Halas Sattasai

Author(s): Hermen Tieken
Publisher: Leiden
Catalog link: https://jainqq.org/explore/006962/1

Page #1

HĀLA'S SATTASAI STEMMA AND EDITION (GATHAS 1-50),
WITH TRANSLATION AND NOTES
HERMAN TIEKEN
REPORail
घटविहिवाधाकार्तिकमयासमपतित्यपहातान्न दिनमानसामंताकतानकडशिपरिक्षितसाहनिदर्षि
झाषितामकारकनिमणमादेहितकानमा अवंशयासारदीवानापनवासवतिकवधाया

Page #2

HĀLA'S SATTASAI STEMMA AND EDITION (GATHAS 1-50), WITH TRANSLATION AND NOTES
Leiden 1983

Page #3

On the cover is reproduced the third leaf (recto) of a MS of the Sattasas at present kept in the Shri Hemachandracharya Jain Jnanamandira in Pāđana in Gujarat, India. For a description of the MS, see A descriptive Catalogue of Manuscripts in the Jaina Bhandars at Pattan, edited by C. D. Dalal, Vol. I, Palm-leaf MSS, Baroda, 1937, p. 178 ff. The leaf reproduced contains a part of the colophon.

Page #4

HĀLA'S SATTASAI STEMMA AND EDITION (GĀTHĀS 1-50), WITH TRANSLATION AND NOTES
PROEFSCHRIFT
TER VERKRIJGING VAN DE GRAAD VAN DOCTOR IN DE LETTEREN AAN DE RIJKSUNIVERSITEIT
TE UTRECHT, OP GEZAG VAN DE RECTOR MAGNIFICUS : PROF. DR. O. J. DE JONG, VOLGENS BESLUIT VAN HET COLLEGE VAN DECANEN IN HET OPENBAAR TE VERDEDIGEN OP VRIJDAG 4 NOVEMBER 1983 DES NAMIDDAGS TE 2.30 UUR
door
Herman Joseph Hugo Tieken · geboren op 1 april 1952 te Lijnden

Page #5

Promotor: Prof. Dr. H. W. Bodewitz

Page #6

CONTENTS
VI1
Acknowledgements List of abbreviations
INTRODUCTION Notes to the Introduction
PART I
CHAPTER 1 1.1 Previous editions (introductory) 1.2 The first edition by Weber 1.3 The MSS edited by Weber CHAPTER 2: 2.1 The three new South-Indian MSS: Ma, Ti and Tp CHAPTER 3 3.1 The stemma, its scope, method and results CHAPTER 4 4.1 The Third South-Indian recension, Jaina-recension and
the Vulgata (introductory) 4.2 The archetype 4.3 The relation between the MSS Ma, Ti and Tp of the
Third South-Indian recension 4.4 The relation between the Jaina-recension and the Vulgata 4.5 The relation between the MSS K, V, B, Y and P of the
Vulgata 4.6 The contaminated state of the text of MS R of the
Jaina-recension 4.7 Diagram showing the relation between the MSS of the
Third South-Indian, Jaina and Vulgata-recensions CHAPTER 5 5.1 The eclectic recensions: Sadhāranadeva, Telinga and
Second Telinga (introductory) 5.2 The Sadhāraṇadeva-recension (S) 5.3 The Telinga-recension (T)

Page #7

VI
5.4 The Second Telinga-recension (W) CHAPTER 6 6.1 Concluding remarks Notes to the stemma. APPENDIX I APPENDIX II APPENDIX III
87
111
119
PART II
143 144 150 151 152 154
CHAPTER 1 1.1 The nature of the text 1.2 The situations 1.3 The audience 1.4 The compiler 1.5 The language of the gāthās Notes to Chapter 1 CHAPTER 2 2.1 The method of reconstruction 2.2 The variants 2.3 The presentation Notes to Chapter 2 CHAPTER 3 3.1 Edition of Gathās 1-50, with translations and notes
161 173 176 178
179
267
271
Index of words discussed Index of grammatical points discussed Bibliography Nederlandse samenvatting Curriculum vitae
273
281
283

Page #8

VII
Acknowledgements
It is to a large extent due to the help and encouragement received from Professor Bodewitz that I have managed to complete my doctoral dissertation on the Sattasaf this year, precisely one hundred years after the last part of the 'first edition' by Weber appeared. It has been a pleasure to finalize my studies under the supervision of the person who also guided my first steps into the study of Sanskrit.
I feel deeply indebted to Dr. G. Schokker who introduced me to the study of Prākrit and who has remained ever helpful with suggestions and criticism,
I wish to express my gratitude to Professor 0. von Hinüber with whom I discussed earlier versions of most of the Gathās edited here and who kindly allowed me to profit from his extensive knowledge of Middle Indic languages.
I feel much obliged to Dr. Th. Damsteegt who read the complete final draught and gave many useful comments.
In preparation to the work on this edition I visited several manuscripts libraries in India and Nepal. A subsidy was granted to me by the Netherlands Foundation for the Advancement of Tropical Research (WOTRO) for the expenses involved during the trip. In this connection I should also like to thank Professor K.L. Janert who kindly allowed me to use his extensive library of catalogues of manuscripts collection's, and the authorities of the Government Oriental Manuscripts Library in Madras and of the University Manuscripts Library in Trivandrum for allowing me to photograph several manuscripts from their collections.
Finally I wish to thank my wife Ingrid. All problems arising in connection with the interpretation of the text were first discussed with her. These discussions made the work on the Sattasat a real pleasure.

Page #9


Page #10

LIST OF ABBREVIATIONS
ABC
Abh.
AIG AKM AMG. Ара. CDIAL CPD DED Desin. Ed. EWA NA
Guj.
Janert, An Annotated Bibliography of the Catalogues of Indian Manuscripts Ueber das Saptaśatak am des Hāla; ein Beitrag zur Kenntniss des Prākrits, by Weber, AKM V 3, 1870 Wackernagel and Debrunner, Altindische Grammatik Abhandlungen für die Kunde des Morgenlandes Ardha-Māgadhi Apabhras a Turner, A Comparative Dictionary of the Indo-Aryan Languages Trenckner, A Critical Pali Dictionary Burrow and Emeneau, A Dravidian Etymological Dictionary Des īnāmamāla Das Saptaśatakam des Hāla, by Weber, AKM VII 4, 1881 Mayrhofer, Kurzgefasstes Etymologisches Wörterbuch des Altindischen Gangādhara('s commentary or his version of the Vulgata) Gujarati Hindi Hemacandra's Grammatik der Prākritsprachen, by Pischel Weber, 'Ueber Bhuvanapāla's Commentar zu Hala's Saptaśatakam', Indische Studien 16, 1883 Journal Asiatique Jaina Māhārāştri Mahārāştrt Marāthi Māgadhi Middle Indo-Aryan manuscript(s) Monier-Williams, A Sanskrit-English Dictionary Raghavan, New Catalogus Catalogorum New Indo-Aryan e.g. Oguj., Old Gujarāti Panjabt Pāiyalacchi
JA
JM.
M. Mar. Mg. MIA MS(S) MW NCC NIA
.
Pāiyal.

Page #11

Pkt pr. m.
PSM
PTSD
PW Retr.
Prākrit prima manus Sheth, Pāias addaman annavo Rhys Davids and Stede, The Pali Text Society's Pāli-English Dictionary Böhtlingk, Sanskrit-Wörterbuch nebst allen Nachträgen Weber, 'Zum Saptaśatakam des Hāla', ZDMG 28, 1874 Sindhi Śaurasen 1 secunda manus Sinhalese Sanskrit The Prākṣtaprakāśa or the Prākrt Grammar of Vararuchi, by Cowell Zeitschrift der Deutschen Morgen ländischen Gesellschaft
sec. m.
Si.
Skt Var
ZDMG

Page #12

Corrections and additions
p.11, 11.27-32 delete. Insert: On the basis of purely external factors Weber arrived at the conclusion that of these six recensions the Vulgata was the most authentic one available (pp. L-LI).
p. 30 after 1. 26 add footnote 119: By definition most retentions are the more difficult readings and are therefore unlikely to be introduced secondarily in a MS for an innovation. This appears nevertheless occasionally to have happened (see, for instance, p. 52), but apparently not systematically or on a large scale. At first the retention could have come to be noted as a variant in the margin of a MS. Subsequently the possibility of choice may have arisen again, which need not always have been settled in favour of the innovation. It should be noted that the source of the secondarily introduced reading need not always have been a second MS. An alternative source could have been formed by the commentaries, which may in certain instances have preserved the original reading and, in explaining it, have turned it into an acceptable variant. These commentaries are also responsible for the introduction occasionally met with of one and the same innovation occurring independently in two different MSS (for a possible instance, see p. 45), especially where it concerns an innovation of the first type, namely a Sanskrit gloss for a typical Prākrit word.
p.32, 11.4-7 delete: Because of this...with some other MSS. Insert: These innovations point to the conclusion that the MSS have suffered from contamination, the extent of which remains, however, difficult to gauge.
p.83, 11.4-7 delete. Insert: The word is found a second time in a context which is not likely to have been confused with the present one, namely in 741: jai vi hu dinnundalia, etc. (Ma duddalia, Ti dinnandalia, Tp diņnuddalia; Madhavayajvamiśra: krtandolika, undaliketi bálikākrTÇIvišeşah; v.1. Tdi(1) lidilia, TC dilandilia, s mandamkillia, w dullia, x vimdiliā; cf. dillindilia 'child' in Pāiyal. 58 and Desin. V 40).

Page #13


Page #14

INTRODUCTION
The present study is preparatory to a critical edition of the complete text of the Sattasai.' The first edition of the Sattasat consists of four successive publications by A. Weber issued in 1870, 1874, 1881 and 1883 respectively, the last one appearing precisely one hundred years ago. Weber used altogether eighteen MSS which he divided into six recensions. As the basis for his edition he selected the so-called Vulgata-recension, which according to him was the most authentic one. However, a closer look at the text of the Vulgata reveals many instances where its readings are obviously secondary compared to those of the other recensions.' Another feature which becomes clear from Weber's edition is the great number of variants available for the text of each individual Gāthā. These variants, besides including the usual number of accidental errors, also, and more importantly, concern entirely different words, Pādas or even hemistiches. These factors call for a new edition of the text, in the course of which the question of the stemmatical relations between the MSS should be taken up anew. .. The present edition thus begins with the study of the stemmatical relations (Part I). In the edition itself, of which in this first volume only a sample consisting of the first 50 Gathās is given, it has been attempted to establish for each Gāthā a text as close to the original as was possible with the MSS available (Part II, Chapter 3).
The stemma involves ten of Weber's eighteen MSS, namely those which contain the Prākrit text. The other eight MSS contain only a Sanskrit commentary or a Chāyā; thus they need not be considered. Consequently the stemma covers five out of Weber's six recensions, the so-called x-recension being available only through a commentary. To these ten MSS could, however, be added three so far unknown MSS from South-India. The latter three MSS contain a seventh recension of the Sattasal, which, after Weber's Telinga and Second Telinga, will henceforth be referred to as the Third South-Indian recension.
One of the results of the stemma is the conclusion that the MSS are basically divided into two branches, a South Indian and a North-Indian one. Furthermore it appears that all available MSS, South Indian as well as North-Indian, ultimately go back to one single North-Indian MS which

Page #15

itself is lost and will henceforth be referred to as the archetype. . The text of the South-Indian branch is in many respects the most authentic one available. As to the order of the Gathās, which differs for each recension, the most original one is found in the North-Indian branch.
The stemma should provide the basic arguments in reconstructing the text of the archetype, in respect of the order as well as of the text of the Gāthās. This does not mean that the reconstruction has become an automatic process. There appear to be several cases where it is impossible to rely solely on the stemma, for instance, in the case of variation in the forms of the words, which appear to have been considerably influenced by Sanskrit in both the branches of the MSS. The method of reconstruction and several recurrent problems encountered will be discussed in Chapter 2 of Part II.
In each case the final test applied to the text is the translation. The Sattasaf has already been translated many times and into many languages, European as well as Indian. Of these only Weber's German translation is critical in that it is verifiable. My translation differs from Weber's in many instances often simply because it is based on a different text. Generally, however, the differences have another reason. : It is to be noted that Weber relied heavily on the traditonal interpretations found in the Sanskrit commentaries which in many MSS accompany the text. For the meanings of words Weber referred in particular to Hemacandra's Prakrit grammar and the same author's Deśīnāmamālā (twelfth century A.D.) and to Dhanapāla's Pāiyalacchi (tenth century A.D.). The latter two works are Prākrit dictionaries. It will appear that often the interpretations given in the Sanskrit commentaries rest on misunderstandings and have been based on faulty divisions of words or compounds, on vague associations, preconceived notions or etymological speculations. Furthermore a considerable number of rules and words included in the above-mentioned grammar and dictionaries appear to be due to their authors' or their predecessors' attempts to account for the derivation and meaning of certain words of, among other texts, the Sattasat. These authors' methods do not differ significantly from those seen in the commentaries.
One of the conclusions that may be drawn on the basis of the above is that between the time of the composition of the Gathās of the

Page #16

Sattasai and that of the Sanskrit commentaries, Hemacandra's grammar and the two dictionaries, there must have been a certain period of relative neglect of, or unconcern with, the text; a period, moreover, long enough to lose sight of the original intention of many of the Gathās and of the meaning of many of the words occurring in them.
The interpretations of the individual Gathās is often difficult. They contain associations or refer to practices and ideas which are not otherwise known. Many Gathās do not have parallels outside the text itself. A further context is provided by the text as a whole, which seems to have been compiled with the specific aim of amusing well-to-do city-dwellers and courtiers. Practically each Gāthā seems to have to be understood in this way, which occasionally may indicate a way out to arrive at a possibly correct interpretation. This broader context, provided by the text as a whole, will be discussed in Chapter 1 of Part II together with a few related problems concerning the compilership of King Hāla and the choice of a Prākrit dialect rather than Sanskrit.
At this stage I have refrained from making a complete word-index of the Gathās edited. I have restricted myself, instead, to preparing an index of words and an index of grammatical points discussed in the Notes.
The numbers of the Gathās refer to Weber's edition of 1881. However, when preceded by an asterisk it concerns a particular Gāthā in the part of the text edited here. In the lists in the stemma (Part I) I refer throughout to Weber's above-mentioned edition.

Page #17

Notes to the Introduction
The text is known by various names, e.g. Gathāsaptaśati, Saptaśataka, Gathākośa and Sattasaí (see Sternbach, 1974: 10). My reason for opting for the last one is in the first place practical: it is the name by which the text is discussed in the most recent book on the anthologygenre, namely the one by Sternbach referred to above. More importantly, it makes immediately clear that the text concerned is written in Prākrit.
For a list of translations, see Sternbach (1974: 10-11).
The oldest Prākrit grammar is very likely the Prākrtaprakāśa by Vararuci. In its original form it dealt exclusively with the so-called Māhārāştri Prakrit (Nitti-Dolci, 1938: 8 et passim). Its main concern is to define this Prākrit dialect vis-à-vis Sanskrit. Thus, it gives rules concerning sound changes in order to be able to derive Prākrit from Sanskrit words, and lists of Prākrit words, Adeśas, which apparently do not have Sanskrit derivations. In the grammar following Vararuci's, written by Hemacandra, there is a considerable increase in rules, often drawn up to account for only one particular word. However, the greatest increase is in the number of Adeśas (for which see the fourth Pāda). The two dictionaries seem to have been concerned with the same problem, including words which scholars at that time were as yet unable to derive from Sanskrit and as such were characteristic of Prākrit rather than Sanskrit.
A case in point is the verb nivvara- mentioned in Hem. IV 92 as an Ādesa for Sanskrit duhkham muñc- 'to relieve oneself from one's grief'. The verb nivvara- means 'to become calm or satisfied' and is related to Sanskrit nirvrt- in the same way as uvvara- 'to overflow' is to udvrtApparently this meaning was no longer available to Hemac andra. ConsequentTy he relied for the meaning on the various contexts in which the verb occurred. The meaning 'to relieve oneself from one's grief' seems to

Page #18

have been based directly on its occurrence in Gathā 204 of the Sattasai. Another good example of the way Hemacandra worked concerns the way he deals with the word omāliam, discussed in the note on Gāthā 32, which probably is merely a corruption for ņomāliam (navamālikām) there.

Page #19


Page #20

PARTI

Page #21


Page #22

CHAPTER 1
1.1 Previous editions (introductory)
The first edition of the Sattasai is the one by A. Weber (see 1.2). Other noteworthy editions are the so-called Nirnayasagar-edition (1889) and the one by A. N. Upadhye (1970). The former is an edition based on several MSS not known to Weber of the Vulgata-recension. The latter is an edition based on a single MS from Madras of the so-called Second Telinga-recension. Beside these there exist numerous other, mainly Indian, editions, which, however, do not bring forward any new material but are mere, more or less faithful, copies of the Nirnayasagar or Weber's edition."
For the present edition only Weber's manuscript material has been used. The material of the two other editions mentioned above has not been taken into account as this would only have resulted in a considerable increase in the number of variants which considering the results of the stemma will play only a marginal role in the final reconstruction of the text. The material could, however, be enlarged by three so far unknown South-Indian MSS.
In this chapter I will first give a description (in chronological order) of Weber's four publications which together constitute the 'first edition' (1:2). This is followed by descriptions of the MSS used by Weber (1.3). The latter descriptions have been kept as short as possible. For further information the literature quoted may be consulted, and especially Weber's publications. The order of the presentation of these MSS does not follow the chronology of Weber's successive publications, but is based on their contents; thus, those which contain the Prākrit text only come first, followed by those which contain the Prākrit text with a commentary, and finally those which contain a Sanskrit commentary or Chāyā only. Next the three new South-Indian MSS will be described (Ch. 2). 1.2 The first edition by Weber
The first edition, by A. Weber, consists of four successive publications, i.e. two text-editions proper (Abh. and Ed.) and two publications con

Page #23

10
taining variant readings and extracts from certain commentaries that were available at the time (Retr. and ISt.). I shall briefly describe each of Weber's publications in their chronological order.
Abh. Weber's first publication on the subject, Ueber das Saptaśatakam des Hala; ein Beitrag zur Kenntniss des Prākrits, AKM V 3, Leipzig, 1870 (repr. Nendeln, Lichtenstein, 1966) is based on a single MS, referred to as K. Abh. comprises a fragment of the text (1-370; one leaf containing the Gathās 94-102 is missing) with a German translation, extracts from Kulanātha's commentary and notes by Weber. The edition is preceded by a detailed grammar of the Prākrit of the Gathās (pp. 30-70).
In an appendix (p. 202ff.) Weber has added in all 67 Gāthās of very similar contents, gleaned from Dhanika's commentary on the Dasarūpa (nos 1-8), from Mammaţa's Kavyaprakāśa (nos 9-65) and from Visvanatha's Sahityadarpaņa (nos 66-67).
In his edition of 1881 Weber referred to this book with the abbreviation Abh., from 'Abhandlung', which will be used here too."
Retr. The above-mentioned edition was followed by an article called 'Zum Saptaśatakam des Hāla', ZDMG 28, 1874, pp. 345-436. In this article Weber compares several MSS acquired only after he had completed the Abh. The MSS are B. P. 14, D, and S (text and commentary) and A and C (commentary only). The former MSS provided variants to the text established in the Abh. Furthermore, extracts from the available commentaries are given: Gangādhara (A, B and C), Pitāmbara (P), Sadhāraṇadeva (S) and anonymous (T"). Where necessary Weber adapted his earlier translations and added notes. The material from these MSS is given only for the Gathās they have in common with Abh. Their additional material appears only in the edition of 1881 (see below).
In his edition of 1881 Weber referred to this article with the abbreviation Retr., for 'Retractatio'. This abbreviation is used here too.
Ed. Weber's third publication on the subject is Das Saptaśatakam des Haia,

Page #24

11
AKM VII 4, Leipzig, 1881 (repr. Nendeln, Lichtenstein, 1966). This is the first 'complete' edition of the Sattasas. However, readings, translations, extracts and certain observations of Weber, found in the earlier publications, have not been repeated here. The edition includes material from a number of MSS acquired only after the completion of the Retr., namely of y, Y, E, E, I, X, W and V.
In the introduction Weber classifies the available MSS in altogether six recensions:
1) Vulgata: 4 and y (text), K, B and P (text and commentary) and A,
C, E,'and (commentary only). 2) Jaina-recension: R (text) (and Bh (text and commentary), edited
only in 1883; see below). 3) x-recension: x (commentary only). 4) Sãdharanadeva-recension: S (text and commentary). 5) Telinga-recension: T (text) and TC (commentary only). 6) Second Telinga-recension: W (text and commentary) and V (Chāyā
only).
This classification is solely based on features in the outward appearance of the MSS. Each recension appears to differ from the others, in the first place, in respect of the order of the Gathās. In this connection the special status of the Sadhāranadeva and Telinga-recensions should be noted: in both the Gāthās have been arranged topicwise in groups, called Vrajyas and Paddhatis respectively. Furthermore, each recension is characterized by the absence or inclusion of certain Gathās. In this connection the Second Telinga-recension stands out: it includes even in its longer form only 104 Gathās.
Next Weber tries to determine which recension would represent the most original or authentic version of the Sattasai (pp. XLV-LI). In doing so he proceeded from the assumption that the most original recension would be the one which includes the greatest number of Gātās occurring in all the other recensions as well. This appears to be the Vulgata with 430 Gāthas that also occur in the other recensions.
In this work Weber has therefore edited in the first place the text of the so-called Vulgata, and in particular as it was commented upon by

Page #25

Gangadhara. This text was followed by the Gāthās of the other recensions or MSS not found in the Vulgata and by 44 Gāthās from several works on Alaskāra. These Gāthās are the same as the 67 found in Abh. (p. 202ff.), minus 23 which happened already to occur in one or more of the six recensions of the Sattasat.
The Gathās have been translated into German from 366 onwards. For the translations of 1-365 one is referred to Abh. and Retr. Variant readings are given below the text. The variants of those MSS, or parts of MSS, already edited in Abh. and Retr. are not given here again.! Beside extracts from Gangadhara's commentary occasionally extracts from the other commentaries are given as well. Where he deemed it necessary Weber has added his own notes and comments.
I shall refer to this work as Ed., which stands for Edition, the abbreviation Weber used in the publication that will be described below.
ISt. Weber's fourth and last publication concerning the Sattasaí is 'Ueber Bhuvanapāla's Commentar zu Hala's Saptaśatakam', Indische Studien 16, Leipzig, 1883 (repr. Hildesheim, 1973), pp. 1-204. It consists of the edition of the variants of MS Bh, compared to the text of Ed., together with extracts from Bhuvanapala's commentary and, occasionally, Weber's own notes. The text of MS Bh is a specimen of the so-called Jaina-recension.
The abbreviation used here for this work, ISt., stands for Indische Studien.
1. 3 The MSS edited by Weber
a. MSS containing the Prākrit text only: R, W, Y and T
Berlin, Staatsbibliothek, MS or. quart. 740 (pp. 139-258). MS or. quart. 740 contains on pp. 139-258 the text of MS v. In 1876 this MS was compared by Weber with a MS from Bombay lent to him by G. Bühler. The variants are designated with the siglum R. The original was a paper MS of 23 folios containing the complete text (1-710), written in Jaina

Page #26

13
Nāgart, of the so-called Jaina-recension. It is dated Samvat 1656 (1598 A.D.). The original MS was badly damaged inside. Lit.: Verzeichniss der Sanskrit- und Prākrit-Handschriften der königlichen Bibliothek zu Berlin, by A. Weber, Band 2, Abtheilung 1, Berlin, 1886 (ABC 20, 2), no. 1593 b; Ed. pp. XXVIII-XXIX.
Present whereabouts are unknown. W is a fairly modern, but undated, paper MS of 37 folios, arranged Europeanwise. The script used is Devanāgari. It contains the complete text (1-709) of the so-called Vulgata. It was lent to Weber by G. Bühler and apparently sent back to the latter while he was still in India. Lit.: Ed., p. XXVIII.
Berlin, Staatsbibliothek, MS or. quart. 740 (pp. 139-258).
the Roman script prepared by Weber in 1875 of MS Wilson 420, Oxford, Bodleian Library. This original is a fairly modern, but undated, MS, written in Devanāgari, of 24 folios, arranged Europeanwise, with 22 lines to a page, each line containing 25 Aksaras. It contains the complete text (1-709) of the so-called Vulgata. Lit.: Verzeichniss... (ABC 20, 2), no. 1593 b; Ed., p. XXVIII; Catalogus 'codicum manuscriptorum sanscriticorum postvedicorum quotquot in Bibliotheca Bodleiana adversantur, Vol. I 2, Oxford, 1864 (ABC 238, 1, 2), p. 381a.
· Berlin, Staatsbibliothek, MS or. quart. 555. MS or. quart. 555 is a modern copy of 70 folios, arranged Europeanwise, written in the Telugu script, especially made for Weber, of the complete text of the Gathās of MS no. 10283 (in the Telugu script) of the Tanjore Palace Library. Above the lines are written the variants found in the Grantha MSS no. 10285 (for 1-400) and no. 10286 (for 400-700) from the same library. These variants are referred to with the siglum To (Weber's To; I have preferred to in order to avoid a possible confusion with T. To distinguish I from To the former is occasionally labelled ta). I con

Page #27

tains the complete text of the so-called Telinga-recension.
For Weber's use a transcript in the Roman script was made of T together with TC (see below) by S. Goldschmidt, who also prepared a complete wordindex of the text: This transcript is kept in the same library, numbered MS or. quart. 706. Lit.: Verzeichniss...(ABC 20, 2), nos 1598 and 1600; Ed., p. XXX-XXXII; .. A classified Index to the Sanskrit MSS in the Palace at Tanjore, by A. C. Burnell, London, 1880 (ABC 306), p. 1740; Retr., p. 347.
b. MSS containing the Prākrit text as well as a Sanskrit commentary: Bh,
K, B, P, W and S.
Bh Berlin, Staatsbibliothek, MS or. quart. 743. MS or. quart. 743, consisting of 122 pages, contains extracts made in 1881-82 by Weber from a MS sent to him by F. Kielhorn from Poona. This original is at present included in the collection of the Bhandarkar Oriental Research Institute in Poona, as no. 245 of 1880-81. This MS, which is undated, consisting of 71 folios, with 17 lines to a page, each line having 46 to 48 Akşaras, contains the text together with a Sanskrit commentary. The script is Jaina Nāgart. The text is incomplete (1-600); the order of the Gathās is that of the so-called Jaina-recension. The commentary, styled Chekoktivicāraltlā, is by Bhuvanapala. Parts of the text repeated in the commentary are designated with the siglum Bho. Where clear distinction is necessary the text proper is designated Bho. Lit.: Verzeichniss...(ABC 20, 2), no. 1597; ISt., p. 1ff.; Report on the Search for Sanskrit MSS in the Bombay Presidency during the years 188081, Bombay, 1881 (ABC 273, 1), no. 245 (p. 65).
Berlin, Staatsbibliothek, MS or. quart. 740 (pp. 1-136). MS or. quart. 740 contains on pp. 1-136 a copy in the Roman script prepared by Weber of an original MS belonging to F. E. Hall. This original, of which the present whereabouts are unknown, consists of 48 folios with the text (1 to 6 lines) in the middle and the Sanskrit commentary (3 to 4 lines) above and below the text. Each line has 55 Akşaras. The script

Page #28

15
is Jaina Nagart. K contains only a fragment, namely of the Gathās 1-370 (folio 17 with 94-102 is missing). The text is a specimen of the so-called Vulgata. The Sanskrit commentary is by Kulanātha. Lit.: Verzeichniss... (ABC 20, 2), 1593 a; Ed., p. XXXIII; Abh., p. 1 ff.
B
.
Present whereabouts are unknown. B is a MS from Bombay. Its form and the script used are not described. It contains a fragment of the text (1-129) with a Sanskrit commentary by Gangādhara (see also below, A, C and E). The text is a specimen of the so-called Vulgata. The MS was lent to Weber by G. Bühler. Weber, who erroneously considered it a copy especially made for him, entered it into the collection of the Staatsbibliothek in Berlin as MS or. quart. 557. Upon being notified of his error Weber sent it back to Bühler in Bombay. Lit.: Ed., p. XXXII; Retr., p. 346.
P. Berlin, Staatsbibliothek, MS or. quart. 365. MS or quart. 365 is a collation of 108 pages of a fragment (1-300) of the text of the so-called Vulgata together with extracts of the Sanskrit commentary by Pftāmbara, prepared in 1872 by R. Pischel of no. 7219, India Office Library, London. This original consists of 63 folios of glazed paper, measuring 9 1/2 by 6 1/8 inches, rather carelessly written in the Kannada script. It is undated but probably goes back to the nineteenth century. The Prākrit text and the Sanskrit commentary are complete up to Gathā 151 (folio 47). From Gāthā 229 (folio 59') there is either only
the Prākrit text of the Gathās or only the Sanskrit translation. :: Lit.: Catalogue of the Sanskrit and Prakrit MSS in the Library of the
India Office II 2, London, 1935 (ABC 164, 2, 1), no. 7219; Verzeichniss ...(ABC 20, 2), no. 1603; A descriptive Catalogue of the oriental MSS and other articles illustrative of the literature, history, statistics and antiquities of the South of India. Collected by the late Colin Mackenzie, by H. H. Wilson, Madras, 1882 (ABC 184), no. 107; Ed., pp. XXXIV-XXXV; Retr., p. 346.

Page #29

Berlin, Staatsbibliothek, MS or. quart. 742. MS or. quart. 742, consisting of 346 pages, is a copy prepared by Weber in 1872 of no. 1750, India Office Library, London. This original consists of 124 folios of grey paper, measuring 12 by 4 3/8 inches, having 11 lines to a page. The script is Devanāgart. It is dated Samvat 1672 (1624 A.D.). S contains the complete Prakrit text (1-700) together with the Sanskrit commentary by Sadhāraṇadeva. The text, which is called Muktavalt and in which the Gathās are arranged in Vrajyās, is a specimen of the so-called Sadhāraṇadeva-recension. Lit.: Catalogue of the Sanskrit and Prākrit MSS in the Library of the India Office II 2, London, 1935 (ABC 164, 2, 1), no. 7218; Verzeichniss ...(ABC 20, 2), no. 1596; Ed., pp. XXXVIII-XLI; Retr., p. 347.
Berlin, Staatsbibliothek, MS or. quart. 590 (pp.1-49). MS or. quart. 590 (pp. 1-49, pages arranged Europeanwise, 22 lines to a page, each line containing 21 Akşaras) is a modern copy in the Telugu script, especially prepared for Weber, of no. 10287 of the Tanjore Palace Library. This original, which is written in Grantha, contains the text
ogether with an anonymous commentary, of the so-called Second Telingarecension. The first eleven Gathās of the original were illegible.
For Weber's use of the MS or. quart. 590 a copy in the Roman script was prepared by S. Goldschmidt. The transcript is kept in the same library, numbered Ms or. quart. 707. Lit.: Verzeichniss... (ABC 20, 2), nos 1601 and 1602; Ed., p. XLII; A classified Index to the Sanskrit MSS in the Palace at Tanjore, by A. C. Burnell, London, 1880 (ABC 306), p. 175.
c. MSS containing only a Sanskrit commentary or a Sanskrit Chāyā: A, C,
E, S, T, X, TC and V.
Berlin, Staatsbibliothek, MS or. quart. 741. A is a copy of 207 pages in the Roman script prepared by Weber in 1872 of no. 944, India Office Library, London. This original consists of 96

Page #30

17
folios, measuring 12 1/4 by 3 5/8 inches with 8 to 11 lines to a page. The script is Jaina Nāgarf. It contains the Sanskrit commentary by Gangādhara Bhatta (cf. B, C and E) on the text of the so-called Vulgata. Lit.: Catalogue of the Sanskrit and Prākrit MSS in the Library of the India Office II 2, London, 1935 (ABC 164, 2, 1), no. 7220; Verzeichniss ... (ABC 20, 2), no. 1594; Ed., p. XXXII; Retr., p. 346 (where the MS is designated with the siglum G(angādhara)).
Berlin, Staatsbibliothek, MS or. fol. 1316. C is a modern copy prepared for Weber through R. Griffith, of a MS from Benares. It contains the Sanskrit commentary by Gangadhara (cf. B, A and E) on a fragment (301-600) of the Vulgata. Lit.: Verzeichniss... (ABC 20, 2), no. 1595; Ed., p. XXXIII; Retr., p. 346.
E
London, India Office Library, Bühler 329. E consists of 75 folios of European paper (watermark 1873), measuring 13 by 4 1/8 inches. The script is Devanāgarf. It contains the Sanskrit commentary by Gangādhara Bhatta (cf. B, A and C) on the text of the socalled Vulgata. Lit.: Catalogue of the Sanskrit and Prākrit MSS in the Library of the India Office II 2, London, 1935 (ABC 164, 2, 1), no. 7221; Ed., p. XXXIII.
London, India Office Library, Bühler 328. & consists of 32 folios, measuring 10 1/2 by 4 3/4 inches. The script is Devanagari. It probably belongs to the nineteenth century. It contains a fragment of an anonymous commentary on the Vulgata. The commentary is the same as the one found in MS 7 (see below). Lit.: Catalogue of the Sanskrit and Prākrit MSS in the Library of the India Office II 2, London, 1935 (ABC 164, 2, 1), no. 7222; Ed., pp. XXXV-XXXVI.
TT
London, India Office Library, Bühler 326.

Page #31

18
T is a modern copy in Devanagari on 137 folios of European paper (watermark 1869) bound in book-form. The pages are written on one side only. Each page has 16 to 23 lines. At the end a date is given, viz. Sahvat 1730 (1672 A.D.), which was copied from the original. It contains an anonymous Sanskrit commentary on the Vulgata. It is complete except for 1-20, 62-78 and 415-430. The commentary is the same as the one found in MS & (see above).
In the margin of the first 29 folios extracts from Gangadhara's commentary (see B, A, C and E) are found, which in Ed. are designated with the siglum B (see Ed., p. XXXIII). Lit.: Catalogue of the Sanskrit and Prākrit MSS in the Library of the India Office II 2, London, 1935 (ABC 164, 2, 1), no. 7223; Ed., pp. XXXV-XXXVI.
London, India Office Library, Bühler 327. x consists of 75 folios. The script is Devanāgart. It is probably not older than the nineteenth century. It contains an anonymous commentary on 133-700 of a separate recension, of which it is at the same time the only specimen. Lit.: Catalogue of the Sanskrit and Prākrit MSS in the Library of the India Office II 2, London, 1935 (ABC 164, 2, 1), no. 7224; Ed., pp. XXXVII-XXXVIII.
TC
Berlin, Staatsbibliothek, MS or. quart. 556. MS or. quart. 556 is a modern copy in the Telugu script, consisting of 74 folios, arranged Europeanwise, with 21 lines to a page and 20 to 23 Aksaras to a line. This copy is based on two originals, namely no. 10283 (written in the Telugu script) and no. 10284 (written in Grantha) belonging to the Tanjore Palace Library. T contains an anonymous commentary on the text of the so-called Telinga-recension.
Of MS or. quart. 556 a transcript was made in the Roman script, for which see above, under T. Lit.: Verzeichniss... (ABC 20, 2), nos 1599 and 1600; Ed., pp. XLI-XLII; A classified Index to the Sanskrit MSS in the Palace at Tanjore, by A.

Page #32

C. Burnell, London, 1880 (ABC 306), p. 1740; Retr., p. 347 (where to was designated with the siglum Ty).
Berlin, Staatsbibliothek, MS or. quart. 590 (pp. 53-62). MS or. quart. 590 (pp. 53-62; pages arranged Europeanwise, with 22 lines to a page, each line having 21 Akşaras) is a modern copy in the Telugu script prepared for Weber of the Telugu MS no. 10289 of the Tanjore Palace Library. The latter MS is itself a transcript of the Grantha MS no. 10288. V contains the text of a Sanskrit Chāyā to the Gathās of the so-called Second Telinga-recension.
Of MS or quart. 590 (pp. 53-62) a transcript was made in the Roman script, for which see above, under W. Lit.: Verzeichniss... (ABC 20, 2), nos 1601 and 1602; Ed., p. XLII; A classified Index of the Sanskrit MSS in the Palace at Tanjore, by A. C. Burnell, London, 1880 (ABC 306), p. 174".

