated.
a. A Jain MS. of the Berlin Library, described by E. Leumann, Saxon Berichte, 1902, 132 ff. (from the textus simplicior).
b. The Buddhist version from Nepal, called Tantrakhyana. It is based on one or several unknown redactions, and augmented from other sources. The prose given by Bendall is not original. Only the stanzas contain the original Sanskrit text.
Cp. Bendall, The Tantrakhyana. Journal of the Royal Asiatic Society of Great Britain and Ireland, new series, vol. xx, p. 465 ff. Hertel, Uber einige Handschriften von Kathasamgraha-Strophen, ZDMG. Ixiv. 58 ff.
In the Jaina Upasrayas of Pophliano pado in Patan (upper Gujarat) and of Dehlano pado in Ahmedabad, there are still numerous Pancatantra
and of their Manuscripts 21
MSS. the use of which unfortunately I was not granted. Cp. Bhandarkar, Report, Bombay, 1887, p. 166 (da° 10, 1.2.44) ; p. 180 (36, 126) ; p. 184 (40, l) ; p. 189 (44, 55) ; p. 190 (45, 2t) ; p. 192 (49, 35) ; p. 195 (55, 3.2) ; p. 217 (18, 4.5) ; p. 226 (31,13) ; p. 237 (86,137); p. 243 (43,32) ; p. 245 (46,23). Nor was I granted the use of the Vienna MS. 17 (Aufrecht, C. C., p. 314).
Two Northern MSS. are preserved in the Palace Library at Tanjore: T1 = 5114 and T = 5115. As I was not granted the use of the originals, I ordered copies to be made of both of them. But the specimens sent to me were executed so carelessly (the copyist did not even copy the single leaves in due order), that not to waste more money I had the copying stopped, r1 seems to be a MS. of the H-class of the textus simplicior, whereas r seems to contain a text of Purnabhadra's recension. I cannot say any more about these two MSS., because I know only their beginnings and because, at all events, the two copies give no fair representation of their originals.
Chapter II. Purnabhadra, his time, his work, and his language.
§ 1. Previous Statements. IN 1891, Aufrecht wrote in his Catalogus Catalogorum, vol. i, p. 344:
revised by desire of Somamantrin the Pancatantra in 1514,1 I. O. 2643.' R. G. Bhandarkar in his Report, Bombay, 1897, p. lix, gave the complete prasasti of Purnabhadra's Pancatantra, with this (faulty) stanza containing a different date of the book :
: u ^ u
Bhandarkar adds : ' This is an edition of the Panchatantra prepared under the direction of a Mantrin or minister of the name of Soma and completed on the 3rd tithi of the dark half of Phalguna of the year 1255 by a man of the name of Purnabhadra. The text of the Panchatantra, he says, had become corrupt, and he corrected every letter, word, sentence, story and verse. Accordingly we find on comparing this edition with the existing text as printed in the Bombay Sanskrit Series that there are differences of
1 The MS. which Aufrecht refers to is our MS. A. See below, pp. 22 and 40. (Aufrecht gives the date A. D.)
22 Chap. II. Purnabhadra, his time, work, and language
reading in almost every line. Some of the prose passages and verses in the latter are omitted and sometimes there are others in the place of those occurring there. Sometimes there are verbose prose passages to which there is nothing corresponding in the existing text. The work might be characterized as Panchatantra re- written. Who the Soma-mantrin mentioned by the author was it is difficult to say. The date in all probability refers to the era of Vikrama, wherefore it is equivalent to 1199 A.D.'
In 1902, 1 proved that both the textus simplicior and the text called by Kosegarten textus ornatior, are Jaina works, and that Purnabhadra, who amongst other sources used the textus simplicior, was the author of the so-called textus ornatior} In a post-card dated Aug. 12, 1902, Geheimrat Jacobi was kind enough to tell me that the date of the stanza published by Prof. Bhandarkar corresponds to Sunday, January 17, 1199 A.D.
§ 2. The date of Fnrnabhadra's Recension.
The date taken by Aufrecht from our MS. A cannot come under consideration, as we have several MSS. of Purnabhadra's work which are much older than the date just mentioned. The author's samvat date given in A is 1571.2 Our MS. Bh is dated sam. 1442 ; bh, sam. 1468 ; P, which mediately goes back to the very old MS. ¥, sam. 1537.
The date published by Bhandarkar is taken from the MS. Il^Decc. Coll. 1894, No. 371. The same date-stanza is given in the MSS. I12 = Decc. Coll., Peterson's Fifth Report, No. 355, II3=Decc. Coll., 1887-91, no. 418, and in the MS. 277 of the British Museum. All these MSS. go back to one common archetype. The British Museum MS. I did not see ; but the others are quite modern copies. They belong to the class of the mixed MSS.3 and are on the whole worthless. But their fifth tantra has been copied from a MS. of the bh-class.4 The text of this tantra as contained in them is inferior to that of the same tantra given in bh. But as in bh the date-stanza is missing, these MSS. apparently derive from some copy older than bh, and the date given in them may be right, if it is compatible with what we know from other sources about Purnabhadra's time. And this is the case.
Klatt-Leumann, The Samacharisatakam, Ind. Antiquary, July, 1894, p. 173, give this information: ' 167 a b Purnabhadra, pupil of Jinapati suri (f Samvat 1277), composed sri-Krtapunyacharitra.'
1 Berichte d. kgl. Sachs. Gesellschaft der Wissenschaften, 1902. ph.-hist. Kl., pp. 92 ff., 97 S.
a The copyist's date is saipvat 1574. ' See above, p. 17. * See below, p. 58 ff.
Chap. II. Purnabhadra, his time, work, and language 23
The %T TJ^TR^t (Hfol' sfirff . sffr %T %dJ«K
> f%sfi*T ^Nrt. <IQ.§M) mentions the following works : p.
30 v^miftHR^ <\t$o ^r^ IS^M f •
I was anxious to procure the prasastis of these works. With respect to that of the former my endeavours failed. But to the kindness of the Jaina scholar, Mr. Keshavlal Premchand Mody, of Ahmedabad, I owe a copy of the Introduction and of the pras'asti to the latter work. Both of them were copied from a MS. of 37 pages [fols. 1] (15 lines to a page), belonging to Maharaj Kanti Vijaya, of Baroda.
The praSasti of the Dhanya$a,licaritra. In his prasasti the author of the Dhanyasalicaritra gives the pedigree of his teachers, calling his gaccha the 'tn^jjm (stanza 2), or ^'s^l^ (stanza 10).1 But he gives only the series of the suris of the Kharatara-gaccha,2 excluding the first teacher peculiar to this gaccha, viz. Vardhamana (fsarn. 1088). All of these names are known to us from Klatt's Extracts from the historical Records of the Jainas, from his Specimen of a lit.-bibliographical Jaina-Onomasticon, and from the Pattavali published by Weber in his Cat., p. 1036 ff. In giving them here from the prasasti of the Dhanyasalicaritra, I add in parentheses the dates from Klatt's ' Specimen '.
1 This gaccha derives its name from that of its founder Camdra-suri. See Dharmasa- gara-gani's GurvavalIsutra,Weher, Cat. 997, p. 1002 (numbered as 15th suri) ; Munisundara- suri's Gurvavali, stanza 26 and p. 15, first stanza (numbered as 16th suri) ; and the Patta- vallvacana of the Kharataragaccha, Weber, Cat., p. 1033 f. (numbered as 18th suri).
1 Called after Kharatara, the 'Severer', the 'Harder', a title which was given to JinesVara-suri of the Candra-gaccha, when, in sam. 1080, in the sabha of King Durlabha of Anahillapura (Anhilvad) he refuted the caityavasins. Cp. the story in Weber's Cat., p. 1037f. ; Klatt, Specimen of a lit.-bibl. Jaina-Onomasticon, p. 46 f. The first two stanzas of the above-mentioned pras'asti allude to this event, comparing JinesVara with a lion (who is <a'^fl"<[ than his opponents, compared to elephants), and saying that in Sripattana (= Anhilvad) and in the presence of King Durlabha the suri proved from the Holy Scriptures (sTT^TT) that monks should not dwell in caityas (temples), but in the houses of householders. The first two stanzas of the prasasti run thua :
But cp. Dharmasagara's (satn. 1629) criticism of this fact in R. G. Bhandarkar, Report, Bombay, 1887, p. 149 f.
24 Chap. II. Purnabhadra, his time, work, and language
The pedigree runs thus : Jinesvara (received the honorary name ' Khara- tara' in Sam. 1080), Jinacandra (composed ^^JK'I^MI in Sam. 1125), Abhayadeva (died Sam. 1135 or H39),Jinavallab&a (died Sam. 1168), Jinadatta (born Sam. 1132, died Sam. 1211), Jinacandra (born Sam. 1197, died Sam. 1223), Jinapatti 1 (born Sam. 1210, died Sam. 1277), Vlraprabha (i. e. Jines"- vara, born Sam. 1245, died Sam. 1331), who was made guru by Sarvadevastiri. The last stanzas of the prasasti I give here literally according to my MS. :
TRANSLATION.*
10. He [i. e. Jinesvara=Viraprabha] who, like the celebrated JineSvara [i. e. Vardhamana, the last Jina], rejoices in the celebrated Candra-gaccha,
1 Thus spelt in MSS. and inscriptions. See Klatt, Specimen, p. 24. In stanza 12 of our prasasti his name is correctly spelt faMMfd. " MS. o^t» for o^fo. s I translate as literally as possible, without any regard to the English style.
Chap. II. Purnabhadra, his time, work, and language 25
who instructs and protects it, and who now renders powerful the Jaina moon [i. e. the Jaina gaccha called Cdndra] as a road [for Salvation], aad who with [his] incomparable words fills the threefold world as with laws, [because he is] trusted [by all the living beings] ;
11. By the command of this [Jinesvara], [when I] dwelt with the teacher Sarvadeva, [who is endowed with] good qualities, in the fortress of Jaisalmer, by the word of those [two men],1 in order to give help to myself and to others, and desirous to concentrate my mind,
12. [I,] the pupil of the teacher Jinapati, the gani named Purnabhadra, composed the story of Dhanya and of Gobhadra's son,2 in this Vikrama year numbered by arrow [5], Vasus [8], Sun [12], on the 10th day in the bright half on the approach of the month of Magha.5
13. The best of the vdcakas, called Suraprabha, who has seen the opposite shore of the whole ocean of pure learning, corrected this story.
14. Whatever religious merit I may have gathered here by composing the pleasant story of the monk Dhanya and of the monk Salibhadra, by this [merit] may this world be the abode of bliss.
15. As long as in the pure lake of the sky the autumn moon holds [or bears] the eminence of the excessive sport of a gander,4 so long is victorious in the world this good story of Dhanya and of Gobhadra's son [i. e. so long may it survive], [and be] read aloud by the well-minded [or : by the wise].
Probable identify of the authors of the PancaTcliyanaka, of the Krtapunya- caritra, and of the DhanyaSdlicaritra. About the identity of the author of the Krtapunyacaritra and that of the Dhanyasalicaritra there can be no doubt. From the passage quoted above, p. 23, it appears that both these works were composed by a monk Purnabhadra in the same year, viz. sam. 1285 ; see stanza 12 of the prasasti given above. In the introduction to the Dhanyasalicaritra, stanzas 10 &., the author says :
1 The text has the plural number.
* In stanza 13 of the Introduction our author says: 3n«ft
* So according to Dharma Vijaya Suri, who refers to Hemacandra's AbhidhanacintS- mani, ed. Boehtlingk and Rieu, p. 26, § 29.
4 So according to Dharma Vijaya Suri. I first thought of separating cftfjj^ 444jhl<,0, and of rendering: 'causes quarrel connected with an excess of great dalliance', i.e. ' causes lore with which joys and sorrows are inseparably connected.'
E
26 Chap. II. Purnabhadra, his time, work, and language
These verses prove that Purnabhadra intended to compose a Krtapunya- caritra, after finishing his Dhanyasalicaritra. In the Samacari-sataka as quoted above, p. 22, it is stated that Purnabhadra, the author of the Krtapunyacaritra, was a pupil of Jinapati, and Jinapati was the teacher of Purnabhadra, the author of the Dhanyasalicaritra ; see above, p. 24, stanza 12 of the prasasti.
The author of the Dhanyasalicaritra composed this work in Jaisalmer ; the author of the Pancakhyanaka must no less certainly have lived in north-western India.1 Like the former, he was a ^vetambara monk.2 His date, as given in the prasasti of the II-class,3 viz. sam. 1255, is compatible with the date (sam. 1285) of the two other works mentioned before. Moreover, the author of the Pancakhyanaka calls himself a suri or guru, i. e. a teacher, whereas the author of the two other works tells us that he is a gani. The author of the two later works accordingly has a higher rank than that of the Pancakhyanaka.4
The identity of our two Purnabhadras would be established beyond any doubt, if it were sure that, in stanza 4 of the Pancakhyanaka prasasti, the author is invoking the tirthakara Candraprabha as a patron of the Candrakula. But this is not sure. The author of the Dhanyasalicaritra invokes Parsvanatha.
Hence we can only say that it is highly probable that Purnabhadra, the author of the Pancakhyanaka, and Purnabhadra, the author of the Dhan- yasalicaritra and of the Krtapunyacaritra, were one and the same person.
1 Cp. ZDMG. lx. 787. Purnabhadra lived in that part of India in which camels are kept as domestic animals, and all the MSS. of his work are written in Nagari characters.
2 For no Digambara monk would have told the story I, xxii, in which the fraudulent monk burnt by the clever minister is a Digambara ascetic. Cp. also the stanza V, 11.
* See above, p. 21 f.
4 JjfW has the same signification as JjfU!^. Wf^ and 34Hl4, according to a kind communication which I owe to i§astravis'arada-Jainacharya 6ri Dharmavijaya of Benares, are the same, whereas ' IJJjfl is the name of the head of the same sadhus' assembly.'
Chap. II. Purnabhadra, his time, work, and language 27
§ 3. Purnabhadra1 s work.
If unfortunately Purnabhadra's prasasti to his Pancakhyanaka is silent about the pedigree of his teachers, it is not so about the work he has done. Let us examine what he himself says about it, in connexion with such inferences as we may draw from a comparison of his recension of the Pancatantra with other recensions of this famous book.
In stanza 2a the author tells us that he revised the whole sastra called Pancatantra at the instance of some minister SrI-Soma.
When King Jayasiipha of Guzerat bade the celebrated Jaina monk Hemacandra write a Sanskrit grammar, he either procured for him MSS. of the eight previous grammars preserved in the temple of Sarasvatl in Kashmir, or, according to another, and more trustworthy source, MSS. of all the existing grammars from various countries.1 According to Biihler even now Hindu princes nearly always provide their court pandits in similar cases with copies, and have these copies fetched from even afar, and at great expense.*
Hence we may safely conclude that a minister, when ordering some literary work to be executed for himself, followed the same manner of proceeding. At all events, as the «Tlffmnfl is a minister's jf'slfa'iJI, he had in his possession the principal works treating of this topic, and doubtless provided the pandit whom he entrusted with the revision of such a work with as copious materials as possible.
Purnabhadra's prasasti, taken in connexion with the evidence of his work itself, shows that our assumptions are right. In stanza 2 of the prasasti, the author says that in his time ' the whole system called Panca- tantra' had lost its original form. Elsewhere3 I have explained, that HUjsWfel^f, 'the whole system,' means 'all the existing recensions'. No doubt, Purnabhadra knew several redactions of this work which are unknown to us.4 But two recensions have been proved to be his main sources, viz. the second recension of the Tantrakhyayika (Sar. /3), and the textus sim- plicior, both in the H-class and in the cr-class.6 None of these recensions, he says, preserved the genuine text, as the author himself had written it down. But Purnabhadra was well aware of the fact that the order of the
1 Cp. Bflhler, Ueber das Leben des Jaina MCnches Hemachandra (Denkschriften der philos.-hist. Cl. d. Kais. Ac. d. Wissenschaften zu Wien), p. 183 ff.
»l.c.,p. 185.
1 In my paper 'Uber das Tantrakhyayika, die kaSmlrische Rezension des Pancatantra1 (=Abh. d. phil.-hist. Klasse d. kgl. sacha. Gesellschaft d. Wissensch. xxii, No. v), Leipzig, B. G. Teubner, 1904, p. IXT.
4 Cp. my edition of the Southern Pancatantra, p. Ixv f.
* See ' Uber das Tantrakhyayika ', p. xiiv, and below, p. 57.
28 Chap. II. Purnabhadra, his time, work, and language
tales preserved in Sar. /3, and no doubt in other recensions which he used,1 was the original one. Hence he adopted this order in his third book, whereas the textus simplicior in this third tantra deviates considerably from it. The last two books were very short in the old recensions of the Pancatantra ; but they were enlarged in the textus simplicior, which in an even higher degree than Purnabhadra's recension is not a mere revision of the old text, but rather a free imitation of it. Consequently Purnabhadra very largely availed himself of this ' remaniement ', not only in these books, but throughout all the text. As to the fifth book, he took it over into his work, with only slight alterations, in the form which it had in the textus simplicior.
Our parallel Specimens I to IV show how he chose his wording, now from the one and now from the other source, according as he was more pleased with the former or with the latter. In most cases it is im- possible to say what principles guided him in his choice. In our Specimen III, however, it is evident why he suddenly abandons the wording of Sar. /3. He does so at the passage where this recension speaks of the brahmanical tirthas. As the textus simplicior replaces the enumeration of these tirthas by a conversation on the dharma, Purnabhadra in this place followed this Jaina recension, and chiefly because its wording was not offensive to his religious feelings.
Purnabhadra's principal aim was to revue the text ; see his prasasti, stanzas 2 and 3.* In stanza 6 he tells us, that of the words of the ' excellent first poet' only 'a handful had remained uninjured'. The very numerous corruptions which can be proved to have existed in the text of the Tantra- khyayika as early as (at the latest) the time of Ksemendra (about 1000 A. D.), and the nature of the textus simplicior, whose wording differs very considerably from that of the more original recensions, confirm
1 Cp. Somadeva, Ksemendra, the Southern Pancatantra, and the Pahlavi version, which all, in this respect, agree with ^ar. Purnabhadra seems to have known Ksemendra's versification of the Pancatantra ; cp. WZKM. xvii. 347. With Somadeva he agrees against all the other recensions in several places. Cp. our parallel Specimen I. ,
* It seems to me that the MS. of Purnabhadra himself (the mulaprati) contained in part original leaves of his sources which he merely corrected: mistakes like vidhaya for pidhaya, drstvapayo for drstapdyo, and others (see the list given below, p. 30 f.), he is more likely to have overlooked in some MS. of the textus simplicior which he revised, than to have copied from it. The scribe of the prathamadarsa or first copy of course preserved the mistakes overlooked by Purnabhadra, and committed some clerical errors of his own. Hence it is possible that the common archetype of bh* is identical with this prathamadarsa. I have not been able to find any other MS. of Purnabhadra's recension which can be proved to go back to another archetype. But cp. our Variants 5, 1 and 33, 22.