Page #33


Page #34

CHAPTER 2
2.1 The three new South Indian MSS: Ma, Ti and Tp
The text of these three MSS forms a separate, seventh, recension of the Sattasaf. Like Weber's Telinga and Second Telinga, ühis recension is found only in South-Indian MSS. I have therefore decided to refer to it as the Third South-Indian recension. It is characterized by a distinct order of the Gathās; see Appendix I, in which a concordance is given of the text of the Third South-Indian recension with Weber's recensions or MSS, and Appendix II, in which a concordance is given of Weber's Ed.with the Third South-Indian recension. The numbers in the following descriptions of Ma, Ti and Tp refer to Appendix I. In none of the three MSS the Gathas are individually numbered. The text, though, is divided into Satakas. The order of the Gathās seems random, or, at least, the principle behind the arrangement, if any, has been left unspecified.
In Ma and Ti the text is accompanied by the commentary of a certain • Madhavayajvamiśra. Specimens will be found in the descriptions of the two MSS. A consistent feature in the commentary is the description of the Nāyika and the specification of the poetical figure or figures present in the Gāthā, but apparently there is no system in relation to the order of the Gathās. The commentary quotes extensively from Bhoja's śrågāraprakāśa and Sarasvatīkanţhābharaṇa. For example, where below in the description of Ma the quotation of the 'beginning' breaks off, the commentary goes on with: uktañca bhojena: nāyikāścatasras, followed by a quotation from Bhoja's Srngāraprakāśa, comprising p. 612, line 19 up to p. 648, line 5 (the examples given by Bhoja are omitted). After that Bhoja is quoted on the sixteen qualities of the Nāyikā, comprising Śrngāraprakāśa, p. 777, line 1 up to p. 779, line 6 (again, the examples are omitted). The commentary to this Gathā closes with a remark on a division of Nāyikās as given in Bhoja's Sarasvat Tkanthābharaṇa: yacca sarasvatīkanthābharaṇe trimś adbhedatvakathanam ... In this case the quotations are found after the commentary proper, which otherwise usually closes mentioning the poetical figure. Generally they are found within the commentary. Moreover, the commentary frequently quotes the Kāmasūtra, for example, in connection with Gathā 12: ukta hi vatsyāyan Tye

Page #35

22
daśa tu kāmasya sthānāni, etc.
Madhavayajvamisra is most probably the same person who wrote a commentary on the Setubandha as well as on the Arthaśāstra; see Handiqui (1976: 89-92). The MSS of these commentaries, like those of his commentary on the Sattasaf, are restricted to, or hail from, Kerala.
The Chāyā found after the Prākrit text seems to have been added only later. It generally has the same lacunae as the text. Furthermore, where the text is lost completely, no Chāyā is found, while in some cases at least the commentary has been preserved.
Ma Madras, Government Oriental Manuscripts Library, no. 3378.. Ma is a Devanāgarf transcript of an original palmleaf MS written in the Malayā] am script. It consists of two parts of which the pages have been numbered separately: 1-11 and 1-225. The main, second, part contains the Gāthās from 11 up to and including 474 of the Third South-Indian recension. The text is accompanied by the Sanskrit commentary of Madhavayajvamiśra. Beginning (p. 1):
Śrt, hālas aptaśatitātparyaţikā sanketasthānīkrtasya śālikşetrasya pariņāmāt pratikşamadhavalibhāvena lavanabhayād rudatím svairinti sakht samāśvāsayati
kim ruyasi oņa amuhT dhavalāantesu chalicchettesu. hariālamandiamuht nadimva cchaņavādiā jāā. kim rodişi avanat amukhi dhavalāyamānesu śāliksetreşu
haritālamanditamukhi nad()īva saņavāţikā jāta, iti. onaamuht avanatamukht, sālicchettesu śālikşetreşu, haritālam manganan yasya taddharitālamandanam, tathābhūta mukham yasyāḥ sā tathā, jāā jātā, saņavāţikā haritālamanditamukht nad Tva jātā, yathā nadT pūrvabhūmikān tyaktvā bhūmikāntaram parigrhyān yevāvatişthate tadvacchyāmāvasthān tyaktva jhatiti puşpitāvasthāṁ gatan(!) sanketabhūmitvānugunapariņāmavatiti phalitārthah, parakTyasvap(/v?)arādh Tra(lacuna)pamā

Page #36

23
End (p. 225):
vidagdhagoşthyāṁ kayościnmattakāśinyos saundaryādinā mithassāmyaprastāve kaścid bhujangaḥ kusumavyapadeśena vaivarna vyañjayati:
gandhena appaņo mālaTe nomālia na cukkihisi(!) anno ko vi haāsāe māsalo parima luggāro. gandhena ātmano mālatyām navamālikā na śokşyati
anyaḥ ko'pi hatāśāyāh māṁsalo parima lodgārah, iti. mālaie jātau, navamālikā saptala, ca nukkihii(!) na hiyate na hina bhavet, māsalo māņsalah, parimalaścirānubandht gandhah vicitraratabhedopamardasahişnutvan tu dhvanyam, hatāśāyā ityatra mālatyā ityanuşangah, veśyāsūttameyam.
In Roman characters the following particulars have been added:
Transcribed in 1920-21 from a MS of Narayanan Nambur ipad (sic) of Kutalur Mana, Nareri, Tritala, Malabar Dt. Compared with the original MS. M. A. J., 19-4-1921.
In the margin is written: 'ac(?) 4-11-21.'
The Gathas have not been numbered separately. The text is divided into Śatakas. The end of each Sataka has been indicated after the commentary on its final Gathā. It runs as follows:
iti vādikavicūļāman imahopadhyāyasrfmanmādhavayajvamiśraviracitāyā hālas aptaśatTtātparyaţTkāyām prathama (dvitfyař, etc.) śatakan sampūrņam.
In the margin of the transcript the scribe kept count of the leaves of the original which appears to have consisted of altogether 105 leaves. The leaves themselves, however, do not seem to have been numbered, as would appear from the fact that the counting of the scribe is consecutive also where the order of the leaves has been considerably disturbed. This is the case from 334 onwards. Whereas Gāthā 334 is found on p. 157, the Gathās 335-341 are found on pp. 187-190, 342-400 are found on pp.

Page #37

24
161-186 (the commentary on 341 on p. 161.); 400-405 are found on pp. 157-160; 406-474 (end) on p. 190 ff. On p. 158 is found the indication of the end of the fourth Sataka, after 333 Gathās (by counting). That this state of the text is due to the disturbance of the leaves of the original -- which were not numbered -- appears from the fact that precisely opposite the respective breaking-points in the margin a number is found indicating the end of a leaf and the beginning of a new one.
At the beginning of this main part a transcript has been added of eleven pages, the original of which must have been very badly damaged, for only part of the text could be recovered. It contains fragments of the text of Gathās 2-10 of the Third South-Indian recension, with parts of a commentary upon them. No part of Gathā 1 of the recension is found. According to a note it was not copied until 1923. No particulars regarding the original are given, but most likely it was the same MS which served as the basis for the main part of the transcript. Possibly copying of the beginning was postponed because of the difficulties encountered in deciphering the text. Lit.: A triennial Catalogue of MSS collected during the Triennium 1916/ 17-1918/19 for the Government Oriental Manuscripts Library, Vol. IV lb, Madras, 1928 (ABC 201, 4, 1b), no. 3378.
Ti Trivandrum, University Manuscripts Library, no. 4763. Beginning (p. 2):
Śrt, halasaptaśatftātparyaţTkā
(lacuna)janasankule pi tuha dañsanan ti(!) magganti ramne va bhamai dimght pava(!)rāide anupingā, cha - sundarayuvanjanasankule pi tava darśanam vimārganti
aranya iva (bhrāmyati omitted) drstir varākyā anudvigneti. ramne va aranya iva, varāiāe varākyā, aņupingā anudvignā aganitaparivādabhayetyarthaḥ, utsavas amājādveva(?) kāmayamānayor mithodarśanam iti prāgeva darsitaṁ, parakTyās vadhire(?)pragalbheyam, upamālankāraḥ.

Page #38

25
End (pp. 282-283):
daravevirorujaalāsu(!) laṁliakesāsu mauliachisu purusdirtsu kāmo piāsu samjāuho vasai.
This text is followed by the Chāyā and a long commentary, which on p. 283 is concluded as follows:
iti vādikavicūļāmaŋ imahopādhyāyaśrīmanmādhavayajvamiśraviracitāyās hālasaptaśatitātparyaţTkāyāṁ saptaman śatakaṁ sampūrnań.
The title-page contains the following information (p. 1):
śrth, hala-saptaśatt-tātparyaţTkā pādi(:)kavicūdāmaņimahopādhyāyaśrTmanmādhavayajvamiśraviracitā. dvitTyasatake ekanavatitamagātham arabhya saptaśatakāntam. lekhakaḥ: (e)n anantakrsnaśarmā
In Malayā]am the following particulars have been added:
Owner of the MS: Ettumānūr Kailasapurattu Pişāroţi. Copying commissioned by: the same. Script (of the original): Malayalam. Material: palmleaf. Date of completion of the copy: August the 8th of 10 )0 (Kolam Era, probably 1914 A.D.). Signature of the scribe. Date of payment of the scribe (not filled in!)
Ti is a Devanagari transcript of 283 pages of an original MS written in the Malayalam script. It contains the text of the Third South-Indian recension from 197 to the end. The text is accompanied by a Sanskrit commentary attributed to Madhavayajvamiśra. The Gathās have not been numbered. The text is divided into Satakas. The conclusion of each śataka has been indicated after the commentary to its final Gathā: iti hālasaptaśatstātparyaţikāyās dvitTyan (trtTyam, etc.) śataka sampūrņam.

Page #39

At the beginning of the MS a group of four Gāthās is found which occur again as 374-377 (see also Tp). Gathā 279 is missing (cf. Tp). Of some Gāthas the text is missing due to lacunae in the original (577, 595, 596, 606; cf. Tp). Lit.: Alphabetical Index of the Sanskrit MSS in the University Manuscripts Library, Trivandrum, Vol. I, Trivandrum, 1957 (ABC 316), no. 4763; A descriptive Catalogue of Sanskrit MSS in the Curator's Office Library, . Trivandrum, Vol. VI, Trivandrum, 1940 (ABC 312, 6), no. 1439.
Tp Trivandrum, University Manuscripts Library, no. 4758. Beginning (1"):
ganapaino namo,
tam namaha ekkadasanam jo vaanek(!)ellavās aditthena sohai munālakabane (corr. into le)ņa via(!)dasaņuggame (rest of Gathā lost through damage).
End (36"):
daravevirorujualāsu luliakesāsu mauliacchTsu
purusāirtsu kāmo piāsu sajjāuho vasai. ŚrT ia.
Tp is a palmleaf MS of 36 leaves measuring 14 by 2 1/4 inches, with 8 lines to a page and 50 Akşaras to a line. The leaves have been numbered with letters. The script is Malayālam. In this MS the Gathās have not been numbered. It contains the complete text of the Gathās of the socalled Third South-Indian recension (1-699; 272 and 694 are missing). The text is divided into seven Satakas. The end of each Sataka has been indicated as follows:
śrt, ia hālakai(also kavi)viraiāņaṁ (for āe) sattas atņa (for Te) puţhamam (vtam, tiam...chatthan; for the seventh Sataka, see above) saań sampunnar, śrt.

Page #40

27
While writing the conclusion of the seventh Sataka the scribe was apparently interrupted and appears to have forgotten to complete it. In between 196 and 197 four Gathās are found which occur again as 374377 (see also Ti). Furthermore, for several Gathās lacunae have been noted (577, 595, 596, 606). The MS formerly belonged to śrf Nārāyanan Govindan Nambūri, Avl., Kumāramangalattu illam, Gangānāśćeri. Lit.: Alphabetical Index of the Sanskrit MSS in the University Manuscripts Library, Trivandrum, Vol. I, Trivandrum, 1957 (ABC 316), no. 4758; A descriptive Catalogue of Sanskrit MSS in the Curator's Office Library, Trivandrum, Vol. VI, Trivandrum, 1940 (ABC 312, 6), no. 1437a.

Page #41


Page #42

CHAPTER 3
3.1 The stemma, its scope, method and results
The stemma includes only those MSS which contain the Prākrit texts of the Gathās, i.e. those ten MSS edited earlier by Weber described in . 3 (a. and b.), and the three new South-Indian MSS described in 2. As such it covers only six out of the seven now available recensions of the Sattasaf, the so-called x-recension being known through a Sanskrit commentary only. The position in the stemma of the remaining eight MSS described in 1. 3 (c.), which contain only a Sanskrit commentary or a Chāyā, can be determined by comparing the order of the Gathās with that found in the first group of MSS. On the other hand, the state of the development of the Prakrit texts on the basis of which the commentaries and Chāyās have been written remains in most cases obscure. For instance, on the basis of a gloss niħsarati it is difficult to determine whether the commented text actually read. (original) nii or (secondary) ntsarai (see Gātha 49).
In the study of the stemmatical relations between these thirteen MSS an important role has been played by so-called innovations. As innovations should be considered those forms and constructions which can be shown to have been secondarily introduced into the text for certain original forms and constructions (retentions) which apparently were considered • unclear or corrupt, were misunderstood or had a graphical form that might give rise to confusion. Typical innovations are synonyms with an easily recognizable etymology. Examples are ntsarai for nti in 49, viltņo for vir(/1)ão in 290 or visarai for pamhusai in 355." It should be noted that many of these innovations, apart from their prākritized forms, cor. respond literally to the glosses found in the commentaries and Chāyās which in many MSS accompany the Prākrit text. Some seem to rest on a misunderstanding of the original text, such as martharam tumam vacca in 161 for maṁtharam pie vacca and kisa auņnāņa me sattt (one phrase) in 384 for ktsa a, puņņāņa me sattī (two phrases). The innovation alimganaduhiāim for "rahiain in 485 seems due to a confusion of the Jaina Nāgart Akşaras -ra- and -du-. Likewise, hoi dāņa(rahiassa) in 136 seems related to original bhoadāņa" through a spelling (b) ho(y)adāņa, with -ifor -(y)a-. On closer examination some innovations have come about

Page #43

30
through corruptions which by a mere coincidence did not disturb the sense. An example is padihāsai in 15, for original padihāai, in which the --- seems to have been inserted due to the influence of the following word sahihi. Other innovations have been introduced in order to remove unfamiliar constructions from the text, e.g. in 176 for the genitives vāeriassa and raassa, constructed with bharia 'filled with, the instrumentals vāeriena and raena have been introduced.
In all cases the secondarily introduced forms and constructions fit, or were made to fit, the metre, and, what is more important, make sense. As a result, in the transmission of the text innovations have proved to be very stable. This is generally not the case with accidental errors that have arisen through, for instance, transpositions of syllables or omissions. The results are forms and constructions which mostly do not fit the metre and do not make sense, so that at some stage they were emended or replaced by innovations.
Innovations are important in two respects. The occurrence of one and the same innovation in two MSS against a retention in the other MSS separates the two former from the latter and indicates that both go back to the same source or that the one derives from the other. As such their function is comparable to that of the so-called separative and conjunctive errors. At the same time they give an indication of the authenticity of the texts of the MSS. A text which contains an innovation against a retention in all the other MSS is less authentic, i.e. farther removed from the supposed archetype, than the latter. The text in, a third MS which contains this same innovation and in addition one of its own, is still less authentic.
On the basis of the innovations it can be concluded that the text of the Third South-Indian recension is in many respects the most authentic one available. The Jaina-recension and the Vulgata are characterized by a number of innovations against retentions in the Third South-Indian recension. The text of the Jaina-recension, at least as found in MS Bh, is, in turn, more authentic than that of the Vulgata, which in addition to the innovations referred to above has a number of innovations of its own (4. 1)
It can be shown that these three recensions go back to one common source, the archetype, which developed into the Third South-Indian

Page #44

recension, on the one hand, and into the Vulgata, on the other, with the Jaina-recension representing an early offshoot of the Vulgata. The archetype was a North-Indian MS, most probably written in Jaina Nāgart. The order of its Gathās must have been very similar to that of the present Jaina-recension and Vulgata (4. 2).
The mutual relation between the three available MSS of the Third South-Indian recension can be described fairly accurately (4. 3). They are divided into two branches, represented by Ma and by Ti and Tp respectively, of which the former seems the more conservative one. The Vulgata MSS go back to one common source which at the same time was the source of the present MS Bh of the Jaina-recension (4. 4). Among the Vulgata MSS it is possible to discern two main branches, represented by MSS K and ♡ and B respectively. These two branches are found combined in a third, represented by the two South-Indian Vulgata MSS, Y and P (4. 5). MS R of the Jaina-recension probably originally was a MS derived from an immediate predecessor of Bh and subsequently was compared with, and revised with the help of, a Vulgata MS (4. 6).
The remaining three recensions, the Sadhāranadeva, Telinga and Second Telinga, represent eclectic traditions. Thus the Sadhāraṇadeva-recension combines readings typical of the Jaina-recension with those typical of the Vulgata (5. 2). The Telinga-recension, which generally agrees closely with the Third South-Indian recension, includes readings typical of. the Sadhāraṇadeva-recension, the Vulgata and MS R of the Jaina-recension (5. 3). The Second Telinga recension is markedly different from all other known recensions in that it consists only of approximately 100 Gathās. Its readings agree closely with the other two South-Indian recensions. Furthermore, it has readings in common with some of the other recensions. It should, however, not be ruled out that its compiler drew Gathās from, or compared the text with, quite different sources, namely Alankaraśāstras. A significantly large number of the Gathās also occur in such works, where they serve as examples. Furthermore, the readings of some Gathās agree more closely to those found in the Alankāraśāstras than to those found in the other recensions of the Sattasaf (5. 4).
However, this does not mean that everything is entirely certain and clear. For instance, while the division of the Vulgata MSS referred to above has been based on the major pattern of the distribution of innova

Page #45

32
tions, it appears that practically each individual Vulgata MS shares at least a few innovations with every other Vulgata MS. Some have furthermore a few readings in common with some of the MSS of the other recensions. Because of this it is doubtful if, for instance, those retentions in v (and B) which separate it from the other Vulgata MSS have been regularly transmitted or have been entered into one of the sources of y after comparison with some other MS. Fortunately these uncertainties only seem to affect the mutual relation of the individual Vulgata MSS and not the position in the stemma of the Vulgata as a whole.

Page #46

CHAPTER 4
4.1 The Third South-Indian, Jaina and Vulgata-recensions
(introductory)
The Third South-Indian recension is known through three MSS (Ma, Ti and Tp), the Jaina-recension through two (Bh and R) and the Vulgata through five (K, V, B, Y and P). Each recension differs from the other two in the order of its Gāthās and in the inclusion or omission of certain Gathās. In all three, however, the order of the Gathās seems random, showing no internal organization. Each recension is furthermore distinguished from the others by a set of specific readings. From a study of these readings and in particular of the innovations, it appears that the text of the Jaina-recension together with that of the Vulgata is characterized by a certain set of innovations against retentions in the Third South-Indian recension (see under a. below); that the text of MS R of : the Jaina-recension together with that of the Vulgata is characterized
by another set of innovations against retentions in the text of the Third South-Indian recension together with that of MS Bh of the Jaina-recension (see under b. below); and, finally, that the text of the Vulgata is characterized by yet another set of innovations against retentions in the two other recensions (see under c. below). .. These facts point to the following -- provisional -- conclusions, namely (1) that the Jaina-recension together with the Vulgata derives from the Third South-Indian recension or its source, (2) that R of the Jaina-recension together with the Vulgata derives from Bh or from its source, and (3) that the Vulgata derives from R or from its source. At this point conclusions 2 and 3 should already be modified. Below it will be shown that R contains a contaminated text, combining readings of a MS closely related to Bh with those of a Vulgata MS (4.6). In this paragraph as well as in the following the Jaina-recension and the Vulgata are considered monolithic wholes, with R either included in the former or in the latter. Conclusions 2 and 3 can consequently be telescoped into one, namely that the Vulgata derives from the Jaina-recension or from its source.
In the lists given below minor mistakes or spelling variants found

Page #47

34
in the individual MSS are not always specified. Occasionally a 'retention' had to be reconstructed. Where this is the case first this reconstructed form is given followed by the forms actually found, between brackets.
In most instances in the three lists given below the decision as to which is the innovation and which is the retention should be clear. Nevertheless, in a number of cases an explanation seems to be required, which, however, would not infrequently involve, for instance, complete word-studies or the consideration of parallel situations. It will be obvious that a discussion of such type can hardly be taken up in the context of the stemma as it rather belongs to the edition of the text. However, as the present edition contains only the first 50 Gātās no more than a relatively small number of instances could be discussed; in these instances a reference is given to the Gathās concerned. Some instances, though, are taken up later in the stemma itself, in which case a reference is given to the page in question. For the remainder I have had to restrict myself to comments on those cases which could be explained by relatively short remarks. For each list these comments, where available, are brought together in one note. Otherwise I have to refer the reader to the edition of the complete text which is planned to follow.
a. Innovations in the Jaina-recension (Bh and R) and the Vulgata (K, V, B, Y and P) against retentions in the Third South Indian recension (Ma, Ti and Tp).
Where the Vulgata MSS K, B and P, and Ma and Ti of the Third South-Indian recension, which consist only of fragments, and the Jaina MS Bh, which misses the seventh Sataka, are not available, this has not been expressly mentioned, but should become clear from their absence among the MSS enumerated. I have, on the other hand, included references to the 'eclectic' Telinga (T), Sadhāraṇadeva (S) and Second Telinga-recensions (w), where these are available. For the latter three recensions, see Chapter 5. It is to be noted that T mostly follows the Third South-Indian recension.
It should furthermore be noted that the Jaina-recension and the Vulgata differ occasionally in respect of the form the innovation in a

Page #48

35
particular instance takes. The following list, though, is in the first place meant to show that the Jaina-recension and the Vulgata forma distinct group by the mere fact that in some cases they have readings which are demonstrably not original, where the Third South-Indian recension has readings which are demonstrably original. The differences in the forms of the innovations in the Jaina-recension and the Vulgata should belong to the discussion of the relation between these two recensions.
..
15 padihāsai Bh, R, K (padihā), , B, Y, P (padiāsahi), S against
padihāai (cf. T) (see Part II, Gāthā 15) 26 suhasutta R, K, B, Y, P, S, pāsutta Bh against sahasutta (cf. )
(see Part II, Gātha 26) 54 sari Bh, R, K, 4, B, P, T, sarisa y, s against siri? 105 dajjihisi Bh, R, K, V, B, Y, P, T, S against bhajjihisi 136 hoi dāŋarahiassa Bh, R, S, hoi kiviņapurisassa K, W, Y, P
against bhoadāṇarahiassa (cf. T: dāņahoarahiassa) 196 'phuddilliyāe Bho, phudilliyāi Bho, papphullide R, upphalliāi y,
uppulidi Y, upphulliai P, daphulliai S against urdaliae (or
uddalie) (cf. T: maridulliae) 12 254 oņiddhae (onivvae Bho) vasio Bh, uņņae vasio R, ulindae vasio K,
oladdhie vasio y, volimdae vasio y, gharālimdae sutto T, animae saio s against olambae vasium (cf. W: vasaim; Ma olappie, Ti,
Tp olabie) 283 nitthāmu Bh, R, K, T, nitthānu , Y, S against niadhāmu 316 suhaa Bh, Ry, Y, S against ahava (cf. T and K!) 318 pabbhattha Bh, R, K, , Y, S against pamhuţtha (cf. T) (pammut
tha Ma, Ti and Tp) 335 jaha na amhe (Bh: amha) musijjāmo Bh, R, na jahā amhe musijjāmo : K, Y, S, jaha ahme vi na hu musijjāmo T against jaha ahme na
amus ijjāmo (cf. W: na amusajjāmo) 360 bhārunnayāhin R, bhārona āhin K, bhāronai 4, bhāronaāhi Y,
bhārubbharāhi T, bharoņain S against bhārāharāni (see Part II,
p. 144) *361 amhāris Tņa Bh, R, amharisThi U, Y, T, assādisthim S against
ahmārisena
:

Page #49

36
visarai Bh, R, S, vtsario X, Y, visumario T against pahmusai 370 anadharayan taṁs avalasta Bha, anisamisatasavalata Bh",
animisam isantasavalanta R, animisam Tstsi valarta , animisatirichadaravalia y, animisas Tsuvvattanakhalaihta T, animisa
valata S against anisa (varnka) valata (see below, p. 50) 384 ktsa au nāņa me sattt Bh, R, , Y, S against kisa a punnāna me
satt (cf. T) 386 nesā gahalanghia Bh, R (1 aviā), T (1amjhiā), na ima gahalanghiā
U, Y, S against esā gahalambia 410 visakaṁd (/d)ali vya muddhā (vaddhatT) Bh, R, gāmanidhūā visa
kafidali bva (vaththia) , vva(ththia) Y, gāmanidhia visalaa
vva (vadia) S against gāmaŋidhūāvisalaa (vaddhartT) (cf. T) 420 paravajjanaccir Thi Bh, R, X, Y, S (°kajja”) against paravasana
naccirthin (cf. T) 431 taha tassa X, Y, S tassa taha T against tastassa 437 atthakkāulavTs athbhaghãi vlocãi), Y, S, atthakkāgaavisambhagāaa T
against atthakkāgaavesam (bhana) 460 duttha Bh, R, W, Y, S against dhavala (cf. T)12 466 maacchi Bh, R, W, Y, S against maaccha (cf. T) 483 aigarueņa vi tammai (Bh khijjai) hiaan bharena gabbhassa Bh, R,
aigaruena vi tammai hiae (s hiaammi) bharena gabbhassa (S gavbhabhāreņa) X, Y, T, S against tamei hiaan ahian bhaggā bharena
gabbhassa (bhaggam Ma, Ti and Tp) 485 alingan aduhiais Bh, R, W, Y, S against alinganarahiāin (cf. T) 487 me hiah tam pi Bh, R, me haań ta pi U, se haan tammi Y,
se hia taṁ pi T, S (5 tammi) against na sahiah tte 514 khijjiavvāin Bh, R, V, Y, S, lajjiavvãi T against rūsiavvāi 541 jalolle Bh, R, X, Y, S and Ma! against jalolla (cf. T) 548 taggaamaño W, Y, T, S against viaħbh amāņo 553 avaņehi Bh, naṁ nehi R, na nehi W; na nesi y against ninneha 575 marid ira Bh, R, K, M, Y, S against siṁdi(/a)ra (cf. T: sand ira) 626 muhapundarta R, X, Y, T, S against suhmarntaria 680 pavālaankurayavaŋŋalohillao R, pavālaankuraavannalohitae ,
paravāla ankuravannalohitae Y, pavvālaṁkur avannalohitae s against pavālaakūralohie niddhe (cf. T) niddhoyadhāurās tti kisa hatthe R, dhā(t)urāe ktsa sahatthe

Page #50

37
V, Y, dhāurae vva kisa hatthes against piddhoadhāurāe vi kТsa
hatthe (cf. T) 692 ahilijjai paumā lehadālivalaena kamalagovīe R, alihijjai paṁkaale
halalibalaena kamalagovte y, ahilijjai pankaale halātlavalaena kalamagovte Y, ahilijjfai pankaaluhiālivalaena kalamao s
against ahilijjai kamalullehadālivalaehi kalamagovte (cf. T) 711 volimto Bh, s (otto), volanto R, Y against bolei (cf. T)
Innovations in MS R of the Jaina-recension together with the Vulgata (K, V, B, Y and P) against retentions in the Third South-Indian recension together with Bh of the Jaina-recension.
3 majjhaārammi R, K, W, B, Y, P, S against majjhaārão (cf. T) 8 mamdan Thữam R, K, V, B, Y, P, T, S against maṁd aŋam bhūam
(maidanaābhūdaṁ Tp, madanabbhūya Bh (see Part II, Gātha 8) 49 nisarai R, K, V, B, Y, P against nii (cf. T, S and W) 78 pailla R, painna K, W. B, Y, P against pahulla (cf. T) (pa(p)
phulla Ti, Tp, pahalli Bh, paala S) (see below, p. 49) 96 parigaa R, W, B, Y, P, T, S, Ma! against pariar as 97 dakkhina R, W, B, Y, P, S against dāhina (cf. T) 106 naanesu R, K, W, B, Y, P, T, S and Tp! against vaanesu
(vaanenasu Ti) 119 ummatta R, K, W, B, Y, P, S against ummaia 129 anunta R, K, X, Y, P, S suvinto T against asunte! 152 pasāalohilla, aigahiagarua K, X, Y, P, T, S (in R the relevant
hemistich is missing) against pasāalohille, aigaruagahia 156 suharasa R, K, X, Y, P against sarahasa (cf. T and s) (saharasa
Bh) (see below, p. 49) 163 jivijjai R, K, V, Y, P, T, S against valijjai Tp, Bho (vilajjai
Ma, Bhd 165 maṁgalan 1 aṁgale R (se hale), K, V, Y, P, S against lagale
maṁgalan (mangale mangalam Bh) 170 palottanta R, K, Y, P, T, S against palotta (cf. v) ; '176 vāeriena, "raena R, K, , Y, P, S against vaeriassa, "raassa
phukkamto R, V, Y, P, antaa K, S against pūmertaa Ma, Tp

Page #51

38
(phūmitti Bh) 184 ujjua nāhan R, W, Y, anuā ņāha K, s (anuă), a pāha P against
annaman āham (unnamanāhan Bh)3 188 pautthamāņań R, K, V, Y, P, T, S against māŋapauttham (see
Part II, Gātha 31) 290 vilino R, K, U, Y, T, S against vill/r)ão 315 gayavaînan R, gaavaian K, , T, S, pautthavaia y against
gaapatam 320 vi na tahā R, K, X, Y, T, S and Tp! against vi na vi taha 325 anu(raimayan, orāaan) R, K, U, Y, S against ana(rāmaan) (cf. T:
ara in rame arti) 353 vi na tahā R (na taha), K, V, Y, T, S against vi na vi taha 368 thana(k)alasu(ppellana) R, T, S, thanaalamuhay, thanajuala
muha y against thanabharu 526 anikkarta R, W, Y, S against anakkaṁta (cf. T: ŋakkaṁta) 532 khokkha R, W, Y, S against p(h)upp(h)a (cf. T and W)
samullihai R, W, y against samalliai (cf. T: samullahai ai,
W: savalliai and S s amulliai) 575 ānandia R, W, Y, T, S against anardira
.
C.
Innovations in the Vulgata (K, 4, B, Y and P). against retentions in the Third South-Indian and Jaina-recensions.
Of the five Vulgata MSS only u and y are complete. Bis only a small fragment and is generally not available below. P contains only the first three śatakas and K the first three and more than half of the fourth. Below I have therefore made two extra categories, namely one where K is not available and another where P is not available.
22 litta K, Y, P and Tp (pr. m.), luppa y, B against tuppa (Tp
sec. m.) (see Part II, Gathā 22) 61 suhāveti K, W, B, Y, P, T, S and Ma! against suhāati (sam
appaħt i Bh) 65 anudiahan K, W, B, Y, P against tā diaharh 104 samuccinai K, U, B, Y, P, S against samuccei (cf. T)

Page #52

39
114 goviāna K, W, B, Y, P, S against goviãe (cf. T)* 160 maṁtharam tumarh vacca K, X, Y, P, S against mastharath pie vacca
(cf. T)(see Part II, the Note to Gathā 20) 203 vahugunavasto jano tassa K, M, Y, P against vahuguno va anno
jaŋo tassa (cf. T and S) 219 na naā K, ņa gaā V, Y, P, S against hmi gaa Ma, Tp (patta Bh,
ettā R, gamiā T) 220 vedhana K, U, Y, P, S against vedhaa (cf. T) 226 suttassa K, 4, Y, P, s against sutte (cf. T)
Innovations in
, B, Y and P where K is not available
95 pabbhasio V, B, Y, pakkhasakht P (cf. S: pabbhattho) against
pahmusio Ma, Tp (cf. T) (pamhusai Bh, paccusio R)
:
Innovations in K, , B and y where p is not available
210
L.
240 añdha(k)ara K,
, adharaa r (cf. S: ardhavera) against arħdhalaal4
4.2. The archetype
The three recensions are related in the following way:
X
Vulgata
Jaina
Third South- Indian

Page #53

40
The two other possibilities can be excluded:
Third South-Indian
Vulgata
Jaina
because the Third South-Indian recension has innovations of its own against retentions in the Jaina-recension together with the Vulgata (see under a. below); and
Jaina
Third South-Indian
Vulgata
because the Jaina-recension has innovations of its own against retentions in the Vulgata (see under b. below).
The two branches, represented by the Third South-Indian recension, on the one hand, and by the Jaina-recension and the Vulgata, on the other, ultimately go back to one single MS, the archetype X. Its existence is proved by certain errors common to the MSS of the Third South-Indian recension, on the one hand, and MS Bh of the Jaina-recension, on the other (see under c. below). The MS of the archetype was written in a North-Indian script, most likely a kind of Jaina Nagarī, as appears from the 'error', found in all available MSS, annāvarāha for āṇāvarāna in 488." The supposed corruption of an- into ann- points to a source in Jaina Nāgart, in which the Aksaras concerned are hardly distinguishable. Furthermore, the MSS of the Third South-Indian recension, which go back more or less directly to the MS of the archetype, have certain orthographical peculiarities which positively point to a source written in Jaina Nāgart (see under d. below). Admittedly, this source is probably

Page #54

not the same as the MS of the archetype. This would appear from the innovation pāada in Ma and Tp (and T) in Gathā 2 (see Part II) for pāua in all the other MSS. Its form shows that it was most probably ad hoc created on the basis of a rule typical of AMg. This seems to suggest that it was introduced in North-India rather than South-India. On the other hand, the word pāada is common in other south-Indian Prākrit texts as well, and in South-Indian MSS of Prākrit texts, so that it may equally well have been introduced in these MSS only at a relatively late stage in South-India.
As already noted above (4.1) all three recensions have a different Gāthā-order. The order of the archetype appears to be most faithfully preserved in the branch represented by the Vulgata, on the one hand, and the Jaina-recension, on the other. In order to make this clear in Appendix III a concordance is given of the Third South-Indian recension (first column), the Vulgata (broken down in hundreds in the following seven columns) and the Jaina-recension (R; last column). It appears that the Third South-Indian recension is based on a selection from a MS the order of the Gathās of which closely resembles that of either the Vulgata or the Jaina-recension, sometimes agreeing more with the Vulgata, Sometimes with the Jaina-recension. It can be observed that the compiler went through this MS from beginning to end picking out individual Gathās or whole strings of them, while rejecting others. The last Gātha selected in this first round is 669 of the Vulgata or, alternatively, 708 of the Jaina-recension (see opposite 425). After that he went through the MS again picking out many of the Gathās rejected first. These were apparently added following the text which resulted from the first round of selection, i.e. after 425. Occasionally, though, some of them were inserted in this first part of the text.
While the Vulgata and the Jaina-recension preserve the order of the archetype they do so only in broad outline. For it appears that in each recension large parts of the text of the archetype have come to be displaced. For instance, in the Jaina-recension (R) the sequence 131-187 of the Vulgata is found in the fourth Sataka between 337-391. A comparison with the Third South-Indian recension shows that the text of the Vulgata is original here (see opposite 135-141 and 172-176). Likewise in the Vulgata whole sequences of Gāthās have come to be displaced (see

Page #55

opposite 321-2, 333-4, 355-6 and 194-5). These changes are most likely due to the disturbance of the order of the leaves of a MS.
Besides, minor changes in the order and position of the Gathās have taken place. Note in this connection the Gathās 188, 189, 190 and 194 of the Vulgata; from their position in the Third South Indian recension as 69, 74, 84 and 70 respectively it appears that the Jaina-recension has preserved their original position here in the first Śataka. On the other hand, the Vulgata has preserved the original order of 332, 333 and 334, which occur as 255, 256 and 257 in the Third South-Indian recension but as 186, 110 (sic) and 287 in the Jaina-recension. It may be noted that the Vulgata has apparently suffered from considerably fewer of these minor changes in the order and position of the Gathās vis-à-vis the archetype than the Jaina-recension. A clear exception is its final Śataka in which the internal order must at some stage have become disturbed considerably (see opposite 10-53). Note in this connection that the first round of selection ended with 669 of the Vulgata, i.e. still rather far removed from the end, and with 708 of the Jaina-recension, i.e. practically at the end. The latter seems more logical."
a.
Innovations in the Third South-Indian recension (only part 197-474, for which all the three MSS, Ma, Ti and Tp, are available, has been taken into account):
476 viakkhanah (vi āamaih), 669 nisajjai (nimajjai), nisajjāsu (cf. T) (pimajjihisi), 724 ukkhippei (cf. T) (utt(n) amghei; ucchappei Bh, utthepei R, utthaggai s).
Innovations in the Jaina-recension:
71 na jano piaamakaamāṇabhamgadukkhakkhamo savvo, 156 hiyayāim, 183 cukkai, 206 taha cci(y)a, 211 hemistiches reversed, 302 va(y) anesu, 385 varis aṁ va, kappam va, 460 sayalagottha, 465 pasannā, 470 atthi Bh, achi R, 505 nayan addhabhallena, 554 tam mama (maha) cc(h)ari(y) an, 715 hatthe.