Chap. II. Purnabhadra, his time, work, and language 29
Purnabhadra's statement. From our parallel Specimens I to III, and from the text printed below, p. 58 ff., it is certain that Purnabhadra had before him MSS. of the H-class as well as of the <r-class of the textus simplicior. Perhaps he also knew the north-western abbreviated recension from which the so-called Southern Pancatantra, the Nepalese recension (v), and the source of the Hitopadesa have flowed.1 He therefore was in a still more puzzling situation than Kosegarten. But the principles he followed in constituting his text were exactly the same as those of this scholar : both of them contaminated the texts they had before them to such a degree that the results were in fact virtually new recensions.
But Purnabhadra's aim was not only to restore the old text ; he also wished to amplify it (prasasti, stanza 6). And this he did in numerous places.
In revising his sources, Purnabhadra, on the whole, abstains from radical alterations. This is clear from our parallel Specimens. These show that — according to what we should expect from the wording of his prasasti — he follows his sources rather faithfully. The story Sar. III. x has been transformed by the author of the textus simplicior, and has been transferred by him to his fourth tantra as No. i. Purnabhadra, who found it in Sar. as well as in the textus simplicior, gives this tale in both places, at first in the form of the Sar. version as his III. xvi (see our Specimen, No. IV), and again in that of the textus simplicior as his IV. i. But even the sources still unknown to us, from which he derived the stories not to be found in Sar. and Simpl., he seems to follow very closely. His story III. viii has been taken either from some text of the Mahabharata, a or from an abbreviation of it, or — possibly, but not probably — it goes back to some revision of the source of the Mahabharata version. Purnabhadra's text is much shorter than that of the Mahabha- rata version ; but nobody will deny that the former, on the whole, goes back to the latter. In contracting the text it was of course impossible for our author — if indeed he and not some other writer before him was the abbreviator — not to change the wording in several places.
Some of the prose stories which he took neither from S^r. nor from the textus simplicior, reveal their origin by their language. Most of the Guzeratums of Purnabhadra occur in such stories : ^ffui^H^ 3 I. xii (73, U); Mftfr ^T I. xxx b (122,18); 3H^j£ <s)<i*HM IV. v (244,18).
1 See my edition of the Southern Pancatantra, p. Ixxxviii.
2 MBh. xii. 143. 10 ff. I can compare only the edition of Protap Chundra Boy. In this edition the story shows several manifest interpolations.
8 See this and the following words in the ' Brief Glossary ' appended to vol. XI.
30 Chap. II. Purndbhadra, his time, work, and language
Similarly we find that a wrong Sanskritization of a Prakrit word, *J occurs in the new story II. viii (166, 2) ; but the same word has been employed by the author himself in the old tale III. ii (184, s). The new story II. v contains the Prakritism xjj[4<<ft (148, 4), and the form
(for °lj^i ; 149, 12.16 ; 151, 2.9). The Prakritism 3?»*J|«*J1 occurs in the katha- samgraha stanza of the new story II. i (127, 16).
Evidently the words oRTT «|f^*i in stanza 6 of the author's prasasti refer in part to the additional stories of his text, and in part to additional stanzas, or to passages in which he strove to imitate the artificial prose style which prevailed in his time (cp. 46, llff. ; 183, 13 ff. ; 185, 12 ff. ; 213, 2 ff.), or even to new features by which he enlarged the old text, as e. g. in the beginning of I. x (66, 10 ff.).
Purnabhadra declares in stanza 2 of his prasasti, that he has done his work gvui<^*Ui and no doubt this assertion is trustworthy. Still he has overlooked several blunders of the MSS. which he used as his sources, or has even misread these MSS.2 I give some instances from his text.
4, 23 f^WT^Vlff , a misreading of Islar. (A 8) 4, 23 TTt ^TK , a misreading of Sar. (ibid.)
4,30 °3rfl\°, a misreading of 6ar. (ibid.) °?iniF0'
11, 23 *inrN! only in Pr and Simpl. MS. I. The MSS. bhN#PM ABh and
Simpl. H h have the blunder IJSli***^. 24,3 H«J*l, a chandobhanga ; also in Simpl. Hlh.
29,20 For arf^l the original reading evidently is 3Tft; but our MSS. of Purn. and Simpl. Hlh agree with us.
39.23 7f for <J?^ ; MSS. of Purn. and HI with us. h correctly 7T*^.
44.24 oqqR^fl: MSS. of Purn. and Simpl. Hlh. The original reading must have been uf^d: •
54, 2 *J^S$*U> the name of the princess, evidently goes back to the which the textus simplicior mentions as the weapon of Visnu.
69,3 *f«raT'l.; but bh*ABh and Simpl. Hh fq«i||«^. I's correct reading
must be a correction. 69, 4 *f?t^*i ; bh* and Simpl. h %^R., A and Simpl. H "%*{*{. The correct
1 Cp. Pischel, Gramm. d. Prakrit-Sprachen (Grdr. i. 8), p. 102, § 130.
1 Or perhaps he used some Nagari transcript of the Tantrakhyayika, in which some
Sarada words were misread.
Chap. II. Purnabhadra, his time, work, and language 31
reading only in the revised MS. Bh, in Simpl. I, and in Kielhorn'a edition (evidently a correction).
163,12 MSS. unmetrical. The chandobhanga seems to be original. Cp. Variants.
207,5 ^cetq-q«fl is here the proper name of the Raksasa. J-lar. /3 has for the reading of Sar. a 5 ^ «r«f<^«t*u^ . As in Sar. /3 —
the recension used by Purnabhadra1 — the subject of the sentence is
missing, Purnabhadra evidently tries to correct this passage. 211,6 *fwni is the correct reading; but Simpl. Hlh with bhN*PPrMA
fqtt(«i. In Bh the passage is altered. 220, 18 fa«i*iMffi<i also SP and v. The original reading of the Pancatantra
must be f«H«<Mf7T7Tt , and so Sar. reads. 220,25 fire also SP/3 and K (oldest MS. of SPa). v and Sar. correctly
fire:; SPNftre:.
235,2 <j for 71^ MSS. of Purn. and HI. 236, 13 *'<jmMl4t ; MSS. of Purn. and HI
236,24 f«*fa<t for «*fo<v; HI the compound fafaaifl; h, corrupted,
240,21 *1TW1ZWT%S ; MSS. of Purn. and Simpl. Hlh 242,11 UZJiNKfo^KI, apparently a blunder for IWU^Kfa^KT, as the *-class reads. Simpl. HI liffm^i^fq^i (I), h
258, 25 ^^ir^^lf^^fo Purn. and Simpl. MSS. HI. See ' Brief Glossary ', vol. XI, s.v. Simpl. h has
262,18 and 263,2 ^ftf^li0 (so MSS.) Purn. and Simpl. HI. 269, 19 *fq<jM<l!*4Hd is right ; but Purn. MSS. and Simpl. HI h correctly, but with a variant, fM
271, 2 *^f7l8 here and in the following text is the correct spelling ; but
Purn. and Simpl. Hlh ^rf°.
These cases show that in several places marked in our text with a star (*) the faulty reading recorded in the variants is surely or possibly that of Purnabhadra himself. Cp. below, p. 77 ff.
§ 4. Purnabhadra's language.
According to stanza 4 of his pralasti Purnabhadra seems to be aware of the fact that his Sanskrit is not quite free from mistakes. The author of the Dhanyas'alicaritra tells us that he has caused his work to be
' Cp. ZDMG. lix. 21.
32 Chap. II. Purnabhadra, his time, work, and language
corrected by a good vacaka. Indeed, Purnabhadra's Sanskrit is neither uninfluenced by the time in which he lived, nor by the vernacular which he spoke from his childhood. Of course it is not quite free from Prakritisms either. Part of these irregularities he took over from his sources.
Some Guzeratisms and Prakritisms of his have been mentioned above, p. 29 f. From the textus simplicior (H-class) he takes the Guzeratisms 3T»frR 285, 21 and 3T^HMI^ 286, 5 (see ' Brief Glossary ', a. v.). To the influence of Guzerati we may perhaps attribute the wrong "^ after the compound in 180, if.,1 and the wrong form °«fi^t<>, as our MSS. write for correct °3fff5B0 in the same passage. In Subhasilagani's Jagadu- sambandha2 we find the wrong compound *ifq*si^i«<i'm'^«4, which Blihler explains as follows: 'The faulty feminine ¥ft«siM has been caused by3 the custom of the Gujaratis to write the parts of a compound separately, viz. qfcHKT *rnrr ^i.' The Hamburg MSS. have ^, not after the compound, but after o^r^ft0 — perhaps a correction of the original wording preserved in Purnabhadra's text — and even more members of the compound are here in the nominative case. To the custom mentioned by Biihler we evidently must attribute the occasional use of f^fif^ for ^if^ (236, 21
1TWH., from the reading of HI fefif^SfTT^; cp. 68,3 ,; 223,15 fsfifq^M). I now regret that I corrected f^f
as the MSS. write in 90,17. Cp. also 3T^T Jl<JN*i5^ for dfloNjfiJ^ or
Prakritisms are fl for Tfit. 39, 23 (or ^ used as a masculine) ; 235, 2 (also HI); 277,13 (HI here correct); cp. 150,17. Other Prakritisms are recorded above, p. 30.* Cp. also the vulgar adjective ^R^f 'belonging to', which, as Prof. Hultzsch suggests, should be written for ^HR 246, l and 3, and which really stands in MS. A.
Orthography and Samdhi. Both are inconsistent even in our best MSS.
Vowels. Post-consonantal ^ occasionally interchanges with 3T- Ex. •5^ best MSS. for usual ^|^ , 222, 13 ; 223, u. fZ*f (rarely MSS. grp), and °3id(J«(<n, 152,8 ; 3T"I^ for usual 3T^P$, 46,5. 5f3 best MSS. for usual
1 Though occasionally ^f occurs also in other Sanskrit texts after a dvandva com- pound. Cp. e. g. Sar. ddimadhyavasanam ca, 133, 14 note.
2 Buhler, Indian Studies, v, No. I (Wien, 1892. In Commission bei F. Tempsky), p. 74. * In Biihler's text ' by ' has been inadvertently omitted.
4 The first member of the frog prince's name ^JI-tfTf 235, 1, might be taken as
a blunder for Prakrit WtSf, i.e. the name of the river iffTIf. But it is more likely a misreading, as HI and h read 'i^TH • Buhler, 8, 16 has
Chap. II. Purnabhadra, his time, work, and language 33
, 120,7. ^Jft for usual ^Ffft, 224, 13. In 225,26 Pr writes in 226, s *PPr ^ril.1 ^ is lengthened in the MSS. in thg^l 74, 8 (by the influence of ^l?).
Wherever in these cases our printed text deviates from the spelling of the MSS., an asterisk refers to the variants.
Consonants. ^ ^ T!t^ ^ before a consonant, and J^ at the end of a sentence or of an even pada, are always replaced by anusvara in the MSS.
*{. stands not infrequently in the MSS. for printed anusvara before «t and J^. Ex.: ^rfaiTflHJ, 28,6; ^fafaMt, 31,20; ^faT^d^M, 29,5; ^, 70,25. fl«4JM and *j««i4«i, 18, l, &c., occur along with
(so 20,5 ; 22,5), W\%<( (20,9), WPTOT0 (20, 13), *?rnfrf (20,24).
T stands sometimes for UJ.
On the other hand, we occasionally find anusvara for correct 1.; so fl fan«lH, 140,18 and * 163, 10; 71 f*JVT*>*<l<IV, 142,5.
^T and ^T, *I and ^ are occasionally confounded2; ^ is often employed for ^§, Z often for "Z (nearly always "Z for ^).
^5 is occasionally confounded with <R ; cp. 15, 18 ; 40, 17.21 ; 41, 4 ; 165, 21 ; 186, 19 (see Variants) ; 229, 9 (Bh) ; 266, 10 (see Variants) ; 280, 8 (see Variants). This produces the variant ^^<^«1 for >d<*U<^«l (see Var. on 40, 17, &c.). Cp. Lanman on Orthographic Prakritisms in Album Kern, p. 302.
S| and ^ (0*Hfll<38 for °<MWI<J0, 53, s), 55[ and JK (this written in the form given by Jacobi, Kalpas., p. 18, note) are occasionally confounded in the MSS. ; cp. e. g. 10, 10 ; 271, 15 (see Variants in both places).
tjj and ^ interchange in ^^T- I follow the Petersburg dictionaries, referring by an asterisk to the variants, where bh^ have X£. The MSS. write «ir«^; I write ^f^t in my text.
Sibilants, especially H and ^J, are sometimes confounded. In one case it is certain that this confusion goes back to Purnabhadra himself:
1 Most of these cases are also found in other north-western works. Cp. even ^t'J'^T
for <^|£-*4| 153, 24, in a stanza not composed by Purnabhadra himself. A
2 T|f^ and ifa are used promiscuously in the MSS., whether they mean ' group ' or
'piece, fragment'. Apparently Purnabhadra pronounced these two words alike, and hence I write in both cases 4gl!£. Cp. also the Petersburg dictionaries and Apte, s.v. TfTIS and *§TTJ*. But etymologically the two words are not identical. 'Group' is
(Purn. 5, 17 and ex conj. 122, 12), Pali and Prakrit sanda; 'piece', 'fragment' is
(Purn. 112, 9.11), Pali and Prakrit khayda.
F
34 Chap. II. Purnabhadra, his time, work, and language
in 56, 12 the wrong form *nSi<T. (for STUFc^) is assured by the pun with
; ' often '.
is confounded with tf in T^pl, 227, 20. Cp. forms like HVJOI, ^i), HT^w(=fi)j TTrgfW^i at the side of *uj»m(gi) and HTFftRi in the Petersburg dictionaries.
Consonants are often doubled after ^ ; but in the great majority of cases the doubling is neglected, jft is often written for Jff:.
Samdhi. As a rule, samdhi is not observed before the apodosis (often in connexion with danda !) ; it is neglected in cases where its observation possibly might cause confusion. Cp. Wackernagel, Altindische Grammatik, § 262, b, S. It is evident that Purnabhadra himself very often neglected the samdhi. Cp. 138, 12 ^§T OT, and 149, is where our MSS. have
3T°. In most cases our MSS. write <Tc^ SH<$\. Before initial ^, samdhi is nearly always neglected in the MSS. An interesting case is 149, 2 f., where the archetype evidently had ctHfJohl^H^yy W- The archetype of bh* wrongly resolved this group into <*Tl<j<*ir^ CS° (instead of «tTlfJ<*iT«t. *JS°) ; and A Bh <I> have a wrong correction of this inadequate reading : ^gtnifaa0. Cp. Prof. Lanman's remarks, vol. XI, pp. xxxix to xlviii.
As our MSS. are inconsistent, samdhi has everywhere been restored in the prose of our printed text except (1) in the case 138, 12, (2) before the apodosis, (3) before and after oratio recta. In the stanzas, in which the rhythm annihilates the pauses in the case of punctuation, we follow our MSS.
In the body of the words, our MSS. are not consistent as to the samdhi in the following cases :
°W or °H° is often written for 6:^°.
°:^i0 and °W, °:*R° are very often, if not in most cases, written for
Punctuation. Our best MSS. are carefully punctuated. They employ danda after the complete sentenceSj and very often before the apodosis. Punctuation before S^Tf after oratio recta is not rare. Before ^7f: in the phrase ^3W ^ i ^TfH they always put danda, or even (^P) double danda. For the sake of clearness, we employ in our printed text ardhadanda before the apodosis, before and after oratio recta, and before *H!5 all this in the prose.
In the metrical parts, we separate the first from the second and the third from the fourth pada of a stanza by ardhadanda, where these padas form one line, i. e. in slokas and aryas. But in these cases we do not destroy the samdhi, which is here maintained in the MSS. even when
Chap. II. Purnabhadra, his time, work, and language 35
they follow our own method. 1 Our MS. bh employs the ardhadanda and the double danda, and these only ; our MS. * employs the danda and the double danda, and these only. Cp. the two facsimile tables in vol. XI.
Gender. ^TRf neuter 233, l (in a stanza taken from the textus simplicior). *J^ masculine, or 7T for <Tf^> 39, 23.
Guna and Yrddhi. *Jeji*uR,<*l0 (also Hamb. MSS.) for «T| 31*11 v«i° (Whitney, § 1222 j, Panini V, 1,133), 258,25. sfrfs^0 for iftf^R°, 262,18; 263, 2 (in both cases with HI).
Verb, fajy+jfd (also HI), 23, 5. Imperative : ^^ sfy for 4*IM l**lf«fj 117,18; 118,2. Infinitive: f«l«if^fj*i, 57,23. Gerund: 3(1 filial, 175,23 (in a stanza). Gerund in -am : ltT^<t-3fi-*X. > 68, 2. Passive for active voice, 205,24 (in a metrical quotation ; also HI).
Noun. A wrong form is the genitive •Jgr^Mt'lt for °3ft, 135, 10.
Nominal compounds. Compounds with proper names:
136,20; 140,15. S|(tM^*!|*gl, 114,20, beside <?|T^(!]%lf8«i, 114,22.
t, 4,6, beside *JT^««i?q>5*iiiif^f^<ri*T, 21,17. «^+i£ttfa<i, 103,3. r: , 266, 10, &c.
A curious case occurs at 283, 23, where I have written * WTT^ %*t with Purnabhadra's source, the textus simplicior, as represented by the Hamburg MSS. Bh, which in the fifth book belongs to the H-class of the textus simplicior, reads *JTTc^ l^t (<ra misread for T^, i. e. %) ; h and Buhler ; (!)• But bh^A and their derivatives write %Trfff%ii, and this
seems to be the old reading, from which Biihler's ungrammatical reading derives, as being apparently an original gloss by somebody not well versed in Sanskrit. I now take ^TTft^'r to be a subttanlive dvandva compound, depending on *|Tdjfd : 'it went to speed and to over-speed', i. e. ' it ran more and more swiftly ' (quicker and quicker).
Prof. Wackernagel, in his Altind. Grammatik, II, § 74 d, gives similar adjective dvandvas. From the Pali I may add Jat. i, p. 160, 3 vamkdtivam- kinam (in a stanza), which the commentator rightly explains as meaning mule vamkdni agge aiivamkani tddiwni sihgdni assa atthiti vamkativamkinam. The compound mancdtimanca, given by Wackernagel from Trenckner, is apparently a substantive 2 formed exactly like our vegativega. Hence %TTf7nfai should be restored in our text.
1 Only the MSS. in such cases, do not separate the combined aksaras. The Hindu manner would be to write, e. g. in our stanza I, 5 (p. 5, 3) ^M4|f4{ | t^^fajo.
a This is also the opinion of Prof. Wackernagel, who kindly pointed out to me this compound.
36 Chap. II. Purnabhadra, his time, work, and language
The rule laid down by Panini iii. 3. 126 (Wackernagel, Altind. Grammatik, II, § 82, a, y) is not always observed (at least not always in our best MSS.). In 131,26 only A — a revised MS. — has the correct form ^H^I ; but Sar. agrees with the other MSS. In 9, 23 the MSS. have our reading. In 227, l ff. Purnabhadra follows this rule, whereas his source, Sar. /3, A 266, neglects it.