Page #56

C.
Common errors in the MSS of the Third South-Indian recension, Ma, Ti and Tp, and MS Bh of the Jaina-recension:
d.
43
49 sartsa Ma, sirTsa Bh (cf. sisira T), for sartra; 167 dharigim Ma, Bh, for dharani; 212 aquhohtaam Ma, Bh, for agahontaan; 292 tanui(d)a (emendation:) Ti, Tp, tanuyaimh Bh (cf. taŋuai and S), for tapuaai; 300 karipim Ti, Bh, for karani; 367 garu(y)ana Ma, Bh, for guruana; 371 bbharata Ma, bbharaftara Bh, for bbhamtara; 379 bhuppaht Te Ma, bhunnamt T(a) Bh, for bhadafitTe; 428 kapphlava Ma, kapalava Bh, for kavvalava; 466 pahia Ma, pasiha Bh, for.pasia; 470 daddham Ma, Bha, for daṭṭhum;
487 pavvasa Tp, Bh, for pavāsa.
Note the omission in both Ma and Bh of kim in 65 peusamattapaddi kim suasi
Orthographical peculiarities in the MSS of the Third South-Indian recension, Ma, Ti and Tp, which show that they ultimately derive from a MS written in Jaina NagarT:
"
1) -mt-for-nn- (in Jaina MSS -nn-): mamtu for manņu in 260 (Ti, Tp), 683 (Ma, Tp), 920 (T1, Tp), 938 (Ma, Tp) and 941 (Ti, Tp). 2) ru- for -tu- or -tu- for -ru-: 189 ruha (Ma) for tuha (cf. R and K; note Weber's hesitation regarding the reading of Bh; Ist., p. 41), 251 tuvah (Ma) for ravah (sec. m.) (cf. K), 476 rūsai (Ti) for tusai, 609 rura (Ma) for tūra.
4.3 The relation between Ma, Ti and Tp of the Third SouthIndian recension
Of these MSS Tp alone is complete (1-699). Ma contains only Gathas 2-474 and Ti 197-699. This means that only for 197-474 all the three MSS are available.
Tp is written in the Malayalam script. Both Ma and Ti are Devanagari transcripts of Malayalam originals. All three MSS have orthographical peculiarities which point to an original in Grantha. Thus, one finds

Page #57

44
-S- for -- in 257 eraso (eram)do) in Ti, and -- for -s- in 70 pā(v)uda (pāusa) in Ma, Tp, 444 Tdam (is aṁ) in Ma, 445 pāuda in Ti, Tp, 538 pauda in Ma, and 578 pā(v)uda in Ti and Tp. In the preceding paragraph I have shown that they also show traces of having been copied from a MS written in Jaina Nāgart. This probably explains the occasional occurrence in Ma and Ti of the Ya-śruti which is otherwise typical of Jaina (Nāgart) MSS. E.g. 2 pāyada Ma, 5 mayulanti Ma, 9 ruyas i Ma and 348 taņuāraya Ti. However, dhāayi in 130 in Ti (dhāvai in Tp), which is an innovation in this particular branch of the Third South-Indian recension (see below) for dhuai found in Ma and the North-Indian MSS, sug- . gests that the -y- may also have been inserted in these MSS independently.
The three MSS have one common source. This conclusion is based on the following two types of evidence:
a.
The order of the Gathās. I have shown in 4.2 above that the present recension is based on a selection from a MS which was the source of the Vulgata, on the one hand, and of the Jaina-recension, on the other. It is doubtful if this process of selection could have taken place three times independently, each time with exactly the same result.
b.
Common errors (only Gathās 197-474, for which all the three MSS are available, are taken into account):
254 olappie Ma, olaħbie Ti, Tp (olambae), 341 tte vi taha (tte vi), 354 parunnena ajja (°enan pi ajja), 367 avilambirte (avao), 373 nihai (nihānai), 395 dararuggaa (daruggaa), 407 aumalana (omarana), 409 pahnu(/a)sai, 462 pahņu(/a)sai (panhuai), 496 soūņa vi (soūņa), 524 kuviassa pāae in Ma, kuviaṁ pasāāeuth Ti, kuvia pasāaeuth Tp * (kuviañ ca pasdeun), 572 qahidattho (gahiattho), 643 ujja (ujja pāmaro, unmetrical), 924 na atthirammi na (athirammi ņa), 932 ehi (ei).
The relation between Ma, Ti and Tp is as follows. Tp and Ma each have innovations of their own (see under a. and b. below), whereas -Ti has none.

Page #58

45
This excludes the possibility that Tp goes back directly to Ma or vice versa. However, it leaves open the possibility that Tp as well as Ma go go back directly to Ti. Tp, though, shares a number of innovations with Ti (see under c. below) and none with Ma. Ti shares one innovation with Ma, namely as aṁs ia for āsamghia in 707. This form āsaṁsia may, however, have been introduced independently in the two MSS. In this connection it should be noted that in the commentaries which in Ma and Ti accompany the text, asanghia is glossed with aśassita. 18 This evidence leads to the conclusion that the first division was between Ma, on the one hand, and Ti and Tp, on the other.
a. Tp has the following innovations (only 197-474):
240 pucchāhito (cheppāhisto), 243 sohai (cf. T) (chajjai), 497 vārei maṁ (cf. T) (vāreha naṁ), 551 unnāmia (uttānia), 607 siae (sāhuli).
b
Ma has the following innovation:
588 vilakkheņa (vilaveņa).
c.
Ti and Tp share the following innovations:
130 dhāayi Ti, dhāvai Tp (dhuai), 246 haliddā (haliddf), 312 gahiūņa (ghettūna), 313 nikkudaakola (Tp sec.m. padohar arkola), 325 na arāmaan (Ma anaa(!)rāmaar) (aņarāmaam), 656 loluva (lehada), 843 digghohņā (d Thu ha), 848 saaṁdhiae, 849 saadhie (cf. T) (saajjhide), sāmalam Tp)muham (sāma muham), 869 rudiūna (rottūņań).
The relation between Ti and Tp still needs to be determined. In this connection some significant similarities in the outward appearance of both MSS may be considered. In the first place, in both Gathā 279 (see Appendix I) is missing. Furthermore, Ti and Tp have a number of identical

Page #59

46
lacunae (290 (Ed. 715), 411 (626), 577 (-), 595 (-), 596 (-), 606 (863), 632 (153), 647 (269), 654 (905) and 690 (530)). The extent of the lacunae is, however, not always the same. In Tp the lacunae of 411, 595, 596 and 632 cover the complete texts of the respective Gathās, while in Ti still one or two words have been preserved.
The Malayalam original from which Ti was copied consisted of a bundle of palmleaves containing the fragment 197-699. To this bundle was added a stray leaf with the Gathās Ed. 492, 481, 482 and 496 together with Madhavayajvamiśra's commentary. In Ti these same four Gathās, with minor variants, are found again as 374-377 (373-376 by counting), embedded be- . tween Ed. 489 and Ed. 488. Ed. 489, however, occurs a second time, namely after Ed. 488. From this it would appear that this group of four Gatha's was inserted here later. In Tp, though containing the complete text (1-699), exactly the same situation is found. Thus, after 196 some space is left blank; then follow Ed. 492, 481, 482 and 496, which are followed by a blank space again, after which the text continues with 197. The same four Gāthas occur again in the fourth Sataka:
Third S-I recension
371 372 373 374 375 376 377 378
Ma Ed. 487
488 489 (492) 481
Ti/Tp Ed. 487
489 (492)
482
(496 503
482 (496) 488 489 503
The addition of the stray leaf containing Ed. 492-496 before 197 is explicable only in the case of Ti, i.e. its palmleaf original. It then follows that Tp is a collation of Ti, or any of its sources, and some other MS. It remains to determine the extent of the text of this latter MS and its relation to either Ma or Ti, if any.
As has been noted above, Tp has the same lacunae as Ti, though

Page #60

47
occasionally the latter has preserved a few words more of the lost text. This would either mean that in preparing Tp the second MS used beside Ti (i.e. its source) also had those same lacunae; or that the second MS, like Ti, was itself a fragment, consisting of precisely the part 1-196 broken off from Ti.
The latter assumption, that the second MS con cained only the fragment 1-196, is refuted by the following instances, drawn from Gathās 197-699, in which the scribe of Tp 'corrects' a reading in accordance with Ti: 333 Ti rakkhido, Tp (pr. m.) rakkhio, (sec. m.) rakkhido, Ti sas arkiru Tp (pr. m.) sasaṁkiu, (sec. m.-) sasank iru, 362 Ti tavijjai, Tp (pr. m.) taņuijjai, (sec. m.) tanuvijjai, 378 Ti aņusikkhirie vi, Tp (pr. m.) aņus ikkharie vi, (sec. m.) aņusikkhirie vi. Probably in these cases the scribe started copying from the second MS and subsequently compared the text with that of Ti. On the basis of this evidence I assume the second to have been a complete MS.
The lacunae which Ti and Tp have in common were apparently also found in the second MS used by the scribe of Tp, for otherwise they would almost certainly have been filled in. In any case, the assumption that the scribe of Tp would on purpose have refrained from fiiling in the lacunae found in his first MS is contradicted by his behaviour regarding some smaller omissions. The latter have occasionally been filled in, either immediately or secondarily, i.e. above or below the line. E.g. 23 Ti aha sa cci (the text breaks off and is immediately followed by Mādhavayajvamiśra's commentary), Tp aha se tti visas imo. This instance possibly shows that the scribe made use of the commentary, which glosses the last word of the lost part of the text, saddahimo, with viśvas imaḥ; 460 Ti uva, Tp id., above the line the syllable ha (uvaha) has been added; 522 Ti suraārambho (...)ņavo, Tp suraārambho uno(!) navo; 688 Ti cadu(...)niunāna, Tp id., above the line the word kamma has been added. Occasionally, furthermore, the scribe of Tp made small corrections. E.g. 333 Ti raian, Tp id., corrected into ramiah, 452 Ti mānan, Tp māl: deleted)nāmaḥ, 873 Ti tae, Tp tte, changed into taha (cf. T), 923 Ti bahala, Tp bahala, corrected into balā.
It follows that the second MS used by the scribe of Tp was a complete M$. As it apparently had the same lacunae as Ti, it may be concluded that it went back to an immediate source of the present MS Ti.

Page #61

48
The relation between Ma, Ti and Tp may be represented in the following diagram. It should be noted that the position of Tp to the left of Ti is arbitrary; it does not imply that Tp is closer to the source of the three MSS than Ti.
1-699
1-699
1-196
197-699
1-699
Ed. 492-496
197-699
Ma
Тр
The third South-Indian recension seems to have existed in relative isolation. As will be shown below the Telinga and Second Telinga-recensions have both drawn their Gāthās from, among others, this recension, but these seem basically to be instances of one-way traffic only. An exception, however, is Tp, which shows traces of having been contaminated by T. With the latter Tp shares the following innovations:
20 desu suhaa majjha oās a Tp, suhaa maha desu oasa T (de suhaa majjha oās aṁ), 49 lagga (cf. Ma lakkā) (lukkā), 243 sohai (chajjai), 497 vārei maí (cf. , Y, S) (vāreha nan), 626 khana (chana), 677 rosaņań (rūsanan).
From the fact that at least one of these innovations concerns à reading

Page #62

49
otherwise typical of the Vulgata, namely 497 vārei mam, and from the fact that of the Third South-Indian recension and T, only the latter also otherwise includes readings typical of the Vulgata (see 5.3), it would follow that Tp was influenced by T, and not vice versa.
4.4 The relation between the Jaina-recension and the Vulgata
In 4. 1 and 2 I have indicated that the Vulgata derives from the same source as the Jaina-recension. The Jaina-recension, which is represented here by two MSS, Bh and R, is characterized by innovations of its own. This excludes the possibility that the Vulgata derives from any of these two MSS directly. As already said earlier MS R contains a contaminated text. The extent and nature of the contamination can only be determined after the main branches among the five Vulgata MSS have been identified, and will therefore be the subjects of 4.6 only. Below some instances will be given which purport to show that the Vulgata, if indeed not from Bh itself, derives from the same source as Bh, irrespective of the relation of R either to Bh or the Vulgata.
a. 78 paiņņa (prak Trņa or prat Trņa) in K, W, B, Y and P. The word be explained as an emendation (i.e. innovation) for pahalla in Bh (yahalla in Bho). This pahalla is a clerical error for pahulla, for which see T and compare pa(p)phulla in Ti and Tp. The development of pahalla into painna may be reconstructed as follows: pahalla + pa(y)alla + pailla (with -i- for -ya-) + paiņņa (with-on- for -11-). For the intermediate form pa(y)alla, see paala in S, and for pailla, see R. Note in this connection that in Jaina Nagart - 11- and -un- are practically indistinguishable, so that the confusion is easily understood. For an attempt to assign a meaning to pahalla, see Bhuvanapāla: pahalladolitam (ISt., p. 51), and Hem. IV 117, where pahalla is mentioned as an Adeśa for Sanskrit ghūrņa-, i.e. as if it is paholla
b.
156. Suharasa (sukharasa) in K, y, Y and P (cf. R). The word may be explained as an emendation (i.e.innovation) for saharasa as found in Bh, which seems to be a corruption, through the transposition of -ra- and

Page #63

50
-ha-, of sara(bhasa found in Ma, Tp, T and S and in the commentaries x and G (Pātha).
In the second hemistich of 237 in), Y and P the following readings are found: u niunasah Tkaralambiakaravalaandoliri, y niunasah Tkaraavalanbiakarabalaaṁdoliri, P niunas aht(ku)karalāvala)valaando. It seems impossible to extract any suitable meaning from these compounds. In the last parts, vPage #64

51
consist of fragments only, namely K of 1-370 (Ed. 1-365), B of 1-129 (Ed. 1-127) and P of 1-300 (Ed. 1-297).
All these five MSS seem to go back to one single source. This follows from a number of things, e.g. the innovations they have in common (see 4.1 c.) and those changes in the order of the Gathas they have in common vis-à-vis the archetype and the Jaina-recension (see 4.2).
In order to determine the relations between the five MSS the innovations may be considered. At this stage the definition of the term innovation as given on p. 29 above may be broadened so as to include those readings typical of the Vulgata MSS against readings common to Bh (and R) of the Jaina-recension and Ma, Ti and Tp of the Third SouthIndian recension. It appears then that the first division has taken place between K and y, each being characterized by a number of innovations against retentions in the other (see under a. and b. below). It is to be noted that K in this respect seems more conservative than y. These two branches seem to have been combined to form a third, represented by Y and P, which is characterized by those innovations typical of K as well as by those typical of v. Besides, y and P have a number of innovations of their own (see under c. below).
MSS Y and P seem to represent a South-Indian branch of the Vulgata, both MSS hailing from South-India. In the case of P this appears directly from the fact that it is written in the Kannada script. Regarding y, which is written in Devanāgarī, Weber (Ed., p. 2) has shown that it was copied from an original written in a South-Indian script. This follows, among other things, from such orthographical peculiarities as -ms- for -SS-, and -hn- and -hm- for -nh- and -mh- respectively.
The short fragment B shares with those retentions typical of the latter (with two exceptions, for which, see under g. below). Furthermore, B and have a few innovations of their own (see under d. below).
From these facts it is possible to conclude that the first division took place between K and y. Both traditions came to be combined in a MS, which formed the basis of Y and P (the possibility that y derives directly from the present MS P, or vice versa, can be excluded as each MS has a few innovations of its own). B would go back to an immediate predecessor of y. In this connection it should be mentioned that the possibility that B was copied directly from the present MS may not be

Page #65

52
ruled out, as of W and B only the latter has innovations of its own (e.g. kujjhai for j(n)ūrai in 38).
These conclusions are, however, of a very tentative nature. For instance, the conclusion that y and P go back to one and the same source is contradicted by such instances where y and P differ, the former agreeing specifically with K and the latter with y, and vice versa (see under e. and f. below). It has also been argued that B derives from the same source as v or was copied directly from the present. MS W. On the other hand, in at least two instances B differs from w and agrees specifically with K, Y and P (see under g. below). These instances would seem to show that the Vulgata MSS or their respective sources were considerably influenced by each other.
Another problem is that K as well as v occasionally have retentions in common with Ma, Ti and Tp (and T) against innovations in Bh, R, Y and P. E.g. 316 ahava in K, Ma, Ti and Tp against suhaa in v together with Bh, R, Y and P, and 26 sahasutta in v together with Ma and Tp agains suhasutta in k together with Bh, R, Y and P. These instances suggest that in K and u or in their respective sources authentic readings were. secondarily introduced, which, in turn, casts doubt on the status of the retentions in general in these two MSS (see a. and b.).
It seems therefore impossible to arrive at an exact definition of the stemmatical relations existing between these five MSS. The diagram on p. 56 is indeed not a stemma; it merely purports to give an idea of the grade of the authenticity of the text of each MS vis-à-vis that in the other four.
a.
U, B, Y and p have the following innovations against retentions in K (i.e. readings K has in common with Bh, Ma, Ti and Tp). Below, of the other MSS only R is occasionally mentioned, namely where it differs from K.
20 puņot, B, Y, P against uņo 84 tujjha kae , B, Y, P against tuha kaeņa 106 dTsai gose v, B, Y, P against gose disai 111 saṁllāvia y, B, paavia y, sallāpalaavisam P against ullavia 120 māruehi(), B, Y, P against mārueņa (Ti, Tp; Gāthā missing

Page #66

in Bh and R) 127 vi niuņo , B, Y, P against a niuņo 153 phālijjaṁtammi, hiaammi , Y, P against phālijjartassa,
hiaassa (hemistich missing in R) 180 sāmalijjai U, Y, P against sāmalte 184 mannus amuppaena V, Y, P against mannusamuppannena 187 karihimi , kāhimi Y, mi p against kāhań
padhamuggaa W, Y, P against padhamullaa 206 kuddalinião X, Y, P against the lihiāe K, tie lihirie R
(tte linião Ma, Tp) tahimmia gae tão W, tahimmi gae tão V, tammi dahmi ta etão P
tahiś cia coriāe K, taha ccia coriāe R 209 diahārambhe cuộvanavalie , Y, diāhassar abhe cumbhanavalie P
against viadiahasavilakkhalakkhie 213 rahasujjalacunvanaṁ , Y, P against rahasuvvellacumvaņań 217 ahilihia , Y, P against alihia 218 padhamuggaa , Y, P against padhamullaa 219 asakkha , Y, P against asajjham 236 ghadiāin , Y, P (Sanskrit) against ghadiāna (genitive),
hiayānań R 263 pattaaphalāņa sarise , Y, P (Sanskrit:ophalānāí sadrse)
against pattakālasāride K, pattayaphalasāricche R :..293 virūo vi W, Y, P (Sanskrit virūpo'pi) against virūo a
b.
K, Y and P have the following innovations against retentions in y:
19 hiaatthian K, Y, P, B(!) against hiaacchia 31 sā K, Y, P against se (B) 33 mae (instrumental) K, Y, P against mahań (genitive) (B)
paliviūņa K, P, oiūna Y, B(!) against palfvium v (+palTvian K) 35 kulavahua nia K, Y, P against kulavahu nia(!) v, B (kulabahū
niaa R) 55 saṁk (h)iņi K, Y, P against kamkhiri (B) :56 tena tană K, Y, P against taha teņā (B) 123 rãeņa K, Y, P against vāena (B)

Page #67

131 kahin pi K, Y, P (T, S) against tahi pi (cf. tah in va Bh,
Ti, Tp, tahin ca R) 205 onaa Y, P, onatta k against oatta
Common innovations in y and P:
72 vahula (vahua) 76 dohaggam (doggaccan) 82 imea (unmetrical) 88 airikkammi vi (vairikkammi vi) 90 eānam Y, enam P (ettahe) 102 parirakkhanekkadinna v(cf. R), parirakhkadissa P (rakkhaṇekka
dinna) 112 tti kira Y, tti tira P (tti ia) 122 dupparigejhjham (-) Y, duparigajja (-) P (dupparialla pi)
arannan (rannam) 132 kim nu Y (cf. S), kim nam P (kim tha) 137 viņimilia Y, viņimtlia P (samTlia) 138 dajhjhihii Y, dajjahii P (khajjihii) 151 rehai (disai) 177 jaha (-) na taha Y, jaha ma na taha P (...taha na) 181 maulain Y, māulāin P (mauain) 201 jaha(m) tań (taha tah) 210 te ahman Y, te appāņa armpām P (amh an te) 262 sundara Y, P (Sanskrit) (vālaa)
Common innovations in ♡ and B:
6 suhaa (--) suhasai U, suhaa hasai tuha B (with suhaa coming
first) 22 luppa (+ tuppa litta) 36 saajjhia a duggaā (saajjhia duggaã a) 41 bharie (cf. R) (maie) 45 aipavas iesu diahesu (cf. I and s) (ņiccapahiesu diahesu)

Page #68

55
68 aippahāammi (aippahāe a) 78 bhasarali , bhasarāvali B (unmetrical) (bhamarāli) 91 vara (pia)
Instances where y and P differ and y agrees with K and P with y:
51 sugandhi Y, k' (cf. 7.) against suamdha P, \ (cf. K?) 178 tujjha virahe'Y, K against tujjha a(j)aso P, (cf. T) 200 genhaha Y, K (cf. s). against mahda P, W (cf. Bh, R) 241 tanut vi Y, tanuï vi k against taņus (-) P, U 274 sabbhāvaramiain Y, K against sabbhāvaneharamiain P, V (cf. Bh;
GTti-metre)
Instances where y and p differ and y agrees with
and p with K:
17 ehai so upaduttho Y, ehii so vi pauttho W, B (cf. Bh)
against ehijjai so pauttho P, K (cf. R, Ma, Tp, S) 22 pariunvana Y, (cf. B, Ma, TP, T) against pariunvia P, K
(cf. Bh, R, S) 72 chattha maggai Y, V (and all the other MSS) against maggai
chattha P, K 76 vimuho Y, V (and all the other MSS) against vimăno P, K
::
Instances where B differs from
and agrees with K, Y and P:
19 hiaacchia w against hiaatthiah B, K, Y, P 33 paltvium v against palTviāna B, K, Y, P
Note furthermore:
7 vania B, vaniāu y against vilaão 46 suņņai K, v?, p against suvvai 42, B, Y

Page #69

56
Diagram showing the relation between K, W, B, Y and P:
4.6 The contaminated state of the text of MS R of the
Jaina-recension
MS R, together with MS Bh, constitutes the so-called Jaina-recension. Bh is incomplete, missing the last Sataka. The text is accompanied by a Sanskrit commentary, called Chekoktivicāraltla, which is attributed to a certain Bhuvanapala. It may be noted that the text quoted in the commentary (Bho) is occasionally slightly different from the text proper (Bha). MS R is complete, having altogether 710 Gathās (but Gathā 14 occurs a second time, namely as 386).
In the first place there is evidence which suggests that R derives from the same source as Bh (the possibility that R derives from Bh itself, or vice versa, can a priori be ruled out as each MS has innovations of its own; for Bh, see, for instance, samappanti in 61, and for R, see 4.1 b.). This appears from the fact that they have in common certain innovations against all the other MSS (see 4.2 b.). It also follows

Page #70

from the changes in the order of the Gathās by which they together are distinguished from the archetype together with the Vulgata (see 4.2).20 Note in this connection also the following peculiar features found in the outward appearance of the two MSS. In both MSS the first half of 482 and the second half of 602 have been combined to form one Gathā, 527 in Bh and 533 in R. The Gathā so formed does not seem to make sense. Note also that Bhuvanapala's commentary on it is lacking. In Bh the last two words of Gātha 187 (Bh 387) are missing: sāhasakāriyae (hiaath tharatharei). The same lacuna must have occurred in the original of R as would appear from the peculiar reading found: sāhasayāe ghariņThiyayam tharatharei.
Above (4.1) it has been shown that Bh, R and the Vulgata are characterized by a number of innovations against retentions in the Third SouthIndian recension. Furthermore it has been shown that R has a number of innovations in common with the Vulgata against retentions in the Third South-Indian recension and Bh. These latter innovations would indicate that what later developed into R, on the one hand, and into the Vulgata, on the other, initially represented one tradition, which must have split off from a MS which was also the source of the present MS Bh. At the same time, though, R has some of those innovations which are typical of
(and y and P) (see under a. below) and some of those which are typical of K (and Y and P) (see under b. below) of the Vulgata MSS.
a... Innovations in R and W, Y and P against retentions in K:
176 phukkato R, X, Y, P against oastaa K, Ma, Tp (phūmitti Bh)
cuivato R, X, Y, P against oamtaa K, Ma, Tp (curħbamāna Bh) • 183 hiyayão R, W, Y, P against hiahi K etc.
233 naharaṁka R, W, Y, P (Sanskrit) against dasaņańkā K, Bh, Ti, Tp
Innovations in Rand K, Y and P against retentions in y:
147 sāsan R, K, Y, P against sāse , Bh, Ma, Tp, T '170 palottata R, K, Y, P (T, S) against palotta y , Bh, Ti, Tp 199 payadiyasineha R, paaliaņeha K, paadiaŋeha Y, padianeha P

Page #71

58
against paadianeha
, Bh, Ma, Tp
Thus R contains material, as does Bh, which anticipates the Vulgata as a whole (retentions with Bh) and material which represents the last stage in the development of the Jaina-recension (innovations with Bh), on the one hand, and material which represents the first and the last stages in the development of the Vulgata (innovations with all the Vulgata MSS and with y and K, respectively), on the other. This state of affairs is explained by the fact that the text in R is the result of a process of contamination. For traces of this process in R, see, for instance, pattayaphalasāri(c)che in 263 which is a combination of pattapphalasāricche of Bh (and Ma, Ti, Tp and k(!)) and pattaaphalāņa sarise of v, .. Y and P (Sanskrit). Besides, in the order of the Gathās R has a feature typical of Bh next to one typical of the Vulgata. Thus, in R Ed. 14 occurs both as 14, as in the Vulgata (14 K, P, E, G, 13 B, 124, Y), and as 386, as in Bh (382). In the former instance the readings agree with the Vulgata (note chittaṁ), in the latter with Bh (note chikka):21 It should be noted that the occurrence of the Gātha in question as 16 in the Third South-Indian recension indicates that the Vulgata has preserved the more original order here.
An explanation of the contaminated state of the text in R should start from a MS closely related to Bh. This would account for the occurrence in R of the retentions it has in common with Bh. It is unlikely that these, as the more difficult forms, would have been entered in a Vulgata MS for the innovations, by definition the easier forms, typical of the latter branch. This assumption dispenses with the necessity to suppose a complete rearrangement of a Vulgata MS in order to make it follow the order of the Gathās of the Jaina-recension. This MS was apparently thoroughly revised by comparing it to one or perhaps more Vulgata MSS. On this point there are basically two possibilities. The first one is that R was compared at least twice, first with a MS like K and next with a MS like y. The second possibility is that was compared only once, namely with a MS belonging to the Vulgata branch represented by Y and P. In the latter case it must have been a MS representing that branch before it became divided into y, on the one hand, and P, on the other, as R does not have any of the innovations common to these two MSS.

Page #72

--
-
-.
..
ψ
Β
γ
PKR
Bh
Ma
Tp
Ti
Diagram showing the relation between the MSS of the Third South-Indian, Jaina and Vulgata-recensions

Page #73


Page #74

CHAPTER 5
5.1 The eclectic recensions: Sãdhāraṇadeva, Telinga and
Second Telinga
These three recensions are eclectic in the sense that the text of each of them combines material of at least two already independently existing traditions. In the following paragraphs I shall try to ascertain which traditions were combined in each recension.
Among the three recensions the Second Telinga stands out as it consists of only 104 (or 102) Gāthas. Weber (Ed., p. XLII) has already suggested that it probably concerns an independent anthology here, which means that the fact that most of its Gāthās are also found in the Sattasat would be accidental. However, as most of the Gathās also occur in the Sattas at it seemed interesting to try to find out from which recension or recensions of the Sattas at they originate.
In contrast to the compilation of, for instance, the Jaina-recension and the Vulgata, the one in the Sadhāranadeva and Telinga appears to have been a very deliberate procedure. Thus, the 700 Gathās of the Sadhāra adeva-recension have been arranged topicwise into altogether 60 chapters or Vrajyās. The extent of the Vrajyās varies from 74 Gathās in the Asat T-vrajyā to 1 Gatha in the Veśyā-vrajya. The principle according to which the Gathās have thus been classified varies too: in some cases it is the type of Nāyikā, as in the Virahin T-vrajyā, in others a specific poetical device, as in the Jāti-vrajyā.
The 700 Gathās of the Telinga-recension are divided into seven satakas. They are further classified into 26 chapters, here called Paddhatis. The extent of the Paddhatis is generally not indicated in the interior of the text, but their names are mentioned, in Prakrit, at the end of each Śataka. In two cases the Paddhati coincides with the Sataka, namely the Viraha-paddhati, which comprises the whole third Sataka, and the Pranayakalaha-paddhati, which comprises the whole sixth Sataka. The other Satakas contain each more than one Paddhati.
While the Paddhatis provide a broad division of the Gathās according to erotic situation, poetical device, etc., within the Paddhati the Gathās have been arranged meaningfully as well. For instance, in the

Page #75

62
Viraha-paddhati, the whole third Sataka, several smaller groups can be distinguished. E.g. 201-213 deal with the time immediately before the actual separation. 214-223 describe the traveller; the Gathās have been arranged according to the season: 215-217 spring, 218 summer, 219-220 rainy season and 221-222 winter. Gāthās 224-237 describe the Virahinī. 238-246 are, again, arranged according to the season: 238-239 spring, 240 summer and 241-246 rainy season. Gāthās 297-299 describe what the husband said to his wife on his return. In 300 after the union separation threatens again. A similar arrangement may possibly also be present in the Vrajyas of the Sadharanadeva-recension, especially in the larger ones.
5.2 The Sadharagadeva-recension (S)
In Weber's material this recension is represented by only one MS, namely S, dated Samvat 1672 (1616 A.D.), which beside the text. of the Gathas contains the Sanskrit commentary by Sadharanadeva. The MS is written in Devanagari, but Weber (Ed., p. XLI) has suggested that it goes back to an original written in the Bengali script. This would appear from the occasional confusion between -r- and -v-, and from the exclusive occurrence of -v- in favour of -b-. He furthermore suggested that the recension itself should originate from Bengal. In support of this thesis Weber refers to the Bengali Viśvanatha (fifteenth century A.D.) who in his Sahityadarpana, verse 565, discusses a type of Kosa in terms applicable to our S-recension, for which, moreover, he mentions as an example a Muktavali, which is the name given to the text in MS S (see Ed., p. XXXVIII).
The eclectic nature of S becomes clear from the fact that it has readings otherwise typical of the Jaina-recension (Bh and R) (see under a. below) beside those typical of the Vulgata (with or without R) (see under b. below). Furthermore, it has readings typical of practically every individual Jaina (see under c. below) and Vulgata MS (see under d. below).
In order to explain the present state of the text in MS S one has to start from a MS of the Jaina-recension. This would account for the occurrence in S of the retentions, however few, it has in common with Bh (see 4.1 b., Gāthās 49, 78 and 156). It is unlikely that these

Page #76

63
would have entered into its text secondarily. Its dependence on a MS of the Jaina-recension becomes furthermore clear from a comparison of the texts of the respective recensions. S has 644 of its 700 Gāthās in common with the Jaina-recension (see under e. below). Subsequently S would appear to have been influenced by all the other Jaina and Vulgata MSS individually. It is, however, also possible that some of the latter MSS were, instead, influenced by the text of S. Only in a few cases can the direction of the influence be ascertained. Thus, the reading pat viesattho in 550 in s would indicate that s was influenced by Y, as the reading pat ghare natthi in the latter MS can be interpreted as an 'emendation' (i.e. innovation) for patņaasurāe (i.e. pat na asure) in y. It is doubtful, though, if this conclusion may also be applied to all the other cases where S and y agree. paviralo, for pariviralo, in 329 in both S and K would on the other hand show that K was influenced by S. It should be noted that in the Bengali script in which the source of the present MS S was written, the Akşaras -ri- and -vi- are practically identical so that the omission of the syllable -ri- can be more easily explained in s than in K, namely as an instance of haplology.
The text of S bears clear traces of this contamination. Consider the following cases:22
354 sunnabhamir īņa vāhulaiāņa + sarahasabhamiriņa vāhulaiāna K .. (surahasa!), W, Y, Ma, Ti, Tp x sunnovās abhamirTņa vāhāna Bh, R 485 hiaan hiaanihian + hiaan hiaanihitta Bh, R, Ma, Ti, Tp *
hiaam hiae nihian W, Y 541 paçaṁtavānasalilanivahena (GTti-metre) + padatavāhanivahena
Ma, Ti, Tp x vāhasalilanivahena Bh, R, etc. .556 pattā niahvapphamsā + pattā ņi ambaphaṁs aṁ Ma, Ti, Tp x patta
n iavapphaṁsā U, Y, R, onva(p?)phaisā Bh
S has furthermore a number of innovations exclusively in common with the Telinga-recension (T). There are reasons to assume that I was influenced by S, and not the other way around. This problem will be discussed in the next paragraph (5.3).
It is noteworthy that s includes practically no readings characteristic of the Third South-Indian recension. Exceptions are: 465 pasijjiri

Page #77

and 616 luliakesāsu miliacchīsu (order). See also 541 paçantavāhasalilanivahena and 556 pattā ņiamvapphaṁsā mentioned above. As some of these instances possibly are retentions it may be that s was, among others, also influenced by a MS of the Jaina-recension the text of which was in certain respects more authentic than that of either Bh or R.
Readings in s typical of Bh and R:
9 mā ruasu, 136 hoi dānarahiassa, 142 dukkhasokkha, 212 accharian, 223 uvvahai vahū jaha jaha, 224 alam me, 238 vikkei, 314 mauliań, 327 vaddhenti S, R, vaftaħti Bh, 393 chippai, 400 rūsai ccia, 416 ajj(h)a, 432 na niruddho, 444 maulan S, R, Bho, muindalaṁ Bho, 465 vahūņa sokkhāna S, R, angāna vi suhāna Bh, 517 visavad (/y) antañ, 524 de dāuń S, je dāum Bh, je dār aṁ R, 579 annanafsu, 583 bhāhi S, R, hohi Bh, 586 siharammi, 691 gavvio (Bh not available).
Readings in s typical of the Vulgata MSS, with or without R:
s includes nearly all the innovations found in K, W, B, Y, P and R (against the retentions in Bh, Ma, Ti and Tp) enumerated in 4.1 b. and nearly all the innovations characteristic of the Vulgata enumerated in 4.1 c. To these may be added the following readings which S shares exclusively with and y, where K and P are not available:
409 ekko panhaai thaṇo vto pulaei nahamuhāviddho, 438 atiujjua S, aiujjua , aiajjua Y, 466 de suaņu, 474 visão S, vesi(/a)nto , Y, 483 tammai hiaammi (locative) (cf. T), 501 vicchāař, 512 ahaan, 520 jo mam aiā(a)rena, 589 taħvira, 595 addaṁs ano, 614 esa guņo, 680 vannalohitae (cf. R vannalohillo), 683 mā eam vvia S, mā etti ccia y, mā ettia bia y.

Page #78

65
Readings in s typical of Bh:
212 so tań, 217 jan mittaṁ vasaņadesa(k)ālammi, 529 paiņo bhuasiharań.
Readings in s typical of R:
196 pari(t)utthā, 586 vialā (cf. Bh viullā), 621 bhamai, parimalai, mu(mc)ai.
d.
Readings in s typical of the Vulgata MSS:
S shares a number of the innovations characteristic of U, B, Y and P (against retentions in K) enumerated in 4.5 a.:
127 vi niuno, 153 phālijjantammi, hiaammi, 180 sāmalijjai, 187 padhamuggaa, 206 kuddaliniai, 213 rahasujjalacunvanaan, 218 padhamuggaa, 219 as akkhan, 236 ghadiais.
S goes with K in 20, 84, 106, 111, 187 (a), 209, 217 and 293.
Readings in 's typical of K:
185 sans itto, 186 sisirāi(), 214 aisochaāla s, aisuira K (ciraāla), 220 dinnatarala, 228 suhaa jaṁ ca, 238 mammurasachahe, 285 vialu', samatthā.
Readings in s typical of :
152 aņuņehi, 360 onaãi, phalahte (singular), 409 āviddho S, alitdho , 441 vijjha ham S, ha vijjā v.