Syntax. Periphrastic present indicative (Guzeratism) : yfafa ?w: , 122,18; qlsiqfH SW, 268, 10 (here also Hamburg MSS.).1 Present indi- cative for imperative: Mfauirl, 37,8; ^<4J|*i:, 92,6; y^W, 92, 11 ; 267,16; 3T^Ti:,265,i7; fa^, 268,5 ; TngwhRTm: , 268, 6 ; **0ft, 271,6; 278,9; 279,24; *R5Tf»T, 282,8; 286, 5; STRlf*, 288,20. Present indicative for conditional: WptfTT, 283,22; »H!j*jfd, 283,23 (in both cases also HI). The conditional occurs 216,8 WtiR^'l.and 3T*rf*TSffl(,, and 230,20
General subject expressed by 3rd person sg. : OT^, 180,20. Cp. 4,21 (but see Uber das Tantrakhyayika, p. 98, 22).
Genitive for instrumental case: °%*rap55f JJcTT, 75,23. Instrumental for genitive case: «H<!n*l€ftlHi!:, 166.7.
Faulty or awkward constructions : f^TS!n!*rffo (for °^fonf*l) • • • , 2, 12 (in accordance with Simpl. Hlh) ; ^T ..... ^Rft n«u*irfl>
109, H (almost literally from Tantrakhyayika, 55, *). jrfflMl<^*jfa for cf?I (the subject being *RT*0, 194,24. ^£1 for ££*{.> 221,29 (or f^l^ <\fitt<\. for f^T?I! U, 1.28). A word like ^: to be supplied 273, i in a stanza (HI have the same wording). 285, l an anacoluthon with HI (stanza). Varia. *iten<l. and JTVf with the dual number:2 OT^ffrTWT^, 43,9; ., 209, s. tr*J and 3T^ in the same sentence, referring
to the same person, 4,sf. faftfa for <*^lt^, only 254,7 (104,11 read H**lfd with *). Superfluous 1^71, 23, 11 ; 42, 6 ; 61, 12 ; 63, 21, &c. for simple ?^7f, 118, 2 (in a new tale). fg}<$ after ^f?f, 94, 19. JTT . . . for »T ... 3^,41,4.
In nin<«i*i>Jr*ifn MTT, 148, 2, MIH^^^H. ^ an adjective. Cp. and f»T^^ (125, 29) beside f*!M5)«ft (3, l), &c.
APPENDIX.
Literary quotations: Salihotra, 279, 9 (also in HI); Karnisutakathanaka, 67, 14. An utterance of the Buddha is referred to in 48, 13.
1 Cp. Campaka^resthikathanaka, ed. Weber, 1. 454 f.; ed. Hertel, § 76. Bflhler, Sitzungsb. d. kgl. Preuss. Ak. d. Wissenschaften, 1883, p. 885.
1 This also occurs in other mediaeval Sanskrit teits, e.g. in Somadeva's KSS. 42, 137.
Chapter III. Account of the Manuscripts on which this
edition is based.
§ 1. Description of the manuscripts.
ALL the MSS. used for this edition are paper MSS. written in Nagari characters.
th = Decc. Coll. x. 190. This MS. originally consisted of 179 leaves, 10 lines to a page. Its first 6 leaves are now missing. Though this MS. proves to be a Jaina MS., as it has the Jaina diagram in the beginning of book V, it has not the square blanks in the middle of the single pages. It has been copied from some MS. which was then old ; cp. Variants 288, 2.4.6.8 ; 290, 8.
According to bh's colophon, this copy was completed in samvat 1468, on the 12th day of the bright half of the month Margasirsa, during the reign of King Sri-Kahnadadeva Vijaya, in Srl-Viramagrama (the modern Vlramgam near Ahmedabad) ' in compliance with the order of the minister Mahamsalasa for the amusement of Josiharadeva, brother to Vadljanarddana of Satyapura ', by Mahamgopala, son of Mahamkesava, of a Gauda family.
The copyist did his work with great care and accuracy. Our MS. contains many glosses, written by several old hands on the margins or between the lines. Most of these glosses go back to one hand, no doubt the hand of some beginner in Sanskrit, who sometimes misunderstood his text. Cp. the glosses on 11, l ; 14, 6 ; 19, 20 ; 22, 25 ; 56, 12 ; 60, 20 ; 62, 16 ; 76,10; 84,17; 154,2; 176,17; 177,9; 194,17; 221,25; 286,3. The same hand entered a lot of vernacular glosses; cp. 11,3.22; 17,17 (twice); 23,3; 46,4,; 65, H; 70,15; 74,5; 104,19; 131,18; 147,8; 175,30; 176,18; 177, 1.29 ; 178, 26 ; 180, 4 (twice) ; 183, 10 ; 188, 22 ; 189, 2 ; 190, 13.18 ; 191, 13 ; 192,7; 193,11; 209, 11; 220,16; 231,2; 247,3; 277,10.11.16.17.18.20; 278,22 (twice) ; 279, 1.2.6 ; 283, 17 ; 288, 13.16. In many cases the text was unintel- ligible for the glossator ; for the marks x and = , which he usually employs in referring to marginal glosses, occur not infrequently without such a gloss. On the first leaves he often separates the words by small vertical strokes, writing initial vowels, and sometimes terminations, over the line. Our stanza I, 52, for instance, looks thus in the MS. :
' 88; 9, so fjctirwui.; 10, n
As I did not succeed in distinguishing with certainty the different hands of the glossators, I mark the glosses in my variants with ' gloss.' or ' corr.
38 Chap. III. Account of the Manuscripts
of bh '. As possibly scribes of other MSS. may have copied from bh the glosses instead of the original readings, I have entered nearlyall of these glosses in my variants. The complete readings of bh are given in my variants.
M" = Decc. Coll. x. 189. This MS. is complete in its beginning, but has a gap extending from 220, 18 of our text to 236, 8 (see Variants). The original number of its leaves was 117. The average number of lines on a page is 17. This copy is written in a hasty hand, but it is pretty correct. Neither the Jain diagram nor the middle squares occur in it.
The colophon tells us that this copy was completed in samvat 1855, sake 1720, in the dark half of Karttika, on the eighth day, a Tuesday, by Harinanda, son of Kasinatha, of a Gauda family. The complete readings of this MS. are given in our variants.
"9 = Decc. Coll. iv. 55. It has 102 numbered leaves, 15 lines to a page. Of these, leaves Nos. 46, 48, and 49 are lost. Moreover, the last one or two leaves are missing. The actual pagination, however, is not the original one ; the original one, written in the margins, has been corrected by a later hand, after leaf 19 had been lost, and hence leaf 20 is now numbered as 19, &c. But another hand writes the correct numbers once more just over the red middle spots of the verso pages ; see Key, above, p. 1, and our Specimens, vol. XI, Table I, No. 1.
fy is a very beautiful Jaina MS., the Jaina diagram appearing in it at the beginning of the Kathamukha as well as of that of books III, IV, V. (The beginning of book II is lost.) In the middle of the single leaves there appear the characteristic blank squares, and the centres of these squares are perforated by small circular holes throughout the MS. These holes appear never to have been used for a string drawn through them, as in palm-leaf MSS., to keep the leaves in the right order ; for such a string would have enlarged the holes or torn the leaves. The single leaves of * show red circular spots, one in the middle of the blank squares of the recto pages, and three on the verso pages, viz. one in the middle and one on each side margin. The original leaf-numbers are written within the red spots of the right-hand margins of the verso pages. The red spots, however, are missing on leaves 83, 84, 89 recto, 90 to 102 inclusive. At the end of the first book, 4| lines of the recto of fol. 45 and the whole verso page of this leaf have been left blank.
This MS. has been written with great care in beautiful characters. The copyist himself corrected it, and added some glosses and various readings. A second old hand added some more glosses.
* seems to be our oldest MS. of Purnabhadra's text. As we shall subsequently see, not only the common archetype of P (dated sam. 1537)
Chap. III. Account of the Manuscripts 39
and L1 was copied from * at a time when * already contained the glosses by the second hand, but — apart from the circular perforations of the blanks, which are unknown in most of the paper MSS. — the forms of the characters in this MS., especially that which ^T has in it, are very old ones.1 This old form of ^1, as it appears in Biihler's Palaographie, Table V, number 13, columns v, and vii to xvii, and Table VI, number 18, columns xv, xvi, xvii, prevails throughout in *. Cp. our Key, p. 2. Only in cases where a vowel, or n and r are written under gh, the modern form of gh is the usual, though not the exclusive one. Cp. also the form of f3?J in our Table II, No. 12, 1. 2 a and that of 3J in our Table I, No. 1, 1. 15 c, with Bahler, Table V, col. xxii, 1. 18. The complete readings and the glosses of * are given in our variants.
P = Decc. Coll. xxiv. 419. It has 96 leaves, 15 lines to a page, and is very beautifully written on fine thin paper. This MS., which is complete, shows the Jaina diagrams as well as the characteristic blank squares. The text on the whole is very correct. From the colophon we learn that this copy was finished in samvat 1537 on the first Tuesday in the dark half of Asadha. The copyist's name is not given. The complete readings of this MS. are entered in our variants.
L1 = Leipzig University Library A. 404. Incomplete. Old. The leaves still extant bear the paginations 2 to 56 (both incl.), corresponding to our text svanama° &c. 2,s to garva te (incl.) 220, 2. 15 lines to a page. No blanks, but Jaina diagram before II and III. Two copyists, the second one (from leaf 11 to 20 incl.) giving a very faulty wording, and leaving out the text between murkhah (67, ii) and sthitavati (74, 17). I only occasionally refer to this MS. in my variants.
Pr is an old MS. belonging to the Jaina Bhandar of Ahmedabad. It was kindly lent me through Mr. Keshavlal Premchand Mody, B.A., LL.B., of the same town. This copy bears the signature TT ^ IT ^0. On its margins the title of the work is given as M-sfiyiMffW . Pr consists of 107 leaves, 13 lines to a page. It is pretty correct. I give the complete variants of this MS.
M = Decc. Coll. iv. 54. 102 leaves, 15 lines to a page. A complete Jaina MS., with Jaina diagrams and blank squares in the middle of the pages. The characters of this MS. are beautiful, but its text bristles with blunders, omissions, and dittographies. Though not dated, this MS. is not modern. In my variants I give the complete readings of this MS., but I have not noted many of its blunders, small omissions, and dittographies.
1 In one case this form of ^f also occurs in Pr, which has flowed from *.
40 Chap. III. Account of the Manuscripts
p = Decc. ColL ii. 46. 93 leaves, 18 lines to a page. This is a complete Jaina MS., though it has not the characteristic blanks. According to its colophon, this copy was completed saimal loka-muni-rasa-SaSi-samvacchare [i.e. sam. 1677] jyesta sustivdAistaml somavdsare Siibhaveldydm "Dhill&naffare [i. e. Dhillo, Thar and Parkar] j)d(isdAa-J&h&mglj:&-rdjye I vd°[i. e. vacaka-]- Matibhadra - tacchwya - vdcandcdryya - dhuryya - vddikarikumbhakamdanamrgdri - sarvaSdstrddlnta-sarasvallkamthdbharana - sakalakaldkaliiagdtra-vidvajjanatilaka -
randmmjama&aramdaldnaddsdnuddsa-Gov&rddh&na.-m'uni-lipd&rtdr iyam prati. The lengthy colophon proceeds to tell in several stanzas that the copyist did his work with the utmost care, and that the good should correct the copy, without blaming the copyist for the blunders he possibly might have committed. Hope is expressed that the Jain community might rejoice henceforth by the favour of the suris (Srlmaj-jinakula^-suri-prasdddtc dram namdatu], and that the MS. might eternally survive and be protected by its owners from oil, water, loose tying-up and dishonest borrowers (' teldd rafaej jaldd raJcse rakxet sithila-bamdhandt parahastagatam rakwt ' evam vadati pudakam ll). The copyist had at least two MSS. before him during his work, choosing their readings as he proceeded in copying, smearing with gamboge whole passages already written in almost every line, and replacing very often the correct readings by inferior ones. Some passages have been copied from MSS. belonging to other classes, e. g. to the Bh-class. Moreover, many corrections and additions have subsequently been added by different hands. Amongst the numerous marginal additions, there are even stanzas in Prakrit. I carefully collated this MS. down to 119, 23 atha inclusive ; but then, seeing that it was of no use whatsoever, I neglected it altogether. In my variants, only occasional references are given to p.
A = India Office 2643, E. 4084, K.R. 9. B. This MS. originally contained 153 leaves, 12 lines to a page. Leaf 92 is missing now. On the first page, which is blank, a European hand has written in English characters : Gaikawar. This copy has been written by two copyists (A1 and A2). The first hand wrote the text of leaves 1 to 93, and of 123 to the end, the second one leaves 94 to 122 (both inclusive). The words from gacchet, &c. to ballyasd I ta ta (inclusive) = our text 177, 12 to 177, 23 are written by both the copyists, and subsequently have been deleted again on foL 93 verso. In this short passage A2 is more correct than A1 ; but the former shares with the latter the mistakes pranadhi* 177, 12, tanna for tatra 13, aparam kdrya° is. Hence it is certain that A2 copied from the same original as A1. A2 leaves blank squares in the middles of the pages.
1 Dharma Vijaya Suri corrects this to -jinakutala-,
Chap. III. Account of the Manuscripts 41
This MS. is rather faulty. A third hand has collated it with some MS. of the text us simplicior, covering the margins with additional stanzas from this text, smearing very often the original readings of A with gamboge, and writing on them inferior readings or downright blunders. In other places the corrector, an ignorant and careless man, restores defective passages ex conjectura.
According to the colophon, the copyist of A1 was one Sieamndara, who completed his work in samvat 1574 dso vadi 9 sukre.
Bh = Decc. Coll. xiii. 86. This MS. originally consisted of 156 leaves with 12 to 14 (generally 13) lines to a page in books I, II, III, and with 11 lines to a page in books IV and V. The first leaf and leaves 132 to 140 (both inclusive) are missing.1 Bh has been copied by two hands, the first one going from the beginning of the MS. down to the end of the third book, the second one from the beginning of the fourth book to the end of V. This copy is a Jain MS., the Jain diagram appearing at the beginnings of books II, III, IV. Besides the usual square blanks in the middle of the pages, which are perforated as in *, most of the recto and verso pages have larger rectangular blanks either at the right or at the left hand margin, or even at both of them.
At the end of the third book, the first copyist gives the date gamvat 1442 var*e without any further information. The rest of the MS. is scarcely younger than its first part.
$ = Decc. Coll. xxi. 719. 97 leaves, 15 lines to a page. Dated tamvat 1661.
§ 2. Value and mutual relations of these manuscripts.
Of the above-described eleven MSS., the first eight form two groups. To the first group, the bh-class, belong the MSS. bh and N. To the second group, the *-class, belong the MSS. *, P, L1, Pr, p, and M. The nature of MSS. A and Bh and * is such that they require a separate and detailed discussion.
The differences between the readings of bh and * are but slight ones. These two MSS. are excellent copies. The cases of the very mistakes taken over from either the textus simplicior or Sar. into Purnabhadra's text and preserved in bh* show how conscientiously the text has been handed down in these two MSS. On the other hand, the fact that N can be proved to go back indirectly to bh, and that PL'PrMp can be proved to go back to * (P, the best and oldest of them, and L1 indirectly), evidently
1 Comprising our teit, p. 244, 10 yadi (incl.) to 260, 2 dustajd (incl.).
Q
42 Chap. III. Account of the Manuscripts
shows that even in ancient times bh* were considered very valuable MSS. I think, indeed, that their common archetype is the prathamddarSa. See above, p. 28, note 2. At all events, their text cannot possibly deviate much from the mulaprati (i. e. amla-pratilipi) ; see our parallel Specimens.
§ 3. The manuscript N goes back indirectly to bh.
For evidence in substantiation of this assertion, see Variants to 27, 10.12 ; 56,12; 57,22; 61,12; 74,*; 75,12; 77,22; 78,9; 80,14; 89,12.13; 101,11; 123, n; 125, 18.19 s1 184,5; 186, a- 208, 2 ; 266, 6.22 ; 270, 22 ; 271,23; 274,17. The MS. N cannot have been immediately copied from bh, for it is evident that a part of N, namely 284,7 to the end of book V, has been copied from a text very closely agreeing with Bh. Cp. the Variants.
§ 4. The manuscripts F L', Fr, p, and M go back to *.
That p goes back to * may be seen from the Variants 72, 22 ; 88, is. Since, however, p is a contaminated MS. (see above, p. 40), and is for this reason critically useless, I have not taken the trouble of collecting further materials in order to ascertain more fully its relation to *. For the other four, the following evidence may suffice.
1. P and L1 go back to *; cp. Variants 3, 10; 4,24; 5, w; 5, si; 7,25; 9, 10;
14,io; 17,u; 23,16; 25,22; 32,24; 33,4; 39,20; 40,4; 42,20; 44,is;
57,18; 76,15; 101, 10; 119,10; 147, 20; 159,19; 164,38; 167, 21; 170, 10;
171, ?; 174, 9; 178, 18.23; 180,14.25; 181, e; 183, 11.20; 184, 9; 185, 6; 198, 15;
199,9; 224,18; 229, 20; 230, n; 231,4; 233,4.14; 247,9; 249, 2; 252,7;
253,24; 255,9; 258, so; 271, us; 275,9; 284,9. Besides P and L1 have a considerable number of corruptions in common.
Cp. 4,25; 5,22; 10,1; 12,21; 19,22; 20,8; 22, i; 23,9; 25,19; 29,6;
31,12; 38,24; 39, 15.16.21 ; 41,7.11; 43,1; 58, i; 59,9; 60.U14; 61,18;
66,7; 78,5; 93,15; 96,28; 99,22; 102,19; 104,20; 106,2.4; 109,25;
114,9; 115,2; 116,3; 128,4; 130,22.29; 131,21; 132,2.6.7.14; 143,11.22;
155,29; 157,3; 160,8; 167,19; 168,3; 169,2.8.10; 170,28; 172,4;
174,19; 175,25; 178,9; 180,4.12; 181,6; 182,11; 183,13; 184,9;
185,13.14; 186,5.19; 188,5; 189,20; 190,5; 191, 20; 194,11; 195,20;
196,3; 197,9; 200,1; 202,7.11; 203,13; 214,10.23; 216,1.10; 219,2. P cannot have flowed from L1, as P neither has I^'s gap (see above, p. 39),
ndr the very numerous mistakes of the second copyist of L1. Cp.
besides 12,24; 66,7; 105,6; 184,8; 191, is; 193,2; 194, 11 ; 197,14;
198, 3.
1 The citations of passages reproduced in Tables I and II of vol. XI are tet in italics. See Variants.
Chap. III. Account of the Manuscripts 43
L1 cannot have flowed from P. Cp. 12, 31 ; 14, 8 ; 15, 18 ; 19, 21 ; 27, 11 ; 32, 23 ;
64,9; 111,8; 155, 28; 169, 7.17; 174, 10; 178,2; 179,29; 190,22; 203,3;
209,18; 217,16. Hence it is clear, that both P and L1 go back to some third MS. which, has
flowed from *. Cp. also 24, 4 ; 190, 10.
2. Pr goes back to * ; cp. Variants 3, 10 ;. 7, 25 ; 33, * ; 101, 10 ; 119, 10 ; 121, 13 ;
147,20,; 159,16.19; 164,5.28.33; 167,21; 169,17; 170, 10; 171, ?; 180,25 ; 184,9; 187,18; 231, 4.
3. M goes back to * ; cp. Variants 3, 10 ; 5, so; 7, 25 ; 9, 10 ; 33, 4 ; 76, is ;
93,9; 101,lo; 147,20; 152, s ; 159,16; 174,9; 181,6; 185,6; 187,18; 231, 4; 247,9; 268,3; 271,28; 284,9.
§ 5. Critical discussion of the manuscript A.
A apparently belongs to the bh-class. With this class it has the author's prasasti at the end of the whole work, and it often agrees with this class in its readings. But very often also it has the readings of the ^-class. As to the gaps, it agrees at 33, 21 with all our MSS. but Bh. It has not the gaps of bh at 25,2; 82, 11, nor those of * at 61,5; 107,25; 113,29; nor has it the gap of N and of the *-class at 265,8, where bh is complete. Again, at 210,15, in the place of the padas missing in bhN, it has a text quite different from that of both the 'f-class and Bh, a cir- cumstance which raises the suspicion that A derives from a revised copy. We shall subsequently see that this suspicion is confirmed by other facts. At 62, i A has a gap which the corrector of A fills in as he pleases. Other gaps of A are 163, 13 ; 164, 15.
At any rate, A goes back to an archetype which was very closely akin to that of bh*. This is clear from the numerous blunders which it has in common with these two MSS.
Blunders common to Abb. 4*.
6,81.33; 9,3.26; 10,2; 11,14.23; 14,16; 18, n; 22,13; 28, 10; 33,12.16; 33,2l(!); 34,4; 35,18; 37,8; 42,8; 43,4.14; 44,6; 46,3; 48,19; 49,13; 51,6; 53, 1.5 ; 56, 3.4 ; 58, 8 ; 60, so ; 64, 3; 65, 8 ; 68, 3.14 ; 69, 3.4.6 ; 71, 10.33 ; 74, 8 ; 76,12; 83,2.6; 87, 16 ; 89, is; 90,17; 91,6.7; 92, i; 93,7.9; 95, n; 96, 10; 97,13; 99,s.6.n; 102,io; 116,13.17; 119, 21; 121,5.7; 123, 12.15 ; 125,i.so; 126,15; 130,4.23; 131,2; 132,16.27; 135,8; 136,4; 142,8; 144,19; 145,8; 150,24; 152,io; 156, is; 158,16; 161,2.13; 164, 10; 168,27; 170,15,20; 172,26; 176,19; 179,i; 180,4; 182, n; 183,6; 186,4; 196,13.14; 197,3; 198,9; 211,6.21; 215,23; 216,1; 218,2.12; 220,17; 223,3; 226, is; 235,16.24; 240,21; 248,14; 253, is; 259,8; 260,24; 266, 20; 269,19; 271, 20; 277,13.19; 278,9.lo; 282,16; 289, 3 (twice).
For these and the following cases, cp. our Variants.
44 Chap. III. Account of the Manuscripts
On the other hand, A has correct readings in many places where bh1!' are faulty.
Right readings of A, where those of bh* are wrong.
7, 16 ; 8, 15 ; 10, i ; 18, 6 ; 25, 19 ; 26, 6 ; 29, 8 ; 32, 23 ; 33, u (with Np) ; 86,20; 37,5; 40,17; 46, 7.21 ; 50, 12.is.ie; 52,23; 58,is.i7; 59, 2S(withPPr); 61, H ; 63, 12 ; 65, 29.30 ; 66, i ; 69, i ; 71, 18 ; 73, 14.21 ; 76, 4 (with pPr) ; 79, 12 (here the copyist corrects the reading of bh*); 83, is ; 85, 18 ; 87, 12.H ; 90,8; 91,19; 93,7; 96,14; 100,8; 101,23; 106, i; 107, 11; 109,5.8; 114,4.25; 116,2.16; 118,14.16; 119, i; 121,6; 122, 11.12; 124, 9.2s; 125,26; 130,9; 131,2 (the copyist corrects here) ; 131, 19.26; 132, 28; 134, i; 135, 21; 138, 6.12; 140,14.18.22; 141,9; 142,6.23; 143,24; 145,24; 147,2; 148,13; 150,20 (with Bh<t>); 152, 9 (with Bh<t>); 154, 2 (cp. 155, s); 155, 17 ; 157, is; 161,22; 162,18; 168, is (with MBh«t); 176,8; 180,7.13; 181,8; 191,19.20; 192,9; 194, 19 (with Pr) ; 197, 10; 200,23; 203,6; 204, 2.s; 206,5.7; 212,12; 216,8; 219, 25.31 ; 220, 7.26 ; 223, 19 ; 224, 16 ; 226, 14 ; 227, 4 ; 230, 20 ; 231, 25 ; 236,12.13 (see Hamb. MSS.); 236,19(1); 244, li(l); 245,13; 247, 7.18 (with Pr); 248,22.30; 250,22; 251,20.26; 254,16; 264,12; 269, 11.20 ; 271,11; 272, 3 ; 276, 7 ; 278, 6.10 ; 282, i ; 284, l ; 289, 10.
If in these passages A is more correct than bh*, this is at least in many cases the result of conjectural emendation. For in other cases the corrections of A are decidedly wrong.
Blunders of bh* wrongly corrected, or even more corrupted, in A. 3, 7 ; 4, 30 ; 13, 16 ; 15, 18 ; 34, 13 ; 35, s ; 39, 6 ; 42, n ; 44, 3 ; 49, ie; 50, 16.22 ; 52, 11 ; 55, 9.10.17 ; 59, 3.28 ; 64, 21 ; 65, so ; 66, 12 ; 70, 2 ; 74, 14.17 ; 80, 5 ; 84,16; 85,19; 99, 15 ; 101,12; 131,18; 132,12; 135, 10; 143,7; 149, 2 (with Bh4>); 155,8; 162,22 (wrong correction by copyist); 165,21; 170,10; 172,3; 173,15; 179,18; 190,2; 199,22; 207,3; 213,5; 214, 21 ; 218,12; 219, is; 222,6; 231, 5 (with M) ; 238,24; 250, is; 251,24; 264, is; 272,11.16; 273,9(!); 281,4; 287,14.
Discussion illustrated by Tale III, viii, Self-sacrificing dove.
Evidently A1 and A2 copied some MS. which had been revised and interpolated, part of the corrections and the interpolations being written on the margins. An interesting proof of this assertion occurs in A2 on fol. 109 a in our Tale III, viii, verses 161 ff. As this passage is of con- siderable critical value, I print it here in four columns. The first column, agreeing with our text, gives the readings of our MSS. bh*, and of Kosegarten's MSS. BCDEFK, which contain this story. Prof. Macdonell most kindly collated for me the following passage with the originals of
Chap. III. Account of the Manuscripts 45
BC;1 and Mr. F. W. Thomas in like manner obliged me by collating the originals of DEF. 2 To Mr. Thomas I owe the confirmation of Kosegarten's supposition, that his MSS. F and L are one and the same MS. 3 The variants I give from K are based on Benfey's collation of this MS., which I owe to the kindness of Miss Emma Benfey, and on Prof. Schmidt's collation of the same. The second column contains the text of A, the third one Kosegarten's text according to his edition of the textus simplicior, p. 180. The fourth column contains an interpolation of the MS. E, which interpola- tion is separated from the other texts by a vertical line. In the footnotes under the first column I give the complete variants (but not all the merely clerical errors) of BCDEFK; in those of the second column, Prof. Schmidt's deviations from A, whose version is given in his German translation ; in those of the third column the complete variants of the editions of Jiva- nanda Vidyasagara (Calcutta, 1892), and of K. P. Parab (Bombay, 1896), who follow Kosegarten's text ; I add Benfey's translation and emendation of Kosegarten's stanza 187. In the footnotes to the fourth column I give the references from 0. v. Bohtlingk's ' Indische Spriiche ' to these interpolated stanzas. It will be seen that most of them occur in the Vikramacarita.
Though Kosegarten prints the story from which the following passage is taken in his edition of the textus simplicior, it does not belong to this recension. It is missing in the Hamburg MSS. HI, in h, in BUhler's edition,4 and in Kosegarten's MS. G, i. e. in Anantabhatta's Kathamrtanidhi, which is an abbreviation of the textus simplicior.8 Besides HI and G, Kosegarten used the MSS. ABCDEFK, and these only. Of these MSS., A B contain Purnabhadra's text ; all the other MSS. belong to the mixed class. In the third book, C and F (which go back to a common source for the last three tantras) and K contain contaminations of Purnabhadra's text with the textus simplicior. D in this tantra contains a textus simplicior interpolated from Purnabhadra's recension. E contains quite a new recension, based on the Jaina recensions and on other sources. It has many interpolated stan/as, and one interpolated story whose wording goes back to Sar. or to some nearly related recension,6 and the order of the stories in E disagrees with that of all the other recensions.7
1 These MSS. are now in the Bodleian ; B •= Aufrecht, No. 337 (written after A.D. 1810j, C «= No. 336 (written A.D. 1800).
1 These MSS. belong to the India Office Library ; D = I. 0. 2790 (E. 4085), E = 1. 0. 1812 (E. 4086), F = I. 0. 2319 (E. 4087).
' Cp. Kosegarten's Praefatio, pp. iv and vi. * As to this edition see above, p. 14.
' See my papers ' Kritische Bemerkungen zu Kosegartens Pancatantra ', ZDMG. Ivi, p. 296 ff., and ' Uber die Jaina-Rezensionen des Pancatantra', Berichte d. kgl. sachs. Ges. d. Wigsenschaften, phil.-hint. Cl., 1902, p. 117 note. Above, p. 18.
• ZDMG. Ivi, p. 317. ' ZDMG. Ivi, p. 326. Above, p. 17.
46
Chap. III. Account of the Manuscripts
I
E
'jfilXdoo &\ uofllppB '
<r
fl-
its 9^< - —
I*
fe"
Chap. III. Account of the Manuscripts
47
o
%
'ft .
V
Is tp, t
If
•f*
oo o
£3 _J- C3
Sr
HA" _ <r
V " *^
£
ft
r
s
*
•-
h
<E *•?
<fe IT
& ft
E J
*
S J
M bo
O • —
s
« §
IB •-
1 a
g
§
r c
5
CO
eo
48
Chap. III. Account of the Manuscripts
®
tfl! tr -
M 3 ^
O • r-i T3
-3 2 oo
Si!
,p 3,) T3
1 1'
CD O
K<E
RT
'IF Iff
to'
,-lap
£ ~ =
r?| JE
r .
l
ID
a
.s •«
^ §
CO 00
0)
._ -a l-l^i
Ifgl
d P
O5
<£>
9 • V .• ** P3 J3 3 ^ , if
* spill rill
£
|p*
C 03 2 J3
fill
rt-i3 a
Chap. III. Account of the Manuscripts
49
•30
i BT II
I i!
s=- &
e
IT
3f X
o-
£
r
e •p"
IB/
s
CM
.
Schmidt (readin 'die Frauen, die nicht opfern ' II
a
•»-
s m
o
50
Chap. III. Account of the Manuscripts
.2 «<K k
-w ~- D «•
1 j ° -r
•+- I-H fc
S
f fc
W |C
V P-
r> !P
wr
•3 «•
2 c
II
1 If
•
00 O O O
M
g^S S,g<- s a
2 o
«w
9
•V
r
- <r
e v
VUl
, ^ .«
s
CO
Chap. III. Account of the Manuscripts 51
From these parallel texts it is clear that both Kosegarten and Schmidt based their texts of this story mainly on A2. But in doing so they were not consistent. Kosegarten omits A2's stanzas 69 and 70, and both scholars omit A2's (first) stanza 65. It will be seen that no other MS. than A" has A2's stanzas 65, 69, 70, 71, 72. In place of 71 and 72, all the other MSS. but K have two prose sentences. In K the first prose sentence is missing. All the MSS. agree completely in the number and in the order of the stanzas, and nearly completely in their wording. Only the contaminated MS. E follows the wording of the MBh. in our stanzas 161 and 162, interpolates as its stanza 5 (i. e. 205) a whole stanza from this text (= MBh. xii. 148, 12), and inserts seven more stanzas, four of which occur also in the Vikramacarita, and one of which (6, i. e. 206) is apparently an imitation of MBh. xii. 149, 13. But even this contaminated MS. has none of the additional stanzas of A2, and has the same two prose sentences as all the other MSS., though in E the wording of these sentences is slightly altered.
Hence it is evident that Kosegarten's text does not agree with any MS. of the Pancatantra, and that Vidyasagara and Parab, who exactly agree in the choice and in the order of the stanzas with Kosegarten's text, simply reprint it with but slight alterations.
Doubtless neither Kosegarten nor Schmidt would have based their texts of our story on A2 if they had been aware of the true nature of A2's wording in our passage.
First of all, it is clear that in our passage the MS. from which the scribe A* copied, contained a wording which agreed with that of our other MSS. as given in our first column. For the scribe first copies his stanzas 64 to 67 inclusive (corresponding to our stanzas 1C1 to 164), and, after doing so, corrects the fourth pada of 64, adds in the margin his stanza 65, and corrects the numbering of the already written stanzas 65, 66, and 67. Some of the copyist's blunders, as '411411 for '41|41(c^ in 69 o, *JT?T0 for ^rflT0 in 70 a, «fi«?ld^q: in 71 a, and the wording of the first pada of 73, prove that he did not himself alter the text which he was copying, but that he copied marginal corrections and additions of his original. This view is confirmed by the fact that in 73 b he writes t^TTTT^^T: within the line. The sign (^ is a hyphen, which in Nagarl MSS. frequently occurs at the end of the lines, and which A2 copied without reflecting. More- over, we shall see that A2 's stanzas 69 and 70 are inserted in a wrong place.
The author of A2's spurious text was shocked by the purport of the genuine one. Whereas the interpolator of E inserts a number of stanzas intended to prove that widows must burn themselves, the interpolator of A2 is an adversary of men's and women's burning themselves alive. Hence he cor- rects in 64 d the true fire of the text to a metaphorical ' fire of penance V
1 Schmidt's ' Opferfener ' is an impossible rendering.
52 Chap. III. Account of the Manuscripts
and substitutes in 73 a mortification for Purnabhadra's forest-conflagration. As he does not think his correction of 64 d to be sufficiently clear, he adds the stanza 65. He strongly opposes the custom of widows' burning themselves with the bodies of their deceased husbands. Therefore in 65 c he points out the •ill^W'S, which, he says, is not so cruel as the prevailing custom, but which, on the contrary, is ^A|H4|. The conse- quence of the female dove's penance is that she beholds her husband in the f^WM. The author of the alteration no doubt takes this word in the sense which it has in Jain mythology ('the highest heaven'), whereas in the genuine text it has the brahmanical meaning (' heavenly car ').
After stanza 164 of our text, A2 adds his stanzas 69 and 70. No doubt the copyist A2 inserted them in a wrong place; for the interpolator himself must have intended their insertion immediately after stanza 65, as they are destined to corroborate his view that a sail burning herself commits a sin. Schmidt has misunderstood these verses. The correct translation of 69 and 70 is : ' She who, being still alive, follows her beloved one by offering her own body in the fire, must no doubt go to a terrible hell. In the law books, in the Vedas, and in other sastras, the following correct rule has been handed down : " Those (wives) who commit suicide will be unhappy in all their following existences." '
After these stanzas the interpolator replaces the prose lines of the original — which in short words gives the purport of Purnabhadra's source, viz. of the Hahabharata version1 — by his slokas 71 and 72. Stanza 71 has been misunderstood by Kosegarten as well as by Schmidt, who have destroyed its meaning instead of restoring it. Kosegarten's *ji|'T«d- in a, and ^T *sf in c, and Schmidt's ^TT% are nothing but wrong conjectures. Kosegarten's alterations have misled Benfey as well as the two pandits, who reprint Kosegarten, not without continuing his destructive work. There is neither a ' dove god ', nor a ' sun-setting ', nor a ' solar heaven of the cock pigeon ', nor a ' proximity of the sun ', nor a ' sun happiness ' in this passage. All these fine non-Indian things have sprung from the bad Sanskrit of the interpolator, from the conjectures of the editors, and from a clerical error of the copyist A2. For ^ftr!^: is nothing else than a clerical error for ^nftfl^T, an expression formed after the analogy of the very frequent word qffj^r , i- e. 'a wife who regards her husband as a god', 'a faithful wife'. *prfa is quite right (*JT^ 3TRf), and so is
' In the edition of Protap Chundra Roy, the only one which is at my command, the story of which Purnabhadra gives an abbreviation stands at book xii. 143, 10-149, 14 incl. To our first prose sentence corresponds xii. 148, 12, to the second one, xii. 149, 1-7 incl. The stanza 165 corresponds to 149,8.11.12.13. In his note 1109 Benfey says: 'Diese Erzahlung stimmt fast wCrtlich zu Mahabharata, xii, Vers 5462-5592.' The mere number of stanzas of the two versions would show that this statement cannot be correct.
Chap. III. Account of the Manuscripts 53
?!T=fi, which the interpolator construes with the genitive
in our passage must not be derived from ^5[, but from ^^. The correct
translation of stanza 71 therefore is : ' Having regarded the male dove
as her god, she became a goddess, and day by day enjoyed godly (i. e. divine)
happiness with the male dove ; for such is the consequence of religious
merit acquired in a former existence.' Her e
in this world causes her «Tl<<H«^ after her death.
Stanza 72 of A2 is designed to replace the second prose sentence (204, 29) of the genuine text, and in stanza 73 a b, the interpolator alters the wording for the reason given above, p. 51 f. Why he altered also the last line of this stanza I cannot say. But it is certain that A's wording is an alteration; for to ^<T^f^f^ *?f^, as the other MSS. read, corresponds MBh. xii. 149, 13: TfTT: taii^^iwii+iM^^m^^J I «t
I have advisedly treated this passage at full length, because it is in several respects highly instructive. First of all, it shows how texts should not be edited. There was not the slightest reason why Kosegarten and Schmidt should leave out one or several verses of A's text, adopting the rest of it ; for all these verses go back to the tame interpolator. As to Kosegarten, our passage shows what critical principles this editor was wont to follow during his work. Not to speak of the fact that books III and IV of his textus simplicior are only an adulterated edition of Purnahhadra's books III and IV respectively, i. e. of the textus ornatior, he follows in our passage in some places one single MS. (A), though all his other MSS. agree against A, and though the purport of the Mahabharata version agrees with all the other MSS. But instead of, at least, following A throughout, he chooses at random the stanzas which he rejects from his text or takes over into it. And this is not only the case in our passage, but throughout his textus simplicior as well as his textus ornatior. It is not only true that both of them are not worth the paper on which they are printed, but also that during more than sixty years they have misled all the scholars who used them, and have made worthless the work of all the translators of his textus simplicior, to begin with that of so eminent a scholar as Benfey.1 The editions of the two Hindu editors, Jlvananda Vidyasagara and Kashmath Pandurang Parab, are even more worthless than Kosegarten's. The passages in which these editors deviate from Kosegarten must induce their critical readers to think that these pandits based their texts on materials independent of Kosegarten's edition. But the passage just examined shows that they mainly reprinted Kosegarten. For the text given by him does not agree with any MS.