Page #79

Readings in s typical of Y:
15 padhamuggaa, 307 gaņaņāhi(m), 323 laggan, 450 suhaa, 487 na dei sivine vi, 550 pat viesattho S, pas ghare natthi Y, 611 va gotthan.
Readings in s typical of P:
105 kajā vi S, karai mia P, 116 dhanummi nisannā S, dunumi nisaņā P, 155 ņa āņaṁti.
Readings in s typical of .y and P:
122 (duppariallam) nei ar aŋŋah (cf. T), 132 kiṁ nu S, Y, kim ņań P.
Readings in s typical of y and B:
45 aipavas iesu diahesu (cf. T).
The numbers in this section are those of s (see Ed., pp. 433-449). A comparison between the Sadhāraŋadeva-recension and MS R of the Jainarecension shows that the former includes practically all the Gathās of the latter. Of the 56 Gathās not occurring in R, 11 are found exclusively in S: 50 (Ed. 800), 69 (801), 105 (802), 150 (804), 223 (805), 365 (806), 404 (807), 446 (808), 681 (812), 699 (813) and 700 (814). 10 are found exclusively in the Vulgata (with x, unless indicated): 4, 5, 71, 74, 131, 133, 313, 409 (-x), 508 and 658; 4. in the Vulgata (with x) and T: 447, 586, 573 and 646; 20 in the Vulgata, Ma, Ti and Tp, and T: 51, 98, 107, 172, 207, 247, 265, 275 (also w), 358 (also W), 366, 368, 471 (also W), 482, 507 (also W), 509, 514 (also W), 582, 637, 688 (also W) and 698 (all also in x, except 688; for 582 x is not available); 3 in the Vulgata and Ma, Ti and Tp: 263, 583 and 635 (all also in x; for 583 x is not available); 3 in Ma, Ti and Tp and T: 148, 444 and 501 (also in x and W); 2 Gāthās are found exclusively in x: 291 and 512 (?); 1 in T: 411; 1 in Y, P and T: 590; and 1 in Y and x: 667.
A comparison of S with the Vulgata yields a practically identical

Page #80

result, s sharing 636 of its 700 Gathās with the Vulgata. Of the 64 Gathās not found in the Vulgata, 11 are exclusively found in s (see above). 13 are found in R: 24, 26, 44, 46, 49, 82, 260 (also Ma, Ti and Tp), 266, 433 (id.), 570 (id.), 627, 688 and 691; 2 in x: 291 and 512 (?); 6 in R and X: 27, 213, 258, 386, 505 and 521; 4 in T (and Ma, Ti and Tp): 148 (Ma, Ti, Tp), 411, 444 (id.) and 501 (id.; also in x and W); 14 in R and T (and Ma, Ti and Tp): 23 (Ma, Ti, Tp), 25 (id.), 45 (id.), 48 (id.), 50 (id.), 190 (id.), 321, 380 (id.), 412, 432 (id.), 497 (id.), 506 (id.), 557 (id.) and 563; 9 in R, X and T (and Ma, Ti and Tp): 28 (Ma, Ti, Tp), 29 (id.), 149 (id.), 224 (id.), 271 (id.; also w), 524 (id.), 626 and 692; 1 in Y and x: 667; 1 in Y, P and T: 590; Gatha 65 occurs in all the MSS minus v (K and P not available), Gāthā 169 in all minusy (K and P not available), and Gathā 191 in all minus y and x (K and P not available).
Above (4.2) it has been shown that the text of the Third South-Indian recension was compiled through a process of selection from the text of the archetype. Traces of the order of the archetype, as preserved in the Vulgata, on the one hand, and in the Jaina-recension, on the other, can still clearly be seen in the Third South-Indian recension. In the same way in s occasionally sequences of Gāthas -- whether interrupted by others or not -- of the Jaina-recension or the Vulgata can be recognized, however, without a clear patter emerging. Though, considering the thorough process of arrangement visible in s this should not be surprising.
5.3 The Telinga-recension (T)
In Weber's material the Telinga-recension is represented by several MSS. The Prākrit text is based on a collation of three MSS, one written in the Telugu script (To), the other two in Grantha (Weber's Te but here Tabelled to to avoid confusion with TC). The text is accompanied by an anonymous commentary.
In the Telinga-recension the tradition of the Third South-Indian recension has been combined with those of the Sadhāraṇadeva-recension, the Vulgata and MS R of the Jaina-recension. Its dependence on the Third South-Indian recension becomes clear in the first place from the fact that T has many readings otherwise exclusively found in the latter recen

Page #81

68
sion (see under a. below) and in the second place from a comparison of the respective texts. It appears, among other things, that T has 101 Gāthās exclusively in common with the Third South-Indian recension (see under b. below). T agrees with the Third South Indian recension in that it, too, has certain forms which point to an original written in Jaina Nägart (see under c. below).
T has a relatively large number of readings exclusively in common with S (see under d. below). Besides, it has a few readings in common with the Vulgata and with some individual Vulgata MSS (see under e. below). In this connection it should be noted that T also shares a number of readings with the Vulgata together with S (see under f. below). Finally, I shares a number of readings with R; and with R together with the Vulgata MSS and S (see under g. below). That T was influenced by these respective traditions and not the other way around would appear from the absence in either S, the Vulgata or R of any of those readings typical of T and the Third South-Indian recension. In this connection it may be noted that during my visits of manuscipts libraries in India I did not come across any North-Indian Devanāgart MS of the Telinga-recension or, for that matter, of the Third SouthIndian and Second Telinga-recensions, while in Trivandrum I was shown a South-Indian MS of the Vulgata with Gangādhara's commentary. This evidence should, however, not be stretched so far as to mean that there never were MSS of the South-Indian recensions in North-India. Furthermore, the manuscripts libraries contain only the top of the iceberg of the available MSS in India.
Above (4.3) it has already been shown that Tp of the Third SouthIndian recension was influenced by T.
T has many of the retentions of the Third South-Indian recension against innovations found in the Jaina-recension and the Vulgata, enumerated in 4.1. Besides, it has many readings otherwise exclusively found in the Third South-Indian recension. E.g.

Page #82

69
2 pāada, kaha te, 104 ciae, 127 ummatthaāi, halão (cf. W), 154 kunejjāsu, 156 bahumagganimmiāiỉ vesāvilaāna pemmā in, 175 golānaikacchakuduga, 177 kalabā, 204 nivvuijjae, addāna va padibimban (order), 250 appian, 253 dummuhao, 255 nivvui an, 257 haliasuā, 259 nadin, 268 sahāvasarala pi thanajuar padian, 284 manań, 291 dimbham anulaṁ(/g)giri, 305 dukkhena, 335 jaggassu (Ma, Tp jaggāsu), 354 tuhņikkaparungeņa pi ajja māŋamsinimuhena, 358 hiaammi, 377 nii, 389 uvvea(m)roirte, 390 māṇassa, 393 kim tattio, 405 kamalahaena jaṁ viasasi, 409 ekkam passasai thana bfam pulauggamena phuriamuhan, 429 agghia, 441 jammastarāi vi saaṁ calanań, 497 ja maā să maā, -516 piavirahe, 521 pagtajualati, 523 tue, 546 vedhă(1)la, 549 maddana, 559. maulantacchãi, 573 ajj(h)ā, 586 vilola, 589 kāūņa, 590 sikkhan, 593 savvo, 598 majjham pat kuanda (m), savattt a (hemistiches reversed), 607 dhakkesu, 643 darapikkakavitthagadhalohena, 647 samahiana T (samihian Ma, sumahia Tp), 649 vipphurai, daiena (cf. W), 651 jāi (cf. W), 653 dfsatadiţthisuhao, ullāventasuisuho, 658 asāvattie, 660 vahai, 677 pasāanaṁ (T paāsaņaṁ), 685 muhalo, 689 annaṁ gāmam pattā, 724 ukkhippei, 747 dhavalia ti kauņa, 756 aghadia, 779 so matthae padai.
The numbers in this section refer to the Telinga-recension, for which, see .Ed., pp. 450-501. Of the 700 Gāthas of the Telinga-recension 28 are exclusively found in this recension (221 also in x) and 14 only here and in W (660 also in x). Of the remaining 658 exactly 101 are otherwise only found in the Third South-Indian recension; some of these also occur in W. Of the reman ing 557, 424 occur in all recensions, i.e. the Vulgata, in which x has been included here, in the Third SouthIndian recension, in R of the Jaina-recension and in S. Of the remaining 133, 35 are found in all except the Third South-Indian recension, 22 in all except R, 16 in all except S and 18 in all except the Vulgata. 3 are found only in the Third South-Indian recension and R, 3 in the Third South-Indian recension and S, 8 only in the Vulgata and R, 4 in the Vulgata and S, 8 in the Vulgata and the Third South-Indian recension, 6 only in the Vulgata, 1 in R (and x), 6 in R and S, 1 in y, P and S, and 1 in y (and W).

Page #83

70
C.
There is evidence in T which points to a source written in Jaina Nāgari; see mastu for mannu in 683, 938 and 941.
Ta which is written in the Telugu script shows numerous traces of having been copied from an original in Grantha. E.g. -i- for -u- in 297 haŋiim ((b)hanium), 298 dijjai (dijjau), bhani iṁ (bhanium), 445 pāisa (pāusa), 454 ukkhuviin (ukkhivium), 484 valiin (valiun), 685 jāmaissa (jāmaussa), viiņei (viuņei), 759 rāmei (rāmeu); --- for --- or -s- for --: 70 pāuda (To pāudū), 538 id. (To paulo), 578 id., 526 paedam (paes am), 392 kudumbi (kusumbhi), 546 kudumbha (To kusumbha) and 388 pasuria (paźduria); -h- for -bh-: 4 hisiņi (bhisiņi), hāaņa (bhāana); and -9- for -bh-: 546 garillTo (bharillto) and 722 garaeņa (bharaena).
Readings T has exclusively in common with S:
146 avah Tri(a), 164 sukha T, sokkhan S, 216 tujjha, 254 sutto T, saio S, 277 te ciraṁ runnah (cf. P), 297 enam, 345 dūmesi, 350 saan cia, 363 sahanijjo, 371 kim ta na, 383 khuha (19), 419 lajjāluint vi vahū, 444 pāvai Tsi vi T, tsi vi pāvai S, bhasala, 487 se hian, 514 kulavahūņań, 527 gao ccia, 531 tattha, 542 hiaammi, 546 mettto, 565 vaaņehin, 573 ditthinah, 586 uppakkha T, purkham S (-kh-!), 647 neii T, naai S, 745 pasāamāneņa, 756 bolaṁte.
e.
Readings T has exclusively in common with the Vulgata:
451 osarasu, 484 tte valiun, 524 (-) vahuvallaha, 566 pikka.

Page #84

71
Readings T has in common with some individual Vulgata MSS:
with K: 235 samaam chia kim vihina, vinimmivia, 265 dese. with Y: 643 alehaladdua.
Readings T has in common with (K), 0, Y, (P) (Vulgata) and S:
142 pariaddhāna, 205 pāsāsakī kāā necchai, 278 sovāṇavant te, 310 mottūņa, 316 kiñ tena sineho, 319 muhāhi, 361 ahmāristhi, 365 mammaraan, 380 sappurisă, 388 samudduttarata, 444 maar andapāṇalohilla, 497 vārei mat (cf. Tp).
Readings T has in common with S and some of the Vulgata MSS:
S and K: 177 (kalaha) ja ma taha na T, S, jam at taha nah K, 354 mănas ini, 356 amha. S and Wand B: 45 (s)aiposiesu diahesu. S and Y and P: 122 (dukkhaallam) nei arang an.
Readings T has exclusively in common with R:
243 jānaṁtassa vi, 268 hiae ko nāma, 291 pălāata, 297 paino turiai, 543 edan Ta, eta To, eyath R, 559 likka, 589 avirala, 620 ekko mao T, ekkamao R, 669 mā pahia, sejjāe.
Readings T has in common with (K), X, Y, (P) (Vulgata), R and S:
90 khamijjanti, 136 horft t, 152 lohillań, aigahiagarua, 153 padihãi, 163 jivijjai, 170 palottanta, 188 pautthamāṇam, 198 jampamāṇā, 290 vilTno, 320 vi na tahā, 353 vi na taha (R vi na taha), 368 thanaalasuppellana, 463 kunai kisa, 575 ānandiali.

Page #85

72
5.4 The Second Telinga-recension (W)
In Weber's material the so-called Second Telinga-recension is represented by two MSS, the one written in Grantha (W), the other in the Telugu script (V). The former contains the text of the Gathas accompanied by a Sanskrit commentary; the latter contains a Sanskrit Chaya only. The beginning of W, comprising the first ten Gathas, was illegible.
The recension is markedly different from the other known recensions in that it contains only approximately 100 Gathas, namely 102 in W and 104 in V. The text bears various names. In W is is called PrakṛtaśṛngāraSataka and in V Prakṛtasataka. Yet another name seems to have been current. In Upadhye's edition of this recension it bears the title SaptaSatisara.23 In this latter edition the commentary, which probably is the same as the one found in W, is attributed to a certain Vema Bhupala. This Vema Bhupala has been identified with a Reddi king of that name who ruled from 1403 to 1420 at Kondavidu in Andhra (Krishnamachariar, 1970: 479; Sriramamurti, 1972: 68-69).
As already said above (5.1). Weber suggested that the Second Telingarecension is possibly an independent anthology, having been compiled independently of the recensions of the. Sattasai. This would, among other things, appear from the fact that the compiler apparently had no intention to create a 'sevenhundred'. The best proof for the assumption that the Second Telinga-recension originated independently of the other known recensions of the Sattasaf, would be to be able to show that there was another source from which the compiler drew the Gathas. In this connection it should be noted that a considerable number of Gathas (17) are also among those Weber collected from the Alamkarasastras (see Abh., p. 202 ff.; collectively labelled 'A' here) (7, 15, 19, 39, 66, 68, 76, 83, 85, 87, 93, 96, 98, 99, 100, 101, 102) and some even exclusively so (68, 93, 99, 101). In some instances the readings agree more closely to the latter. E.g. 875 anaddamana W and A (+Sanskrit; anollamana T), 860 dūna W, düni A (+Sanskrit; domia T), 939 nimiliaamh W, milia A (pimiam T), 951 kaccT a W, kattT a A (kattTo T).
From whatever source or sources the compiler of the Second Telingarecension drew the Gathas, most of these also occur in the available recensions of the SattasaT. A relatively large number of them otherwise

Page #86

occur exclusively in the Telinga-recension, some of which are found in the Third South-Indian recension as well (see under a. below). The readings generally agree closely with those of the two other south-Indian recensions (see under b. below). Occasionally though it has readings typical of the other recensions (see under c. below). The text of W shows clear traces of contamination, for which consider the following examples:
33 unna viņisasamto as in Ma, Tp and T(?) beside kisa hma para
muh Te sayaņaddhe as in all the other MSS 237 sahiaasah Tanapasaria as in Ti, Tp and T beside karajuala as
in and y 688 vaccasu as in Ma, Ti, Tp and T beside imāna as in U, Y. and S.
a. In this section the longer form, with 104 Gathās, of the Second Telingarecension is considered (see Ed., pp. 502-508). Of the 104 Gathās 11 are found exclusively in the Second Telinga-recension: 41, 67, 68, 89, 93, 94, 99, 101, 103 and 104. 14 Gathās otherwise occur only in T: 3, 17, 19, 39, 66, 76, 81 (also in x), 82 (also in the South-Indian Vulgata MSS Y and P), 83, 85, 87, 96, 98 and 102. It shares 7 Gathās exclusively with T and Ma, Ti and Tp: 4, 8, 14, 29, 56, 77, 78. The remaining Gathās are found in all the other recensions or MSS.
b. Readings common to W and T:
33 petthan, 91 pasāerti, 167 bhana kaha ajja vi dhammia, 149 tassa saāsaṁ ghanandhaāre tti, 335 bahula, 372 dinno stsammi saha juānehin, 583 dhania(in) uvagūhia mae, 655 navarāaratta, kaaloa, gammiasavvaria, 664 anumannai jaha jāram, 669 attā atta Yamajjai, 938 tanno (Ma, Tp tannāi).
Readings common to w and the Third South-Indian recension:
33 unna vinisas asto, 190 devaullaań, amha, 237 sahiaasah Tana

Page #87

74
pasaria, 254 olambae vasais, 335 na amusajjāmo, 649 paripucchia na jampai.
Readings common to W, T and the Third South-Indian recension:
42 dharan, 127 ummattaāi, halão, 499 katto khemaṁ katto, mattaovari. (compare khujjamjao with kuvvajjao in Ma), 649 vipphurai, daiena, 651 jāi, 688 vaccasu.
C.
Readings in W not found in the two other south-Indian recensions:
33 ktsa hma par anmuhte sayaņaddhe K, 4, B, Y, P, S, 85 nikkaiavena jānam gao si k nivvur tānam Vulgata, R, S, 173 raia y 175 goātadaviadakuduñga Vulgata, s (Gātha missing in Bh and R), 476 ciasaan (+vidāaman Bh, R, Y; transposition of the Akşaras -da- and -a-, and -S- for --- (Grantha:)), ettha abhavvāe...kaha nu kāappa Vulgata, Jaina, s (Ma, Ti, Tp: kaha nu abhavvāe...ettha kāavvam; Gathā missing in T), 550 pas vi annatto Y, S, 554 eddahamettan(/g) game , Y, 688 imāņa V, Y, S (Gathā missing in Bh and R).

Page #88

CHAPTER 6
6.1 Concluding remarks
One of the results of the above study is the conclusion that all the available MSS ultimately go back to a single MS. While the text (readings) of this MS appears in many respects to be most faithfully preserved in the Third South-Indian recension, the order of the Gathās is in the Jaina-recension and the Vulgata. The remaining three recensions were compiled by selecting Gāthas from the former three and thus suppose their prior existence. This conclusion is of considerable importance for the history of the genre. The Sattasaf is generally considered the oldest anthology (Kośa) of the classical period (Sternbach, 1974: 10). In all later kośas the verses are divided into sections on the basis of categories developed in the Alamkāraśāstras and in the Nāyikābhedaliterature. A similar arrangement is also found in the Sādhāraṇadeva and Telinga-recensions, which, however, have been shown to be relatively late revisions of the text. However, in its original form, as represented by the Jaina-recension and the Vulgata, which is probably older than any of the other known Kośas, the order of the Gathās appears to be random.
In these two recensions occasionally two or more Gathās are connected by a similar content, idea or motif. Thus, Gatha *6 and *7 refer to two different trees, the Kurabaka and the aśoka, which blossom only after having been embraced and kicked by a young woman respectively. *13 and *14 describe a newly-married wife's clumsiness in the kitchen; and in *41 and 42 a woman who does not want to hurt the feelings of her husband and her neighbour respectively. However, no overall organization on the basis of content is visible. In fact, the Gathas seem to follow each other associatively, in which the occurrence of identical words seems to have played an important role. Thus, it appears that consecutive Gathas or larger sequences of Gathās are linked together through the occurrence of identical words (Tieken, 1978). A similar type of concatenation has also been recognized by Bloomfield (1899: 39) in Vedic literature, while Schubring (1955) has demonstrated its existence in classical Kavya.

Page #89

76
Another conclusion which follows directly from the above study concerns the problem of the authenticity of the tradition which ascribes the Gathās to individual authors. The occurrence of the names is restricted to MSS P (Vulgata) and Bh (Jaina), while a few names are also found in the beginning of MS K (Vulgata).
The earliest reference to the Sattasat is probably the one found in the thirteenth verse of the introduction to Bāna's Harşacarita (seventh century A.D.) (see Ed., p. XI). In so far as the names found could be identified they belong to persons living before Bāna. Thus, a significantly large number of names refers to kings; note those ending in -rāja, -adhipa and -sena. Some of these have been identified by Mirashi (1947) as kings of the South-Indian Sātavāhana-Andhrabhrtya, Vākāțaka and Rastrakūţa-dynasties respectively. The latest datable name appeared to be that of the Rastrakūta king Māna or Mānāňka (420-455 A.D.). The only positively post-Bāna author is Vākpatirāja, if he is indeed the same as the author of the Gauqavaho. Vākpatirāja lived between approximately 715 and 750 A.D. The name Vākpatirāja is, however, restricted to MS Bh, MS P having a different name.
On the other hand, there is no evidence for the presence of the names in any stage of the transmission of the Sattasaf earlier than the Vulgata. This appears from the following comparison between the names occurring in the Jaina-recension and the Vulgata.
In P the names, occurring in Sanskrit and in the genitive, are available up to the middle of the second Sataka, i.e. only as far as PTtāmbara's commentary is available (see Ed., pp. LIII-LV). The names are found after the Gathās. In Bh practically each Gathā is provided with a name, in Sanskrit and in the genitive (see ISt., pp. 9-16). They are found before the Gathā (ISt., p. 17).
Whereas P and Bh generally agree on the names they differ in the attribution of a particular Gathā to a particular author. It appears that in Bh the names have shifted one -- sometimes more than one -- Gātha forward. This is most probably the result of connecting the name with the Gathā which follows instead of with the one that precedes. That P and not Bh has retained the correct ascriptions becomes clear from those names which were created after some incidence described in the Gātha. E.g. Kālasāra, 'Black antilope', found in Bh with 24 (Ed. 26)

Page #90

should belong to 23 (Ed. 25).
In 4.2 I have shown that the Jaina-recension and the Vulgata go back to one source. Each is characterized by changes of its own in the order of the Gathās. Below I will consider two instances where the flow of the text was thus interrupted.
In Bh the nickname Pavana, 'Wind', is found with 521 (Ed. 498). This name more likely belongs to 450 (Ed. 497), i.e. the Gathā which precedes in the Vulgata, than to 520 (Ed. 627), the Gathā which precedes in the Jaina-recension. It looks as though the names found in Bh were borrowed from a Vulgata MS. It is also possible to maintain that when the order of the archetype, which is preserved in the Vulgata here (see Appendix III, 454, 337 and 412), came to be changed in the Jaina-recension, the name which originally belonged to the preceding Gathā was taken along with the following Gathā to the new position in the text. This latter explanation does, however, not hold in the case of Sthirasaha, 'Strong branch', found in Bh with 99 (Ed. 168). This name should belong to 367 (Ed. 167), the Gathā which precedes in the Vulgata, rather than to 98 (Ed. 96). From Appendix III (117) it appears that in this case it is the Vulgata in which the original order is disturbed, while it is preserved in the Jaina-recension. It follows that the names of the authors found in Bh were indeed borrowed from a MS of the Vulgata. 24
This also means that there is no evidence for the presence of the names in any stage of the transmission of the text prior to that of the Vulgata. This latter conclusion should, however, be treated with caution. In this connection it is to be noted that the names were, or were considered as, part of the commentary. Thus, in P they are available only as far as PTtāmbara's commentary is; in v and y, which contain only the text, they are completely absent despite their final Gathā which specifically says that the Gāthās are accompanied by the names of their authors (see Ed., Gātha 709). Furthermore, they were not considered an essential part of the text, as appears from the fact that in K they are given only for the first few Gathās. Therefore the absence of the names in the Third South-Indian recension, on the one hand, and in the Jainarecension, on the other, cannot be taken as absolute proof that they were absent in the archetype as well.
If despite these uncertainties the names were indeed added to the

Page #91

78
text only in the Vulgata, the question arises what may have motivated this addition. Above we have seen that in the Sadhārazadeva and Telingarecensions the Gathās are arranged topicwise, as is the case in most other texts of the genre. This revision of the original text most probably took place in an attempt to imitate those other kośas. Only a few of these Sanskrit kośas are original works. Most of them are mere compilations of verses from works of different authors, in which case often with each verse the name of its author is given. The addition of the names to the Gathās of the Sattasat may be explained as the result of an attempt to copy yet another feature of the Sanskrit kośas.

Page #92

79
Notes to the stemma
The Gathās aptasatt of Sātavāhana. With the Commentary of Gangadharabhatta, ed. by Durgāprasad and kasinātha Pandurang Parab, Kāvyamala 21, Bombay, 1889.
Saptaśatīsāra with BhāvadTpikā of Vema Bhūpāla, along with the Chappannaya-Gahão (text and chāyā), ed. by A. N. Upadhye, Shivaji University Sanskrit and Prakrit Series Vol. III, Kolhapur, 1970.
3
E.g. Hindt Gāthāsaptaśatí, ed. by Narmadeśvara Caturvedt, Vidyābhavana Saṁskṛta Granthamālā 55, Benares, 1961.
E.g. Hindi Gāthās aptaśatt, ed. by Jagannātha Pathaka, Kashi Sanskrit Series 192, Benares, 1969.
5
Weber's Abh. has been extensively reviewed by G. Garrez in JA., 1872, pp. 197-202. In 1872 Weber published in ZDMG 15, pp. 735-45, a short article called 'Ueber das Saptaśatakam des Hala'. It mainly served to announce the discovery of some new MSS, which were edited later in his 'Retractatio' (see below).
6 I use here the siglum T for Weber's Te in order to avoid confusion with T. I use C for Weber's G as he does himself in his edition of 1881.
For the first 365 Gathās of the Ed. the variants of the MSS edited earlier in respectively the Abh. and the Retr. have to be gleaned from these two publications themselves. This can be done only by a roundabout way, for the variants given in the Retr. refer back to the text

Page #93

of the Abh., which occasionally differs from that in the Ed.
8
80
Bhuvanapala is apparently otherwise called Tribhuvanapala. Another MS containing the text of the Sattasaf accompanied by the commentary by Tribhuvanapala is no. 76 of the Kaiser Library in Kathmandu, Nepal. The MS, consisting of 230 palmleaf folios, is written in the Old Newari script. It is at the same time the oldest datable MS of the Sattasaf, dated 341 N.S., i.e. 1221 A.D. Tribhuvanapala is the father of a certain Ajada, another commentator on the Sattasat. A commentary by Ajada is found in MS 385 of 1887-1891 in the Bhandarkar Oriental Research Institute in Poona. A short description of the MS is given by Sandesara (1972). The MS was filmed by me in the summer of 1976. It consists of a bundle of altogether 48 folios not forming a consecutive text but being a collection of parts of what once was a complete MS. The text is a specimen of the Jaina-recension. After Gatha 200 the following particulars regarding text and commentary are given:
9
iti mahāpi()ditasrtdallanipṛthivтpalasutatribhuvanapalasunun trividyena SrTmadajadena kṛtāyah mahārājādhirajasratavahanagāthākosasaptasatītтkäyä dvittya Satam samaptar.
As far as can be judged from Weber's edition of Bhuvanapala's commentary, which consists mainly of extracts, Ajada's commentary is merely a slightly shortened version of that of his father. A MS containing this commentary is mentioned in the Brhattipanika (published in the Jainagranthavalf; see NCC V, p. 343 b). The Bṛhattipanika is an old (Safvat 1440, i.e. 1383 A.D.) list of MSS in the possession of several Jaina Libraries of that time in Gujarat (Tripathт, 1977).
Kulanatha also wrote a commentary on the Setubandha. For particulars of this commentary, see Handiqui (1976. 96-99 et passim)
10
I have been able to locate five more MSS which may contain the text of

Page #94

the Third South-Indian recension. It should, however, be noted that the catalogues and lists from which the references have been taken with one exception (see 1 below) do not quote the beginnings and ends of the texts. The conclusion regarding the recension to which the texts belong has therefore been solely based on the fact that in each MS the text is accompanied by a commentary which otherwise is exclusively found in connection with the Third South-Indian recension. Note that all the five MSS mentioned below are written in the Malayā]am script. Compare Ma, Ti and Tp which are written or go back to a MS which was written in this script. It seems that the Third South-Indian recension has been preserved solely in Kerala!
MS no. 4762 in the University MSS Library, Trivandrum (see Alphabetical Index... (ABC 316), no. 4762, and A descriptive Catalogue...(ABC 321, 6), no. 1438). The MS was formerly owned by Mitran Avl., Cempakaśćeri 1llam, Kudamā lūr, Eţtumānūr. It is a palmleaf MS of 213 leaves of 7 by 1 3/4 inches. The script is Malayalam. From the quotations of the beginning and end given in the last-mentioned catalogue it appears to contain the text' of the Gathās of the Third South Indian recension together with Nadhavayajvamisra's commentary, from 1 to 681. This MS was shown to me while I visited the library in Trivandrum. I was, however, not allowed to film it as it was feared that the MS, which was already very much broken, would deteriorate further.
MS no. 294 mentioned in the Lists of MSS collected by the Curator for Publication of Sanskrit MSS, Trivandrum, Vol. I, p. 30 (ABC, UC 32); described as Hālas aptaśat Ttātparyatīkā by Mādhavayajva. The MS was formerly owned by Govindapisharodi from Kailasapuram (cf. Ti). The material is not specified; the script is Malayalam and it is said to contain 3650 Granthas.
3 MS no. 279 mentioned in the Lists of MSS collected by the Curator for Publication of Sanskrit MSS, Trivandrum, Vol. IV, p. 25 (ABC, UC 32);

Page #95

82
described as Hālas aptaśatītātparyaţika by Madhavayajva. The MS was formerly owned by Mitran Namburi, Kumaranallur. The material is not specified; its script is Malayalam and it is said to contain 4200 Granthas. It lacks part of the beginning.
MSS 2 and 3 should at present be in the possession of the University MSS Library in Trivandrum but could not be located.
MS no. 4304 mentioned in the Catalogue of the Sanskrit MSS in the Panjab University Library Vol II, Lahore, 1941 (ABC 145); described as Gāthasaptaśatīvyākhyā tātparyadipika by Madhavayajvā. The material is palmleaf, the script Malayālam. It consists of leaves 4 to 106, comprising the fourth and the fifth Sataka.
MS no. 4305 mentioned in the same catalogue as '4'; described as Gāthasaptasattvyākhyā tātparyaţTkā by Madhavayajvā. The material is palmleaf, the script Malayalam. It consists of 79 leaves, said to contain altogether 678 ślokas(?).
11
Theoretically it could be assumed that in each case the typical MIA form, e.g. nti, vir(/1)ão and pamhusai, is the innovation, secondarily introduced in order to give the text a more pronounced prākritic appearance. However, the distribution of the sanskritic forms in question, namely nisarai, vilino and visarai respectively, coincides with that of the other categories of innovations mentioned on the same page, which suggests that they are innovations as well.
12
54 Without going any further into the problem of the variation of siri and sari I should like to note that commentator Bhuvanapāla (MS Bh) must have read siri (Sanskrit śrt) as well (see ISt., pp. 45-6). 196 For udalie or uddali e Madhavayajvamiśra in his commentary gives the following explanation:

Page #96

83
dramilair uddalir iti yā krīda vyapadiśyate soddalika...utkuñcitobhayajānukās īnayā sajaghanotkamp am rabhasaikaikapādotsāranena nivartaniyatvād uddalikāyāḥ puruşāyitopabhogitvam iti mantavyam.
Prākrit umali or uddali seems to be connected with Tamil untal (pronounce undal) 'to play hop-scotch' (for information regarding this meanning, which is not found in any of the dictionaries, I am indebted to my colleague Dr. A. Govindankutty). 431 tamtassa stands for tāntasya, the genitive of the past participle of tam-. Misunderstood, it was divided into tah tassa, which triggered the innovations given. 460 In Bh this Gāthā is attributed to Dhavala (ISt., p. 161). 487 vābeņa...ņa sahian tie, 'her tears did not allow her...' 626 suhmamtaria, '(shade) interspersed (antaria) with small (suhma, Sanskrit sūkşma) (patches of light)'.
13 96 pariaram in vaivedhapariar am is apparently a third synonym for 'fence! As such pariara, from parikr-, 'to surround', would be typical of MIA. For the type of compound, see the note to Gāthā 4 in Part II. 129 asuņte, 'without listening'. aņuņTa is a nonsense-word 184 ann amanāha consists of anyamanas and aham. 315 gaapaiam stands for gataprad Tpām, but was misunderstood.
14
114 Read sirigovide, i.e. Rādhā. siri (Sanskrit śrt) is found in the MSS of the Third South-Indian recension and in MS Bh. 240 amdhalaa consists of andha and -laa, for which suffix see Pischel $ 595.
15 For 488 Weber in his Ed. established the following text:
annāvarāhakuvio vesattaņāvarāhe
jaha taha kālena gacchai pasan kuvia kaha tah pasāemi.

Page #97

84
Ein wegen andrer Vergehen Er zürnter wird wohl mit der Zeit irgendwie wieder gut. Wie soll ich aber ihn wieder gewinnen, der darüber zürnt, weil ich ihm feindlich gesinnt sei?!
At first sight the text and translation seem in perfect order. No significant variants are available except for pesattaņā(varākena) in Bho, i.e. Sanskrit preşyatva (but Bhuvanapāla dveşyatva). In my opinion the text would in fact improve considerably by adopting the word pesattana and emending annā (anya) into ānā (ājmā). Then, instead of saying that there is an offence consisting of showing one's animosity and that there are other offences, the speaker (a woman) would now distinguish gradation: in offences of essentially the same order, namely between an occasional command (ājñā) and treating someone as a servant, the victim being her husband. I should like to propose the following translation:
Someone who is angered by being offended with a command always comes round again, even if it takes some time. (But) how will I be able to pacify someone who is angered by being offended because he was treated as a servant?
16
In this connection a few peculiar omissions in the Jaina-recension should be mentioned, in each case Gathas the contents of which may have been offensive to Jainas. Their absence shows the working of censorship. The following Gāthās are left out: 120, 121 and 598 which mention a hunter, 175, 669 and 688 which mention a dog (175 also 'killing'), 431 which mentions the felling of a tree, and 405 which mentions a Navakāvāliā. Note in this connection also the absence of 548 which contains a word which, among others, means 'getting erected' (viabh amāņo). In the Vulgata this same word was changed into taggaamano.
17 It may be asked if these changes in both the Vulgata and the Jaina-recension, many of which appear to have come about by accident, do not at the same time indicate that the Gathās were originally not numbered.

Page #98

85
In any case they are not in the MSS of the Third South-Indian recension:
18
See also Gātha 76 in which āsaṁs ia in K, W, B, Y, P and S beside asaghia in Bh (cf. as adhia in R).
19 The present MS must have been compared with another MS, as appears from the many marginal additions. Occasionally mistakes are corrected: 419 voras avādin (cf. T), v vorasaṁghādin, 461 ramiavva (cf. Y), ramiavvaa. Other instances concern variants. In a number of cases these variants seem to have been drawn from a MS belonging to the K-version; 'e.g. 214 maída, v2 suņha (K), 289 tti paripucch°, tti ia pucch (K); in other cases from a MS like.y: 46 suņņai, v suvvai (v, B), 369 phaggucchana, 2 phaggucchava (Y, T). In some cases the source is not known: 474 nakkhāņa kāmiņio surae rehāith vi gananti.
20
The outward appearance of the text in Bh and R is practically the same. The MSS differ, however, on the following points. In R two Gathās of Bh are missing: 282 and 551. Bh 140 is found in R as 109. In R, Bh 382 occurs twice, namely as 14 and as 386. Furthermore, in R, Bh 353 and 354 have accidentally come to be combined into one Gatha: 358. In Bh the following Gātrās of R are missing: 2, 101, 144, 304, 315, 414, 416, 548 and 569 (569 in R in the margin). Note that several of these Gathas, namely 101, 144, 304, 416 and 569, are not found in Ma, Ti and Tp either.
21
It is not clear how much importance should be attached to the addition of Ed. 552 (562 W, Y, 563 1, missing in Bh, Ma, Ti and Tp) in R in the margin after (?) 568 (Ed. 550: 560 V, Y, 561 , 562 Bh, 446 Ma, Ti, Tp). It is, for instance, not clear if the addition is to be taken as a correction for an initial oversight on the part of the scribe, or whether We have to do with a marginal addition made by a later reader of the MS. Note that the Gathā in question, like R 568, mentions a buffalo (mahisa).

Page #99

86
22
The present copy of S has numerous marginal additions which indicate that it was compared with one or more other MSS. E.g. 61 samāniãin, sa samāpiāith (cf. B), 50 jura, s2 juvara (cf. Bh, R), 120 mārueņa (cf. K), sa mārueh it (cf. We B, Y, P), 184 manqusamuppanneņa (cf. K), s? mannusamuppāena (cf. , Y, P), and 441 a(9) ghissaṁ (cf. Bh, R, Ti, Tp), accissan (cf. W, Y, T).
23 Saptaśat Tsara with Bhavad Tpikā of Vema Bhupala, along with the Chappannaya-Gahão (text and chāyā), ed. by A. N. Upadhye, Shivaji University Sanskrit and Prakrit Series Vol. III, Kolhapur, 1970. The book 'has unfortunately not been available to me. See also Saptaśattsāra, ed. by D. V. Sitaramaswamy (with Telugu translation), Andhra University, Waltair, 1951, mentioned in Sriramamurti (1972: 68).
24
Apparently only the names were borrowed, as the Jaina-recension as represented by Bh does not show traces of the influence of the Vulgata on its text.
25
As for the two recensions no certain dates are available, theoretically also the reverse may be assumed, namely that they themselves were the first kośas thus arranged and that they set the standard to be followed in all subsequent kośas. It is unlikely, though, that they did indeed excercise such an influence, considering the rarity and geographical limitations of the occurrence of the MSS containing these two recensions.

Page #100

87
APPENDIX I
Concordance of the Third South-Indian recension with Weber's Ed. and his MSS.
The first column refers to the text of the Third South-Indian recension. The order of the Gāthās is basically that of Tp, the only complete MS of this recension, Ma breaking off after 474 and Ti beginning only with 197. Furthermore, the order of the Gathas of the transcript Ma was considerably disturbed as a result of the confusion of the leaves of the original (see above, p. 23). The second column refers to Weber's Ed. The third column to the Jaina-recension as represented by R, the other Jaina MS, viz. Bh, missing the final Sataka. The next four columns refer to the vulgata MSS K, V, Y and P. The fifth Vulgata Ms, B, which contains only a very short fragment (1-129) is subsumed in the second column. MSS K and P are likewise fragments, containing 1-370 and 1-300 respectively. The eighth column refers to the x-recension, the only MS of which, namely x, misses the first and part of the second Sataka. The final three columns refer to the Telinga (T), Sädhäranadeva (5) and Second Telinga-recensions (W), respectively.

Page #101

APPENDIX I
Concordance between the Third South-Indian recension and Weber's recensions.
Third S-I
recension
1) 1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
2) 16
17
18
19
2 2 2 2 2 2 2
20
21
22
23
24
25
26
27
Ed. R(Bh) K 4
1
2
3
893
5
6
7
8
766
9
10
11
236
12
14
61
15
637
16
1
13
23
3
7
~~3
1
615
9
10
11
186
2
5
56
667∞
567 co
667∞
8 8
240
12
12
14
14
65
60
15
15
16
16
13
13
17
17 17
19
19
19
767 616
624 619
680
626
42
45
. 683 627.
9
10
88
11
123
123
YPX TS
9
10
11
239
13
5
8 8 8
9
10
11
240
13
12
12
62 61
17
16
16
14
18
20
15
14
17
19
1
2
3
634
633
690
689
42 43 43
. 693. 692.
5
6
7
PRESELLER 22
9
10
11
239 240
12
14
61
15
16.
13
17.
19
43
1
7
5
501
148
1
7
8
480
643 692
646
173
233
40
238
442
479
455
292
590
209 284
313 427
388 203
26
663
653
611
261
356
389
139
263
367
489
506
13
89 146
30 600 26
487
380
488
276
322
w(v)
659. 565. 410.
1
2
3
4
1) In Tp only; Ti begins only with 197.
2) In R a second time as 386; in Bh missing here; only found as 382.

Page #102

Third S-I recension
28 . 821.
29
1) 30
31
23 365 A I I I I G
33
34
37
38
39
40
41
42
43
44
45
Ed. R(Bh)
46
47
48
49
50
51
2) 52
53
677 628
626 631
605 629
685
630
693
638
63 67
640
646
647
769
745
774
732
499
648
661
681
664
636
783
676
735 651
796 689
641690
692
637
647 697
653
809
.
699
704
701
K Y P XT S
64
49
20
687
686
636 635
615
614
695
694
704 702
64
63
377
646
89
686
377
645
1) Occurs a second time as 411. 2) In Ma after 55.
685
651
650
703
701
657
656
663 662
63
49
648
647
662
20
607
686
651
377
622
654
626
606
90.
170
115 84
369
138
236
423 500
187222229 +
186
335
424
511
222
390
266
671
593
459
392
136
215 408
124 147
94 617
90
23
649
659
658
651
661
660
652
662
661
54
49
53
50
49
57
55
20
20
21
20
174
56
564
613
574
574
87
57
21
21
21
22
21
21
488
175
58 . 270 222. 273. 273. 274. 273. 272. 594. 312 .
100
184
185
437
140
47
50
224
444
320
.
-
w(v)
14
64
82
32
22
23

Page #103

90
T
S
W(v)
173
272
322
36
Third S-I Ed. R(BH) K ♡ Y P X recension
59 . 24. 24. 24. 25. 24. 24 . 60 25 25 25 26 25 25 61 117 132 120 119 122 120
22 22 2223 22 22 26 26 26 27 26 26 27 27 27 28 27 27 28 28 28 29 28 28 29.29
30 30 30 725 611
31 191 190 192 198 194 32 197 196 198 31 33 31 32 31 32 33 35 33 34 33 33 855 18936 192 191 193 199 35 38 35 36 35 35 36 39 36 37 36 36 38 41 38 39 39 39 40 43 40 41 41 41 39 42 39 40 40 40 41 44 41 42 42 562 617 572 572 828
43 46 43 44 37 37 190 47 193 192 194
50 46 47 46 46
54 50 51 50 50 53 52 53 54 53 53 54 58 54 55 54 54 57 61 57 58 57 57 58 62 59 58 58
5963 58 60 59 59 92 . 60. 64 . 59. 61. 60. 60.
. 632. 530 . 205 176
60 649 649
324 529 325
490 368 543
502 589 326
420
653 527 309 354 346 410 394 526 404 527 384 293 305 437
499 : 537 697 10 161 579 342 425 343 160 336 611 235
39 277
62
21
499
656
343 429
209
395 528 . 227 . 344

Page #104

Y
P
X
T
S
w(v)
Third S-I Ed. R(BH) K recension
93 . 65. 69. 62. 94 70 74 69
51 55 51 55 59 55 66 70 63 71 75 70 72
71
66 · 65 . 65 . 71 70
70 52 51 51 56 55 55
66 66 72 73 72
. .522
290 450 614 294 498 657
523
76.
854
101
76 75
75 74
75 74
102
103
296 561
239 353 695 79 588 333 534 602 278 194 647
157 657 589 601 242
104
105
139
16
106 107
82 129 81 84 86
81 132
80 83 85
81 132 80 83 85
108
53
355
109 110
75 79 74
74 78 73 913
81 85 80 127 · 142 132 80 84 79 838782 85 89 84 856
89 91 86 87 92 87 90 94 89 93 97 92 95 99 lac. 96 100 lac. 253 202 256 168 102 169 97 103 lac.
111
87 88 90
112 • '113
114 115 116 117 118
90 88 91 94 96 97 256 170 98
87 88 90 93 96 97 257 171 98
96 94 256 170 97
10 596 623 178 580 320
47 500 265
644 327 334 273 280 445 685 244 345
27
119
120
46
121
122 123 124 125 126
100 104 105 107 108
345 · 861 .
106 116 117 118 120 297
.
lac. 105 106 108 109 347
.
102 106 107 109 110 349
.
101 107 108 110 111 349
.
99 105 106 108 109
28 421 487 415
193 348 583
. 361 .
330 454 588 458 389 542
.
8

Page #105

T
P
X
T
S
W(V)
.
284
285 112 113
. 114. 112 .
286 285 115 113 116 114
Third S-I Ed. R(Bh) K recension
127 . 700 . 124. 112 . 128 282 235 285 129 110 123 113 130 111 125 114 131 857
112 126 115 133 118 133 121
121. 124 123 136 126
129 140 130 137 128 141 131 138
135
. 413. 19 669 440 455 206 335 356
9 642 497 579
132
115 121 124
134 135
114 120 123 125 131 130 134
117 123 126 128 134 133 137
128
126
136
130 131 135
136 137 142
659 533 217 277
613 310
32 358
139
132 850
299
140 141 142
143
254 149 143 148 151 152 156 157
254 148 143 149 151 152 157 158
255 150 146 151 153 154 158 159
254 149 144 150 152 153 157 158
128 548 347
- 539 168 607 288
144 145
254 29
608 152.234 157 607 159 . 160 113 165 96 166 482
317 312 34 169
146
251 200 146 352 141 347 147 353 149355 150 356 155 360 156 361 852 310 253 159 364 160 365 162 367 163 368 165 370 167 372 164 369 166 371
147 148 149 150 151 152 153
111 477
90
201
314 160 161 163 164 166 168 165 167
314 161 162 164 165 167 169 166 168
314 162 163 165 166 168 170 167 169
161 162 164 165 167 169 166 168
154
155 156 157
173 175 177 174 176
448 453
511 510 578 491
43
158
159 160
186 390 189 188 190 189 6939 169 . 373. 171 · 171 · 172 · 171 . 178. 426 · 571 .

Page #106

Third S-I Ed. R(Bh) recension
K
U
r
P
x
T
S
W/v)
58
185
273
61
943
420
38
161 162 163 164 165 166 167 168 169 170
171 1) 172
173
174 2) 175
176 177 178 179 180
400
829 . . . . . .
· 171 . . 174 378 176 176 178 177 183 484 215 179 380 181 181 183 182 188 634 216 175
177 177 179 178 184 447 514 176 379 178 178 180 179
480 177 154 180 179 181 180 186
662 180 381 182 182 184 183 189
481 181 382 183 183 185 184 190 301 388 182 383 184 184 186 185 191 347 425 183 384 186 185 187 187 193 172 394 696 388 708 705
138/696 192 108 195 194 196 192 197 552 618 193 127 196 195 197
416 456 195 109199197 199 194 199 513 248 198 147 202 200 202
202 319435 199 148 203 201 203 202 203 340 414 . 201 150 · 205 204 205 204 205 604 544 202 151 206 205 206 205 206
540 204 153
207 208 207 208 281 108 206 157 210
209 210 225 451 210 161 213 214 213 214 179643 207 158
211 210 211 210 211 111 152 444 446
455 455 482 61 562 938
612 223 173 228 226 227 226 227 330 211 162 215 214 215 214 215 122 216 166 220 219 220 219 220 567 299 2 18 168 223 221 222 221 222 401 - 459
219 169 224 222 223 222 223 433 31 : 221 . 171 : 226. 224. 225. 224. 225 . 339. 417.
208
209
210
181 182 183
214
184 185
186
187 188 189 190 191
1) Missing in Tp. 2) In Bh as 140.

Page #107

W(v)
70 -
194
57
199
35
203
523
381
205
254
Third S-I Ed. R(Bh) K W Y P x T S recension
192 · 225 . 175. 230. 228. 229. 228. 229. 463 . 460 . 193 229 179 234 232 233 232 233
251 228 178 233 231 232 231 232 436 250 195 231 181 234 235 234 235 450 503 196 232 182 236 235 236 235 236 492 638 197 235 185 239 238 239 238 239 129 220 198 240 189 244 243 244 243 244
185 243 193 247 246 247 246 246 33 126 200 246 195 250 249 250 249 249198 253 201 247 197 251 250 251 250 250 555 683 202 249 206 253 252 253 252 252 477 462 486
302 204
252
201 255 255 255 255 297 392
203 257 257 258 257 256 405 398 206 914
536 207 819
265 217 267 268 269 268 267 17 674 209
220 271 271 272 271 270 490 · 162 210 271 224 274 274 275 274 273 125 136 211 267 219 270 270 271 270 269 23 676 212 273 226 276 276
276 275 211 187 213 274 227 277 277 278 277 276 156 · 463
276 228 278 279 280 279 277 155 249 215 924 216 460 467 471 471 495 78 597 217 275 229 279 278 279 278 278 629 314 218 280 233 283 283 284 283 282 543 315 219 287 240
290 291 290 289 76 285 220 289 277 291 292 293 292 291 646 624 221 290 278 292 293 294 293 292 654 587 222 291 279 293 294 295 294 293 408 120 223 .292, 280 294 295 296 295 294 135.
53
208
268
214
1) Ti begins here.

Page #108

95
298
297
206
306
230
232
234
315
237
578
Third S-I Ed. R(Bh) K Y P X T S W(V) recension
224 294 . 282 297 . 297 . 298 . 297 . 296 . . 468 225 869
397 226 532 493
542 542
51980 119 227 295 283 298 298 299 298
469 228 302 245 306
306
304 541 316 229 297 285 300 300 301 300 299 409 470 298 242 302 302 302
300 254 651 231 300 244 304 304 304
302 107 219 304 247 308 308 308 306
286 233 312 255 317 316 316 314 174 274 313 256 318 317 317
465 235 316 259 321 320 320 318 364 106 236 849
296 578 572
588 588 580 683 68 238. 239 318 261 323 322 322 320 326 466 240 324 267 327 328 328
326 374 67 241 322 265 326 327 324 239 36 242 325 268 201 329 329
631 243 932
585 244 328 272 334 332 334
330 337 467 246 247 912
532 248 579 573
581 84 594 249 896
505 250 331 275 350 335 341
333 56 326 270 329 330 330 328 295 252 434 618
445 445
444 687 253931
582 254 255 332 286 335 336 335 334 467 472 28 256 333 110 336 337 336
154 452 75 · 257 334 287 337 338 337 1 335 473
258 . 563. 620 . . 573 573 . . . 694. 92.
251

Page #109

Third S-I Ed. R(Bh) recension
K
V
Y
P
x
T
S
W(V)
525
341
332 342
342 344
342 344
340 342
418 114
262
345
475
516
189
259908 . 260 338 291 261 340 292
910 263 341 293 264 342 294 265 359 314 266 343 295 267 347 300 268 351 305 269 348 301 270 349302 271 354 309
355 310 273 360 315 274 361 316 275 362 317 276 365 320
257
343 344 363 345 351 352 358 359 355 356 364 365 366 370
295
347 352 356 353 354 359 360 365 366 367 370
345 346 364 347 352 356 353 354 359 360 365 366 367 370
343 197 345 . 364 372 346 180 351 355 147 352 210 353 359571 360 365 375 366 187
352 301
272
250
464 262
367
370
418 - 478
277
278 1) 279
280 281 282 283 284 285 286 287 288 289 290 291
367 323 368 327 369 344 371 330 373 332 376 335 523 562 377 336 378 337 706 392 707 393 714 394 715 395 379 . 396.
372 373 374 376 379 382 533 383 384 385 386
372 373 374 376 379 382 533 383 384 385 386
372 332 421 373 531 393 374 468 · 486 376 331 423 378 378 396 381 640 546 576 564 305 382 236 357 383 323 440 384 620 319 385
395 51 386 119 169 17 387 574 191 388 . 406.483. 65
387
388
. 389. 387
1) Missing in Ti and Tp.

Page #110

T
SW(V)
.
389
61
.21 . 666.
44 689 274 359 70 95
327
51
395 396
390 392 394 396 397
398 401/373
402
130
397
311 371 454
121 267
301
404
89
303
206
207 406
581 268
408
80
413
Third S-1 Ed. R(Bh) KŲ Y P recension
292 .380. 397. . 390. 388. 293 688
699 697 294 381 398 391 389 295 383 400 393 391 296 385 402 395 393 297 387 404
397 298 388 405
398 299 389406 - 399 300 392 411 402 400 393 412
403 401 302 395 414
405 404 203 152 207 206 207 304 397' 415
407 406 305 400 418 410 409 306 404 . 415 414 307 407
418 417 308 711 504
412 419 423 423 310 311 416 424
427 427 312 417 425
428 428 313 689 634 700 698 314
429429 315 419 426 430 430 420 427
431 317
432 432 318 724
429 319 425
432
436 436 320 431
442 442 321 433 441 444 444 322 565 325 575 575 323 759 564 324 554 598
564 564 325 266 218 . 269 269 270 269
595 361 171 172 265 65 371
37 67 175 243 270 516 422 457
416 419 422
309
902
426 427 611 428 429 430 431
485 520 598 275 268 170
418
321
316
586
421
428
221
432
436
441
443
357269 116 165 578 292
88 609 551 297 637 557
449 498 . 18. 675.
463
54
454 268

Page #111

327
329 330
331 332
334
336
450
Third S-I Ed. R(Bh) K Y P X T S w(v) recension 326 353. 308. 354.358 . 358. . 358 . 626 . 300.
566 435 576 576 464 219 403 328 556 600 566 566 456 697 83 570 440 580 580
46969691 572 443 582 582
471 310 384 571 442 581 581 470 120 693 574 156
584 584 473 181 333 575 204 315 585 585 474 673 33 438 241 449 449
476 167 231 335 439321
450 450
477 108 217 445 447
456 456 547 615 559 337
458
461 461 485 577 322 338 779 659
598 45 339 452 444 463 463 487 519 525 71 340 451 459
462 462 486 625 323 341 454 461
465 465 489 71 100 455 462
466 466 490 2 2 343 457
468 468 492 645 492 344
468 472 472 496 417 493 1) 345 462
469
473 473 497 101 469 476 480 480
504. 694 91 2) 347 463
474 474
498 478: 194 348 472 479
483
507 655 532 349
480 485 485 508 350
482
486 486 510 276 373 351 935 476 483 487 487
178 69 353 691 636
702 700 605 698 17 354
488 488 512 150 181 355 478 486
489489 513 241 356 531 491 541 541 518 298 385
357 .534 . 496. . 544. 544. . 521 . 308 . 430 . 1) Ed. p.198: R 470; see, however, ISt., p. 162. 2) Ed. p.199: R 469; see, however, iSt., p.162.
342
464
461
346
470
483
603
474 475
352
511
477
485

Page #112

99
Third S-I Ed. R(BH) recension
K
T
P
x
T
S
W(V)
358 359 360
. 536 . 498.
538 500 541 505
. 546 . 546. 548 548 551 551
. 523. 279 . 374 .
525 427 64 528 244 72
361
539
502
549
549
526
241
63
553 * 490 491 494 496
553 490 491 494 496
543 507 479513 480 514 483 515 485 522 782 663 484 521 487 524 488 528 489 529 492 531 481532 482 533 496 455 503 537 504 538
624
362 363 364 365 366 367 368 369 370
371 1) 372 2) 373
374 375 376 377 378 379 380 381 382 383 384 385 386 387
552
495 495 498 498 499 499 500 500 503 504 492
492 493 493 507 508 513 513 514 514
546
530 672 42 535 93 208 536 650 622 537 656 202 539 348 387
675 48 538 342 436 541 309 439 542
551 543 545 338 441
625 344 328
536 555 213 338 556
623 278 557 105 143
414 524 559 336 560
640 561 99 212 562 . 555 563. 584.541
488
- 72
848
515
515
558
31
505 539 720 540 506 541 507 543 508 546
509 548 . 510. 549.
516 516 517 517 518 518 519519 . 520. 520.
1) In Ti and Tp after 377. 2) In Ti and Tp after 371 and, a second time, after 372 (see previous
note).

Page #113

100
Third S-I Ed. R(Bh) recension
K
W
Y
P
T
S
W(V)
1) 388
389
390 2) 391
392
529
393 394 395
. 523. 523. 526 526 527 527 528 528
529 531 531 535 535 592 592 595 595 598 598 601 600 603 602 604 603 608 607
396
397
398
590
399 400 401
634
402
. 513 . 552.
516 556 517 557 518 519 558 521 560 525 565 582 566 585 577 588 580
585 592 589 593 593 597 717 389 610 517 609 495 612 544 617 591 871 618 619 595 621 602 626 631 627 526 760 574 643 692 839 644 693 . 654.
403 404 405 406 407 408 409
410 3) 411
412 413 414 415 416 417
620 619 619 618 622 621 627 626
495
. 566 . 382 · 615.
569 285 206 570 293 289 571 494 637 572540 290 574 481 525 578 575 519 584 218 587 685 75 590 315 443 592
169 610 594 63 566 595 131 144 629
162 190 676 . 674 474 494 678 345, 442 663 642
400 664 189274 597 379 600 52 567 647 115 84 615 199
441 497 627 54 568
240
92 . 604. 38.
628 629 631 636 637
627 628 630
635
636
653
652
628
654 653 . 664. 663.
1) Missing in Ma. 2) In Bh as 551. 3) In Ma and Tp also as 30.

Page #114

101
Third S-I Ed. recension
R(Bh)
K
Y
P
x
T
S
W(v)
418
.
419
.645. 695
656 658
655 666 668
654 665 667
. 644. 451. 665 152
264
420
421
670 669 6.71 670 674 673 679678 247 248
599 262 689572 639 407 688 439 247 286
507
37 15
426
248
247
348
111 133
116 132
119 135
117 133
140
373
8 139 460
245 649 501
74
421
421
142
132 233 66
416
420 . 414 417
415
419
418
263
422 660 423 661 424 664 425 669 708
244 · 194 427 428 747 112 429 114 129 430 130 · 143 431 741 432 433
410 512 434 837 435 405 436 408 437 792 673 438
422 583 440
428 442 834 443 429 444 437
549 446 550 568 447 551 571 448 553 586
449 559 606 1) 450 356 311 1) Missing in Ma.
72
409
420
419
387
418 433
423
433 434 439
433 434 439
434
441
434
438
117
164
229
107
445
92
440 440 448 448 559 559 560 560 561 561 563 563 569 569 361 361
439 447 366 449 83 98 450 429 513 451 595 453
459678 198 361/466 289353
357

Page #115

102
P
x
T
S
w(V)
. 501 . 644. 306.
503 68 162 551 256 41 552 200 205
214 . 691 57 564
407 554 587 376
303. - 5 627 556
459
460
563
585
159 674
406
464
Third S-I Ed. R(Bh) K Y recension
451 466 473 . . 477 477. 452 468 475 479 479 453 497 456 508 509 454 498 527
509 510 455 832 456 615 582 625 624 457 501 535
511 511 458 502 536 512 512 522 561
532 532 524
534 534 461 577 570
587 587 462 583 567
593 593 463 586 578
596 596 843 465 589 584 599 599
594 592 605 604 467 595 706 606 605 468 602
612 611 607 545
617 616 470 614 581 624 623
620 596 630 629 472 263 215 265 266 267
473 279 232 282 282 283 1) 474 281 234 284 284 285
283 236 286 286 287 476 847 477 309 252 313 313 313 478 319 262 324 323 323 479 296 284 299 299 300 480 335 288 338 339 338 481 358 313 362 363 363 482 803
483 · 366 . 322 . . 371 . 371 . 1) After this Gathā Ma breaks off.
466
469
471
93
591 75 571
144 255 596 77 572 632 472 - 509 672 166 211 661 79272
598 666 266 265 69 · 86 282 281 65 284 283 60 636 286 285 251 340
271 311 559 317 321 14 298 257 354 336 431 474 363 370 259
121 148 . 371 . 110 . 141 .
475
679
44

Page #116

103
486
267
884
462
Third S-I Ed. R(BH) K
Y P X T S W(V) recension
484 .440. 324. . 451 . 451 . 478 . 303 . 232. 60 485 749 328
260 845 487 370 329
375 375
375 333 422 488 152 3580 153 154 155 154 162 489 154 359 154.156 157 156 164 485 490 883
461 491 598 · 609 608
661 492 493 187 391 190 189191 191 196 402 482 494 386 403 396 394 395 242 370 495 567 436 . 577 577
67 496 817 497 498 830
190 499 823
106 .500
301 305 305 305 303 471 49 502 756 519
334 432 503757 520
433 504 842 505
109 506 86 90 85 87 86 86
192 243 507 859
359 88 93 88 89 89 89
328 509 99 105 lac. 100 100 100
27 665 103 104 105 106 104
453 120
123 125 125 123 662 582 512 133
136 135 138 136 143 228 369 513 143 349 145 145 148 146 154/641 207 38 514 145 351 147 147
148 156 87 608 515 1 84 385 187 186 188 188 194 330
835
230
501
255
824
508
115
510 511
446
516
517
. 196. 145.
198 . 200. 195 . 200 . 123 . 180 .

Page #117

104
w(v)
840
252
529
945
Third S-I Ed. R(Bh) K
Y P x T S recension
518 · 197 . 146. 200. 199. 201 . 196. 201 . 403 . 457 . 519 215 165 219 218 219 218 219 643 298 520 220 170 225 223 224 223 224 329 415 521 226 176 231 229 230 229 230 398 391 522 233 183 237 236 237 236 237 491 639 523 277 230 280 280 281 280 279 349 329 524 305 248 309 309 309 307 284 186 525 526 306 249 310 310 310
308 535 308 527 170 374 172 172 173 172 179 245 612 528 217 167 221 220 221 220 221 25 109 323 266 326 327 326
325 103 140 530 430
441 441
294 115 531 841 532 573 111
583 583
472 101 151 533
633 534 441 326
452 452
479 260 355 535 520 559
530 530
496 641 858
358 898
510 538 868
· 385 539 941
617. 540 915
537 541 224 174 229 227 228 227 228 41 672 542 124 137 127 126 129 127 133 282 282 543 125 138 128 127 130 128 134 158 246 544 126 139 129 128 131 129 135 638 238 545 142 348 144 144 147 145 153 283 130 546 957 547 237 187 241 240 241 240 241 411 461 548 258 210 261 261 262 261 260 476 158 549 259 212 262 262 263 262 261 658 182 550 920
546 551 278 . 231 . 281 . 281, 282 . 281 . 280 . 127 . 137 :
536 537
5 2
59

Page #118

Third S-I
recension
552
553
554
555
556
557
558
559
560
561
562
563
564
565
566
567
568
569
570
571
572
573
574
575
576
1) 577
578
579
580
581
582
583
Ed. R(Bh) K
921.
415 423
427
433
900
465
472
546
510
320 263 325
545 509
212 163.
293
281
307 250
838
950
872
891
876
862
890
255 207
903
825
885
864
918
919.
930
923
. 895.
216
296
311
258
1) Lacuna in both Ti and Tp.
426
438
476
5.56
324
105
258
Y P XT S w(v)
426
438
476
556
324
555
555
215
216 215
296
297
296
311 311
259
425
437
500
533
322
532
216
295
309
258 257
547.
473
489
514
504 270
362
204
15
670
141
199
134 661
383
529
226
350
237
651
412
486
432
365
483
40
517
133
464
367
484
163
544
545
576
549
503.
127
29
78

Page #119

106
K
P
x
T
S
W(V)
Third S-I Ed. R(Bh) recension
584 • 909 . . 585 894 586 56 60 587
.
.
.
.
. 526 .
.
502
56
57
56
56
168
606
588
904
589 590 591
287
287
289
287
287
237 .
-520
20 351 641 618
588
-
284 853 948 350 576
592
593
360
94
33
303 205
355 586
355 586
354 475
594
154
413
1) 595 1) 596
597 598 599 600 601
873 927 836 7983 32 34 44 48 45 49
78 32 44 45
80 33 45 46
79 32 44 45
95
79 31 44 45
602 603 604
863
47
51
47
569 232
596 518 - 533
125 553 682 366 201 - 337 151 695 523 177 240 145 390 690 376 431 352 210 157 229
52
605 606 607 608 609 610 611 612 613 614 615
52
77 72 74 73 73 78 82 77 79 78 78 952 84 88 83 85 84 84 91 95 90 92 91 91
98 104 lac. 99 99 98 916 . 122 . 135. 125. 124 . 127. 125.
538
.664 . 580
1) Lacuna in both Ti and Tp.

Page #120

107
w(v)
377
925
623
9
627
399
13
Third S-I Ed. R(BH) K Ų Y P x T S recension
616 101 . 113. lac.. 103 104 . 102. 506 · 346 617 109 121 110 111 112 110
18 618 865 619 106 119 107 108 109 107
164 179 620 119 134 122 121 124
585 621 131 338 134 133 136 134 524 534 622
556 137 342 139 139 142 140 148 288 234 624 140 · 346 142 142 145 143 151 325 167 625 139 345 141 141 144 142 150 55 593 626 144 350 146 146 149 147 155 597 629
870 628 161 366 162 163 164 163 171 195 264 629 205 155 209 208 209 208 209 630 148 354 150 151 152 151 158 102 214 631 946
635 632 153 3586 155 155 156 155 163 364 633 173 377 175 176 177 176 182 663 581
208 159 212 211 212 211 212 224 339 635 209 160 213 212 213 212 213 165 197 .636 937
213 164 217 216 217 216 217 143 155 638 222 172 227 225 226 225 226 231 347 639 238 237 238 237 238
153 640 238 188 242
242 241 242 241 242 699 641 926 642 250 199 222 253 254 253 253 12 673 643 256 208 259 259 260 259 258 466 644 257 209 260 260 261 260 259363 645 260 211 263 263 264 263 262 118 159 646 917
539 647 269 221 272 272 273 272 271 112 156 648 262 214 265 266 265 264 324 649 315. 258. 320. 319. 319
317. 246. 66.
55
634
605
637
19

Page #121

108
Third S-1 Ed. R(Bh) K Y P x T S W(v) recension
650 .288 . 276 . 290 . 291 . 292 . 291 . 290 · 67654 . 651 136 341 138 138 141 139 147 32 602 652 48 52 48 49 48 48 191 592 18 786 667
204 654 905
521 655 23 23 23 24 23 23 163 200 10 656 299 243 303 303 303 301 657
522 658 246 307 307 307
305 128 138 659 314 257 319 318 318
316 188
653
677
53
256
660
661 662 663 664 665 666
391
667
668 669 670 671 672
317 260 322 321 321
368 576 336 289339 340 339
247 351 901
515 382 399 392 390
98 160 344 296 346 348 348 347 146 604 384 401 394 392
393 619 311 933
591 .. 699 122
111 475 281 514 553 524 524
567 396 616 844
259 391 410
401 399
400 82. 374 333 380 380
563 304 396 420 406 405
· 216 39 897
509 399 417 409 408
218 402 334 381 381
545 772 643
25 413 421 424 424
423 182 362 432 409 443 443
442 568 648. 949
647 569438 579 579
468 85 575 453 460 464 464
488 91 372 . 512 . 551 . . 522 . 522 . . 565 . 636. 554
673
674 675 676 677
375
700
678
679 680 681 682 683

Page #122

109
K
r
P
T
S
W(V)
470. 470 481 481 484 484
84
558
558
688 689
Third S-I Ed. R(BH) recension
684 .459. 466. 685 470 477 686 473 481 687 763 597
548
822 1) 690 530
691 493 692 831
693 587 579 2) 694 390 407
695 526 449 696 697 743 542 698 529 489 699 .616.590
494. 386. 614. 505 104 103 509680 195
570 448 149 698
97 517 652 548 249 375
212 589 208 378 399 566 307
381 531
540 • 504
540 505
597 400 536
597 398 536
62
539539 626. 625 .
684 508 271 516 648 620 . 666. 153. 696.
1) Lacuna in both Ti and Tp; Gātrā has been identified with the help
of the commentary. 2) Missing in Tp.

Page #123


Page #124

APPENDIX II
111
Concordance of Weber's Edition (1881) with the third South-Indian recension.
Ed
Ed
1.
Ed 32. 601 33 72 34 - 35 75 36 76
63. 34 64 -
Odvor own
66
97
94. - 95 114 96 115 97 118 98 613 99 509 100 120
1 616
11
2
-
12 13 14.
13 15 20 16
72 73 74 75 76
99 608 102 101 -
42 26 43 83 44 602 45 603
85 47 605 48 652 4954 50 86
3 510 4 121 5 122 6 619 7 123
124 9 617 110 129 11 130
132
17 .18
21 ..
78 609 79600 80 106 81 104
1922
55
57
83
59
60 63
107 611 108 506 111 508 110
53 87 54 88 55 96 56 586 57 89 58 90 5991 60 92 61 17 62. -
85 86 87 88 89
26
17 61 18 133 19620 120 511 21 134 22 615 23 135 124 . 542
112
30 67 31. 71
91 612 92 - 93. 113

Page #125

112
Ed
Ed
Ed
Ed
125. 543 26 544 27 105 28 137 29136 130 430
621
159. 150 160 151 61 628 62 152 63 153 64 156
154
194 · 70 95 175 96 517 97 518 98 176 99 177 200
178
228 . 194
29 193 230 .
31 195 32 196 33 522 34 639 35 197 36 14 37 547 38 640
65
138
157 155
67
179
68
303
69
117 160 527
651 623
170
4 5 6
180 629 181
71
240
198
183
140 41 42
73 633
162 75 164 76 165 77 166
43 .44
199 426
625 624 142 545 513 626 514 141 143 630 144
78
47
201
79163
168
634 9 635 210 182
11 187 12 561 13 637 14 - 15 519 16 188 17 528 18 189 19 190 220 520
81
169
46 47 48 49
82
170 171
83
145
515
191
488 632
159 493
49 202 250 642 51 140 52 204 53 116 54 205
571 56 643 57 644 58 548 59 549 260 645 261. -
489
55 146 56 147 57 - 158 . -
8974 190 84 91 - 92 173 193. 174
638 23 186 24 541 25 192 26 521 227 . -

Page #126

113
Ed
Ed
Ed
Ed
330.
250
32
262 .648 63 472 6465 208
325
1211 68 209 69 647
| 58
296 .479
97 229 198 230 199 656 300 231
1 501 2.228 3 658 4 232
524
255 256 257 480
| 662
| 210
39
.
526
340
261
364 .. 65 276 66 483 67 278 68 279 69 280 370 287 171 281 72 -
282 1672
676
1283 177 285 178 286 | 79 291
292
294 82 664 83 295
666 296
212
74.213
41 263 42 264 43 266 44 665 45 125
西石刀
217
| 214 77 523 178.551
79 473 280" 218
1474 128. 475.." 590
563 8 . | 9 477 310 149 11 - 12 233
234 659 649
235 117 661
239 478 558
267 | 269
49
| 270
350
1593
51
268
494
297
326 54 271 55 272 56 450 57 . 58 481
241
290
219 | 650 | 220 1221
222 192 223 193 562 194.224 295.227
529 240 242 251
88 298 189 299 390 694 191 671 192 300
93 301 | 94 . 195 302 | 96 673 397,304
59
265
26
27
245
360 273 161 274 162 275 363 ..
28 244 329 ..

Page #127

114
Ed
Ed
Ed
Ed
398 ..
99 675 400 305
432.679 133
321 252
500 .. |1 457
2 458 3 378
379
466 .451 67 68 452 69 346 470 685 71 .
348 686
349 75 350
444 334 335 484 534
39
440
2乃乃阳门四
352
184
336
79
480
307 436
438 410 433 11 . | 12 309 | 13 678 14 .
553 16 311 17 312 18 314
315 316
354 355 365 366 375 376 367 370 1368 203 371
51
340
| 381 16 383
7 384, 18 385
386 510 387 11 . | 12 683 | 13 388 | 14 669 ( 15 . 16 389 17 390 | 18'391 | 19 392 520 535
21 393 22459 23 284
460
394 26 695 127 .
28 . 29 698 530 690 131 356 | 32 226 1533 ..
52
339 682 341
86 87
53
as
54 55
317
1342
56
439 440
490 91
57
343
58
319
59
554
441
684 160
216 61 344
345 163 347 164 · 465,556
62
93 691 94 . 95 . 96 377 97 453 198 454 499 . 39.
29 443 430 530 431,320

Page #128

115
Ed
Ed
Ed
Ed 534 · 357
35 -
602
468
636. 40
568. -
69 681 570 329
36
358
331
38 39 640 41
- - - 45
359 362
38 39 540 41 42 43
73 74 75
330 532 332 333 594 461
360
404
43 44
414 416
403
364
44
237
405
248
45 46 47 48
560 557
- 688
16
470 456 699 406 408
49
445
83
462
446 447
51
19
409
52 53 54
53 48 417
448
396 463 693 397 465
620 471 21 410 22 -
55
-
324
56
419
328
398
58
399
411(/30)
660
422
59 449
400
412
466
467
62 81 63 258 64 56 65 322 66 327 567. 495
97 401 98 491 99 - 600 - 601. -
669 . 425

Page #129

116
Ed
Ed
704 ..
1.677
| 287 7 288
738 . | 39 740 . 41 431
| 37
| 36
a 乃乃形刀昭四wa
7933
anu乃乃形刀昭四waw8w8858
13
.
32
293
502
57
503
89 313 690 91 92 46
33
353
700
127
| 2 . 703 ..
771.
.
805 ..

Page #130

117
Ed
Ed
Ed
.
840.525 41 531
874. - 75 - 76 568
42
8 - 9 · 49
43
464
45
670 486
476
908. 259
9584 910 262 11 - 12 247
103 14 206
540 16 614 17 646 18 579 19580 920 550 21 552
380
.49 850 51 52
490 492 575
17 18 192
236 139
. 148 591 100 73 109 131 536 507
55 56
23
28 689 499 505
890 91 92
570 567 - 5
23 582
215 622 641 598
58
59
860
585
126
581
82 161
61 62 569
604 576
63
253 243
498
65
618
692
667
455
583 96 249 97 674 98 537 99 - 900 555
1 663 2 310 3 572 4 589 5 654
6 657 907 . 607
351
37
636
34 442 35 500 36 599 37 434 * 38 564 839. 415
68 69 225 870 627 71 407 72 566 873. 597
38 185 39 - 940 941 . 539

Page #131

118
Ed
Ed
Ed
Ed
952 . 610 53
942. - 43 167 44 - 45 533 946. 631
947. - · 48 592
49 680 950 565 951. -
957. 546 58 .59 - 960. -
54
-
55 956.
- -

Page #132

119
APPENDIX III
For an explanation of the purpose of this concordance, see p.
Third S-I recension
Vulgata (Gangadhara-version)
2 3 4 5 6
Jaina-recension
(R)
1
7
236
19 616 619 626
45 627
628 631
.
.
.
.
.
.
. 626.

Page #133

120
Third S-I recension 30 .
1
Vulgata (Gañgādhara-version)
2 3 4 5 6 . . . . .
Jaina-recension 7 (R) . 626. 631 605 629 685 630 693 638
67
63
640
646
647
499
648 661 681 664
636
676
651 689
641
690
692 647 653
637 697
649 651 652
699 704 702 53
.
20
564
613
270
222 24
117
22
• 26
.
.
.
.
.
.
.

Page #134

121
Jaina-recension
Vulgata (Gangādhara-version)
2 3 4 5 6 . . . . .
7
Third S-I recension 1
. 27
28
(R)
64
.
.
562
190
66 .
.
.
.
.
.
.

Page #135

122
Third S-1 recension
1
Vulgata (Gangadhara-version)
2 3 4 5 6
7
Jaina-recension
(R) 75
98
99
72
127
1
253
168
103
106 116
104
105
117
107
108
118 120 297
345
124
282
110
235 123 125
111

Page #136

Third S-I recension
132 33
34
* ទី ១១ ១ ១៩៦៦ ១៩៖១ ៦ ១.១៦១៩៩ នន ៩៩៩ ។ ៩ ៖ នគ
35
36
37
38
39
140
42
43
44
45
46
48.
47
150
52
53
55
56
57
58
59
160
61
62
63
-64
165
Vulgata (Gangadhara-version)
1 2 3 4 5 6 7
. 112.
118
121
123
129
128
132
146
141
147
149
150
155
156
159
160
162
163
165
167
164
166
186
169
174
179
175
176.
123
251
310
Jaina-recension
(R)
126
133
136
140
141
200
352
347
353
355
356
360
361
253
364
365
367
368
370
372
369
371
390
373
378
380
379

Page #137

124
Third S-1 recension 166
1
Vulgata (Gangadhara-version)
2 3 4 5 6 . 177. . . . .
7
Jaina-recension
(R) .
154
68
180
381 382
181
182
383
183
696
384 388 108 127
192 193 195
109
198
199
201
147 148 150 151 153 157
202
204 206 210 207
161
158 446
444
223
211
173 162 166
216
218
219
221
168 169 171 175 179
225 229
228
178
231
181
232 235
182 185
240
189
.
.
.243 .
.
.
.
.
193

Page #138

125
Third S-I recension
1
Vulgata (Gangadhara-version) 2 3 4 5 6
7
Jaina-recension
(R)
200
.
.
.
.
.
.
. 246 .
247
195 197
249
206
486
523
201
252 254
203
265
217
220
268 271
224
267
219
226
273 274
227 228
276
460
467
229
275 280 287 289 290
233 240 277 278 279 280 282
291
292
294
532
493
295
283
302
297 298
245 285 242
300
244
304 . 312 .
247 255
.
.
.
.
.
.