1 Of course, Benfey's introduction to his translation is even now very valuable.
54 Chap. III. Account of the Manuscripts
in the stanzas adopted or rejected, but it completely agrees in this respect with the text of the two Hindu editors. Parab's reading of Kosegarten's stanza 187 c, moreover, is a fair illustration of the way in which he endeavours to correct a meaningless passage. His ' correction ' seems to be based on Vidyasagara's quite impossible explanation.
This much on the untrustworthiness of A2. But the text of A1 is not more trustworthy. At 211, 21, for instance, A1 shows foolish alterations. The point of the story Purn. III. xii (Sar. III. viii, Old Syriac VI. vi, SP. III. viii, Simpl. IV. vi HI = IV. vii Buhler) lies in the circumstance, that the clever wife/«//y reaches her aim, i.e. the cohabitation with her "dMMfir. In the original version of our tale, the adulteress, answering a question of her paramour, tells in a loud voice that all women are unchaste by nature, but that she truly loves her husband exclusively. Thereupon her husband is convinced that he has got the most faithful wife in the world. The author of the textus simplicior evidently thought that no husband would allow himself to be convinced by any such trick. Accordingly he alters the text.1 In his version, the faithless wife tells the adulterer that Candika has pointed out adultery to her as the only means of preventing her husband's death which hangs over him by Fate and which, by sexual union, goes over to him who plays the husband's r61e. The words of the goddess, according to the Hamburg MSS., run
thus: *rf^ ^yviW *ff k<*ifal'V m«i«n*f H*fHunif5R«i =fiTffa cTrT^ HiiJ: f ?) SJM^rtJ^T^I *N<f?T I Wrrf »j«iv«^fanT ^faft I Buhler, p. 19. 12 has
the same wording, except ^J%, H^^HTlM^rejtl,, w-grtTj and
Purnabhadra's text 211, 21 comes very near to the wording of the Hamburg MSS. Cp. also the wording of Bh in our variants. Instead of OTfW*FT> A1 has 3TOtf'rf5f*TWlN, which compound apparently was first intended to mean 'touching [by the limbs] except the male and female organs.' But as the copyist (or some previous glossator) feels that this word is not clear, he makes it an adjective by adding in the margin fsnj^T. Now the passage means ' a cohabitation without touching of the male and female organs ' (Schmidt, p. 232 : ' Wenn du mit einem fremden Manne auf gemeinschaftlichem Lager ruhend den Beischlaf ausfuhrst, ohne dass sich dabei die Geschlechtstheile beriihren '). I am at a loss to say how the interpolator imagined an snftf'Tf^flWlI'i f'J^Wl to be possible. But his alteration, which is proved to be such an one by Purnabhadra's source, the textus simplicior, destroys at the same time the point of the story.
In the same story, the genuine wording of Purnabhadra, as given in our text, p. 212,6, is nearly identical with the wording of the Hamburg
1 Apparently in following some other source, whether literary or oral. Cp. Cbauvin, Bibl. des ouvrages arabes, ix, p. 39, no. 34.
Chap. III. Account of the Manuscripts 55
MSS. (flgifa'K* »rf
Biihler's text ins. ^J before ^R^, oin. Tf$ after 7T*T., and ins. 3TOV after . But A1, in consequence of his first alteration of the text f%°), alters again, continuing after OTfSR : W ^wf+TWT*TT
^ IfrT^fft I ifll*l«l*Jc*l I 4l%S*<ll^if*I<l<»T«l. I ^fi^T rfTRTTfal I f*|VU*J cj ^=(<Tj*j«<I^, &c., 1. 8 (Schmidt, p. 232 : ' "Du bist die Erste unter den Frauen, die ihrem Gatten anhangen, darum dass du selbst bei der Vereinigung mit einem Fremden die Keuschheit so bewahrt hast. Um meine Lebensdauer z\i verlangern und den Tod abzuwenden hast du so gehandeltl " Nach diesen Worten umarmte er sie liebevoll, nahm sie auf die Schulter, tanzte mit ihr herum und sprach dann zu dem Herrn Warst- du-besser,' &c.).
And again the conclusion of our tale is awkwardly amplified in A1, which for «j«i»v to ^fTT (212, ll) reads : «pzf SfWT I
I
| (Schmidt, p. 233 : ' und nachdem er darauf umher getanzt war, sagte er : " Ja, du Vordermann unter denen, die Keuschheit iiben, auch du hast mir einen Dienst geleistet I " und liess ihn von der Schulter nieder. Vor alien seinen Angehorigen pries er dieser Beider Tugenden. Wo er immer an die Hausthiir von Angehorigen u. s.w. kam, da pries er auch deren Tugenden ').3 I need scarcely add, that here too the textus simplicior confirms the wording of our text 212, ll. The Hamburg MSS. read : dfl^^R^*! 2«!-«*l«ri^M*Ji{diTXg WW (!) ; Buhler (19, 0t)
Cp. also A2's interpolation 122, 5, and the transpositions in A1 3, 18 and 201, 18. These cases show that the reviser, or the revisers, did not shrink from even serious alterations of the text which they copied. Moreover, our parallel Specimens show that all the variants of A1 A2 can be proved, by the testimony of the sources, i. e. the textus simplicior and the Tantrakhyayika, to be alterations. Wherever A has the evidently right reading, it is not to be made out with certainty whether this correctness is due to conjectural criticism, to collating some other MS., to A's going back to some MS. older than bh, or even to mere chance. Though I very carefully collated the whole MS., I am not able to decide this question. The only thing quite sure is that A is the copy
1 Read $Tl». The same mistake in A1, p. 211,22 of our text. This shows that the alteration does not go back to the copyist himself, who did not understand the text which he was copying here. 2 Bead °*<(<Ji«u»ii .
' Schmidt's second MS. K has a gap, by which the whole story has been lost.
56 Chap. III. Account of the Manuscripts
of some revised and adulterated MS. For the constitution of my text A was almost useless. Good readings of A, not confirmed by bh*, have only the value of conjectures, or of various readings the sources of which we do not know.
In order to allow the reader to form a judgement of his own, I give the variants of A from the beginning of the work to 12, 13 inclusive, from 126, i to 134, 23 inclusive, and for the prasasti. Besides, I have entered the readings of A into my variants at all the places marked in the Sanskrit text with an asterisk, and occasionally in some other passages.
In our parallel Specimens I to IV all the readings and all the more important blunders of A are given in the notes. The reader will see that not even one reading more original than those of bh* is to be found in these parts of the MS. A.
§ 6. Critical discussion of the manuscripts Bh and $.
The MS. Bh at first puzzled me very much, and it cost me consi- derable time before I was aware of its true nature. Its age of course prepossessed me in its favour, and this impression was strengthened when I collated the fifth book, which more closely agrees in Bh with the Hamburg MSS. (textus simplicior) than any one of my other MSS. On the other hand, Bh deviates considerably from bh*A in the rest of the work ; transpositions of words are very numerous ; synonyms appear in very many cases for the words used in bh^A. Again this MS. bristles with blunders of every kind. But Bh has exactly the same stories, and these stories in exactly the same order, as bh^A. It was not until I got the MS. 4> and the Sarada MS. P, that I found out the worth, or rather the lack of worth, of Bh.
Bh and <f> belong to the class of the mixed MSS. The greater part of their first three books has been copied from a fragmentary Purnabhadra MS., containing the text from 6, 2 dhdramdtrdrthi down to the end of book III. The kathamukha and the beginning of book I contains in Bh the text of Purnabhadra from 1,1* na vidvdn inclusive to p. 3, 25. The text between dhurvodhdrau and dhdramdtrdrthi 6, 2 has been supplied from some MS. of the textus simplicior. In <t>, the text to dhdramdtrdrthi has even been twice supplied from MSS. of this recension.
The text of Bh agrees very closely with :that of * in its readings and in nearly all of its blunders, down to the end of book III. From the beginning of book IV to the end of the work, the two MSS. disagree in a most remarkable manner. Whereas, in book IV, Bh contains a faulty text of the bh-class, <t> in this book contains an equally or even more faulty text of the *-class. In book V, both these MSS. contain a textus simplicior. But here again the difference is evident. For Bh contains
Chap. III. Account of the Manuscripts 57
a very valuable old specimen of the H-class of this text, agreeing in many blunders, but not in the interpolations, with the Hamburg MSS. The MS. <J>, on the contrary, contains a text of the <r-class of the textus simplicior.
Discussion illustrated by text of Tale V, v, Ass as singer.
In order to prove what has just been said, I beg to refer the reader to the following specimen, Tale V, v, Ass as singer. In this specimen I give the textus simplicior according to the Hamburg MSS. H I. The notes contain the complete variants of the following texts : —
Textus simplicior, H-class : 1 | the Hamburg MSS.
Textus simplicior, <r-class :
o- = Decc. Coll., Peterson's Fifth Report, No. 356.
s = Decc. Coll. i. 17.
B = Biihler's edition.
pr = the MS. of the Ahmedabad Bhandar, lent to me through
Mr. Premchand. h = a recent copy of the MS. Bhandarkar, Report Bombay 1907,
p. 55, § 46.
Purnabhadra's recension :
bh*A, the MSS. just mentioned. Mixed recensions :
t the MSS. just mentioned.
n1 = Decc. Coll., Bhandarkar, Report 1894, No. 371.
ns = Decc. Coll., Peterson, Report V, No. 355.
n3 = Decc. Coll., Bhandarkar, Report 1897, No. 418. The variants of the o--class MSS. of the textus simplicior, and those of 4> are given on the left-hand pages, the variants of all the other MSS. on the right-hand pages. It will be seen at once, that all the MSS. whose variants are given on the left-hand pages form one group, and that those whose variants are given on the right-hand pages form a second group. Nobody who compares the various readings will doubt that the text represented by H I is on the whole older than that represented by the
<7-claB8.
The cases in which Bh agrees with HI against bh* are set in italics in the text ; the cases in which bh* agree with the cr-class against HIBh are set in fat Italics in the variants.
I
58 Chap. III. Account of the Manuscripts
Text of Hamburg MSS. HI corresponding to our text 270,17 to 272,21.
270, 17 ' sadhu, matula, gltena ! ' varito na maya sthitah.
18 'apurvo 'yam manir baddhah: sampraptam gitalaksanam." 19 cakradhara aha : ' katham etat ? ' BO 'bravit :
21 asti kasmimscid adhisthana Uddhato nama gardabhah. sa ca diva 22 rajakagrhe bharodvahanam krtva ratrau svecchaya paryatati.
Variants of ho-sprB*.
270, 17 $ </J, om. Una II hers prB maya proktdpi na sthitah, $ maydty ukto na sthitaih II 18 pr bamdhah II 4> samprdpta II 19 a- guvarnnasiddhir
abrawt II 21 prB om. asti II <£ uddhamo; <r uddhatandma\ gardabhah; <J> ra-
sabhah for gardabhah II After gardabhah hcr<J>prB ins. prativasati sma II hprB om. ca II a- om. diva ; s daiva, h$B sadaiva, pr samdaiva for diva II 22 ho-s<E>prB karma for bharodvahanam II After paryatati pr ins. tata, B tataA ; then ho-s*prB ins. pra- tyuse (4> pratyusam, ho-$ add. vamdhanabhayat, pr barndhanabhayat, B bandhanabha- yat) svayam eva, then o- bamdhanasthane samasrayati, hs grham ya, h adds <t, $ ro/a- kagrhe ydti, prB rajakagrham dyati; then ho-s*prB rajakopi tarn (hprB <a<a« <a7W, s ia(as feawi na for <am) bamdhane, <r niyukte, s yunaktih, <I> «a yurnkte, h na yukti, pr °na niyuktih, B °na niyunakti II 271, 1 h<rs$prB at/ia for athdnyada II
o- om. tasya; B tasmin II hs$ om. ro<ra« II o-$prB om. ksetresu; hs ksetrani II After paryatatah $ ins. ksetre, pr ksetrani satrau (read ratrau), B ksetrdni II ho-s om. kaddcic II ho-s<£prB saA« for sdrdham II ho-s$prB samjdtd for babhuva II 2 o- sa uddhato, hs4> sa ca plvaro, prB sa ca pwaratvat (B "toaa") for iaw ca II ho-pr wMt° ; 4> vaditamgam II 4> om. karkatikdksetresu praviiya', ho-s prB karkatikaksetre (pr °ksotra for °ksetre ; prB add. srgdlasahitaK) praviiati ; then $ ins. karkkatika- bfiaksyanam karoti | pratyuse sthagrham thati I <a<A<Z srgalai ca ; <r ins. : iasya c« prstato lagnah I trgalah praviiati ; hs ins. : <aiAa trgalah ; h adds ca ; then ho-s ecant (s etadaiai for «raw» ; h adds tow) drov opt ratrau (hs yathecchayd for rdirow) karka- tikdbhaksanam krtvd pra° svasthanam vrajatah; prB ins.: ewam <aw yadrcchaya virbhatikabhaksanam (B ci° for vi°) Artea pratyaham pratyuse svasthanam vrajatah ; $ om. this sentence II 3 $ a<Aa kadacin madoddhetena rdsdbltam tena ksetara- ma4idhyasthitena irgdlam abhihitam; hs a^a kadacit tena (h adds safia) madoddha- tena (h madoddhata) r&sabhendbhihitam I s adds ksetramadhye, h adds ksetramadhye after bhagnisuta (sic!); prB ai/ta kadacit tena madoddhatena rosobhena ksetrama- dhyasthitena 6rgalo 'bhihitah ; o- a<Aa kadacit tendbhihitam H S fcpa&ya 2, B paiya patya II <E> pafyat II atwatirmmald rajanl, s pasyeyam nirmmala rajani, h paiya- yam nirmmalarajam, cr paiyaitam nirmaldm rajanlm II Before <a<, s ins. sa aAa II 6 4> karisydmiti II ho-s*prB ins. kathaya before katamena \\ s tamenana, * Aeraa for iaiemena II o- ins. gitam before karomi II hs karomiti II <r srgdla for sa II o- ins. &/w after a/ta II cr mamaka, s tana for mama II h aZam for mama II s £i for
; ho- om. Aim II
Chap. III. Account of the Manuscripts 59
271, 1 athanyada tasya ratrau ksetresu paryatatah kadacie chrgalena sardham maitrl babbuva. 2 tau ca vrtibhangam krtva karkatikaksetresu pravisya tatpbalabhaksanam 3 svecchaya krtva pratyuse ya^asthanam vrajatah. atha kadacit ksetrama4dhyasthitena tena coddhatarasabhena srgalo 'bhihitah : ' bho bhaginlsuta, 5 pasya ! atlvanirmala raj am. tad aham gitam karisyami. tat 6 katamena ragena karomi 1 ' sa aha : ' mama, kim
Variants of HIBh, 270, 18 A laddha \\
22 Il'iril' rajakasya grhe II
bh bharodvalianam II H radrau II
271, 1 ffn'H' tathdnyada II
2 HIbh* vrttibhamgam, A vrtti\bhamgam, II'II'II* varttibhamgam II H karka [new line] kaksetrem, I karkketikaf II A tatphalaip bhaksanam II
3 bh*rPll1II'A svasthiaam II IIln3 ksetramadhye sthitena II
4 bh^ffn'H'A om. tena II bh* madoddbatara&abhena, A madoddhataruhliM-
bhena, WlP mahoddhatardsabhena, II5 mahodatarasabhena II A bho bhaginiauta I pafydtlvamnirmalarajani II
6 HI karisyami II HI kathamena II ^H'n'H'A praha II
60 Chap. III. Account of the Manuscripts
anenanarthapraca 7 lanena ? yatas cauryakarmapravrtta vayam. cauraja- rair nibhrtair eva 8 sthatavyam', iti. uktam ca :
9 kasl vivarjayec cauryam, nidralus carmacaurikam, 10 jihvalaulyam ca rogadhyo, jivitum yo 'tra vanchati. 11 tatha ' tvadiyagitam sankha&z&fanuvadi, na madhuram ', iti durad
Variants of ho-sprB <£.
B nendndrthapraldpitena, $ anendnarthe jrraldpena, or anendrthacdlanelam, h vrtha- pralipitena ; prB anena, then pr vrthdrthapravalapralapitena, B vrthdrthapracdla- nena \\ 7 4> caurakarmmapravrttd, B caurakarmmaprakrtd, a- cauryakarmapra-
vrttair, prB caurafcarmapravrttdv II prB avam for vayam; or om. vayam I! <1> ins. tan, hs tarn after vayam II ho-s$prB transp. : ni° (h nibhrtam, pr nivrttaU; prB add ca) cau° ; <b caurai, h caurair jdraih II prB atra for «w II h<rs$ om. era II 8 ho-s<£ stheyam II prB om. iti II h<r ins. nayah, s waya after »<t II pr om. uktam ca II hs ins. yataA after uktam ca II Do- kdsam, e prakaiam, * hdsyam, pr fe^i, B &d/t || s varjjayaf II <E> caurya, a-s cauro II First p&da in h : cauranam varjjayet kasam II $ nidralubdhaA II s carwwnacoraiaA, B so ca cawrikam II 1O 4> jihva- lolyam II tr rogarto ; pr rujakramto, B rujakranto for ca rogadhyo II cr4>prBi/Mtt<am, BJivamtam II 11 ho-sprB aparam, $ param for <a<Aa II <rs*prB tvadlyam;
4> om. gitam II prB om. iankhaiabdanuvadi ; 4> iamkhaidbdanukafi, s iamkhasabda- nukaram, h iamsa&avddnukaram; <r kafhoram for fankhaiabdanuvadi II ho-s om. na madhuram; <E>prB na madhurasvaram ; prB add samkkasdbdanukaram II h<rs4>prB om. t<t |) <!> ins. ca after opt II 12 <rs*hprB iruyate for drutvotthdya II s^prB ins. <arf a<ra, ho- <a£ra before ksetra0 ; then o^ ksetrarakfakah purusah prasuptas ttstati, h s ksetre raksapurusah (s raksd0) suptas ti °, prB ksetre raksdpurusdh suptd I sam<t (B suptah santi), 4> ksetrapdldh purusa prasuptas tistamti ; then o- «a, prB <a, $ te ca; then ho-s$ samutthdya, prB utthdya; then o-s bamdhanam, h vadhavamdhanam, * bamdham \ badhatn va, s vamda vd, prB vadham bam- dham vd; then ho-s<J> vidhdsyati, prB karisyamti II 13 o- <awi for tdvan II
(rs amrtakalpd, h amrtakdlpdi, $pr amrtamayd6, B amrtamayii; then o- karkati, h cirbhidydh, s cirbhadydh, $ cirbhi{ikdh, pr cirbhatya, B cirbhatih II ho-s$prB om. nibhrtah ; then er avydpdro bhava, h ma avydpdraparo bhava, s mdvydpdro bhava, prB wia <vam avydpdraparo bhara, $ vydpdraparo bhavdn II <r om. <ac chrutvd II o- gardabhah, ® rasabha, hs «a II hs$prB d^a II s aAo for 6Ao ; then h wa, o-4>prB na traTra, s tozm na; then ho-s^B wist, pr c«tet'; then ho-s $ vandsrayatvdd ; then 14 s$prhB gitarasam, <r gitasukham; then B vand'srayatvdt, pr vind&rayatvdt II o- om. <e° 6Aa° M° ca II s'I'prB tenaitad, h tenaivam, s$prBh bravlgi \ tdctam; then hsprB ca, $ caA; then s* yatoA II 15 o- om. this and the following line II
pr iaratyotsndhate II pr dwra, <E> rfwre II 16 hs$prB jdyate for v&afo* II
hs«I>pr karnne, B karne II prB gitajhamkdrajd, h gitajhamkdrayd, s gitasamskarajd, * gltddhyamkdrajd II
Chap. III. Account of the Manuscripts 61
api 12 srutvotthaya ksetraraksa ' bandha ! band A ! 'awam vidhdsyanti. tad bhaksaya 13 tavan nibhrtah ! ' tac chrutva rasabhah praha : ' bhoh 1 vana- sr&y&tvdd 14 ^itarasam na vetsi ; tenaitad bhanasi. uktam ca :
15 ^arajjyotsnahate duram tamasi, priyasamnidb.au,
16 dhanyanam visati srotre gitasamskaraja sudha.'