Page #139

126
Third S-I recension
1
7
Jaina-recension
(R)
Vulgata (Gangadhara-version) 2 3 4 5 6 . . . 313 . . .
316
234
.
.
.
256
259
578
572
318
261
324
267
265
322 325
268
328 327
272 271
579
573
331
275
326
270
434
618
286
332 333
110
334
563
338
340
341
342
359
343 . 347 .
293 294 314 295 300
.
.
.
.
.
.

Page #140

127
Third S-I recension 268 .
1
7
Vulgata (Gangadhara-version)
2 3 4 5 6 . . .351 . . .
348
Jaina-recension
(R) .
301
305
69
270
349
354
302 309 310
355
360
361
n仍仍叫四四四四四
315 316 317
25
320
367
323
368
280
327 344 330
369 371 373 376
332
335 562
523
377 378
336 337
392
393
394
290
191
379 380
395 396 397
688
381
383
385
398 400 402 404 405 406
387 388 389
392 .393.
300
411 412
301
.
.
.
.
.
.

Page #141

128
Third S-I recension
1
7
302
.
Vulgata (Gangadhara-version)
2 3 4 5 6 . . . . 395 . . . 203
397 400
404 407
Jaina-recension
(R) . 414
152 415 418
412
416 417
689
418 419 420 421
426 427 428 429 432
425 431
433
565
554
266
353
566 556 570 572
441 325 564 598 218 308 435 600 440 443 442 156 204 241 321
571
574
575
438 . 439.
.
.
.
.
.
.

Page #142

129
Third S-I recension
1
7
Jaina-recension
(R)
Vulgata (Gangadhara-version) 2 3 4 5 6 . . . 445 . .
450
336
.
.
.
447 458 659 444
459
461
462
452 451 454 455 457 461 462 469
464 468 469 476 470
463
479
472 474
480 482
475
476
483
691
636 485
477
478
486
491
531 534
496 498
536
500
538 541
505
539
502
543
507
513
514
479 480 483 485
515
522
663

Page #143

130
Third S-I recension 370
1
7
Jaina-recension
(R)
.
.
Vulgata (Gangadhara-version)
2 3 4 5 6 . . . . . 484. .
487 488 489 492 481 482 496
521 524 528 529 531 532 533 455 537. 538
503
504
505
539
506 507
540 541 543 546
508
509
548
510
549
513
552
556
516 517
3
557
518 519
521
558 560 565 566
525
582
585
577
588
580 585
590
592
589 593
593
597
389 517
.
.
.
.
.
.
. 610,

Page #144

131
Third S-1 recension 404
1
Vulgata (Gangadhara-version)
2 3 4 5 6 . . . . .
Jaina-recension
(R)
.
7 : 609. 612 617
495 544 591
618 619 621 626
595 602
627
526 574 692
643
644
693
654
695
645 656 658
660 661
664
669
708
244
194
114 130
112 129 143
410
512
405 408
.
.
.
.
.
.
.
.
673

Page #145

132
Third S-I recension 438
Vulgata (Gangadhara-version) 2 3 4 5 6
Jaina-recension
(R)
1
7
.
.
.
.
. 409.
.
.
39
422
583 430
440
423 428
434
429 437
549
550
568 570
551
586
553 559
356
606 311 473 475
466 468 497 498
456
527
615
582
501
502
522 524
535 536 561 563 570 567 578
577
583
586
589 594 595
584 592 706
602
545
607 614 . 620 .
581
.
.
.
.
.
.
596

Page #146

133
Third S-I recension 472
1
7
Jaina-recension
(R)
.
Vulgata (Gangādhara-version)
2 3 4 5 6 . . 263. . .
279 281
.
.
73
215 232 234 236
252
319
262
284
288
335 358
313
366
322 324
440
328
370
329
152 154
187
391
386
403 436
190
106 230
301
519
520

Page #147

Third S-I
recension
506
7
8
9
510
ཋ ཌ — ྋ ྋ ཋ ཋ ཌ — 2 8 ཀླུ 8 མ ¢&8སེསྣུམ་ཊསྐྱསྐ࿄88རྐྱ ྂ
11
12
13
14
17
19
520
22
23
27
28
29
530
31
32
33
34
35
37
38
539
Vulgata (Gangadhara-version)
1 2 3 4 5 6 7
86.
88
99
103
120
133
143
145
184
196
197
170
215
220
226
233
277
134
217
305
306
323
430
441
573
520
Jaina-recension
(R)
90
93
105
115
349
351
385
145
146
165
170
176
183
230
248
249
374
167
266
223
111
326
559

Page #148

135
Jaina-recension
Third S-I recension 540
1
Vulgata (Gangadhara-version)
2 3 4 5 6
7
(R)
224
174
124
137 138
125
139
126 142
348
237 258
187 210
259
212
278
231
415 427
423 433
465
472
546
510
320
263
545
509 163
212 293
281
307
250
255
255
207
573
.
.
.
.
.
.
.
.

Page #149

136
Third S-I recension 574
1
Vulgata (Gangadhara-version)
2 3 4 5 6
7
Jaina-recension
(R)
56
284
237
350
303 205
576

Page #150

Third S-I
recension
608
9
610
11
12
18
19
2 2 2 2 2 2 2 ≈≈≈ 8 ☹ ≈≈ ----≈≈
620
21.
23
24
27
28
630
32
640 641
1 2
73.
78
8888
84
91
98
Vulgata (Gangadhara-version)
122
101
109
106
119
131
137
140
139
144
161
148
153
173
3 4 5 6 7
205
137
208
209
213
222
234
238
Jaina-recension
(R)
77
82
88
95
104
135
113
121
119
134
338
342
346
345
350
366
155
354
358b
377
159
160
164
172
184
188

Page #151

138
Third S-I recension 642
1
7
Jaina-recension
. (R)
Vulgata (Gangadhara-version) 2 3 4 5 6 • 250 . . .
256
.
.
199
44
257
208 209 211
260
269
262
315
221 214 258 276 341
288
136
48
52
299
303
246
314
257
317
260
336
289
382
344
399 296 401
384
699
122
514
553
410
391 374
333
396
420
.
.
.
. 399 .
.
.
.
417

Page #152

139
Third S-I recension
1
Vulgata (Gangādhara-version)
2 3 4 5 6 . . . 375 . . .
7
Jaina-recension
(R)
676
.
.
334 643
77
413
421
79
432
409
569
438
453
512
460 551 466
459 470 473
477
481 597
548
530
690 91
490 534
493
587
579
390
407 449
526
542
529
489
.
. .
.
.
.
.
. 616 .
590

Page #153


Page #154

PART II

Page #155


Page #156

CHAPTER 1
1.1 The nature of the text
The SattasaT is a compilation of 700 (saptaśata) Gāthās, or āryā-verses, in Prākrit. Their content is erotic, depicting, for instance, lovers' attempts to meet, various ways of sexual intercourse, unfaithfulness, lovers' quarrels and their reconciliation. The situations in the majority of the Gathās are set in the country-side, descriptions of city and court-life being conspicuously absent. The people depicted are those inhabiting the country-side such as farmers, hunters, tribals and warleaders charged with the protection of the villages. The Gathās are monologues, mostly spoken by, or describing, these peoples' women. The compilation of the text is traditionally attributed to a certain king, named Hāla, to whom I shall return later. This and the fact that the Gathās are composed in a language which, but for its Middle Indic appearance, is basically a form of Sanskrit, and in a style which is highly literary or Kāvya-like, suggests that the text was intended for a cultured audience of 'connoisseurs'. This conclusion, which is amply borne out by internal evidence, for which see below, raises the question what interest this audience could have had in a text which basically describes life and in particular love and sex in an environment so completely different from that of their own. Most scholars in fact seem to recognize in the Gathās a kind of romantic attitude on the part of the poets and the audience towards life in the country-side. Thus, according to Weber the text was the product of 'feingebildeter, wo nicht überfeinter, Dichter aus der Stadt, die ihrerseits das Landleben, wie die Poesie der Natur, die Liebe speciell, verherrlichen' (Ed., p. X). Keith maintains that 'the prevailing tone is gentle and pleasing, describing simple loves set among simple scenes, fostered by the seasons, for even winter brings lovers together, just as a rain-storm drives them to shelter with each other' (Keith, 1928: 223). Compare, finally, Warder according to whom the life in the villages depicted in the Gathās 'was full of hardships, but not without compensating pleasures, especially those of love' (Warder, 1:974: 182). As I will show below these views are widely off the mark. For instance, the people in the villages are almost throughout depicted

Page #157

144
as simpletons and misfits, especially where sexual matters are concerned, while the tone is in fact far from gentle, but rather superior and patronizing. In order to clarify this some of the situations found in the Gathās will be considered more closely below. This should at the same time give a clearer picture of the circumstances, ideals and ambitions of the audience for whom the text was evidently intended.
1.2 The situations
A recurrent theme in the situations of the Gathās is the foolishness of the people depicted. Thus, the Pulindas, tribals living in the Vindhyamountains, mistake dark rain-clouds for elephants (116). Frequent references occur to the pāmara, a poor farmer." In 238 he barters his cloak for an ox because he is fooled into thinking that his wife's dark breasts, which remind him of heaps of smouldering ashes, will keep him warm during the winter. His stupidity furthermore consists of acquiring an ox when the ploughing-season is still far off. In fact his foolishness is almost proverbial. Compare Gāthā 643:
kamalan muata mahuara alekkhaladduaṁ pā-
darapikkakavitthagandha lohena maro vva chiviūna jānihisi.5"
Bee, if you abandon the lotus for the kapittha, attracted by the fragrance of their as yet half-ripe fruits, you will soon find out (your mistake) after you have touched them, like a pāmara touching a sweetmeat in a painting.
The halia or 'ploughman' is equally foolish. Compare Gathā 360:
ajjań mohaņasudhia daraphudiavetabhārā-
maa tti mottu palaie halie harāni hasian va phalah Thi.
When the ploughman ran away from his genteel wife (or mistress), who had fainted from love's extasy, thinking her dead, the cotton plants seemed to laugh with the lips of their bulky fruits slightly opened.

Page #158

The people depicted in the Gāthās have in common that they need to work in order to make a living, which leaves them little time and energy for love. Their women suffer accordingly. A case in point is Gāthā 324:
cikkhillakhuttahalamuha
appattamohagasuha
145
kaddhanasudhie paimmi päsutte ghapasamaan pamar
savai.
The poor farmer's wife who failed to reach her climax
as her husband had fallen asleep exhausted from drawing the ploughshare through the thick mud, cursed the rainy-season.9
On this theme several variations occur. For instance, the hunter, having married recently, discovers that he is no longer able to draw his bow, weak from making love with his wife. He has to shave the bow thinner. See, for instance, Gāthā 120:
ukkhippai mandalina- ruega gehangatahi vähTe sohaggadhaavaḍāa v- va uaha dhanurufparicho1T.
Look at that! From the courtyard of the hunter's house a whirlwind throws up a line of shavings from his bow, which looks like a banner advertising his wife's (conjugal) happiness.
ekkallamao dīvaapiajaassa jaha dhanush
As a result the arrows will, of course, lose much of the required force. Furthermore, love has made him weak-hearted. He comes to lack ruthlessness which is necessary in order to kill. Compare Gāthā 620:
mate taha lalio saanhãe padiah vahassa hatthão.
10
The stag wandering alone was fondled by the deer in heat, used as a decoy, in such a way that the hunter, who had a loving wife himself, dropped his bow.
As a result the hunter's family is threatened by poverty and starvation.

Page #159

146
In Gātha 951 a trader in skins has to be sent away empty-handed as skins are no longer available in the hunter's house after his marriage.
The scars on the chest of the war-leader, contracted in performing his duty of protecting the village, make him an unfit lover as his wife cannot fall asleep on account of their roughness (see Gatha *31). In Gathā 687 the husband's victory in a wrestling-contest indicates his lack of interest in his wife:
āņard amtena tumam malli na lajjasi naccasi
paino pahaeņa padahas addeņa dohagge pāadijjarte.
.
Woman, shouldn't you be ashamed instead of dancing, as the drum beaten to celebrate the victory of your husband in a wrestling-contest proclaims you unhappiness.
A large number of Gathas concern the itinerant merchant (pahia) and, his wife. They deal invariably with the husband's absence from home, actual or imminent. This is a common theme in Indian erotic poetry. In this text these Gathās provide yet another example of the above-mentioned theme, namely of the impossibility to combine work and love. The travels of the merchant are for him a necessary means of earning a living. He is reluctant to leave his wife (see Gātha *25). During his travels he has to undergo numerous hardships including a ruthlessly hot climate. Furthermore he misses his wife. See, for instance, Gātha 641:
agghāi chivai cuộvai jāakavolasaric
thavei hiaammi janiaromasco chi e pahio mahuapuppham.
The traveller smells at the madhūka-flower, touches it, kisses it and presses it against his heart, while all the time his hair stands on end because of its resemblance to his wife's cheek.
The majority of Gāthas concerned depict the fate of his wife and especially her utter unhappiness during his absence from home. Not all details need be treated here. One example (670) should suffice:

Page #160

147
pariosasundarain tais cia una virahe
suraesu lahaṁti jāi sokkhāi khāugginnai k Tramti.
The pleasures of love-making, which gave so much satisfaction, are during his absence turned into stale memories.
Other Gathās deal with farmers, notably gahavais and halias. The gahavai or 'lord of the house' seems to be a member of the landed gentry. The halia or 'ploughman' is apparently a small farmer who like a sharecropper partly makes a living by hiring out his labour. Despite the obvious differences in their social positions, they are seen to exchange brides. The gahavai gives his daughter in marriage to the halia, thus placing the girl in a very difficult position as appears from, for instance, Gathā 602:
maṁdan pi na ānai hali- gahavaisua vivajjai
anamdano iha hi daddhagāmammi avejjae kassa sāhāmo.
The ploughman does not understand (a thing, not) even slowly (or: does not even know what illness is). His genteel wife is dying but who can we tell, for in this wretched village there is no healer for her ills.
The gahavai's daughter is complaining in Gathā 164:
vaink a ko pulaijjai kena samaṁ va hasijjai
kassa kahijjai suhaṁ va dukkham va pāmar apaure haaggāme."
Who is there to glance at, who is there for me to talk to about my pleasures and my griefs, who is there to have fun with, in this dreary village full of yokels?
After her marriage the halia's daughter, coming from a simple family, encounters problems in her new environment. One thing she lacks is coquetry." In order to attract her husband's attention she behaves rather clumsily. E.g. Gātha 107:

Page #161

148
golāadatthian pec- adhatta uttarium
chiūņa gahavaisua haliasonha dukkhuttārāi paavte.
When the farmer's daughter saw her husband, the son of the householder, standing above the sloping bank of the Godavart, she began to clamber upwards along a (steep and) difficult path. 16
The gahavai thus marries beneath his social position, while for the halia the marriage with the gahavai's daughter means a step upwards along the social ladder." However, the gahavai also marries his daughter to the gāman I or war-leader. For the gamant this marriage proves a misalliance as his wife is unable to fulfill his ambitions, which consist, among other things, of having her commit Satt at his funeral. Thus, in Gatha 407 she, however inadvertantly, spoils this ceremony by sweating and thus extinguishing the fire. It should be noted that in the context of lyrical poetry this sweating is in itself quite legitimate:
vijjhavijjai jalano anumaranagh analingia-
gahavaidhui vitthaas iho vi piaamasuhasijjirang Te. 19
The gahavai's daughter extinguishes the fire, its flames already shooting up high, as she sweated from the pleasure of embracing her dearest one at the occasion of following him in death.
In this Gāthā the gamant is not specifically mentioned. In the following Gātha (449) he is, again at his funeral which was meant as a spectacle during which all his wives were expected to commit Satt. This time the SatT-ceremony fails because the villagers among whom he is living are unable to appreciate the ceremony, having only eyes for his favourite wife:
gāman ino savvāsu vi mammaccheesu vi val
piāsu anumar anagahiaves asu lahāi uvar in valai ditthT.
Though all the war-leader's wives are beautifully

Page #162

149
attired for the occasion of following their husband in his death, at the most painful moment all eyes are directed towards his favourite.
Most of the other situations are common to Indian erotic poetry in general. However, they agree with those discussed above in that they likewise deal basically with the impossibility of lovers to have a satisfying relationship. Before marriage a girl is closely watched by her parents who do not allow her to go out. She peeps through the holes in the fence at the boys passing her house but experiences considerable frustration as they do not respond. If she succeeds in attracting one boy's attention and they decide to meet, they have to do so in the fields (9) or in the dense shrubbery along the river (103). All this, however, provides hardly a romantic experience. The lovers are in fact constantly worrying about the fields which will one day be harvested or about the shrubbery which will in time become less dense. The rainy season is supposedly the ideal time of the year for meeting in secret because then the nights are really dark (249). However, the paths leading to the place of meeting are slippery and dangerous in that season (231, 445). • Married life is likewise represented as full of problems. After the marriage ceremony the girl is taken by her husband to his own house where he lives with his parents, his brothers and their wives in a joint family. This living together is dictated by economic circumstances. The woman is not supposed to entertain extravagant wishes, which would put too great a strain on the resources of her husband's family. This appears to be difficult when she is pregnant and suffers from cravings for rare and expensive things (dohada) (172). The woman is expected to restrain herself in every other respect as well, thus becoming a kulavahū.
Sex itself fails to provide real pleasure. For instance, the woman soon discovers that her husband is more quickly exhausted than she is, leaving her unsatisfied (155, 413). When she wants to continue, taking the initiative (viparitasurata), she merely annoys her husband (see *23).
A considerable number of Gathās deal with the husband's unfaithfulness which in the first place causes great grief to his wife. She takes revenge by sulking (māna), causing, in turn, considerable frustration

Page #163

150
to her husband, who in the end is forced to throw himself at her feet in an attempt to propitiate her.
1.3 The audience
It will be clear that the picture given of the life in the country-side is far from flattering. The text would thus most likely have been intended for people who were not themselves depicted in the Gathās and who, intelligent and living in completely opposite, i.e. better circumstances, could afford to have an interest in the situations described. This audience must have belonged to the cultured and refined leisure-class living in the cities and towns, which brings to mind the so-called Nāgaraka as described in Kāmasūtra I 4. The audience must also have included kings, as :) the ideal warrior, which appears from the descriptions of the gāmaņ1. For, these seem to have been intended for a more succesful member of the warrior-class, namely one who has no problems in having performed the proper rituals and who, furthermore, has established peace, thus having time and energy for things other than waging war, namely love and, as the existence of this text serves to show, enjoying literature.
It is evident that the text served in the first place to amuse the audience, which raises the question as to what the humour consisted of. It would seem that it was in any case based on a certain measure of contempt for the people living in the country-side. The main purpose of the text, though, would seem to have been that of providing an outlet for the audience's own frustrations, presumably caused by the painful but inevitable clash of their ideals and ambitions with the realities of life. The text may have acquired this function in that it projects the audience's problems onto a completely different world with different people and different customs. In this way these very problems could become objects for amusement. Whatever the precise nature of the perception of the text may have been, it contains numerous specimen of often crude humour, which is otherwise extremely rare in Indian erotic poetry. An instance of this humour may be found in Gathā 324, quoted above (p. 145), where the farmer's exhaustion which makes him unable to satisfy his wife's desires is attributed, of all things, to his having dragged the plough through thick wet mud. Another example may be

Page #164

151
found in Gathā 360 (see above, p. 144) where the farmer is surrounded by slightly opened cotton-fruits which seem to refer to the state of his mistress from whom he is trying to run away.
1.4 The compiler
As already mentioned above the compilation of the SattasaT is attributed to a certain king, called Hāla. This Hāla is traditionally identified with a sātavāhana king of that name, who is supposed to have ruled for a brief period in the beginning of the first century A.D.C4 Compare verse 13 of the introduction of Bāņa's Harşacarita, which seems to contain the earliest reference to the Sattasaf (seventh century A.D.):
avināśinam agrāmyam akarot sātavāhanaḥ (v.1. sālio) viśuddhajātibhiḥ kośaṁ ratnair iva subhāşitain.
Sātavāhana made an immortal and refined kośa with Subhāşitas consisting of pure characterizations, as if it were a treasury filled with the purest pearls.
It is commonly assumed that Bāņa refers to the Sattasas here. This assumption finds support in his use of the word agrāmya, which with reference to the Sattasaf acquires a double meaning, referring to the subjectmatter of the text, which is village-life (grāma), on the one hand, and to the result, which is a kośa for the cultered city-man (agrāmya), on the other.
The ascription of the compilation of the text to a Sātavāhana is supported by internal evidence. For instance, the Gāthās describe precisely the realm of this dynasty, namely the area roughly coinciding with the present state of Maharashtra. Thus, they often refer to the Vindhyamountains (70, 115, 116, 117, etc.) and to the rivers Narmadā (549, 760) and Godāvart (58, 103, 107).25 The Sātavāhanas are indeed traditionally represented as patrons of Prākrit literature. It is on the other hand also possible that this tradition was in fact based on their association with the Sattasaf. In this connection it is interesting to note that several of the names of the 'authors' which in some MSS are found with

Page #165

152
the individual Gathās belong to kings of the Sātavāhana dynasty and their successors, the Vākāțakas and subsequently the Rāştrakūtas. There are reasons to assume that the names were added to the text in a relatively late stage in its transmission (see above,p. 76 ff.). Their presence does, however, show the persistence of the traditional association of the text with Maharashtra kings.
1.5 The language of the Gathās
The Sattasaí is generally assumed to be the earliest text showing the use of a Middle Indic or Prākrit dialect for purely literary purposes." This particular dialect later became known as Māhārāştri. Initially Māhārāştrī was closely connected with the erotic-lyrical genre to which, the Sattasai belongs. Thus, it is found in the dramas, in which those characters who otherwise speak SaurasenT Prākrit, used Māhārāş trt for their verses. These verses concern mainly descriptions of nature and its effects on the characters' moods. The practice of using such Maharaştrt verses was begun by Kalidasa. The fact that verses in Maharastrt were absent from the earliest dramas by Aśvaghoşa suggests that they are indeed innovations in the work of Kalidasa. Subsequently, however, Māhārās trī developed into just another literary language beside Sanskrit; it came to be used in epic Kavyas (e.g. Setubandha and Gauqavaho), while the Jains used a mixture of Māhārāştrt and Ardha-Magadht for their stories.
In this connection it may be asked what determined the use of a Prākrit dialect for literary purposes in the first place. In the case of the SattasaT its use is mostly taken for granted, the text having been compiled, if not composed, by a king of the Satavāhana dynasty, which for their inscriptions and coins used a Middle Indic dialect rather than Sanskrit. However, the fact that the Sātavāhanas used a Middle Indic dialect for administrative purposes does not mean that they were ignorant of Sanskrit or that they would not have had the possibility of choice between Sanskrit and Prākrit.” Presumably they merely did not consider the use of Sanskrit for administrative purposes."
I would therefore like to put forward the following alternative and in many respects speculative explanation. Above it has been argued

Page #166

153
that the Gathās were intended for the amusement of a rich and cultured audience. The humour was at the cost of the simple and somewhat foolish people inhabiting the villages of Maharashtra. The Gathās themselves are mainly monologues spoken by these villagers. In this connection the use of a Prākrit dialect may well have been determined by the intention to give their words a rustic and local flavour. This explanation presupposes pride in the knowledge of Sanskrit on the part of the audience as well as of the authors." The above view seems at least partly to be supported by the way Prākrit dialects were used in the dramas, providing an indication of the character's status being lower than that of one speaking Sanskrit. -

Page #167

154
Notes to Chapter 1
It goes without saying that this observation is based only on the Gāthas that can positively be traced back to the MS of the archetype, i.e. those that are found in the Third South-Indian recension, on the one hand, and the Jaina-recension and/or the Vulgata, on the other. Nevertheless there is a small number of Gāthās, which can be traced back to the archetype as well, that appear to be exceptional, namely those which mention prostitutes and thus seem to describe life in the city (e.g. *21, 156 and 474). However, their number is so small that they need not affect the argument.
Exceptionally a Gatha contains a dialogue, as, for instance, in 384, 676 and 950.
This example and the following contain instances of the so-called Bhrāntimadalamkāra. Other instances of this figure of speech are found in 531, 532 and 640.
The word pāmara literally means 'having a skin-disease (pāman)'. In Gatha 331 the man is indeed scratching his head.
Pada bas found in Ti, Tp and T. For alekkhaladduan Ti and Tp have alekkhe ujja (unmetrical); compare alokhkan ujjaya in R and alekkha ujjua in X and S. Perhaps the text should read alekkhe ujjam, in which ujjam represents Sanskrit urjām 'refreshment'.
The reading is based on Ma, Ti and Tp. bhārāharāhi consists of bhāra and adhara.

Page #168

155
For ajjā 'genteel woman', from Sanskrit aryā, see Thieme (1938: 99-100). For the halia being married to a woman of a higher social position, see below.
sudhie as found in Ma, Ti, Tp, Bh and R.
Note that the rainy season is otherwise associated precisely with conjugal happiness. Yet another point in the Gāthā is the stupidity of the woman. Though she curses the rainy season it is, with the harvesting season, the only time of the year when her husband has ample work and consequently enjoys income.
10
The text is based on Ma, Ti, Tp, Bh, R and T.
11
This Gathā is absent in the Third South-Indian recension.
12 . săricchide as found in Ma and Tp.
13
See Fick; 1897: 164.
14
The text is based on Ma and Tp.
15 This applies to the simple villagers in general. Note the complaint in the preceding Gatha: vaṁk a kopulaijjai...pāmar apaure haaggāme.
16
The meaning of the Gathā becomes clear only from a comparison with 193:

Page #169

156
golāvisamoārac- anuaṁpān iddos aṁ
chalena appā urammi se mukko teņa vi sã gādham uvaudhā.
Under the pretext of going down to the Godāvart along a rough path she throws herself against his chest, and he embraces her tightly which seemed innocent as he did so (ostensibly) out of pity.
17
The gahavai gave his daughter in marriage to the halia to secure his labour during the ploughing and harvesting seasons, thus sacrificing his daughter's happiness for his need for labourers.
18
The gahavai and gāmant, though obviously representing two different life-styles, would seem to be socially equal. In any case the gahavai. will need the services of the gāmaņt to protect his interests and where necessary to enforce his rights. To be certain of the services of the gaman t the gahavai apparently has to establish a relation with him through marriage.
19
alingia as found in Ma, Ti, Tp, T and S.
20 My interpretation of this Gatha differs considerably from Weber's, according to whom it is the eyes of the gaman which even after his death were drawn to his vallabhā. On the other hand, the ceremony is a clear instance of potlatch. Note that the women are especially dressed up for the occasion (aņumar anagahiavesasu). Therefore ditthí should rather 'refer to the eyes of the spectators. Their special interest in the vallabha lies in the fact that when they want to ask a favour of the gamans they will usually address him through her as his favourite wife

Page #170

157
21
With this assumption I disagree with, for instance, Warder who is of the opinion that the Gathās were in the first place composed by the villagers, and by implication were meant for an audience consisting of villagers. Subsequently the Gathās would have been copied and imitated by poets and scholars from the city (Warder, 1974: 182). In fact, the exclusive preoccupation seen in the Gathās with the villagers' poverty and stupidity and with the awkward situations arising from these factors, make it hardly likely that the Gathās were ever meant for their amusement. It should be noted that Warder on the same page-writes that the Gathās are 'songs of the villagers, in which their joys (sic) and sorrows are recorded'.
Compare Kosambi who in the introduction to the Subhasitaratnakoşa (p. XLVIII) wrote that 'the new poetry... must have aroused amusement and that tinge of good-natured contempt for the hālika, which we find even in the Sattasat. However, in my opinion Kosambi presents only a limited view of what the text is about.
23 See Gathā *3. The word kaivacchala (kavivatsala) would indicate that Hāla is a king.
24 The name Hāla figures in the Purāṇas in the lists of the Andhrabhrtya kings, for which see Pargiter (1913: 35-43). The Purānas call the Sātavānanas Andhrabhrtyas, referring to the time when their realm beside Maharashtra included Andhra as well (see Gairola, 1956: 161-2). The dates of Hala are computed on the basis of the Purāņic lists. Two Sātavāhana coins have been found bearing the name Hala (Rama Rao, 1971: 37).
25 Other rivers mentioned are the Yamunā, the Tāpt and the Murală, but in Gathas which are restricted to the North-Indian (Yamuna in 671,

Page #171

158
Tāpt in 239) or the South Indian branch of MSS (Murală in 876).
26
Lévi, 1902: 111-2.
27
See, for instance, Keith (1928: 30): '... the second century A.D., in which century Māhāraştrt lyric began its career, made known in the anthology of Hāla: Earlier literature in a Middle Indic language is invariably of a religious nature, either Buddhist or Jain. Occasionally these religions produced 'literary' literature as well. See, for instance, the Thera and Ther T-Gathās, from which, however, religious ideals are never completely absent. On the fusion of religious or ascetic tendencies and the erotic in the Thera and TherT-Gathās, see Lienhard (1975).
28 The name Mahārāştrt turns up only relatively late. It does not occur in the probably oldest description of the dialect in Vararuci's Prakrtaprakāśa, but is mentioned only in chapter 12 of that grammar (sutra 32: śeş am māhārāştrivat), which is almost certainly a later addition (Nitti-Dolci, 1938: 19).
29
Keith (1928: 30) writes: 'Of the Prākrits Māhārastrt held pre-eminence by its use in the dramas, whence it was introduced by Kalidasa from lyric poetry' (by which he refers to the type of poetry as found in the Sattasaf). The influence of the Sattasas on the works of Kalidasa has been the subject of an extensive study by G. L. Hart III, The Poems of Ancient Tamil. Their Milieu and their Sanskrit Counterparts, Berkeley, 1975.
30
For the Satavahanas exclusively using Prākrit for their official documents, see, among other, Lévi (1902: 110)
31
In this connection a legend has been preserved which seems to have an

Page #172

159
historical basis. It is told that the Kātantra grammar of Sanskrit was composed for the benefit of a Sātavāhana king who knew only Prākrit and had difficulties in conversing with his wife who spoke Sanskrit. Lévi (1902: 112) argues that this legend could actually refer to king VāsişthTputra, who was married to a daughter of the Mahākşatrapa Rudradāman'. His wife composed the only Sātavāhana inscription in Sanskrit whereas all the other inscriptions of this dynasty are exclusively in Prākrit.
32
In any case, as pointed out by Lévi (1902: 112) the language of the Sātavāhana inscriptions is, but for its Middle Indic phonology, extremely close to Sanskrit: 'Les scribes qui rédigeaient en prācrit les panégyriques royaux...n'avaient à faire qu'un effort insignificant pour tourner la louange en sanscrit; ils effleurent le sanscrit de si près qu'ils semblent s'en garder plutôt que d'y prétendre.'
33
In whatever way one looks at Māhārāştrt Prākrit it is fundamentally a form of Sanskrit, differing from it almost only in pronunciation. Contrast Apabhraća, which represents a thorough breach with Sanskrit, having an entirely different morphology. In the Sattasaf the local, i.e. Maharashtrian, element is mainly present in the vocabulary. Garrez (1872: 203-7) has shown that the Gathās contain several words which are typical of Marāthi only. Some of the words collected by him appear much more widely distributed in NIA than he supposed. E.g. bhana-, for the NIA derivatives of which, see Turner, CDIAL 9383. Its use in the Gathās .is, however, indeed foreign to Sanskrit. Garrez was correct on, for instance, dāve- 'to show' (CDIAL 6276) and rumda 'large, broad' (10781). To these words may be added saajjhia 'neighbour' (see Gātha *39), Marathi sejf 'female neighbour' and sejār 'neighbourhood' (13077); acchabhalla 'bear' (Gathā 109; see also the note to Gatha *4), Marāțht ăsval, ăsvil, asol 'id', and Konkant āsvel (f.) 'id' (2446); and tuppa 'ghee' (also found in AMg. + M.?), Marāght and Gujarātī tūp 'id' and Marāthi tupat 'oily' (5864) (possibly a loanword from kannada; see DED 2685. Turner, loc. cit., suggests that Prākrit tuppa goes back to RV trpra 'hasty?').

Page #173


Page #174

CHAPTER 2
2. 1 Method of reconstruction
In the edition (Gathās 1-50) I have attempted to reconstruct a text as closely to the original as the available MSS allow. The arguments used in the reconstruction are basically derived from the stemma. The MSS, which ultimately go back to one single MS, the archetype, are divided into two branches, consisting of the Third South-Indian recension, on the one hand, and the Jaina and Vulgata-recensions, on the other. Another result of the stemma is the conclusion that the order of the Gāthās is most faithfully preserved in the latter two recensions. However, each of the two is characterized by certain transpositions of larger as well as smaller sequences of Gathās of the text of the archetype. This text can only be established again by identifying these sequences and restoring them to their original positions. In doing so the Third South-Indian recension plays an important role as its text was based on a selection from that of the archetype. Traces of the order of the Gathās of the archetype can still be clearly perceived. However, such a reconstruction can only be partly successful. For example, in the Vulgata a group of Gathäs somewhere after 433 was apparently displaced and transferred to the sixth Sataka (see the point opposite 321 in Appendix III. Note that the order in the Jaina-recension is considerably disturbed as well). It is impossible, though, to determine whether the part of the text which was displaced began immediately with 434 or only with 435. A suitable principle for dealing with such uncertainties remains to be found.
In the part of the text edited here no serious problems in reconstructing the original order of the Gāthās occur. The differences between the Jaina and Vulgata-recensions consist only of some transpositions of individual Gathās in the Vulgata. Thus, Gāthās *31, *32, *36 and 47, which in the Jaina-recension (R) occur as 31, 32, 36 and 47, are in the Vulgata (G) found as 188, 194, 189 and 190 respectively. Their occurrence in the Third South-Indian recension as 69, 70, 74 and 84 respectively shows that their position in the Vulgata at the end of the second Śataka is the result of some independent development in that recension.
A different matter is the question which Gāthās belonged to the text

Page #175

162
of the archetype and which represent later additions. Only those which occur in the Third South-Indian recension, on the one hand, and in the Jaina-recension and/or the Vulgata, on the other, can positively be traced back to the archetype. One way of editing the text is by first giving only the Gathās which are clearly authentic, with the remaining Gathās following in appendices arranged according to the position in the stemma of the MS or MSS in which the Gathās occur. The basic text would then probably contain not more than approximately 500 Gathās. The alternative, which I have opted for here, is to take one complete text which is close to that of the archetype as the basis, in this case a combination of the Jaina-recension and the Vulgata. This text will include Gathās of which the authenticity is doubtful. Some of these may possibly go back to the archetype. In the part edited here altogether four Gathās are missing in the Third South-Indian recension. It is almost certain that one of these, viz. *4, originally formed part of the text of the Third South-Indian recension and consequently of that of the archetype (see the note given there). It is equally certain that Gathā *40 is a later addition to the text in the North-Indian branch. This follows from the occurrence of a relatively modern word, bh amāgana, which is otherwise found only in Apabhraía. This argument will turn up again in connection with 77 (absent in Ma, Ti, Tp and T (and W)) which contains the Apabhransa verb nada- 'to nag', and with 92 (absent in Ma, Ti and Tp (and W); but found in T) which contains the word phaladiā (see Bha phalagayā), a formation, like bhamāçana, typical of Apabhraía. The status of the two remaining Gathās, *18 (absent in all the South Indian recensions) and * 37 (absent in Ma, Ti and Tp (and W); but found in T), remains unclear.
As already noted earlier, for the text of each individual Gathā often numerous variants are available. The elimination of variants, involved in reconstructing the text of the archetype, is generally settled with reference to the stemma. Thus, the text almost automatically contains that which Ma, Ti and Tp have in common with Bh and R and/or with the Vulgata MSS. Not infrequently two different, but stemmatically equivalent, variants are available. In such cases the basic argument in the elimination is that of the lectio difficilior.
Occasionally, however, matters may be more complicated. Certain pro

Page #176

163
blems or ambiguities of the Prākrit text may have been solved identically in the different branches. A case in point is found in * 26, with kunati ccia... jānastT in Ma and Tp, and R and y. From the point of view of the stemma this reading should have been adopted in the present text. A closer look, however, shows that probably it is only an 'emendation' introduced independently in the two branches, for kuņamta ccia...jāņaṁto found in Bh. Another case in point concerns the variation between e and de or te, the personal pronoun of the third person plural and the enclitic forms of the personal pronoun of the second person singular. The variante is usually restricted to only one or two MSS (see the note on Gātha *15). Only once, namely in 683 is it found in practic cally all the MSS. However, this instance differs from the others in that the word e was misunderstood. In mā e mannun, e was combined with mā to form mãe, the vocative of māā 'mother'. As such it apparently escaped 'emendation' into clearer de or te. It follows that in all instances e
is a retention. A similar situation exists with regard to the variant : i beside vi. The former occurs each time only in one or two MSS, the
other MSS having vi (see the note on *17). Only in 149, in jTviasesa i, is it found in practically all the MSS but, again, presumably because it escaped 'emendation' into vi, being misunderstood as the case ending of the instrumental of jTviasesā, i.e. jfviasesai.
Moreover, certain changes in the text may have taken place more or less unconsciously in all the branches independently. A case in point is the .corruption of ruvvau (MS Bh) in Gātha * 10 into ruvvasu (Ma, Tp, R, etc.).
In all stages of the transmission the text appears to have been influenced in one way or another by Sanskrit. This is apparent in the case of variants consisting of different spellings of one and the same word. Consider the variation between deara and devara as found in, for instance, *28. Whereas Tp and and B have dearo, Ma and R, K, Y, P, T and S have devaro (Bh deuro). Considering their peculiar distribution the variation must have an external cause, which then most likely is Sanskrit. This means that devaro is secondary. In any case there is no positive evidence to assume the opposite tendency here, namely that the
text was given a more pronounced prākritic appearance, as, for instance, • by eliding the medial -v-. See in this connection what has been said

Page #177

164
above about the innovations consisting of sanskritic words against retentions consisting of typical Prākrit words (p. 29 and the note given there). A more complicated case is that of the variants chāhin (a typical MIA formation) and chāaṁ (sanskritic) found in * 41. The former is restricted to and B only. Following the stemma we would have to adopt chāaṁ found in Ma, Tp, Bh and other MSS. However, there are reasons to assume that chāhim is original, chaam having been introduced secondarily as the easier form. In this particular case another external factor, namely the influence of the indigenous grammatical tradition, may have. been involved. Where there are two variants which differ only in spelling it seems justified then to adopt in the text the one which deviates most from Sanskrit, irrespective of the position in the stemma of the MS in which it is found. Thus, in * 29 deņa and rao, restricted to T, are adopted instead of tena and ravo found in the other MSS, and in * 30 nikkia found in k rather than nikkiva in the other MSS.
The same principle is applied to such instances as bhisiņīvattammi (only in S; also in Y but in its South-Indian original -v- and -2- are hardly distinguishable; beside opattammi) in *4, pāavadiassa (only Bh; beside opadiassa) in *11, and naarahara (only K; beside .ghara) in * 32. Usually the original form of the initial consonant of the second member of a compound is maintained. In these compounds the rules of internal sandhi have been applied. While the above instances are restricted to one or two MSS only this is different in the case of pautthavaid beside pautthapaia. The latter is found only in Ma, Ti and Tp(?). It should be noted, though, that in the Malayālam script 'as in most other South Indian scripts the distinction between v and p is paleographically almost insignificant. The transcription of a word with either v or p can often only be established with reference to the etymology. Other compounds showing this internal sandhi, which can be positively traced back to the MS of the archetype, are, for instance, mahilana (Ma, Bh, R, etc.; beside jana in Tp and B) in *48, Golāada in 107 and vaṇaava (Ma, Tp, Bh; beside odava in R, K, etc.) in 117. In all these cases we probably have to do with more of less fixed compounds inherited from Sanskrit.
In this connection the following two instances may be considered in which an initial consonant is apparently lost, namely

Page #178

165
aṁdharahian in *32 in K (for gamdharahian) and icchena in * 33 in Tp (for kicchena). In both cases, though, it concerns a free-standing word. In the Note to *15 several other words have been collected which have lost their initial consonant, but those are all either particles, enclitic forms of the personal pronoun or personal pronouns. In this respect the case of icchena, which is an adverb, may be different from that of aħdhao, which is a substantive. Furthermore, amdhao in k may well be due to a clerical error; note the context: suhaa aṁdhao. As such it has a parallel in the text of Gathā 193, which in K immediately precedes it. MS K reads să ațham uaūthū (for sā (9)ādham uaūdhā). In this respect the case of icchena is, again, different. On the basis of the above consideration I have decided to place adhao of k in the critical Apparatus, but to adopt iccheņa of Tp in the text.
In what follows I will discuss several sets of variants, which occur comparatively, frequently, in which the two main branches are diametrically opposed. This is the case with, for instance, the variants didha and dadha, patthi and puttht, and silahaņijja and salahanijja, of which the former are found in the South-Indian branch and the latter in the North-Indian one. Occasionally some North-Indian MSS, however, have didha and pattht as well. The first two pairs concern different but equally possible developments of Sanskrit -r-. The third pair concerns a difference in the colouring of the Svarabhakti vowel. If the variation is indeed a purely phonetic matter the South-Indian branch should probably be followed, as I have subsequently decided to do. The South-Indian scribes would in any case have been less influenced by the actual pronunciation in their own language of the words concerned.
Another type of variation concerns the particle u(v)a(-ha) 'look' (see, e.g. *4), which occurs as uva(-ha) in Ma, Ti, Tp and T and, where available, w, and as ua(-ha) in Bh, R and the Vulgata MSS. In this connection the question arises whether in uva(-ha) we have to do with an instance of the secondary insertion of the -v- or in ua(-ha) with an instance of its secondary loss. It should be noted that in the MSS of the South Indian recensions one does at first sight not otherwise find evidence of the automatic insertion of a -v- in the combination ua (e.g. uaa (udaka), sua (śruta) and lua (*luta, beside Sanskrit lūna)). If the -v- was secondarily inserted here the etymology must have played a role.

Page #179

166
However, for u(v)a(-ha) no etymon is or ever was available (traditionally it is glossed with paśya(-ta), which is a translation not an etymology). This seems to point to the conclusion that uva(-ha) is original and that the loss of the -v- in the North-Indian branch is the result of a secondary development.
However, the case of u(v)a(-ha) does not stand on its own. The same type of variation with exactly the same distribution of the variants can be seen in u(v)a (Sanskrit upa), ru(v)a (rupa), dhu(v)a- 'to shake, to wash' (dhuvati), su(v)a- 'to sleep' (Pali supati, Sanskrit svap(a)ti) and ru(v)a 'to weep' (Vedic ruvati). The instances of u(v)a and ru(v)a need not be specified here. Those of dhu(v)a- are the following:
130 dhuvai T, dhuai Ma(!), Bh, R, K, V, Y, P, S (means here 'to shake' the toes of the husband to wake him up gently. Wrongly interpreted to mean 'to wash'; from there dhaayi in Ti and dhavai in Tp), 280 dhuvai Ma, Ti, Tp, dhuai Bh, R, K, V, Y, T, S, 433 dhuvai Ma, Ti, Tp, S, dhuai Bh, R, , Y, 680 dhuvasi Ma, Tp, dhuasi, Y, T, S (in R the word is missing), 767 dhuvai R, T, S, dhuai Ma, Tp (remarkable reversal:). Note 532 dhuai in R, all the other MSS having dhunai.
The instances of su(v)a- are:
*33 suvai Ma, Tp, S, suai Bh, R, K, , B, Y, P. 65 suvasi Ma (-vv-). Tp, K, Y, P, S, suasi Bh, R. v. B, 66.suvasi Ma, Tp, Bh, suasi Bhb, R, R, K, , B, Y, P, S, 412 suvaha T1, Tp, T, suaha Bh, R, (suha in Ma, suhaa in Y, and suhaha in S), 717 suvai Ma, Tp, R (-vv-), T, suai Ti(), Bh, S, 804 suvasu S.
The instances of ru(v)a- are:
*9 ruasi (su) all, 143 ruvasu Ti, Tp, T, ruasu Bh, R, K, 4, y, P, S, 147 ruvamti Ma, Tp, T, ruamti Bh, R, K, W, Y, P, 307 ruvai Ti, Tp, T, ruai Bh, R, K, V, Y, S, 334 ruvasi Ma, Ti, Tp, ruasi Bh, R, , Y, S, 379 ruvatT Na, Ti, Tp, ruafitt Bh, R, W, Y, S, W (ruat T), 400 ruamtam all, 503 ruvant? Na, Ti, Tp, T, ruafitT (ti)-Bh, R,

Page #180

167
W, Y, S, 517 ruvasi Tp, ruasi Ma, Ti(!), Bh, R, W, Y, S, 556 ruvati Ma, Ti, Tp, T, ruamti , Y, S (rovasti Bh, R), 569 ruai Ti, Tp(!), Bh, R, V, Y, T, ruvai S, 596 ruaṁtammi , Y, 649 ruai Ma, Tp(!), W, Y, W, ruvai R, S, 784 ruai T, S, ruvai R, 837 ruaṁtie Ma, Ti, Tp(!), ruvat Te T, 848 ruvai Ma, Ti, Tp, ruai T, 885 ruvasu Ti, TP, T, ruvasi W, 895 rua Ti, Tp(!), T, 909 ruvasi Ti, Tp, ruvasu T, 915 ruvasti Ti, TP, T, 941 ruāmi Ti, Tp(!), T, 960 ruasi W, 963 ruai W.
Contrary to u(v)a(-ha) these words have comparatively obvious derivations from Sanskrit. For the -v- in Prākrit the corresponding Sanskrit words have either a -v- or a -p-, which may have provided the basis for the secondary insertion of the -v- in the Prākrit words. For ru(v)a- the situation is different. As already indicated above it most probably goes back to ru-, ruvati (Vedic; Sanskrit has rauti, but also ruvanti). However, in the commentaries and, Chāyās on the Sattasat it is almost invariably glossed with rud-, in conformity with Sanskrit, which uses rud- to denote 'to weep' (of human beings) and ru- to denote 'to howl' (mainly of animals). It is significant that precisely for this word the exceptions (rua-) in the South-Indian branch are numerous. A comparison of ru(v)a- with the other words involved suggests a different approach to the problem. It is possible to argue that in the MSS of the South· Indian recensions in ua a -v-, probably suggested by a pronunciation of ua as /uwa/, was inserted unless the etymology of the word prevented it. Thus, we find dhuva-, suva-, etc., but not * uvaa (adaka). For u(v)a(-ha) there simply was no etymology to prevent the insertion of -v-. Whereas the form rua- would be expected throughout on the basis of the traditionally accepted etymology, the occasional occurrence of ruvamust be explained as an instance of a scribal lapse.
The problem of the variation between uva(-ha) and ua(-ha), etc. seems still far from definitely solved. Nevertheless some principle has to be decided upon with regard to the text to be edited. I have decided in each instance to adopt the form without the -v- (unless not available, as in, for instance, 909 and 915), thus in the main following the North-Indian , branch. In doing so I adhere in a way to the principle of the lectio difficilior: for most words concerned the form without -v- is the one

Page #181

168
deviating most from Sanskrit.
A frequent dilemma occurs in connection with the endings of the nominative and accusative neutre plural, -āi(n), the genitive plural, -āņam) etc., and the instrumental plural, -ehim) etc., when found at the end of the hemistich. This position is metrically neutral so that it does not make any difference whether the text reads, for instance, -ãi or -ain. On this point, though, there is a significant difference be- . tween Ma, Ti and Tp, on the one hand, and Weber's MSS, on the other. The former, with only a few exceptions, throughout have the endings without Anusvāra, while the latter favour those with Anus vāra. The actual situation in Weber's MSS is admittedly unclear. K, edited in the Abh., indeed largely favours the endings with Anusvāra, but this is to be expected in a MS in which the Anusvāra is inserted apparently at random and without regard for the metre (see salak ārāņań in *3 and land aha in *8). In Retr. all deviations have been noted for B, P, T and S; from K, which for these MSS gives the same picture: generally the endings occur with Anusvāra. However, in Ed. Weber refrains from noting this type of variation for the MSS, or parts of MSS, edited there. Moreover, he seems rather careless on this point. Thus, in 354 he reads muheņań , against all the MSS, which have muhena, except possibly Bh, for which in ISt. he does not note a variant (p. 109). This may, however, also be taken as an indication of the fact that he did not note this type of variation of Bh from the text of Ed. either.
Before continuing the discussion of these endings in hemistichfinal position a few remarks may be made regarding their occurrence within the Gathā. There the variation between, for instance, -āi and -āin is entirely regulated by the metre. The endings without Anusvāra are considerably more frequent than those with Anusvāra. The latter are specifically found at the end of the uneven Padas where u-u does not occur. In this position one also finds -ena (relatively frequently) and -es un (once, in 77, a Gathā which probably is a later addition anyhow). -es un is otherwise not found. -enam is otherwise found only once, namely in 525. -ain, -āņam and -ehin are otherwise rare too, and a comparatively large number of these instances is found in particular before the enclitics cia, pi and va, -ehin in fact only so.
While within the Gathā the occurrence of the short and the long

Page #182

169
endings is entirely regulated by the metre, Weber notes that his MSS are highly irregular in that they add the Anusvāra at random. They do so in particular in the case of the plural endings which end in -a and in -i. This agrees with the state in the hemistich-final position where these Same MSS would often have -āin, -āņaṁ and -ehin, while -esuṁ (plural, but ending in -u) and -enań (ending in -a, but singular) are conspicuously absent. It would follow that the presence of the Anusvāra in -ain, -āņań and -ehim in hemistich-final position in Weber's MSS is secondary.
However, this is only part of the picture. Ma, Ti and Tp at the end of the hemistich indeed generally have the endings without Anusvāra, but exceptions occur, which are, however, usually restricted to one or two of these MSS. See, for instance, naaņāin (qaaņālt) in Ma in *5. The addition of the Anusvāra can hardly be explained as secondary, as otherwise, i.e. within the Gathā, it is not found at all. A more likely explanation is that the source of Ma, Ti and Tp occasionally did have -ain, -āņań and -ehim, but that most instances have been removed in a process of normalization in favour of the endings without Anus vāra. The reason behind the preference for the endings without Anusvāra is unclear. It might be a case of extending the endings without Anusvāra which prevail within the Gathā to the instances at the end of the hemistich. Note in this connection that in Tp the endings -e and -o, which mostly are to be scanned short, are represented by the characters ě and throughout, i.e. also where they are metrically long and also where the distinction does not matter, as at the end of the hemistich. A further illustration of the extent to which the MSS of the Third South-Indian recension were normalized will be given below in connection with the Ya-śruti.
Apparently the situation in the source of Ma, Ti and Tp was irregular. It is in any case no longer possible to determine whether in a particular case the archetype read, for instance, -āi or -ain. In the edition I have decided to follow the South-Indian MSS, which means that I have throughout adopted the endings without Anusvāra, except where Ma, Ti and/or Tp have an ending with Anus vāra.
As to these variants the following procedure has been adopted. First those MSS are mentioned of which the existence of a variant is beyond doubt, and then I have added between brackets those MSS of which the reading cannot be definitely ascertained, mostly R, V, Y and Bh, and T

Page #183

170
and S in so far as edited in Weber's Ed. Those MSS which are not specified do not deviate from the text adopted.
A further problem concerns the opposition between v and b. Ma, Ti and Tp have both v and b, much in the same way as we are used to from the Sanskrit dictionaries. The same is the case with R, B, Y, P, T and W. For B, P and T, see Retr., p. 349. In Ed., p. 5, Weber notes that R, Y, T and W usually maintain (Sanskrit) b initially. He does not mention anything about b after Anusvāra. At least T also has b after Anusvāra. The remaining MSS, i.e. Bh, K, and S have only v.
In this connection the question arises which group of MSS most faithfully represents the state of the MS of the archetype. As far as Ma, Ti and Tp is concerned, which neatly distinguish between v and b, there is evidence to suggest that the b was introduced only secondarily. A clear instance is bola- 'to pass' for vola- in, for instance, MS Bh. The past participle boliņa suggests that it consists of the verb 11-, in which case the first syllable is a prefix. No prefix bo- exists, while vo-, from Sanskrit vyapa- or vyava-, is relatively common. It follows that the b- in bola- is secondary. From the introduction in this word of bfor original v- it appears that its derivation was not known. It is to be noted that in Sanskrit vyap (/v) alf- does not occur with the meaning 'to pass'. It is not clear which other word the scribe had in mind when introducing the spelling bola-. One possibility is that b was introduced almost automatically for v unless the derivation of the word in question prevented it. volaia in Ma in 21 most likely is a survival of this process. Another instance is vāha 'tear' in Ma in 25: Ma, Ti and Tp otherwise have throughout baha. The instance in *25 differs from the others in that the word was wrongly supposed to contain a pun on the words vyadha 'hunter and başpa 'tear'. This pun is recognized by all the available commentaries including the one by Madhavayajvamisra found in Ma. It should be noted that the idea of the pun could only have come up in a text which did not distinguish between v and b! In the South-Indian MSS a dilemma arises. Whether one reads vāha as in Ma or baha as in Tp the pun is lost. It would seem that in Ma the dilemma was solved by leaving the text as it was.
It follows that the MS of the archetype most situation has its parallels in the NIA languages.
likely only had v. This Only a few of these

Page #184

171
languages distinguish between V and b; most have either v or b. It is a well-known fact that this phenomenon has greatly affected the transmission of Sanskrit texts, in which originally v and b were distinguished (see AIG I,S 161). It is in itself not surprising that it is precisely in the South that apparently a tradition was preserved with regard to writing. v or b in particular words. This almost certainly has to do with the fact that most South-Indian languages do distinguish between v and b.
The decision to write in the text v throughout introduces yet another problem in connection with the variants. MS K, edited in the Abh., has only v. For B, P, S and T, edited in the first place in the Retr., the deviations from K in respect of this feature have been carefully noted; at least, Weber does not state explicitly that he makes an exception for the variation between b and v. The text in Ed. has only v. On page 5 Weber remarks that he does not specify the variants, unless the word in question should happen to be mentioned for other reasons. Bh, edited in ISt., seems only to have v; a few instances of b in this MS have been noted.
My procedure is the following. Where I have adopted a form with v I first mention those MSS which positively have a variant with b, and subsequently add within brackets those MSS of which the reading cannot be definitely ascertained.
As I have shown above (p. 40) the present manuscript tradition goes back to one single MS the text of which was not entirely flawless. Some errors in the MSS can be traced back to this particular Ms. Where possible an emendation has been suggested. In some cases the way to emend the text is pointed out in the MSS themselves. See 292 in which the archetype seems to have read taņuãi (cf. Bh tanuāi and Ti and Tp tanuai(d)a) which in all the other MSS was corrected into tanuāai. In other cases the emendation could be based on peculiarities of the script of the MS of the archetype, which was Jaina Nāgart. In yet a few other cases the emendation could be based on an informed guess, for instance, gammihii for gammihisi of the MSS in 609 (see the note to *10). In a few cases, though, no convincing emendation seems available.
The present manuscript tradition seems to have started in a milieu of Jains. This explains the occasional occurrence in the MS of the arche

Page #185

172
type of a word in a form otherwise typical of AMg. An instance is piva, the particle of comparison. In such instances I felt free to adopt in the text what seems the regular M. Prākrit form, namely via (see the note on *1).
It appears that the archetype was furthermore characterized by certain orthographical conventions typical of, and otherwise exclusively found in, Jaina MSS, e.g. the Ya-śruti, the use of i and u for ě and ŏ respectively, and the use in certain positions of the dental nasal in favour of the retroflex one. Within the transmission of a text these features seem basically interchangeable. That is to say, a Jain copying a non-Jaina MS will add, for instance, the Ya-śruti where appropriate, and, vice versa, a non-Jain copying a Jaina MS will delete the Ya-śruti: 'Because I wish to avoid the impression that the Sattasat is the exclusive property of the Jains I have decided to divest the text of these orthographical features.
The Ya-śruti is regularly found in Bh and R. Its occurrence in the MS of the archetype is proved through occasional instances in Ma, which are to be explained as remnants after it had been generally removed from the text.
For ě and © in a closed syllable Jaina MSS commonly write i and u respectively (see Pischel & 109). This spelling seems to show that in this position the phonemic contrast between i and ě and between u and o had become less distinct. There are indications that this spelling was also found in the MS of the archetype. In the course of the transmission, however, e and seem to have been restored in most MSS except for Bh and R. In one instance this restoration did not take place because of a misunderstanding of the derivation of the word in question. The instance is found in 782, which reads:
durapainpapar imala angacchitt a piva vam-
sapallava muddhapupphapaṁguranan mahena di na mahusirTe.
It is almost certain that chittań (found in all available Mss, including Ma, Ti and Tp) stands for chetta (Sanskrit kşetram), and that we should read agań chic/e)ttam. The word chittañ, however, was misunderstood as the past participle of chiva- 'to touch', which prevented the restora

Page #186

173
tion of e for i. I should like to suggest to emend the instrumental vammahena (found in all the MSS) into a genitive vammahassa:
The nymph of Spring has given her body (amgaṁ), of which the fragrance spreads very far, which is decorated with sprouts and which has tender flowers for its cloak, to the god of Love (vammahassa) as his abode (chi(!e)ttaṁ).
This example concerns an instance of the writing of i for è in a closed syllable. In Bh the writing of i for ě is also frequently found in an open syllable at the end of a word; e.g. tfi for tTě in *22. The situation in the MS of the archetype is unclear. The instance of jiviasesā i in 149, discussed above (p. 163), for which Ma and Tp have introduced jivias addhãe, shows a tendency on the part of the latter MSS to introduce e for i. However, it does not become clear if and to what extent i was written for è in open syllables in the archetype.
For initial n- and medial -nn- MS Bh consistently writes n- and -nn- respectively, the latter combination being often written as -rinThis spelling is typical of Jaina MSS. As such, it might have been a feature of the MS of the archetype, too. The spelling seems to be old and to have a basis in some spoken languages (Bloch, 1934: 79). On the other hand, it is possible to discern a development within the Jaina tradition itself, the old palmleaf MSS having n- also initially. It is possible that we have to do here with a locally (Gujarat?) developed convention.
The Ya-śruti, i and u for ě and o in closed syllables and n- and -nn- for n- and -nn- respectively, are more or less consistent features in Bh and/or R and need not be specifically mentioned in the critical Apparatus. An exception is made for i for čas in tti, which is irregular in Bh.
2.2 The variants
In general all available variants have been specified irrespective of their nature. For those of Weber's MSS I make use of his published material only. The text edited here in many cases differs considerably

Page #187

174
from the one in Weber's Ed. Often the readings of his MSS have to be ascertained by inference, i.e. where Weber does not mention any variant on his text his MSS presumably do not deviate from it. Above I have already mentioned several types of variants for which this method does not work, as in the case of the variation between -ai and -aim at the end of the hemistich and of the variation between v and b. On the other hand, a certain number of consistent variants need not be mentioned. Some of these have already been discussed above: the Ya-śruti and the use of i and u in closed syllables for ĕ and respectively, in Bh and R, and the use of n- and -nn-. in favour of n- and -nn- respectively in Bh. A few others will be discussed below, which seem basically determined by the various scripts used.
One group of peculiarities concern the South-Indian MSS Ma, Ti, Tp, Y and T. It should be remembered that of these five MSS only Tp is actually written in a South-Indian script, viz. Malayalam. Ma, Ti and y are Nagari transcripts and T is a Roman transcript. For the two other South-Indian MSS, P and W, Weber relied on Roman transcripts as well. In these the peculiarities mentioned below seem to have been silently corrected.
Tp throughout writes 1, which is typical of South-Indian MSS (see Lüders, 1940: 548). This 1 was apparently also found in the source of Ma and Ti, which confuse 1 with i or u. For 1 in the source of the transcript T, see Ed., p. XXXI, note 2.
In most of the South-Indian scripts a dot or circle in the line is used to indicate the doubling of the following consonant. This sign is often confused with that for the Anusvāra, a dot or circle written above the line. This confusion is apparent in Ma, Ti, T (Ed., pp. XXXXXXI) and y (Ed., p. 2). In Ma and Ti this 'Anus vāra' was subsequently replaced by the homorganic nasal (see below); e.g. agupinga (anuviggā).
In Ma and Ti initial i, u and 1 (1) are often interchanged due to the similarity of the characters in the Malayalam script.
Ma, Ti, Tp, y and T write ho and hm throughout (in Tp hp and hm are often preceded by the dot or circle indicating doubling). This spelling is typical of South-Indian MSS in general (see Printz, 1921: 24). The spelling represents an interpretation of the Nagar ligatures for ph and mh, in which the nasal is indeed written within and below the h.

Page #188

175
Note in this connection that aspirated nasals are unknown in South-India.
Regarding the variants of Bh given in Ist. the following should be observed. Weber notes that in Bh the ligature -jj- is represented in two ways, namely by the usual symbol in Jaina Nāgart, which slightly modified is also used for -jjh-, and by a symbol which resembles the ligatures for -şv- (-şb-), -bb- or -mv-. The second ligature Weber transcribed by any of the latter combinations or by -yy-. Unfortunately Weber does not reproduce the ligature in question so that we have to guess what it actually looked like. It may have been the ligature , which, however, in the appendix on paleography in L'Inde Classique (p. 695) is classed among these for -jjh-. Alternatively, it could have been the ligature , which according to Jacobi (Kalpasūtra, p. 18, note 1) stands indeed for -jj-. To distinguish this ligature for -jj- from the other one I have transcribed it as -jjh-; thus, kuppijjha in *17 for kuppiyya of ISt., p. 35.
Weber (Ed., p. 3) notes that his MSS practically throughout write -ch- for -cch-. In y as well as T -Śch- is found, which occasionally is very similar to -chch-. Likewise his MSS often have -khkh-, -jhjh-, -thth- and -dhoh- for -kkh- etc. The same is the case in MS K, for which see Abh., pp. 26-7. W and y write - thţh- throughout. In Ed., p. 3, Weber furthermore notes that his MSS rather randomly add the Anusvāra, in particular after the case endings for the plural which end in -a and -i (see also above, p. 169). In his Ed. Weber did not specify such variants unless the word in question should already be mentioned for some other reason. On this point I had to make do with what Weber cared to mention. This type of variation is exactly known for K, through Abh., and for those parts of B, P, T and S edited in his Retr.
The majority of the MSS writes -ink-, -ńc-, etc. Some, though, use in these combinations the homorganic nasal instead of the Anus vāra. This is consistently the case in Ma and Ti, both Devanagarī transcripts. One wonders if this feature was introduced secondarily in these MSS at the time of their transcription into Devanagari. It is in any case significant that Tp, an original Malayalam MS, has the Anusvāra. -nk-, -nc-, etc. is also found in some of Weber's MSS, but as the exception rather than the rule. For the instances in K, see Abh., p. 20, note 2, and p. 21. Note furthermore tantatatattiñ in *2 in (but in the same Gathā

Page #189

176
pa ananti) and dant in 23 in T and S. I have the impression, though, that in the edtions following that of MS K in Abh. Weber neglected this type of variation. For instance in 1 K reads samkanta. In Retr. Weber did not give any variants, which would mean that B, P, T and S has samkanta as well, which I have strong reasons to doubt. I follow the practice found in the majority of the MSS, writing in these combinations the Anusvara. The relevant variants from Ma and Ti need not be specified. As far as Weber's MSS are concerned I have mentioned their variants only when they are actually available; thus, samkanta in 1 for K, but no variants for B, P, T and S.
2.3 The presentation
The presentation of the text is as follows. Above the text of the Gatha its position in the various MSS is indicated. The relevant MSS are given in the order of the authenticity of their texts, i.e. Ma, Ti, Tp; Bh, R; K, V, B, Y, P; T; S; W, with W, which strictly speaking is probably not a recension of the Sattasat, put at the end. As the order of the Gāthās in this edition is basically that of the Jaina-recension and the Vulgata, a different order in the enumeration might be expected. This would, however, result in two systems, one above and another for the Critical Apparatus below the Gatha. For practical purposes it has been decided to use only one system. For completeness' sake also those MSS which contain only a Sanskrit commentary have been given, namely G (Gangadhara; stands for A, C and E), and T, and X, which contains what seems to be a separate recension. G, E, π and x are mentioned in this order after P of the Vulgata. W and V, the latter containing only a Sanskrit Chaya to the Gathas, are treated as a single witness: when a Gatha is illegible or absent in W but its Chaya is found in V, the latter is mentioned. The MSS are broken up recensionwise by semicolons.
Below the text the Critical Apparatus is found, preceded by a description of any larger lacunae. The order of the variants follows the text. A variant may concern a single word or a whole part of the text. Two dashes indicate that a following part of the text is referred to. When more than one variant is available these are given recensionwise and when an identical variant occurs in the MSS of two 'consecutive'

Page #190

177
recensions, the variant in question is not repeated. Note in this connection that this meant that the order K, 4, B, Y and P could not always be maintained.
The Critical Apparatus is followed by an English translation. The translation in a way had to be a compromise. I have attempted to keep as close as possible to the Prākrit text in order to help the reader to identify the words. On the other hand, in several instances a literal translation fails to convey the intention of the Gathā as a whole, so that recourse had to be taken to certain idiomatic expressions. In several cases even then further comment on the meaning of the Gathā was deemed necessary, which is found in the Notes.
I have dispensed with introducing specific symbols which would indicate where the text adopted is doubtful or conjectural. Problems concerning the reconstruction are discussed in the Notes.
These same, Notes are furthermore intended to clarify all other problems which may arise in connection with the text. A recurrent theme in the Notes is the discussion of the relation between the text and the indigenous grammatical tradition, with special emphasis on the grammars by Vararuci and Hemacandra, and the relation between the text and traditional lexicography, as exemplified by Dhanapala's Pāiyalaccht and Hemacandra's Deś Tnāmamālā.
The Gathās of the Sattasaf have often been quoted as illustrations in Sanskrit works on poetics. The earliest work of this type in which Gathas have been quoted is the Dhvanyaloka; e.g. Gatha 669 on p. 20, 110 on p. 119, etc. Another work in which Gāthās from the Sattas at occur is Bhoja's Śrngaraprakāśa. In an appendix (Abhang) to his Abh. (p. 202 ff.) Weber collected the Prākrit Gathās quoted in Dhanika's commentary on the Daśarūpa, in Mammaţa's Kāvyaprakāśa and in Visvanatha's Sāhityadarpana. A number of these Gathās is also found in the Sattasat. Besides in works on poetics, Pādas or even smaller parts of Gathās have also been quoted in Hemacandra's Prākrit grammar. In this edition the problem of the testimonia is not taken up. I have made an exception for the quotations collected in Weber's Abhang and those found in Hemacandra's grammar. They are referred to immediately below the Critical Apparatus; only important variants are noted.

Page #191

178
Notes to chapter 2
The -v- in dhuva- and ruva- features in the Prākrit passives' dhuvvata (564) and ruvvau (*10). It is, however, unlikely that the -v- was secondarily inserted in dhuva- and ruva- on the basis of these passives. Traditional scholarship was not so much interested in relations existing between words in Prākrit as in the relation between Prākrit and Sanskrit.
It should be noted that in Pkt ru- and rud- have partly coalesced, roissam, roiavva and runna (+*rudna) most likely going back to rud
Weber refers to two editions of the Kavyaprakāśa, viz. the Calcutta editions of 1829 (Kb) and of 1866 (ka). None of the two was available to me. I refer to the Poona edition by R. D. Karmarkar; of the two editions of the Sāhityadarpaņa, namely the ones of 1828 (Sb) and of 1851 by E. Roer (Sa), only the latter was available to me.

Page #192

CHAPTER 3
3.1 Edition of Gātās 1-50, with translations and notes
11 1 Ma, 2 Tp; 1 eh, R; K, \, B, Y, P, G, 5; T; S; V) I starts only with *21; x only with 133. The first ten Gathas of W were illegible.
pasuvaiņo rosāruņa- gahiagghaparik aaṁ via
padimās ark amtagor imuhaaṁd am samjhāsalilaṁjalin namaha.
In Ma lacuna for the text; in Tp for pasuvaino...padima (saṁk anta). pasuvaiņoh K, vasuvaino P -- sak anta k -- muyacand am Bh -- gehia Tp; gahiagdha K. -- paṁkayaṁ piva Bh; okaammia , ommia B, í bia y; kaa via T.
Bow before the moon-face of Gaurt which by her anger at Pasupati has become the counterpart of the morning sky; it looks like the cupped hands filled with Sandhyā-water, holding a (red) Arghya-lotus.
The position of this Gāthī at the beginning of the text follows the convention that a literary work should open with a verse praising a god or goddess, or, as here, exhorting the reader to do so. (Cf. Kavyādarśa I 14: sargabandho mahākāvyam ucyate tasya lakşaņam/ astr namaskriyā vastunirdeso vāpi tanmukham. Note also the presence of a Nāndt at the beginning of a play.) These verses are often characterized by particularly intricate descriptions. Here Gaurt is described as being enraged at her husband Pasupati, a situation which is fully in accordance with the general content of the text. It may, incidentally, be noted that the Gathā, in containing the word for 'moon', follows exactly the prescription given for the content of a Nandt (see Lévi, 1963: 132).
This Gathā, together with the following two, forms an introduction to the text. Gāthās *2 and *3 give the impression of having been composed for the occasion and have as a matter of principle to be set apart from

Page #193

180
the other Gāthās. The same would seem to be necessary for Gatha *1. Thus, the description of a goddess and her husband seems out of place in a text which is mainly concerned with common people such as farmers and hunters, despite the parallelism in their respective love-lives. A further argument is its form, which is that of a Stotra. It should be noted that most of the other Gathās which have the form of a Stotra, praising either Siva, Vişnu or Ganesa, are in all likelihood later additions to the text, as follows from their absence in the Third South-Indian recension. See, for instance, 448 (Siva), 151, 406 and 411 (Vişnu), and 403 (Ganesa). In some of these, 403, 406 and 411, erotic connotations are lacking completely. However, Gathā 455, a Stotra praising Siva, like the one under discussion, goes back to the archetype.
My translation of the Gathā differs considerably from Weber's. The latter scholar connects the first word of the Gātha, pasuvaiņo, with sanjhāsalilańjalin which he takes directly as the object to namaha, and arrives at the following interpretation: 'Bow before the Sandhya-water in Pasupati's hands' (see 'Retr., p. 347). He may have been led to do so by the position of the particle of comparison, which is here, contrary to common practice, not found after the Upamāna but after a compound which describes the Upamāna. It should be noted that this is not uncommon in Pkt (see Gathā 93: siviņaap Tena va pāņieņa), nor in Skt (see Speyer § 430).
The -e- in Tp gehia, for gahia, is probably due to the South-Indian pronunciation of -a after certain consonants as -e. This phenomenon has been described for Malayā] am by Godavarman (1935-37) and by Kunhan Raja and Ramachandra Sarma in their edition of the Prākrtaprakāśa (p. IX). These authors restrict it to -a occurring after voiced plosives and after C-, y-, r- and 1-. Indeed in the South-Indian MSS of the Sattas at one comes across *7 lehada Tp, 60 deia Ma and hypercorrect 132 davvena Ma and 532 gahnai Ma, Ti, TP, T. But in these same MSS -e for -a or hypercorrect -a for -e is also found after voiceless plosives and after h-. E.g. 189 teha Ma, 412 teio Ma, 642 kendofța Ma and 183 pecchanna Ma (but in the last instance probably under the influence of the following palatal plosive; cf. 373 pacchaîto Ti, 457 pacchai Ti); hypercorrect -a in *17 kuppajja Ma, Tp, 4?0 uvakkhio Ti, Tp (written upakkhio), 445 uppakkhimo Ti. After h-: 914 hero Ma (+(b)haro) and 930 hea Ma..Further

Page #194

181
more one comes across *39 e Ma (for (c)a), 93 aviehņa Ma and *17 anunajja TP (probably under the influence of following -jj-). In 365 aggapaapallianaṁ Ti and in 578 padipallio Ti, pallio for pellio may be due to the preceding syllable pao.
For via MS Bh has piva and MS v ia. This variation needs to be discussed. Corresponding to Skt iva the Pkt of the Sattasaf has commonly (v)va, which Pischel ($ 143) explains as an apocopated form of iva, as (t)ti of iti or pi/vi of api. Besides, but considerably less frequently than the latter, a bisyllabic particle is found, via after vowels, in the SattasaT -- possibly accidentally -- only after -o, but, for instance, in the Setubandha also after - and -e: Gathā 337 parivāho via, 443 kuruņāho via, 612 maņoraho via, 767 pariamo via; cf. Setubandha V 77, VIII 101, XII 95 (v.1.), XIV 80; after -a in I 7, VI 71, and after -e in I. 47.
The same form is also found after Anusvāra, but has, as already briefly indicated, several variants there: piva in the Jaina MSS Bh and R as well as in the South-Indian MSS Ma, Ti and Tp which ultimately derive from a Jaina Nāgart MS. MS T, which is otherwise closely related to the MSS Ma, Ti and Tp, usually has via. The transcripts of the SouthIndian MSS often read viva, showing a confusion between pand v, characters which are very similar in the South-Indian alphabets. The Vulgata MSS and s usually have iva (or ia), in which case the preceding Anusvāra is, replaced by -mm (in the Vulgata MSS) or by -rim (in s). The instances are the following. Note that viva of the South-Indian MSS has been silently corrected into piva:
1 park aan via TP, R, K, B, Y( - bia), P, T, S; piva Bh; -mm ia y, 125 pihin via Ti, Tp, Y, P, T; piva Bh, R; -mm iva K, , -im iva S
(Most probably, however, the text should here be emended into
niht via, the accusative nihim being ungrammatical) 290 duvvaanaṁ via Ti, Tp, Y (- vio); piva Ma, Bh, R, T; -mm iva K,
V, -rim iva S, 331 vand in via T; piva Ma, Ti, Tp, Bh, R; -mm iva K, W, Y, S,
336 os ah aṁ via T; piva Ti, Tp, Bh, R; -mm iva K, W, Y, -ím iva S, · 363 das anaṁ via T; piva Bh, R; -mm ia K, -imm iva , Y, -ím iva S,
388 lacchin via U, Y, S; piva Ma, Bh, R,