Variants of HIBh, bh*nlnJnsA.
Bh "pracdlenena ; II1 anendrthapralapane, corr. to anena vyarthcf, which is the reading of II* ; II* anena vyarthapralapitena II
7 HI (not Bh) °pravrtya, IIs °pravrtto, bh °pravrddha II II1 II2 IIs caurajatair II HIbh nirbhrtair II
9 HI* A kaM II Bh cauram for cauryam II II1 II5n* "corikam II
10 A rogadhye II VWU''Il3jivitam \\
11 bh* tada II bh tvadlyagatam, II1 11* II3 tvadiyam gUam II A iamkhavada- nuvadi, bh II1 II' II3 iamkhanadanwvadi, * iamkhananudanadi, corr. to iamkhana- nunadadi \\ HIBh II1 IIs IIs nam for na II A ayi for a/« II
12 A ksetrajm/rasa, n1!!2!!* ksetraraksakah purusa, Bhbh>J> ksetraraksapurusa II bh*A bamdham vadham ca vi°, Il'II'n3 vadham vamdham ca vi° II
13 bh*A aha II
14 Bh gltam ratam II bh*A ins. tvam, II1!!3 ta tvam, TP tat tvam before glla° II bhn'II'II' ins. na between tvam and gita°, om. na before vetsi II II1!!2!!8 jdnasi for vetsi II Bh na »«<fjru II bh II1 11*11' vravlsi, * bramei, A 6awm II 15 HI
tjyotsnahate, Bh (aratyotsndhate, * iaratjjotmahate, II1 £sa[corr. from deletedjj'yofyyo corr. from some other aksara]feraoAa<e, II3 ksarajotsnahate, II2 drarajnayotiahate II A j9Mram, n* «?wr« II II1 IIJ jyriyarnsannidhau II ISbh^rofe, corr. from /ro<r« II bh A II1 II* II8 gitajJutmAdraja; * gitajhamkaraja, jharp, being very similar to Sam ; hence P gitaiamkaraja II
62 Chap. III. Account of the Manuscripts
17 srgala aha : ' mama, asty etat. param kathoram unnadasi. tat 18 kim tena svartbabhramsina 1 ' rasabha aha : ' dhig murkha ! kim 19 aham gitam na janami ? tac chrfiyatam, tasya bhedah : tad yatha :
20 sapta svaras, trayo gramah, murchanas tv ekavimsatih,
21 tanas tv ekonapancasat, tisro matra, layas trayah II
22 sthanatrayam yatinam ca, sad bhedas ca, rasa nava,
23 varnah sat, trimtat bhasas, catvarimsat tatah smrtah II
Variants of hers prB <£.
17 h mana, a mamaka II <$ asyaitat II a- param na vetsi tvam kevalam anudiSate kim Una, &c. ; s param na vetsi gitam I tac chrdyatam, &c. ; $ param gltah kalam annatasi ta kim, &c. ; h param na vetsi glta tvam kevalam unnadasi I tat kim, &c. ; prB param na vetsi tvam gitam I kevalam unnadasi II tat kim, &c. II
18 cr«l>h svarthabhramsena, prB svarthabhramiakena; a adds kim II h sdvravlt II er gardabha ; <£ rdsabhah II $ om. a/to II B dhig twice II Qjanann 19 hcr^prB om. gitam after aham, inserting it after janami II * tarhi for tac II h bhedah II h<rs om. tad yatha; prB tad yatha tasya bheddn 6rnu (B °n chr°) \\ 20 o-prB murchanat caika0 II h "vimiati II 21 trs tana ekona°, h tonany ekonapamcasa, pr tana tv ekona° II o- tisras tola for tisro matra, II o- layas II In $ the fourth pada runs thus : ity eta irutimamdalamm ; in prB ity etat svaramamdalam II 22 cr yajinam II $ (transp.) ca jdtinam II s om. ca II Second pada in <r : sat kdvydni rasai ca sat, pr sadgasya, then one aksara left free, then «a rasa niva; B sad asydni rasa nava; hs satsasydni (s ins. ca) rasani ca; O satsvaida rasa navd II 23 s varnna, $ varsd, B (not pr) rdgdh for rarwaA II ho-s<fcprB trlmsatlr II s<l> bhdsd, o- bhdvdh, B (not pr) bhdvds II Fourth pada in o- : satca- tvdrimiatih smrt&li, hspr dvicatvarimiati (pr adds A) smrtah (h om. A), B catvdrimiat tatah smrtah, * dvicatvaraviiatis tathd I ?na<ra II 272, 1 h pamcdsihyadhike ; O pamcasityadhikam II a caitad II Second pada in ho-prB: gitamgdnam satam smrtam, B gitdmgam satatam smrtam, $ gitam II naganam 6atam II (om. smrtam) II After line 1 o-$prB insert a half sloka; first pada crh$prB: svayam eva jmra, proktam (h iastre for proktam) ; second pada : o- svayam eva sruteh priyam, * Bharatena &rutam iriyam, prB Bharatena sruteh (pr few0 for £rwc) param, h vedena ca iruteh param II 2 sB om. this line II h yKamg^A safia samvrtam II cr$pr vrtam \\ 2 a B om. this line II ho-s$pr kartie II h saradi II 3 <J> nanya- hdtapriyam lake II s param for priyam II o- sasyate, h durlabham, for drsyate II
4 o- suskasnayurasasvadas, hsB fuskasnayusvarahladdt (h sw°, and °wica0 for °«va°), pr iuskasnayuh svarahladat, $ suskasnacchuradblwdat II Fourth pada in o- : tyakta dksena Ra°, pr tyaktas Tryaksena Ra°, B Tryaksam jagraha Ba°, s ya(or yu)ktas Tryaksena Ra°, h paktas Tryaksena Ravanah, * paktas Tyaksena Rdmanah \\
5 "3? tvam; ho-s om. <mm; prB bhaginisuta for iram II prB vadan for vadasi II <E> manasamti for «ac nt° II cr^prB om. ca II 6 o- mamaka, O mam II <l> mady for i/arfy II <r om. <ad aham II s4>prB ins. tavad after aham II h dvaradeiasthah, tr vrttidvirasthitah, B vrttedvaridesasthah, pr miter, B wter, prB dvarasthitah, $ vddidvasthitah II her ksetram, prB ksetrapam for fee<
Chap. III. Account of the Manuscripts 63
272, 1 pancasityadhikam hy etad gltanam ca satam smrtam, 2 suvarnaracitam suddham gitangaih sakalair yutam II
2 a dhanydndmjdyate karnaih viSesdc charadi zthite II
3 nanyad gitat priyam loke devanam api drsyate ;
4 Buskasn&yuravdhldddt Tryaksamjagdda Rdvanah II
5 tat katham tvam mam anabhijnam vadasi, nivarayasi ca 1 ' si-gala 6 aha : ' mama, yady evam, tad aham vrtti(fflrasthah ksetrapalam 7 avalokayami ;
Variants of HIBh, 17 n'rPH' tat for asty etat II n'n'IT1 kathorasvaram nadasi II
18 n1!!2!!* ten&rihablvramiina II bhA dhig 2, * dhig dhig II
20 HIbh*A "vimsati II
21 bh*A tan&ekoaa0, TL1U" tamos cekona0, W tana$ caikona? ; Bh tv enako0 for to ekona0 II
22 bh^An1 sad dsyani for sad bhedas ca (op. the reading of s), corr. in n1 to sadjasyana; II2 sad jasya ca; H3 sad gasyana II HI bhedd II Bh «arf akard rasa na I om. va II
23 ^ffn3 trlmsatlr, bhAH2 vimfatir II bh Iharyos, A 6/*aw/ II H'H'n3 (om. tatah) smrta, H1 mtdhaih, njns budhai II
272, 1 HIbh °tyadhikam, in bh corr. by cop. to our reading; A "lyadhekam II HI spatpat for % eiod II
2 bh'HI1!!5!!3 vftain, A drtam for yutam II
2 a bh^An'Ifn3 om. this line I) Bh karnne II 3 bh giiatvaram or gtta-
dva/ram for gltdt priyam, corr. by cop. to gitakaram; & gitadvdram, An1!!1!!3 gltad varam II 4 bh*!!1!!1!!3 °snayuravair Ham, A. °smaywrawaeiaiam II Bh °ra«a- Ihddat II Fourth pada in bh^AlI'II2!!3: raramje Ravanah pwra (II1 II3 pwrah) II
5 Bh om. iwwi II Bh nivarayisi II
6 bh^An'n'n3 vrtlidvlradefast/iah, Bh vrtipwradeiasthah (see the corrupt reading of HI in the text) II
64 Chap, III. Account of the Manuscripts
tvam punah, svecchaya gitam kuru ! ' 8 tatha eonusthite gardabha utkan- dharo bhutva sabdayitum arabdhah. tatah. 9 ksetoa.rafoafcd rasabhasabdam $rutvd krodhad dautan dantaih pllOdayanio lagudaru uddisya dhdvildh. sametya ca tavat tdllditah, yavad bhumiprsthe patitah. tatas ca sacchi- faolukhalam gale 12 baddhva ksetrapalah. suptah. rasabho 'pi Ja&'svabha- vagatavedanah 13 ksanenabhyutthitah. uktam ca :
Variants of ho-sprBS".
1 h om. svecchaya II 8 ho-s#prB tathanustlte (B °s(hi°) II After tothdnustite a- todyltam akarnya I talo lakutam utpddya pradhdvitah (1. 10), s rdsabharafitam akarnya ksetrapah krodha damtdn lagudam udyamya pradhavltah (1. 10), h utkamdharam krtva rimkatwrn dravdhah I iato rdsabharimkitam samakarnnya ksetrapala krodhdnudattanusaya lakutam udgamya pradhavltal) • $ uktam- dhararikimtum arabdhah II tato rdsabhamnkimtu samakarnnya ksetrapalah krodhd dumtd£ carvvayan I lagudahastah pradhavltah (1. 10); prB rdsabharatanam akarnya ksetrapah (pr adds l) krodhad damtdn gharsayan pradhavltah (pr °to, om. |; 11. 10, 11); ydvod rdsobho dr«(as (pr hrstah l) tdval (pr tdvat) lakutapra- hdrais tathd hato yathd pratddito bhiiprsthe (pr °s(e) patitah (pr om. A; 1. 11) II
10 h samastakena for sametya II s$ om. ca after sametya II 4>h pratadlto \\
11 (r bhumau, $ bhamau, h bhuprste, s bhuprstJu) II 4> patatitih n prB tatai ca sacchidrolukhale baddvd (pr badhva) gato murso (B om. mursd) bhuyo (pr adds l) pi (12) prasuptah; a- tato grivdydm udusalam baddhvd bhuyo 'pi (12) prasuptah; B tatah swxhidrodukhala I vaddho gatamarso bhuyopi (12) suptah; h tatah cchidro- dmalatn vavdhd ksetrikah prasuptah ; <fc tata£ ca tacchirodhdtudusalam gale badhvd (12) ksetrikah prasuptah II 12 $ om. 'pi after rasabho n s svajd- tiavabhdvdn gatavedandt ; a- svajdtisvabhdvdt ksanenotthitah, h svajdtiprabhdvdd gatavedana ksa°, prB svajdtisvabhdvdd (pr °prabhd° for °svabhd°) gatavedanah (pr 0<a»a° for °tave°) ksa°; $ jdtisvabhdvdd gatavedandm ksanena utthitah ll 13 o-s om. uktam ca II 14 $ sdrameyasvards'vdndm, <rsB sdrameyasya cdsvasya, pr sara- inayasya vd$vasya, li sdrameyasya ddsasya II smiesyatah II B pr rdsabhasya vUesatah (pr °naA) n 15 h ^>a<i, s parajo \\ lis °janita II 16 a- tato^ ca rasabho 'pi tad evodusalam adaya vrttlm curnayltva pa° d° ; s tatodev&dusalam, $ tatahi ca deva udukhalamm, then s$ with o- (only s vrtim); prB totos tern evolnkhalam (pr °fa0 for °kha°) adaya vrtinj (pr vrttini) curnayltva (pr °rnna°) paldyitum a° ; h to<M ca II terf evodusalam adaya vrttlm bhurnayltva pa° dravdhah II 17 ho-sO etasmina a°, prB atrdmtare II ho-sprB ins. ';;t after 6rgdlo II hcrs$prB durad eva (pr adds »n) tern dr$tva (s drstam) sasmltam (h savismitam) (18) aAa I (19) siidhu mat u la, spr giten(21)eti, o-4>B gitena muyd prokto ($ j/w/fcto for prokto) 'pi na sthltah, h ^ftena nivarito aa tnaya sthitah, om. the second part of the sloka ; 2O cr*B apurvo 'yam manlr baddhah \ (B om. I) sampraptam gitalaksanam (B °nam) n
Chap. III. Account of the Manuscripts 65
14 sarameyakhar«&wz«»z, gardabhasya visesatah,
15 muhurtat parato na syat praharajanita vyatha.
16 tatas ca vrtim bhahktvd kantkastham ulukhalam dddya palayitum 17 arab- dhah. agminn antare srgalo durdt tarn avaloky19edam uvdca: 19 ' sddhu, matula, gU'21 eti.
Variants of HIBh,
8 bh * II1 II2 IIs tathanustlte, A tatha (corr. by cop. from tethd) anustite II A bhuyd \\ II1!!'!!3 sabdayitum II 9 bh*An1IIan3 ksetrapala II Bh tatoh ksetraraksakaras tat idbdam II bh*A ralabhasdbditam II bh*!!1!!2!!3 sama- karnnya, A trutva satnakarnya \\ bh* damtair || *A nipldayamto, bh nihpi- dayamto, n1!!2!!' nispidayamto II
1O II2 lakutam II bh^Bh udyamya, A udyasya, n1!!2!!3 udgrhya for uddiiya \\ 3 pradhavltah, IT1 pradhovita II bh*An1n2ns pratadlto II 11 A
yava ruviprste ya (ya del. again) II II1 II2 bhuprs(e, corr. in n1 to bhupraste, which is the reading of II3 II HI sacchidrolusalam; bh gacchidrodusalam; * sachidrodusalam ; A sacchidraudusalam, corr. to °lam; W II2 II3 sacchidram udukhalam II
12 HI ladhah II HI kselrapaldh II ITU2!!3 prasuptah II bh^n'H3 sva- jatisvabhdvagatavedanah, A svajatisvabha/va/vagatavedanah, II2 svajdtlyagatasvabha- vavedanah II
13 n1!!2!!' ksanendpy utthitah li 14 Bh °khard6vdnam, corr. by cop. from "khakhdtvandm ; bh* A °khardivasya II II1 sdrameyasya c&svasya, corr. from other aksaras, the last of which being svanam; IIs sdrameyasya cA&vasyam, n2 saro- meyasya vdsvasya II 15 bh prajdrajanitd, A prahdrajanitavyethd II 16 HI w«twi II HI ulusalam II bh^An1!!2!!3 «a<a^ ca tem evbdusatam (A em wdw- ealam/m) ; in II1 corr. to evddukhalam (which is the reading of II2 IIs) adaya vrttlip (Wll3 vrtim) curnnayitva pa°d° II 17 bh^Affn^3 etasmlnn \\ Bh durattardt for dwm« II Bh giteneti II bh^AffH^3 <fr° rfurad eva tern (H1n2ns enani for eva etom) drstva sasmltam (A sasmidamtom) (18) zcZam aAa II (19) s&dhu matula gitena varlto na maya sthitah \ (20) apurvo 'yam manir baddhxh (A baddho) sampraptam gitalaksanam n
21 bh*A add t'/i II * adds toAa 6 II
K
66 Chap. III. Account of the Manuscripts
From the specimen just given it appears that in not a few cases Bh comes nearer to the text of the Hamburg MSS. than bh4T!1n2n3A. But in some of these cases Bh and the Hamburg MSS. are decidedly wrong, viz. 271, 13.16.23 (a gross chandobhahga) ; 272, 2 a (an interpolated half sloka) ; 6 (the reading of Bh being a corruption of that of HI, and that of HI being an obvious corruption of that of bh^AlT1!!2!!3). Besides, in 271, 11 BhHI read nam for na, as apparently some previous copyist, who did not understand the wording, thought xabddnwvddinam to be the adjective neuter. But as n1 n2 n3 have the same blunder, this case is not conclusive.
These are serious mistakes which Purnabhadra, who tells us that he has corrected the text aj^rrn^U! (289, 2e), is not likely to have overlooked. It is true that in some cases he has taken over blunders from his sources.1 But these cases are comparatively rare. And not only in the few lines of our specimen, but in the whole text of book V, Bh has many mistakes in common with both of the Hamburg MSS. or else with one of them. Cp. our variants at 264, 6.14 2.24, 266, 10 (HI blunder: venivacchardjah, Bh wrong correction thereof : venuvatsardjah)? 269, 22 (paldyanavisayah BhH and jaldxrayam BhHI). 272, 22 (the number of the preceding tale being inserted in a wrong place). 275, 10 (same gap in BhH; corrected in I). 277,3.15.17 (wrongly corrected in I). 278,8 (BhHI vayam for vanam; but vanam must be the original reading, as it forms the contrast to grftam ; cp. also 278, is). 279,11 (original reading vyathd, as in our text; Bh corrupted to yathd; HI — a correction of this corruption based on the end of the fourth pada: — tatha). 280, 20 (Jco'pi for kam api; h also has this blunder!). 281, 9 (our text : trsnaika tu ; corruption in Bh : trsmkd tu ; correction thereof in HI : trsnd [I ir&nau\ kdpi). 12 (deva for yena, which is necessitated by the con- struction). 14 (Bh H). 282, 4 (our text : hatah Satruh ; H corruption : hatah Satrum ; corruptions thereof in Bh and I ; Bh : hatafatrum, I : hatah Satru). 6 (same gap in BhHI). 283,13 (a$vamadhyastho, corrupted to madhyastho in the archetype of BhHI ; this is corrupted to madhydsthdmi-n. H, and wrongly corrected to madhyasthdm in I). 284, 12 (BhI °gatirt Hh "gati, for °malir). 285, 21 (BhHI arddhodite for anuddhdnah ; but cp. 286, s).
If Purnabhadra had not been aware of all these gross blunders, he must indeed have been a blind man or a *J<sN^mnS!:. Certainly no such man would have been entrusted with the revision of an old celebrated work by
1 See above, p. 30 f.
3 Simpl. MS. h has a compound: °mahamamsavikrayasadhakav>ltij>rdbht-tindm, but it adds ekatamah li
s As to VInavatsa, cp. Speyer, Studies about the Kathasaritsagara, Amsterdam, 1908, p. 5.
Chap. III. Account of the Manuscripts 67
a minister.1 Hence we must conclude that, in the fifth book, not Bh, but the bh *-class has preserved the genuine text of Purnabhadra, and that the text given in Bh is a copy of some old MS. belonging to the H-class of the textus simplicior.
This view is corroborated by the fact that many — and always good — readings, in which bh* deviate from the Hamburg MSS., are to be found in the <r-class of the textus simplicior. Hence we may conclude that Purnabhadra used at the same time MSS. of both the H- and o--classes, preferring in most cases the H-class.2
The wording of the textus simplicior as contained in Bh's fifth book is of a high critical interest. In 1902, when I was not yet aware of the true nature of this part of Bh, though I saw that BhHI formed a clearly distinct group of MSS.,3 I thought it probable that the stories V, xv, xvi (Biihler and HI) did not originally belong to the textus simplicior, though they stand in all the MSS. of this recension I had, and have up to this day, examined.4 Now these two stories are missing in Bh. This shows that my view in this respect was correct.
As in the fifth book Purnabhadra follows the textus simplicior much more closely than in the rest of his work, I give the complete variants from Bh for this book from 260, 2 onwards. The beginning of the fifth book unfortunately is lost in Bh.
§ 7. Books I to III in Manuscripts Bh and $.
In order to show the relation between Bh and 4> in that part of the two MSS. which contains Purnabhadra's text, I give their readings, and nearly all of their even insignificant blunders, from the beginning of book II, p. 126, to p. 134,23 inclusive. It will be seen that * cannot go back, in this part of the text, to Bh. Both Bh and <!> must go back to some previous MS. Cp. Variants 127, ll.H.26. 128, 8 (here it is evident from <J>'s reading that, at the time when the source of 4> was copied, a small bit of the vowel under ^ was still visible in the original) ; 128, 7.12 (where the difference between the readings of Bh and 4> must go back to some marginal addition) ; 128, 19.30 ; 129, 1.9 ; 131, 8.9.10 (the interesting interpolation of f^lfe(/l«KW inserted only in Bh in due order) ; 132, 8
1 See pra&wti, 289,18.
2 See our parallel Specimens I to III. There, indeed, nearly all the text of Purnabhadra' a recension is to be found in HI, or Kielhorn-Buhler and h, where he follows the textus gimplicior.
• Berichte der kgl. Sachs. Ges. der Wissenschaften, phil.-hist. Kl. 1902, p. 68. 4 1. c., p. 68 f.
68
Chap. III. Account of the Manuscripts
(<I>'s reading more correct than Bh's, the case being such that no copyist would have been aware of Bh's blunder).
The fragment of Pumabhadra's text which forms the stock of books I to III in Bh<I>, does not contain the genuine wording, but an adulterated one. In very numerous cases, words have been transposed, omitted, or replaced by synonyms, without any evident reason, and other texts, especially the textus simplicior, have been compared by the reviser to whom Bh's text goes back. This occasionally causes disorder. For instance,
Discussion illustrated by text of Tale I, xiii, Lion's retainers
outwit camel.
In the following parallel texts, the words taken into the text of Bh from the textus simplicior are set in italics in the columns of Bh and HI.
Our text p. 75, 18.
18 bahavah panditah ksudrah, sarve mayopajlvinah I
19 kuryuh krtyam akrtyam va, ustre kakadayo yatba
20 Damanaka aha I katham caltat I so 'bravit I 22 asti kasmimscin nagare vanik Sagaradatto nama I sa ustrasatam 23 bahutnttlyacelakasya bhrtva kasyams'cid di& prasthitah I atha tasya 24 Vikatanamflstro 'tibharena nipidito visrasta- sarvaiigo niscestah 25 patitah I tato vanik ce- lakabharam aiiye.sustresu vibhajya ksipt-
va 26 ' aranyabhOmir iyam visama, asmin sthanc aa sakyatc sthiitmn' 27 iti Vikatam vihaya prasthitah I tasmims ca sarthava- lie gate Vikatah 28 sanaili sanaih samcaraii ^aspam bhaksayitum arabdhah I evam asau 76, 1 katipayair evahobhir balavan sam- vrttah I tasmims ca vane Madotkato 2 nama siinhah pratiTasati snia I tasyanucara dvi- pivayasagomayavah I 3 atha tais tad vanam bhramadbbir drstah sarthavahaparibhrag- tah sa ustrah I 4 tarn cavijnatapurvarnpam hasyajanakam drstvs simhah prstavan I idam 6 apurvam sattvam ilia vane prc- chyatam I kas tvara asi I tato 6 'vagatatattva- rtho vayaso 'bravit I ustro 'yam 'o'56 Pra" khyatanama I 7 tatah eimhena prstah I lihoh, kutas tvam iha I tena catmano yatlia-Svrttavi- yogah sarthavahat samakhyatnh I &c.
Bh (exactly as in the MS.).
vahavah pamditah ksndra sarve mayopajlvinah I knryuh krtyam akrtyam va ustre kakadayo yatha II 806 Damanaka aha II katham etat II so 'vruvit II asti kasmimsci nagare vanik Sagaradatto nama I sa ustrasatam bahumulyasya celakasya bhrtva kasyamcid disi prasthitah I atha tasya Vikatanama ustro 'tibharena pidito vi^rasta- sarvamgo niscesta patitah I tato vanik ce- lakabharam anyesu ustresu vibhajya ksipt- va aranyabhumir iyam visama "sinin s thane na sakyate sthatum iti Vikatam vihaya prasthitah I tasmin sarthava- he gate Vikatah sauaih sanaihr utthaya samcaran sispsm bhaksayitum aradhvah I cva ca sau katipayair evahobhir vvalavan su- vittah I tasmims ca vane
kaddcit tair Has tatah
paribhramamanaih sdrthdd bhrastah Kraihanako nama ustro drstiify atha simhah aha V aho apitrcam idam iatvaip\
tajnayatdm I kim etad aranyalc&m grdmyam va tata, irutra vdyasah aha II bho svdmin grdmyfyam ustrandmo jlvavitesah I tava bhojya\s tad vyapadyataip simhah II na ffrhsun dgatam hanmi I ulctam ca II grhe iairum api prdptam vifoastam akutoyam I yo hanydt tasya papam syae chatavrahmayaghatakam II Jtc
Chap. III. Account of the Manuscripts
69
in the following four parallel texts of our Tale I, xiii (Lion's retainers outwit camel), Purnabhadra follows Sar. ft, i. e. the secondary recension of the Tantrakhyayika, with an enlargement at the beginning of the story, in which our author, following the beginning of the frame-story of book I, narrates how the camel came to the forest. Bh has this selfsame be- ginning ; but from 76, 3 onward, this MS. copies a textus simplicior of our story. The consequence of this awkward contamination is a double one ; (1) the camel is twice introduced into the story, and (2) it bears two different names in the different parts of our tale. In the beginning it is called Vikata with Purnabhadra's text, whereas in the subsequent part of the fable its name is Krathanaka as in the textus simplicior.
Hamb. MSS. (Text exactly according to H ; in the footnotes readings of I).
bahavah pamditah ksudra I * sarve mam^opajlvinaji I J knryu ' krtyam akrtyam va I * ustre kakadayo yatha I Damanaka aha I katham etat so 'bravit I '
asmi * kasmimscid vanoddcse7 Madotkato nilma limhah prativasati ema I tasya ciinucarah anyepi dvl- pivayasagomayavah8 samti I atha Jeaddcit tavr ita» talo bhramadbhih 9 earthad bhrastah 10 Krathanako " ndmfftro (JfffaA I la atha gimha aha \ aho apurvdyam satiiai
tatjftayatam I Icim ayam aranydko ls va gr&myo va I tat frutvd " vdyaia aha I evamin 15 gramyoyam uttrah nama jlvaviiesah tava1' bkojyaj ca vydpadyatam I " eimha aha 1 18 «4ham
kagatam hanmi \ '• vMum ea 1 18 gfhe iatrum api praptaip vi&vaitam vihitagamatn " yo hanyat taiya pdpam syd \ t »atamahmat)aghatajam I M
Skr. ft.
bahavah panditah ksudraa sarve tnayopajlvinah I kurynr dosiam adosam va ustre kakadayo yatha II Damanaka aha I katham caltat I so 'bravit I
asti, kasmimscid vanoddcse Madotkato nama
fiimhah prativasati sma I tasyanucaras trayah piaitasino dvi-
pivayasagomayavah I atha tair
bhramadbhir drstas sarthavahaparibhras-
ta ustrah I tarn cajnatapurvarupam
hasyajananam drstva simhah prstavan I
idam apurvam sattvam iha vane prc-
chyatam I kas tvara iti I tato 'vagatatattvS-
rtho vayaso 'bravit I a-
khyatanamostro 'yam iti I tatas tena simhasa-
ka^am viivasyanltah I tenapi yathavrttam
atmano viyogas sarthavahat samakhyatah I &c,
* I "in. danda I 10 I °dbhastolt II
13 I ins. ma II " I tnchrufvd \\ " I mdmin II "I 'xastava \\
I double dan^a II " I niAitdyatarp with following danija II ao I tydchatabrdhmava? II
1 I ksudrah, om. danda II om. danda !l "I anti II 11 I double danda ||
* I mdnsopafivinah II 3 I kuryuh II
I vano&cie II » I drnpimva" II ' I "dbhi II
5 I sSbravlt,
11 I Kratha II
17 I om. danda II
70 Chap. III. Account of the Manuscripts
In the Tale II, vi, the two genii Karman and Kartr are confused, p. 157, 21 and 24 ; but in the second place Bh reads Kartrn for Karman, and in the following part of the story the mistake is not maintained.
In 49,14. the reviser shows his pandityam inasmuch as, after ^?R ^, he inserts tR|f<>m% micjid^l Hli^i (see Variants). But his pandityam did not prevent him from believing that tortoises are covered with hair; for
in 170,10 he makes fTT^:<*<Asn«*)4'i SMTWH.1 out of
And again, his pandityam abandons him in 218,12, where write 3T^t f^T$ I 3T^f f^3 I3 Purnabhadra here observes with his source Sar. /3 the rule laid down by Panini in his sutra viii. 2, 84 : <^{,1^ ^, i. e. ' (Pluti takes place) also in calling from afar.' This sutra was unknown to our reviser who, knowing that the figure ^ is frequently used in the MSS. to imply repetition (ex. jft^ for aft jft), and not seeing why this sentence was put twice in the text, writes it only once in this form :
3Tft f^I f^f fa*T IrtJt^iT, &c.4
Evidently this reviser used still other sources than the textus simplicior. For after the kathasamgraha '-stanza 125,30 he adds:
•T
%flf ^lRnid*Tl (!) «IWT I As to this stanza, see our 'Variants'. It is not the only one which has been interpolated in this revision.
A comparison of the other MSS. of Purnabhadra's text with his main sources, viz. the textus simplicior and the Tantrakhyayika, shows that the numerous deviations of Bh4> from our text go back not to the author, but to one of those awkward revisers who, in India, have so frequently destroyed the works of the poets. The text of Bh* is much more disfigured than that of A. Still the archetype of Bh must have flowed from a MS. whose text came very near to that contained in bh't'A. For in books I to III Bh has numerous mistakes in common with these MSS., or wrong corrections of their blunders. Cp. 9, 26; 10,2; 11,2.23; 13,16; 18,11; 28,10; 35, 18 (wrong correction); 42,8; 43,4; 44,6; 59,3; 60,30; 62,1 (MS. A correct) ; 64,3; 69,3; 74,8.17; 83,6; 93,9; 96, 10; 99,5.6.11; 101,12; 119,21; 121,7; 122,12; 123,12; 131,18; 132, 12 (wrong
1 So Bh*.
2 But cp. Variants.
* Our text spells f«(HI$ with Bohtlingk in his second edition of Panini. 4 MS. A reads 3Tlff (=MdJ 3Tt (!) f^5RJ, I, taking 9 for an old-fashioned form of <^. 6 This expression is to be found in Merutunga's Prabandhaeintamani (Bombay, 1888), p. 25.
Chap. III. Account of the Manuscripts 71
correction); 136,4; 138,12; 143,24; 145,24; 147,2; 152,10; 154, 16 (wrong correction) ; 161,2; 162, 13 ; 163, 13 (see Variants) ; 170, 20 (wrong correction) ; 179,18; 180,4; 186,4; 192,23; 198,9; 203,6; 204,5; 211,21; 212,22 (wrong correction).
There can be no doubt that Bh, in its Purnabhadra part, contains a very much adulterated text. Nevertheless, it has right readings in some places where bh* are defective. Cp. Variants on 33, 12.15.21 (cp. 6ar. A 39 to A 40. This passage is not to be found in the textus simplicior) ; 49,16; 83,2; 86,11; 102,10; 55, 10; 66,20; 71,10; 79,12; 80, s; 83,2; 132,27; 155,8; 156,15; 172,3.26; 183,6; 187,10; 193,2; 194,19; 197,10; 214,21; 220,7.
A great number of these passages contain trifling cases. Only in 33, 21 all our MSS. have a gap, which Bh — and KL2Mii2 — evidently fill in correctly. Our restoration of this passage is based on the consideration that the copyist's eye probably skipped from a first fRI^I^i: (1. 22) to a second fqJI«a<*: , such aberrations being the most frequent causes of gaps. If this view is correct, all these MSS. must have filled in this gap from some other MS., for they omit the first f^RRi:.
As in the case of A, it is not to be made out with certainty whether the Purnabhadra fragment contained in Bh* goes back to some MS. older than the archetype of bh^, or whether the right readings in Bh* in places where bh* are wrong, are due to revision. At any rate the blunders which Bh<J> have in common with bh^, show that such a MS. could not have been much older than the archetype of bh*.
Our parallel Specimens I to IV show that the text has undergone many alterations in Bh. It is true that in some cases Bh goes with either the Hamburg MSS. or the Tantrakhyayika against bh*. But none of these cases is such that we must conclude that Bh* have flowed from some more original archetype than bh*. In Specimen I, 1. 151, e.g., Bh has the same blunder as blutA, viz. Jjf*T for *J*ft. This blunder evidently goes back to a misreading of ^JI^T at the end of the pada. The copyist of the archetype of bh'J'ABh* took the second aw-stroke for a danda, and misread 1% as fa. In the same specimen Bh inserts T in 1. 45, makes ^nri^ out of *J^*ll<t. in 1. 60, and omits ^ in 1. 136 ; in all these three cases he destroys the metre. In this specimen the cases are especially frequent in which Bh goes with the Hamburg MSS. against bh'I'A. But it is quite certain here that these coincidences are due to the collation of some copy of the textus simplicior.
72 Chap. III. Account of the Manuscripts
In line 115 ff. our parallel texts run as follows :
HI athava so'tra raja, tad visVasasthane caturah feSakan atra dhrtva h athava yadi so'tra raja, tad vigvasaithane caturah SaSakan atra dhrtva Kielh. atha yadi so'tra raja, tato vis"vasasthane caturati salakan atra dhrtva Purn. tatas
Bh atha yady asau iha raja, tad visVasasthane caturalj fo.s'akan dhrtva tatas
H I tarn ahuya drutataram agaccha ; yena yah kaScid dvayor tnadhye
h tarn ahuy- agaccha ; yena yah kas'cid dvabhyam madhyad
Kielh. tarn ahuya drutataram agaccha ; yena dvayor madhyad yab kafoit
Purn. tarn ahflya drutara agaccha ; yena yah ka^cid avayor madhyat
Bh tarn ahuya drutam agaccha ; yena yalji kas'cid avayor madhyat
HI **.)&> Ba sarvan etan bhaksayisyati.
h raja bhavisyati, sa sarvvan etan bhaksayisyatiti.
Kielh. parakramena raja bhavisyati, sa sarvan etan bhaksayisyatiti.
Purn. parakramena raja bhavisyati, sa sarvan evaitan mrgan bhaksayisyati. Bh parakramena raja bhavisyati, sa sarvan ava(!) etan mrgan bhaksayisyati.
The sentence tad, &c. (Hlh), or tato, &c. (Kielh.), is grammatically incorrect, inasmuch as the subject of dhrtva is the lion, and that of ahuya the hare. Purnabhadra, for this reason and for a reason which we shall consider hereafter, deletes the words athava, &c. But it is quite clear that his tatas corresponds to the tad (Hlh) or the tato (Kielh.) which in these sources begins the apodosis. The author of Bh's archetype must have had before him Purnabhadra's text as given in bh*A; but besides he must have compared some MS. of the textus simplicior. For in his wording, the apodosis is twice introduced, first by tad, as in Hlh, and secondly by tatas, as in Purnabhadra (and in Kielhorn's text). This faulty construction can only be explained by the supposition that the author of Bh's archetype was not aware of the fact that in Purnabhadra's text tatas corresponded to tad of the textus simplicior, which he had before him, and that he only saw that in this text there were some more words (atha to dhrtva), which accordingly he inserted, without reflecting, before Purnabhadra's tatas.
As to the purport of our passage, I cannot believe that the text of the (T-class is here more original than that of HI. The wording of the Hamburg MSS. means : Bhasuraka is an usurper. Or else, if he is indeed the legitimate king, let him come, in order that that one of both of us who is the legitimate ruler may eat all the animals. This passage lacks wit; for evidently there is nobody to decide as to the lawfulness of the kingship of the two lions. The <r-class as represented by Kielhorn's text improves the sense, saying that the usurper proposes a single combat '
1 But the single combat is not even mentioned in the old MS. h of the tr-class •which only has the future tense bhavisyati with Kielhorn.
Chap. III. Account of the Manuscripts 73
in order to decide who, in the future, shall be the king of the forest. Accordingly Purnabhadra deletes the words athava, &c., which contain a conditional acknowledgement of the lawfulness of Mandamati's rajatvam.
Nobody will doubt that the reading of Bh is a contamination of Purnabhadra's genuine text and of the textus simplicior. Here, as in the case treated above, p. 68 f., the interpolator was not clever enough to avoid the traces of his activity. In the first case, he preserved the camel's two differing names from both the sources which he contaminated ; in our passage, he preserved, from these different sources, two different words — tad and fatat — which, though differing, correspond to one another.
Although these cases, taken with many others which of course I cannot treat here, have firmly convinced me that Bh does not go back to an archetype independent of that of bh*A, I give nevertheless the readings of this MS. throughout from the beginning of page 126 to 134, 23 inclusive, and for the passages marked with an asterisk in books I to III inclusive, and quote Bh occasionally in some other places.
Of book IV, Bh has only the text from the beginning (p. 228) to tatha hi (inclusive), p. 244, 10. In order to show the difference between Bh and <t> in this book, I give the complete variants of these two MSS. from the beginning of IV to p. 229, IT. From 229, 17 to 244, 10 the readings of Bh are given only in the passages marked with an asterisk. The readings of <£ I have neglected altogether.
Chapter IV. Principles which guided the editor in the construction of the text.
§ 1. Basis of the text of our edition.
INDIA is the ' classical ' country of interpolation and adulteration of texts. The more celebrated a work became, the more it was disfigured by copyists and revisers. Not even texts which, like the Mahabharata, are held to be sacred, have escaped this lot. A work so widely spread as the Paficatantra in its numerous recensions has undergone the most important changes in respect of its wording and of its contents, and that continuously, even to our own time. New editions quite different from the old work were prepared, and these new editions, after some time, were compared with older ones and melted together with them into new texts. In Purnabhadra's time there existed several redactions of this work, and Purnabhadra was well aware of the fact that none of them contained any
L
74 Chap. IV. Principles which guided the editor
longer the text as written down by the adyakam. In revising what had grown in the course of time to be a ' whole sastra ', he collected the different recensions and contaminated them, as shown above, not without inserting new materials.
This was the Hindu manner of philological work, which to our days prevails amongst the old style pandits. European scholarship has arrived at other methods. Whereas a Hindu wants before everything else a most readable text, we want a text that comes as near as possible to the wording of the author himself. But when Kosegarten gave the first edition of the Pancatantra, he followed not the European, but the Hindu manner of proceeding. Instead of separating the various recensions of the work which he was editing, he contaminated them; with what result has been shown above, p. 44 ff.
On p. ix of his edition of the textus simplicior he says : ' Utrum editio omatior, an simplicior, sit habenda vetustior, vel primae Pantschatantri formae propinquior, de ea re sententiam ferre certam non audeo; magis perspectum hoc habebunt posteri. ... Si quid video, editio ornatior, quanquam in earn ipsam recentiora multa recepta esse crediderim, in universum ad antiquam libri formam propius accedit, proptereaque cum libro Kalilae magis quam altera convenit. In editionis meae volumine hoc primo scriptura potissimum (!) ad editionem simpliciorem accommodata est, quoniam codices H.I.L. qui mini obtigerunt primi, illain editionem exhibent, eoque factum est, ut ad earn primam ex illis codicibus eruendam me adplicarem. Qui codices ubi nimis vitiosi vel mutili esse mihi vide- bantur(I), ex ceteris meliora vel pleniora supplevi. (Hence he gives in books III and IV a disfigured ' textus ornatior ', imagining the text of the Hamburg MSS. to be mutilated in them.) . . . Versiculos recepi nimis multos (I), ut lectores critici eorum, quos aut retinendos, aut eiiciendos esse censeant, ipsi instituere possint delectum.' An editor who renounces the critical examination of the text which he is editing to his readers instead of taking this duty upon himself, should abstain from editing altogether.
Translators who followed Kosegarten propagated the error about the true form of the Pancatantra among all the philologists and folklorists who were forced to base their research on translations. The Hindu editors, in reprinting Kosegarten's text, not without new alterations and additions, settled the opinion amongst Indianists that on the whole Kosegarten's text corresponded to the MSS. of this work. The only edition of the textus simplicior which has been prepared in a critical spirit is that of Kielhorn and Biihler. Though of course this school-book is not a critical edition in the strict sense of the word, it offers to us
in the construction of the text 75
the corrected text of one single MS., namely, of one which, belongs to the (r-class of the MSS. of the textus simplicior.1
Kosegarten's publication of a small portion of the textus ornatior, i. e. of Purnabhadra's recension, is as uncritical as his edition of the textus simplicior. I need not expatiate here on this topic, as any one can easily compare Kosegarten's text with our text and with onr variants as well as with the other recensions of the Pancatantra. I point out only the fact that the characteristic passage 4, 21 to 5, 2 is missing in Kosegarten's text.
That texts like these of Kosegarten are not only useless (this negatively), but also (this positively) a great and, effectual obstruction to the progress of philological and historical research, is a fact that no one is now likely to deny. Hence my first aim was to clear up the following questions : —
(1) How many different recensions of the Pancatantra are still existing?
(2) In what genetic relations do these recensions stand to one another ?
(3) Which MSS. are the most faithful representatives of their respective
recensions ?
The pedigree of the old Pancatantra recensions down to that of Purnabhadra has been established in the Introduction to my edition of the Southern Pancatantra. It is also given at the beginning of this volume, p. 5. The Southern Pancatantra and the Tantrakhyayika are critically edited. The genetic relations existing between the old Pancatantra texts are minutely studied in the Introduction to my translation of the Tantrakhyayika. Moreover, I have shown that Purnabhadra based his text mainly on the secondary recension of the Tantrakhyayika (Sar. /3), and on the textus simplicior which, as our parallel Specimens show, he used in MSS. of both the H- and the (r-class.
Now I have examined all the available MSS. of the Jaina recensions of the Pancatantra. All the MSS. of Purnabhadra's recension had to be classed roughly under two heads, the original class, and the mixed class.
To the former class I allot those MSS. which are most consistent at the same time with the Tantrakhyayika and with the textus simplicior, but do not share the provable interpolations of the latter. To the second class belong all the other MSS. The MSS. of the mixed class based on Purnabhadra's text show with especial frequency interpolations from MSS. of the textus simplicior and contaminations with them. Other mixed MSS. are based on the textus simplicior and contaminated with Purnabhadra's recension. But there are also MSS. which contain new recensions based
1 See above, p. 12 (pr), p. 58 ff., our parallel Specimens, and ZDMG. Ivi. 298 f.
76 Chap. IV. Principles which guided the editor
on the Jaina recensions and contaminated with Sar. /3, with the Southern Pancatantra, with the Hitopadesa, and with other sources.
The main criteria for the classification of the several recensions are : (1) the number, choice, and arrangement of the single tales and stanzas, and (2) the wording of the texts. In the arrangement of the tales con- tained in book III, Purnabhadra follows the oldest texts (Sar., Som., Ksem., S.P., Semitic recensions). Mixed MSS. in most cases deviate from this arrangement. For the first book, moreover, there is a sure criterion in the story of the Weaver as Vishnu. Mixed MSS. based on the textus simplicior have this story as I, v, but follow Purnabhadra, throughout or partially, in the arrangement and in the number of the stories of books III and IV. Mixed MSS. based on Purnabhadra' s text agree with him in the first and generally in the second book, but deviate from him in books III and IV. But there are even mixed MSS. which agree throughout with Purnabhadra' s genuine text in the number and in the arrangement of the stories ; cp. above, p. 56, § 6.
For the stanzas, also, I compared Purnabhadra's main sources, and this enabled me to find out easily the interpolations of single MSS.
Basis of the edition i bh, N, A ; vp, PL1, p, Fr, M ; Bh $.
The comparison of the number and arrangement of the stories con- tained in the single MSS. showed that only the following MSS. came under consideration for an edition of Purnabhadra's recension : bh N A, ^PL^PrM, Bh*. As shown above, ABh<I>p contain revised and con- taminated texts, which are very likely to have flowed from the same archetype as the other MSS. mentioned. Of these MSS., N goes back to bh, PL1 p Pr M go back to *. Consequently the only possible basis for our text must be the two equally excellent MSS. bh and *, which agree very closely in their wording. I generally follow bh, unless its readings are clearly wrong. Wherever both bh and * have a wrong reading, the emendation of which was not evident, I compared Simpl., Sar., A and Bh. In most cases this comparison affords sufficient evidence. There are, however, some rare cases, in which the same difference which appears in the best MSS. of Purnabhadra's text is to be found in the most trust- worthy MSS. of the textus simplicior ; x cp. 22,25; 181,2; 214, 19 f.; 242,11. In 68,11 a gloss in the margin of * gives the reading of Simpl. HI h. If some copyist would have preferred this variant, copying besides exactly the wording of his MS., this variant could induce some editor to prefer it and to reject what is Purnabhadra's genuine text. All the passages
1 In later MSS., owing to constant collations and contaminations, such cases are extremely frequent.
in the construction of the text 77
in which our text deviates, even in trifles, from bh and * are marked with an asterisk.
Though of course, except in passages where bh* are incomplete to-day, the MSS. NPPrM cannot come under consideration for the constitution of the text, I not only give the complete variants, but even most of the blunders of all these MSS. For these blunders are of the highest im- portance for critical work, as nothing is more useful to prove the relations which exist between kindred MSS. than the mistakes which they contain. Only of M a great many of the blunders have been omitted in my variants, as this MS. is extremely faulty. It bristles with misreadings, small gaps, and dittographies. I got this MS. before I had seen *. Otherwise I should have jotted down still more of its mistakes, and the relation between M and * would appear even more clearly than now. To the contaminated MSS. pABh<J> and to L1 only occasional reference has been given, except in book V, where I give the complete readings of Bh, which, as stated above, p. 56 ff. and p. 67, in this book contains an old and very valuable textus simplicior. From these variants it will be seen with how insig- nificant alterations Purnabhadra took over the textus simplicior of the fifth tantra into his own recension.
Manuscripts bh and ty differ very little from Furnabhadra's autograph text.
The very fact that so many MSS. can be proved to go back to bh* shows that in ancient times these two MSS., which I had the good fortune to use for my edition, were renowned for their value. As stated above, p. 37, the MS. bh goes back to a MS. which already was old when bh was copied from it. Our parallel Specimens also prove the excellence of the text of bh*. Indeed, I am convinced that neither of these copies deviates to any considerable degree from the text as written down by Purnabhadra himself, and that consequently our printed text comes as near to the author's genuine wording as any one of our current editions of say Goethe's prose works does to Goethe's own autograph text thereof.
§ 2. Emendation of the text.
Inferior MSS. sometimes have more correct readings than our oldest and most authentic ones. The question arises, whether in these cases we should conclude that these MSS. go back to some source independent of the archetype of our best MSS., and whether, if this be denied with good reasons, we should tolerate evident blunders in our texts.
In order to settle these important questions, I beg to be allowed to consider some standard examples, the nature of which we are able to
78 Chap. IV. Principles which guided the editor
determine with certainty. I take these instances from writings of eminent modern scholars, whose learning as well as whose accuracy is far beyond any doubt; and only to avoid the possible charge of malignity, I add some instances from my own writings.
Paul in his fundamental work ' Principien der Sprachgeschichte V p. 86, last line, gives hortibus as the dative case of the plural ofhortus. Hillebrandt says on p. iv of his ' Vedachrestomathie ' :2 ' Der leidige Druckfehler, welcher S. 38 entstellt [viz. Atharaveda], ist meine Schuld und von mir trotz dreimaliger Correctur ubersehen worden.' To the kindness of Miss Emma Benfey I owe the MS. of her celebrated father's translation of Christoforo Armeno's ' Peregrinaggio di tre giovani figliuoli del re di Serendippo' the beginning of which translation he published in the third volume of his periodical ' Orient und Occident '.3 This MS. is very carefully written in its author's fine and sympathetic hand. In this most authentic arche- type I read, amongst other slips of Benfey's pen, this sentence on leaf xvi, first page : ' Da aber der Jiingling beschlossen hatte, sich auf jede Weise an dem treulosen Minister zu rachen, ging er . . . in das Schlafzimmer der jungen Dic/iter des Ministers und umarmte sie alle drei mehrere mal.' L.v. Schroeder's excellent works are remarkably free from misprints. Still he writes, on p. 514 of his celebrated book ' Indiens Literatur und Cultur in historischer Entwicklung ' : 4 'Am Bedeutendsten und Selbstandigsten sind unter denselben zwei Dichtungen, welche dem Kalidasa zugeschrieben werden . . . : der Kaghuvan^a . . . und der Kumarasambhava, d. i. die Geburt des Liebesgoties ..." In the pedigree of the different recensions of the Pancatantra, p. Ixxix of my edition of the Southern Pancatantra, I wrote ' 7Ww<7?<-Fassungen ', and later on, as my attention was concentrated on inserting the newly discovered recension v, I repeated this blunder on p. xci. In my essay on the origin of the Hindu drama and epic, WZKM. xviii, p. 165, I wrote : ' Die dramatise/ten Beziehungen, die zwischen dem Epos und dem Drama bestehen, sind langst erkannt worden.' This, I am bound to confess, is unmitigated nonsense. What I wanted to write, was of course : ' Die Beziehungen, die . . . '. Likewise I wanted to write Tamil- for Telugu-. Paul intended to write horlis, Hillebrandt Athanaveda, v. Schroeder Kriegsgoites. Like Hillebrandt and no doubt the other scholars mentioned before, I had again and again revised the printer's copy and the proof-sheets without seeing my blunders. For there is not only a ' Druck- fehlerteufel ', who disfigures the words written by the author, but there
1 Halle, Max Niemeyer, 1886.
* Berlin, Weidmannsche Buchhandlung, 1885.
3 See Chauvin, Bibliographic des ouvrages arabes, VII, p. 160.
4 Leipzig, Verlag von H. Haessel, 1887.
in the construction of the text 79
is &pisaca much more malignant, the ' Schreibfehlerteufel ', whose deviltries are infinitely more dangerous to the author who has his text, i. e. the wording he intended to write down, firmly impressed upon his mind, and who very often does not discover the fatal slips of his pen until, the work being printed off, these rakwas stare at him from amongst the lines with devilish grimaces.
Benfey, of course, would have removed from his MS. most of the slips which his hand had committed while his mind was intent on finding an adequate rendering of the text he was translating. But the blunder Dichter for Tochter is one of the very kind which would escape the scrutinizing eye of the author, when Tochter is impressed on his mind. Thus even modern authors on philological topics, who in the course of their studies are trained to philological d/cpi/3eia, and whose metier it is to jot down, in preparing their editions, the very smallest clerical errors of their texts, are liable to overlook in their own wording evident blunders which at least Lord Macaulay's school-boy would detect at the first glance of his eye. And yet these authors certainly will revise their works again and again, first in their printer's copy, and afterwards in the proof-sheets. The old Hindu writers, owing to the circumstance that their works were not printed, had no occasion of revising them as often as modern authors. Moreover, a clerical error would much more easily escape their attention, as there is not the least separating of words in Sanskrit MSS., a circumstance which certainly does not tend to render revising more easy.
Hence the very thing which we should expect is that the Hindu archetypes (i. e. the authors' own autograph copies) should hardly ever have been free from mistakes, although these authors were men of un- doubted learning and thoroughly acquainted with the Sanskrit language in which they composed their works.1
But for the work of Messrs. Fischer and Bolte,* I probably should have published Benfey's translation of Christoforo Armeno's ' Peregrinaggio '. In doing so, I should of course have corrected the slips of Benfey's pen wherever the words he intended to write could be settled with certainty. I should have caused to be printed ' der jungen Tochter', and in order to give my edition of Benfey's work a diplomatic value, I should have mentioned his clerical error in my notes.
The case is different wherever authors can be proved to have sinned against the rules of the language. So even Schiller, for instance, uses
1 Cp. also above, p. 28, note 2, and p. 30, note 2.
2 Die Reise der SOhne Giaflers aus dem Italienischen des Christoforo Armeno fibersetzt durch Johann Wetzel 1583 herausgegeben von Hermann Fischer und Johannes Bolte. Tubingen, 1895 ( = Bibl. des Litt. Vereins in Stuttgart, CCVIII, Tubingen, 1896).
80 Chap. IV. Principles which guided the editor
the decidedly wrong form umrungen for umringt, as if this word derived from the verb ringen (' to wrestle '), and not from the substantive Ring. In his ' Jungfrau von Orleans ', verses 947 f. (I, 9), he says : — Umrungen sahn wir uns von beiden Heeren, Nicht Hoffnung war, zu siegen noch zu fliehn.
The slip certainly originated in the idea of struggle (Heeren) which was in the poet's mind. In a similar passage of the same tragedy, verses 2399 f. (Ill, 8), Schiller uses the right form :—
Umringt von Feinden ka'mpft sie ganz allein, Und hilflos unterliegt sie jetzt der Menge.1 And this same right form occurs in verse 447 (I, i) : —
Und find' ihn — hier 1 umringt von Gaukelspielern . . . The same holds true in India. Even so scholarly a work as the classical author Dandin's KavyMarsa is not quite free from anomalies; see Bohtlingk's edition,2 p. vi. In another sastra, whose aim was in part to teach standard language, viz. in the Tantrakhyayika (A 266), we find blunders against the rule laid down by Panini III, 3, 126 (Wackemagel, Altind. Gramm. II, § 82, a, y). In a very interesting paper read before the Twelfth International Congress of Orientalists,3 Prof. R Leumann proved from old palm-leaf MSS. that down to about the seventh century A. D. the Sanskrit written by most authors was not the pedantic one which had been laid down by the grammarians. On the contrary, Brahmans as well as Buddhists and Jainas wrote a Sanskrit more or less incorrect and in- fluenced by the popular languages. It was not before the time of the commentators, amongst whom Samkara and Haribhadra hold a prominent position, that the usual standard of Sanskrit was raised to a higher level, and that many things which before that time used to be tolerated, began to be avoided by good authors. But even after this time we find grammatical mistakes in the works of excellent authors which are critically edited. Hemacandra's so-called shortcomings in grammatical and lexicographical and metrical things have been pointed out by Jacobi, p. 9 f. of his edition of the Parisistaparvan.4 As to occasional mistakes committed by another learned author, viz. Pradyumnasuri, cp. Jacobi's edition, p. 3 f.6 Jacobi says that, ' as an epitomator and poet, he [Pradyumna] has done his
1 In the first passage (947 f.), the battle has not yet begun. This shows that Schiller cannot, by any means, have intentionally chosen the abnormal form umrungen.
1 Dandin's Poetik (Kavjadarca). Sanskrit und Deutsch herausg. von 0. Bo'htlingk. Leipzig, Verlag von H. Haessel, 1890. Cp. also Biihler, WZKM. viii. 29 f.
8 See Bezzenberger's Beitrage, 1900, p. 125 f.
4 Sthaviravalt Charita or Parisishtaparvan . . . Calcutta, 1891 (B. I.).
6 Shri Pradyumnacharya, Samaraditya Samkshepa, Ahmedabad, 1906.
in the construction of the text 81
task well. His language is concise in the narrative parts, pathetic in the moralising portions, and poetic in the descriptive passages which offer an opportunity of showing his proficiency in Alamkara. Still his work is not free from faults, even against grammar, the worst of which, a Prakritism, ^f|: instead of T[f*n, occurs VIII 520. And so he occasionally does not conform to the nicer metrical habits with regard to the Sloka as observed by the classical poets. But these shortcomings he has in common with most Jain writers during the period of their greatest literary activity and excellence (about 900-1300 A. D.).' As early as 1877, G. Biihler expressed his view about Jaina Sanskrit as follows : ' Die Kenntniss des Sanskrit ist bei den Jainas nicht weit her und hat auch wohl nie den Grad der Vollkommenheit erreicht, der sich bei den Brahmanen findet, obschon es nicht zu leugnen ist, dass sie in der Glanz- periode der Jaina- Wissenschaft vor etwa 700 Jahren hbher gestanden hat als sie jetzt steht. Selbst die grossten Jaina-Gelehrten wie Abhayadeva, Hemacandra und Malayagiri, welche unter den Caulukyas von Anhilvad- Pathan 943-1304 p. Chr. lebten, waren nicht im Stande, ein vollstandig richtiges und idiomatisches Sanskrit zu schreiben. Auch bei ihnen kommen hie und da wirkliche grammatikalische Fehler vor, und von dem Prakrit beeinflusste Redeweisen sowie vom Prakrit ins Sanskrit /urlickiibersetzte Worter sind haufig ... Es giebt [viz. to-day] deshalb unter den Yatis sehr viele, die wohl etwas Sanskrit lesen, es aber nicht