Page #195

182
507 janań piva Ma, Ti, Bh, R; -mm iva W, -m iva Y, -im iva S, 609 duddha piva Ma, Ti, Tp, Bh, R, T; -rim ia , -mm iva Y,
-rm iva S, 782 chittań piva Ma, Ti, TP, R, T; -mm iva S.
Of these variants iva, which is found exclusively in the Vulgata MSS and S, is almost certainly an innovation. The change of the preceding Anu- . svāra into -mm or -rim represents an attempt to preserve the metre. Before iva, which begins with a vowel, the Anusvāra may, as in Skt, have changed into -m, in which case the syllable is short. As such the phrase vard imm iva has an exact parallel in Pāli mamm iva (see CPD II, p. 331; for alternative mām iva, see Geiger 71). The status of iva in Pāli, however, cannot be discussed here.
The MS of the archetype seems to have read via as well as piva. The former is not restricted to M. The single prevailing form in ś. and Mg. is via. It is, furthermore, undoubtedly identical to pali viya. In all these dialects and languages the particle occurs after vowels and after Anusvāra.
Geiger ($ 66) suggests that Pāli viya, through metathesis, goes back to *yiva, which would be a frozen sandhi-form. He proceeds from this form *yiva in order to be able to account for the medial -y-, the origin of which, however, need not detain us here. I largely agree with Geiger's explanation but should like to add some further remarks concerning the circumstances which may have been responsible for the supposed metathesis. An indication for this may be found by comparing the other most common enclitic particles, namely (t)ti, (v)va, (c)cia and vi/pi. It should be noted that all these begin with a consonant, some after having lost their original vowel. This apparent tendency of enclitics to begin with a consonant may have been responsible for the metathesis in via.
The second particle, piva, is otherwise exclusively found in Amg. and JM. texts. It occurs only after Anus vāra; after vowels these texts have viva. According to Pischel ($ 337) viva would consist of a kind of proclitic v- plus iva. The existence of such a proclitic v- is, however, doubtful. In the same paragraph Pischel has collected several more instances of this v-, most of which, however, should be explained differ

Page #196

183
ently. E.g. v- in vuccai (Skt ucyate) is from vac-. Alternatively viva may be a frozen sandhi-form. It is difficult, though, to separate viva from the form via of the other dialects. Admittedly, the origin of the medial -v- presents a problem, for which I am unable to offer a satisfactory explanation. One might suppose contamination with iva, found in AMg. texts besides viva. I am in any case hesitant to connect it with Gāndhārt Dhammapada Pkt viva, about which much is still uncertain; see Brough (1962: 110-2).
The conditioning factor for the variant piva beside viva is strikingly similar to that for pi and vi. It has even been suggested by, for instance, Weber (Ed., p. 1), Goldschmidt (Setubandha, index, s.v. via) and Jacobi (Kalpasūtra, p. 100) that it is a compound particle made up of pi/vi plus (i)va. This derivation is doubtful. Admittedly, compound enclitic particles are not entirely unknown in Pkt, which has (c)cia, from caiva, but the latter combination is already known in Skt, whereas api plus iva is not. : This derivation has earlier been dismissed by Pischel (5 336) who, instead, suggests that the equation pi : vi may in fact have been responsible for the origin of piva from viva. A logical extension of Pischel's reasoning would be that piva is in reality but a hypercorrect form or a scholarly invention. For, an original initial v- remains vafter Anus vāra, and in this connection it does not matter whether it is a proclitic consonant, a glide or a medial -v- transposed through metathesis. Note that in the case of pi/vi the situation is reversed, the P- being the original and the v- its phonologically conditioned variant.
The fact that it is well established from the oldest Amg. texts onwards would seem to speak against this view of piva. This may, however, also be taken as an indication that these texts were subjected to rigorous redactional interventions. That this is indeed the case I shall try to demonstrate elsewhere.
It remains to explain the occurrence of piva in M. Pkt, in which, at least in the Sattasat, viva does not occur. It should be noted that at the beginning of the present manuscript tradition of the Sattasat stands a MS, or a line of MSS, written in Jaina Nāgars, which most likely will have been written by someone well acquainted with AMg. and/or JM. It is not unlikely that at this stage piva was entered into the text for via oc

Page #197

184
curring after Anusvāra, either unconsciously or consciously because via disturbed the scribe's notion of correctness. Ag. piva beside viva was, if not invented, at least supported by the parallelism with pi : vi.
The case of via after vowels is slightly different. The same scribe may have compared it to AMg. viva. The absence in via of the medial -vneed not have disturbed him, its loss not being uncommon in most of the literary MIA dialects.
It should be clear that a definite conclusion regarding the status of piva in M. Pkt needs a closer study of the text-transmissions of, for instance, the Gauqavaho and the Vajjālagga, in which it occurs likewise. In the Setubandha no instances of piva are found, the particle not occurring after Anusvāra.
Not all the instances of piva in the MSS need automatically go back to the MS of the archetype. This may be especially relevant in the case of Bh and R, which are Jaina MSS. It may also be relevant in connection with the MSS of the Third South-Indian recension. The case of 290, where Ti and Tp read via but Ma piva, should make us aware of the possibility that piva, but also via, may have developed into an independent variant. It may anyhow not be without significance that T, which otherwise agrees so closely with the MSS of the Third South-Indian recension, mostly reads via for piva in the latter.
Finally the evidence of the traditional grammars may be considered. It should a priori be noted that not much value should be attached to absence or presence in the grammars of a particular form of the particle, which may merely indicate that the grammarians' observations were restricted to certain texts only or to certain MSS of texts. Thus, Var. IX 16 mentions via, miva (apparently abstracted from -im iva) and mmiva (from -mm iva) but does not mention piva. The latter is given in X 4 for Paiśct. Note, however, that according to Pischel ($ 336, note 6) mmiva of IX 16 is a clerical error for piva. Hem. II 182, on the other hand, mentions via and miva as well as piva and viva, but it should be added that in his Sutras not always a clear distinction between M. and AMg. is maintained.
In the edition I have throughout adopted via, even where piva can be ascertained for the MS of the archetype. My reason for doing so is twofold: if AMg. piva is indeed a ghost-word the decision is an obyious

Page #198

185
one; and even if it would be linguistically acceptable for AMg., its occurrence in one dialect next to via, which is so much different, is in my opinion very suspect. Its occurrence in the MS of the archetype beside via can be explained through redactional interventions.
21 2 Ma, 3 Tp; 2 R; K, V, B, Y, P, G, 5; 7 T; S; 2 V) Missing in Bh.
ama ar pāuakavva kāmassa tatatattin
padhiu souṁ ca je na ānanti kun anti kaha te na lajjaṁti.
In Ma lacuna after kāma(ssa). amian R; K, V, B, Y; T; S -- vāyada Ma, pāada Tp; pāian R; pāada T; pāia S -- pathiin Ma, pathium Tp; v; pathidum T -- sodum P; T -- a K, V, va P; yi. T -- ja Ma -- ņa jāŋanti Tp; R; K, B, Y, P; T; S -- tattatatti R; taħttataṁtti K, tantatatattiñ V, tattaciỉtan B, tattattattin Y, tattatatti P; s', tattauttin s2 -- te kaha R, W, Y; te kaham K, B, P; te kaha S -- lajjiṁti To.
Those who do not know how to read and listen to (this) nectarlike Kāvya in Prākrit, but are solely
concerned with scholarly works (in Sanskrit) on love, .how can they not feel ashamed?
The readings adopted for, and my interpretation of, the compound in c differ from Weber's. The final member of the compound should read tattin, which word Deś in. V 20 translates with, among others, tatparatā 'being devoted'. Indeed it seems to have this meaning as well as that of 'worry, concern' (see Gathās 51, 272, 833 and 866). Turner, CDIAL -5683, derives tatti from Skt tapti 'heat'. For the development of the meaning, note especially Si. täta 'effort, aim' quoted there. The first member of the compound should read tamta (tantra), which has the pregnant meaning 'scholarly work in Sanskrit' here, thus opposing pāuak avva 'literary work in Prākrit'. The study of the latter is considered a necessary adjunct here to that of the former.
The form pāada (prāksta) found in Ma, Tp and I seems an independent

Page #199

186
variant turning up in the South-Indian MSS of Pkt texts or in Pkt text's written in South-India. See, for instance, Karpūr amañjart I 7, MSS S and U, which have a variant pāuça, and MS T, which has yāțaya, and Candralekha I 11 and Kamsavaho IV 48h. The latter two texts were written by South-Indians, Rudradāsa and Rāma Pāṇivāda respectively. Hence pāada could be an innovation in the South Indian MSS of the Sattasat. On the other hand this does not explain how pāada could have become prevalent in the South and how the notion that it is the correct form could have arisen there. In this connection it should be noted that pāada, from prākrta, is formed after a sound-change, -rta- + -ada, typical of AMg. and JM. (Pischel S 219). Its occurrence in the South-Indian MSS of the Sattasat could then be explained with reference to the Jaina. Nāgart MS from which they ultimately derive. The word may have been introduced in this MS by its scribe who very likely was familiar with AMg. and JM. or with the sound-changes accounting for the derivation of these dialects from Skt. Furthermore, from its occurrence precisely in these South Indian MSS of the Sattas as it may have become known in the South as the only correct form for Skt prāksta. However, M. Pkt usually has pāua. Also the earliest AMg. instances of the word, found in late canonical texts, which, however, are considered in any case, older than the Sattas at, read pāgaya or pāyaya, showing a sound-change similar to that in pāua. The word pāada therefore most likely seems to be a learned, hypercorrect Alg. form.
It may not be without significance that pāada, which is thus well known in South India, is not mentioned in any of the grammars (pauda in the pratyādigana in Prakrtasarvasva II 10 is a corruption, through paduga (MSS U, I, O and G), for pāhuda (cf. Hem. I 206)). This would indicate that the grammars did not use or know the South Indian MSS of the Sattasaf.
After na the verb jāņa- (Skt jñā-) invariably loses its initial j-; see, for instance, 347 na āņimo. This loss is usually restricted to -j- found within a word. Apparently the phrase na āņa-, which occurs frequently, has come to be considered as one word (cf. na ittho in 502). All the MSS, however, except Ma and y, have na jānanti here, which nevertheless is to be rejected as the easier reading.

Page #200

187
31 3 Ma, 4 Tp; 2 Bh, 3 R; K, U, B, Y, P, G, 5; 5 T; 8 S; 3 V)
sattasaāin kaivac- hūleņa viraiaiṁ
chalena kodte majjhaarão sālakārāṇa gāhāna.
In Ma lacuna between h(ālena)...rāņa gāhāņa. saīsin K -- kaivachalena K, vatthalena , kavatsalena y -- kotiye Ma; kodta U, B, Y; koţte S -- majjhaārammi R; U, B, P, masjhaārammi K, majhjhaărammi Y; majjhaārammi 5 -- sālaiņa T -- viraiāna Bh -- salamkārāna K -- gāhāņa R; K, U, B, P; T; S; (Bh).
From among ten millions of Gathās full of Alankāras, sevenhundred were collected by Hala, the patron of poets.
The ending -e in kogte is to be scanned short. In most NagarT MSS this -ě, for lack of a separate character, is represented by -e. In some MSS, and practically throughout in B, Y and S, it is represented by -i after -ā (-ai) and -ū (-ūi) and by -a after -T (-Ta). Bh, which usually has -te, sometimes has -Ti; see, for instance, * 22' muhti and tfi. Similarly, - að and -1o are written as - du and - Tu respectively. -ūu for -ūő is not found in the Sattasaf.
i and u as graphic representations of ě and Ở are also found in closed syllables (Pischel $ 119). The origin of -a in, for instance, kodta, is as yet unclear. It may, however, have arisen from confusion between - i- and -ya- : kodTe + kosti → kogfya + kogfa. The recognition of this ending -a was facilitated by the fact that an ending consisting of a vowel similar to the stem-vowel, as in - Ti, was apparently considered awkward. Thus, while Var. V 22 accepts the ending -a, the commentator Bhāmaha, ad Var. V 23, explicitly forbids it after a-stems, where it would give forms like mālāa.
In the Gathās of the Sattasat the endings -e and -o occur practically only in positions where they have to be scanned short. Long -e and - are found at the end of the hemistich, which is metrically neutral. Furthermore they are found at the end of the uneven Pādas where (u) - U does not occur. Otherwise they are rare: see (*22), *43, 226, 228, 322, 492, 513, 593, 599, 739, 789, 797 (ex conj.) and 810. The same is also

Page #201

188
the case with the ablative ending -o as in -ão.
In S the endings have been normalized in favour of -a, -i and -u, which are often also found at the end of the hemistich.
For a possible derivation of the Dest-word majjhaāra (Deśīn. VI 121: majjhammi majjhaāra), see Emeneau (1968: 32-3).
4( 3 Bh, 4 R; K, W, B, Y, P, G, 5; 444 T; 77 S; 97 W) Missing in Ma, Ti and Tp.
ua niccalanipphamdā nimmalamar agaabhāana-
bhisiņTvattammi rehai valāā pariţthiā sarkhasippi wa.
uva T; W -- nippada Bh; nippandă T; S; pipaņdā w -- bhisinā K;.. hisint T; visint S; bisiņt W -- pattammi Bh, R; K, 4, B, 'P, vattami Y; pattammi T; pattami W -- (lacuna)yā Bh; balaa R; B; T; W -- hāana T; bhāana missing in W -- bhāyaṇuvariţthia R; parittiā k, padithi ya (vt illegible), pariţthidā B; padiţthia W -- samkhasutti Bh, R; K, , B, Y, P; S; saṁkhicchippi T.
Kavyaprakāśa, p. 30 (sutti); Sāhityadarpana, p. 20 (osutti); Pāda a quoted in Hem. II 211.
Look! (Standing) absolutely motionless between the lotusleaves, the (female) heron looks like a (white) sippi, a shell, set in (a setting of) spotless emerald: ::
In the Gathā the speaker draws a second person's attention to something in his environment, composing as he does a striking comparison. As such the Gathā has exact parallels in, for instance, 62, 63, 64 and 75. The interpretation of the Gathā raises problems which I have been unable to solve so far. According to Weber the speaker draws attention to a quiet place where lovers can meet without being disturbed. This interpretation does not take account of the second line, which refers to something unlikely, namely a shell in a setting of precious stone, viz. emerald. Curiously, the setting is worth more that the thing set. As such the comparison in the second line seems to contain a comment on the situation

Page #202

189
described in the first line. It may, for instance, imply that the balākā spoils the beauty of the place or that it has no right to be there. Apparently the bird is preying, standing absolutely motionless in order not to scare away the fish. It remains unclear to me, though, what precisely the speaker wishes to convey to the person addressed.
The particles ua and uaha are used to draw someone's attention to a particular object or person. They are common in the Sattasat (ua in 49, 75, 120, etc. and uaha in *18, 62, 109, etc.) and in the Pkt verses in the Skt dramas (see Pischel in his translation of Hem., pp. 97-8). In the other genres, e.g. the Setubandha and the Gauḍavaho, they are absent. Usually ua and uaha are translated with pasya and pasyata respectively (see Hem. II 211 and. Pischel, loc. cit.). These glosses were undoubtedly inspired by their mutual relationship which is similar to that of the imperatives vatta and vattaha. Weber suggests a derivation from the verb uh- (Abh., p. 29, note 4, Ed., p. 4 and word-index, s.v. uh-(2)), which is hardly tenable (see Pischel, loc. cit.). Whereas the North-Indian recensions have ua(-ha) the South-Indian recensions have invariably uva(-ha). For the reason behind the decision to follow the former here, see above, p. 167.
'lotus-plant from Skt
Var. II 38 and Hem. I 238 derive bhisin b/v)isini, which seems to consist of bisa 'lotus-stalk' followed by the (possessive?), suffix -int. Skt has bisa, Pali, Pkt and NIA bhisa (for NIA, see Turner, CDIAL 9249). The variant b(/v)isin (here S and W; in 8 T and S) is due to the influence of Skt. According to Mayrhofer (EWA II, p. 433) the existence of bisa beside bhisa would indicate that we have to do here with a loanword from an indigenous Indian language.
For the reasons to adopt (bhisint)vattammi, found in S (and y), rather than pattammi found in the other MSS, see above, p. 164.
For reha to shine' (Hem. IV 100), from Skt ribh- (rebhati) 'to sing' and 'to shine', see Mayrhofer, EWA III, pp. 73-4.
For the meaning of valaa (Skt balaka; originally invariably feminine), 'small white heron', see Thieme (1975: 15-22).
The retention of the intervocalic -g- in maragaa 'emerald' in Pkt as well as Apa. (see Hem. IV 349) is unusual. This may have to do with the fact that it concerns a foreign loanword; see Mayrhofer, EWA II, pp. 557-8. A comparison with Greek (o)uápaydos and Tibetan mar-gad, seems

Page #203

190
to show that the voiced -g- in Pkt is more original than the voiceless -k- in Skt marakata. The latter may be hypercorrect.
Whether the text reads samkhasutti in accordance with Bh, R, etc. or sankhas ippi with. T and W (for (c)chippi in T, see below), in both cases it concerns a compound of two synonyms for 'shell'. The context suggests that the compound denotes a single entity, which leads to the translation 'a suttt (or sippt), namely a sarkha' (Karmadhāraya). As such only samkhasippt makes sense, consisting of saikha, a word for 'shell' which is already long established in Skt, and sippt which is not found in Skt and is thus typical of MIA. For instances in Pāli, see PTSD s.v. sippt and sippikā, and for Pkt, see Gathā 62 and Setubandha, index s.v. The word sippt is a loanword from Dravidian (see DED 2089). Moreover, the compound sak hasippt seems to have several parallels, namely (vannag)ghaatuppa 'ghee' (see *22), vaivedha(a) 'fence' (see 96, 220, 221, 564) and acchabhalla 'bear' (see 109). Each compound consists of a word commonly known in Skt, ghrta, vrti and rksa, followed by a MIA. synonym, namely tuppa, vedha(a) and bhalla respectively. It would seem that the latter words were still considered new and strange and needed explanations.
It is significant that the more original reading, namely saikhasippi, occurs in the two South-Indian eclectic recensions T and W; this points to the conclusion that the Gathā was drawn from a text which was closer to the archetype than that of the North-Indian branch. It is extremely likely that the Gathā originally formed part of the Third South-Indian recension, from which it is now lost.
For chippt in T compare *9 chāli in Ma, Tp and chana in Ma, Tp and T. ch- in chali may be due to immediately following chettesu. It may, however, not be without significance that in all these cases it concerns an original ś: śāli and sana, and for sippt, compare Mar. stp or sip.
The variants hisiņi (for bhisiņt) and hāaņa (bhāaņa), both found in T, seem to be due to confusion of the Grantha characters for bh and h.
5( 5 Ma, 6 Tp; 4 Bh, 5R; K, , B, Y, P, G, 5; 148 T; 173 S) Missing in W.
tāva ccia raisamae
mahilāņa vibbh amā viraati

Page #204

191
jāva na kuvalaasac-
chahāi maulati naaņāin.
In Ma lacuna from the beginning upto sacchahāi. cciaṁ K; via S-- samaye B -- vimbh amās k -- sacchabhāi Tp; sachahā iņ R; K, V, B, Y (ohāiń), P; T; s-, Sachāhāií s -- mayulati Ma; maulenti. w'; enti st -- ņaaņāi Tp; ņaiņāiń R; (, Y).
Dasarūpa, p. 79.
At their love-play women go on coquettishly flashing their eyes until they (, the women, have their orgasm and) close their eyes which resemble the petals of the blue lotus.
The derivation of sacchaha 'similar to' (Pāiyal. 74: sacchaho samāņo, and Deś Tn. VIII 9: sacchaha...sarise) is uncertain. Those suggested by by Bühler (Pāiyal., index, s.v.) and Pischel (Deśmn., index, s.v.), from *sākşaka, and by Weber (1872: 742), from sa-chāya, are phonetically unsatisfactory.
In the language of the Sattas at the original distinction between the verbal e (+ aya) and a-stems seems in the process of becoming blurred, in favour of the a-stems (for this tendency, see Pischel & 491). Most often this is the case before -mti and mta, where the metrical distinction between long e and short a is annulled. Thus, we find here maulanti (from mukulayanti). Otherwise a fore is rare; see, for instance, kahai (kathayati) in 59. Some MSS, and S with great regularity, show a tendency here and there to reintroduce the distinction. See maulenti in v and s, and in 59 kahei in s, which, in fact, spoils the metre.
For the reasons to adopt na anain as found in Ma, see above, p. 169.
6( 6 Ma, 7 Tp; 5 Bh, 6 R; K, U, B, Y, P, G, 5; 233 S) Missing in T and W.
nohaliam appaņo kin eań khu tuha hasai suha-
na maggase maggase kuravaassa a valiamuhapankaan jāā.

Page #205

192
In Ma lacuna upto (e) am in c. Apparently the variants readings of P have been overlooked by Weber. nohaliyan Bh; dohalia B; noalian sa (nohaliam in Ed., p. 6 probably is a misprint.) -- ayyaņo K -- kuruvaassa S -- eā K -- kha Ma; tu Bh; vu K, kkhu Y; vvia kkhu S -- hasai tuha suhaa Bh, R; tuha suhaa hasai K, Y, suhaa suhasai U, suhaa hasai tuha B -- viliyamuha Ma; valiāņaņa , B, valiāna Y -- paṁkaja(lacuna) Ma.
Pāda c quoted in Hem. II 198 (ean khu hasai).
Why don't you look for a 'yield of new fruits' for yourself instead of looking for a 'young woman' for the kur abaka-tree? Thus, you know, handsome man, your wife is laughing, her lotus-face turned aside.
This Gātha and the following are closely related. The present Gathā refers to the belief that the kur abaka-tree, a red amaranth, blossoms only after having been embraced by a young woman; in * 7 reference is made to the aśoka-tree, which needs to be kicked before it blossoms. For these motifs, see, among others, Karpūramanjarf II 43. Furthermore, in the following Gathā a pun is made on the word asoa, meaning 'aśokatree' as well as 'griefless'. Likewise in the Gathā under discussion a pun is made, namely on the word nohalia. According to Hem. I 170 this word goes back to Skt navaphalikā. In Skt it is restricted to the lexicons, which translate it with 'a girl with whom menstruation has recently begun' and 'a newly-married woman' (PW, s.v.). It has been analysed as a Bahuvrthi-compound, consisting of nava and phala, the latter having the ad hoc meaning 'menses'; see AIG II 2, S 199 dy. The commentaries available on the Gathā give a totally different translation, namely navaphalodgama 'a yield of fresh fruits' or 'a fresh yield of fruits (see Abh., p. 74 for Kulanātha). The word probably is a feminine abstract noun ending in -ika, for which type of noun, see AIG II 2, $210.
The situation underlying the first line seems to be the following. The man addressed was in search of a young woman (nohalia), ostensibly for the kurabaka-tree. It is also possible that he looked for one for his own pleasure and used the kurabaka as an excuse. When his wife

Page #206

193
found out she said he had better look for a nohalia for himself. In the latter case we would have to do with the abstract noun, which evidently is used in a metaphorical, or ad hoc, sense here. As such it can have several meanings (note in this connection Omar. navhält, quoted by Turner, CDIAL 7004, which means 'first fruit of the season' as well as 'prime of youth'). The first sentence may, for instance, contain a reference to the man's old age or even impotence. Hopefully for him the humour is of a milder kind, navaphalika referring to 'children'. The woman says then that her husband does better to invest his energies in her than in the kur abaka-tree.
In the second line I have adopted tuha hasai suhaa of Ma, Tp and S (see also Ę: tava hasati subhaga). The order in Bh, R, K and y, tuha suhaa, with tuha immediately preceding the vocative suhaa, appears more logical, which, however, at the same time serves as an argument to consider it as secondary. Note in this connection eya tu (for khu) hasai tuha suhaa in Bh, which still contains traces of the reading of Ma, Tp and S.
In some of the other cases where such a broken-up sentence is found (note also tuha (c)...jaa (d) here) it appears to be due to the particuIar circumstances under which the words are spoken (see the note to *20). The form of the sentence in this Gāthā might then be explained by assuming that the messenger, who is laughing about the 'joke' made by the wife, is incapable of formulating a coherent sentence.
For valia turned (as ide)' Ma reads vili(y)a. Evidently this is a scribal error or a reading mistake made by the copyist of the transcript Ma. The meaning of vilia 'disappeared' (see 53 paccakkhadifthavilia 'clearly disappeared') hardly makes sense here.
7( 7 Ma, 8 Tp; 6 Bh, 7 R; K, V, B, Y, P, G, 5; 238 T; 40 S) Missing in W.
tāvijjanti asoeki sahai ko vi kassa vi
hi ladahavilaão daiavirahammi pāapahār aṁ pahuppato.
In Ma lacuna upto pāapahārań. tāvijjintiñ K -- asohim Tp; asochin K
--
ledaha Tp; land aha K
--

Page #207

194
vilayāu Bh; vilaā k, viraão V, vania B, vaniāu Y, virahão P -- virahasmi k -- paguppan to Ma; pahuşpasto Bh; pahuppatto k, pahupanto y, pahupanto P.
Even irresistable temptresses let themselves (occasionally) be tormented, (namely) by men who experience no grief when separated from their mistresses. (For) why would someone who is his own master (is the master) suffer to be kicked by someone else?
Whereas the usual place of the finite verb is at the end of the sentence, in this Gathā it is found at the beginning. For similar instances, see 562, 692, 693 and 769. According to Gonda (1952: 67) this results in the accentuation of the affirmative nature of the verb. Thus: 'contrary to what one generally believes laqahavilaās can be tormented'. Gāthā 817 shows that this indeed happens only rarely. Usually the ladahavilaas themselves are the tormentors. In fact only saints are deaf to their seductive charmes or, as here, people who are completely indifferent to female company: asoehi...daiavirahammi, 'people who experience no grief...when separated from their mistresses'. Gātā 817 reads as follows:
lagahavilaāņa loaņa- jhijjaħti mahāsattā
kadakkhavikkhevajanias aftāvā cittuvveanas ahā hosti.
The syntax of the Gātha is misunderstood by Weber. I should like to propose the following translation:
Tormented by the side-long glances of the irresistable (ladana) vilaās one (inevitably) becomes thin. (Only) great saints are able to bear the agitation of the mind (such women bring about).
In the second line of the Gathā under discussion a pun is made on the word asoa of the first line. Asoka-trees are believed to crave for a kick from a young woman. Only when satisfied in this desire do they bring forth flowers. For this motif, see Karpūramañjart II 43. The

Page #208

asokas of the first line are strange aśokas indeed as they do not show the usual reaction to the kick (charms) of the ladahavilaas.
In Hem. II 128 the word vilaa is mentioned as a Pkt substitute for Skt vanita. In Paiyal. 12 it is enumerated in a list of synonyms for 'woman', ranging from vahua to niyambint. In both Gatha *7 and 817 the women denoted by ladahavilaa are sexually extremely attractive. Most likely they are prostitutes as they are actively seducing men (see 817). It may be asked if these two connotations, attractive and prostitute, are conveyed by either ladaha or vilaa.
The word vilaa is relatively rare. In the Sattasat it occurs three more times and in the Gaugavaho only once (166 vāravilaa). It does not occur in the Setubandha. For instances in later texts, see PSM, s.v. The following discussion is restricted to the instances in the Sattasat.
195
In 156 and 890 the compounds vesa- and vesavilaa occur. The meaning 'prostitute' for. vilaa would lead to a redundancy; compare vesavilaa in 890 with Skt vesastri, -yoşit, etc., in which the final member denotes simply 'woman'. vesavilaa may be taken as a Karmadharaya; in theory it could also be a Dvandva or Tatpuruşa. However, the vesăvilaas are no common prostitutes. To the purely perfunctory performance of their profession they have added a special attraction by showing some sort of personal involvement or interest. The clients are fascinated by this as appears from the following two enraptured exclamations.
156 namdamtu suraasarahasa
vahumagganimmiāiṁ
taghapahari saalaloassa vesavila pa pemmih.
Long live the affectionate feelings which the vilaas who are prostitutes know how to show in so many ways, which (alone) are able to completely satisfy the vehement desires for love-making everybody has!
For the connotations of the word permai, see the expression premabandhura a prostitute who has, or is able to convince her client she has, a real affection for him', found in Kumarapalapratibodha (ed. Alsdorf), p. 109, verse 76.

Page #209

196
890 sacchadar amanad as ana
suviaddhavesavilaa
samvaddhiagaruavammahavilās am rami aṁ ko vanniu tarai.
Who can describe making love to the skillfull vilaās of the prostitutes' quarters, during which love's joviality, which is already vehement, is further enhanced as they show they do it all spontaneously?
The only instance in which the word occurs by itself is 233. It is not clear whether the woman is a prostitute here or not. It is clear, though, that it concerns a woman who at least in her youth was very exciting, her body being compared with the temple of the god of Love. She brought the men she made love with to complete frenzy as appears from the fact that they bit her practically everywhere.
thanajahanani arvovari uvvasiānanganivā
dasaņańkā gaavaāna vilaana samūlavandha vva dfsaṁti.
The tooth-marks on the breasts, the thighs and the buttocks of old vilaās look like the foundations of a deserted (delapidated) temple of the god of Love.
The foregoing instances show that the attractiveness of the vilaās does not depend on the presence of the adjective ladaha. Furthermore, 156 and 890 show that the word vilaā does not by itself mean 'prostitute', though it may be used to refer to such a woman. It would seem that the word means 'woman' but with special reference to her great attractiveness and to the effect she has on men. As an English translation I have therefore adopted 'temptress'. In this connection also the derivation of the word may be considered. I would venture to suggest that it'simply is the feminine of Skt vilaya 'destruction'.
In their derivation of pahuppa- 'to be (one's own) master' Weber (Ed., p. 6) as well as Jacobi (1886: 253) proceed from the past participle pahutta, which would go back to Skt prabhūta. Analogically after litta 'smeared' : lippa- 'to be smeared', a passive pahuppa- would have been formed, which has acquired an active meaning. This derivation is

Page #210

197
questionable on two grounds. In the first place no certain instances of hutta from bhūta are known. The latter regularly becomes (b) hūa. Secondly, it assumes the existence in Pkt of deponentia of which, however, no certain instances seem to exist either (see the note to *10). Pischel (5 286) suggests instead that pahuppa- is a denominative verb based on the abstract noun prabhutva. Admittedly the sole instance of this noun in the Sattasaī, in 91, has pahutta. However, the alternative development of -tv- to -pp- is known dialectically; see Pischel § 300. In this connection it may be interesting to note that Hemacandra, who in IV 63 mentions pahuppa- as an Adeśa of prabhū-, specifically adds that it is used only when the subject or agens (Kartrka) is a prabhu'. However, this derivation creates only a new problem, as no explanation for the origin of the past participle pahutta beside pahuppa- is available.
8( 8 Ma, 9 Tp; 7 Bh, 8 R; K, y, B, Y, P, G, E; 442 T; 487 S) Missing in W.
attā taha ramanijjaħ Tuatilavādisariccha
amh am gāmassa madanam bhūam sisirena kaam bhisinisaṁdań.
In Ma lacuna from gāmassa onwards to the end. amha, Bh -- gomāsa Ma -- mamdaņaābhūdaṁ Tp; mama anabbhūyan Bh, maño an Thūań R; K (haa), V, B, Y, P; mamdan Tbhūdam T; S' -- lūyatilavāda Bh -- sarichan K, sarisan B -- bisiņT T; visant S -- mand an K, sandhan Y.
Mother-in-law, the beautiful cluster of lotuses that used to be the ornament of our (my parents') village has been treated by the winter in such a way that it has come to look like a harvested sesame-field!
According to Weber (Abh., p. 75) the bhisiņisaṁd aṁ and the tilavādī would be places where the woman speaking used to meet her lover, an interpretation apparently also found in the commentaries, which, however, have not been quoted specifically by Weber. It seems unlikely, though,

Page #211

198
that a woman would discuss such matters with her mother-in-law (attă, from Tamil attai; see Turner, CDIAL 222, and Pāiyal. 253 attā sāsū, but see also below). It is more likely that she complains in guarded terms to her mother-in-law about the behaviour of her husband, and especially about his coldness or indifference towards her, which has physically ruined her 'who used to be the ornament of her parents' village'. The misunderstanding as evinced by the commentaries and Weber may rest on the fact that fields are indeed often referred to as lovers' meetingplaces (see, for instance, Gātha *9). The gloss saht 'friend' for attā in Pāiyal. 108 (māuyā sahs attā) and in Deśīn. I 51 (mā i-piuccha-sāsūsah Tsu attā) may go back to a similar misunderstanding.
mand an an bhūam has been reconstructed on the basis of Tp 'mand anabhūdam and Bh maid anabbhūyam. A comparison with the phrase unnaehi hoūņa in 83 suggests that bhūam functions here as a kind of emphatic tense marker: 'which was the ornament'. maíd an Thuaṁ of R etc. is probably an innovation for the uncertain form in Bh.
taha in a has been constructed with sariccham in c, '(it) has been treated in such a way... that it has become like...' (cf. taha...vva in 413).
lua is the past participle of Skt lū- 'to cut'. For the short -uin lua, see Jacobi (1898: 569), who explains it from the present stem lunāmi. Compare dhua from dhū-, dhunoti.
For bhisint, 'lotus-plant', see *4.
9( 10 Ma, 11 Tp; 8 Bh, 9 R; K, Ų, B, Y, P, G, 5; 455 T; 488 s) Missing in W.
kiñ ruasi onaamuhT hariālamand i amuht
dhavalāatesu sālichettesu nadi va saņavādia jāā.
kin ruyas i Ma; mā ruasu Bh, R; S -- chalTcchettesu Ma, chalichettesu Tp;
cettesum T; sālikkhettesu S -- muhā Bh; mukht B -- pada Bh; in Tp the final vowel of muht and the first syllable of nadi are illegible -- cchana Ma, chana Tp; cchana T -- savāțuyā Bh (saņavādiyā Bho).
Why are you crying with your face bent down now that the .

Page #212

199
rice-fields are turning white? The hemp-field (its surface decorated with yellow flowers) has become like an actress whose face is decorated with yellow paint!
The girl is crying because the rice-fields are about to be harvested after which it will be impossible for her to meet her lover there. She is comforted by her friends who point out that the hemp-field, which is in full bloom, could be used instead.
For hari(t)āla, 'yellow orpiment', used by actors as paint in dramatic performances, see MW, s.v. Unfortunately no sources are given there.
For the variant chali (Ma, Tp) for sāli, and for chana (Ma, Tp, T) for saņa, see *4.
10( 11 Ma, 12 Tp; 9 Bh, 10 R; K, M, B, Y, P, G, 5; 292 T; 276 S) Missing in W.
sahi erisa ccia gaT eāņa vālavālun
mā ruvvau tasavaliamuha adań kitamtukudilāņa pemmāņa.
erisi Tp; R; K, M, B, Y, sa Pi sārisi T; erisi S -- vviya Bh; via S -- gain K, P(?) -- ruvvasu Ma, Tp; R; K (ruccasu), , B (-cc-), rua . v, ruvasu P; ruvvasu T; ruccasu s-, ruchasva st -- tankhalia B -- eāņa K -- bāla Ma, Tp; R; B, Y, cāla K, P; bāla T -- bālukki Tp; vālukvi K, vālukki , vălukki B, vāluki P; vălukki T. -- bhumtu (for tantu) Bh -- kudilāņań k -- pemmānań K, B, P; T; S; (R, W, Y).
Pada a is quoted in Hem. II 195 (erisi, v.1. erise, ccia).
Friend, stop crying with your moon-face turned as ide: this is the course of these affectionate feelings which 'curl' (indiscriminately around everyone who is available) like the tendrils of the young cucumber-plant.
In the case of pronominal adjectives such as erisa the feminine marker, as in Skt, generally is -1. Pischel ($ 245) quotes jādisi,

Page #213

200
tādist, etc. Pāli, too, has tādist. Berger (1955: 42) explains the development of -r- of tādrśa into -i- instead of into -a- in tādisa as an instance of assimilation to the final -I in the feminine tādist. The feminine marker -ā is extremely rare in these words. Pischel (loc. cit.) quotes only sarisayā (Nāyadh.). To this may be added Pāli ktdisā in yādisikTdisā jivikā (PTSD, p. 553C) and Apa. erisa in erisa joda na jutta (Kumār apālapratibodha, p. 107, verse 66).
In the text therefore erisa of Ma and Bh is adopted, the change of the latter into erisi (TP, R, etc.) being much more likely than that of erisi into erisa. We should, however, be aware of the possibility that erisa is a clerical mistake for erisi with -sa from the preceding word sahi.
In the Sattasas the word tassa occurs twice more, namely in 148 where it denotes a certain enticing way in which the eyes move (tasavalaṁt addhatāraāloā) and in 692 where it denotes a particular position of the feet (heels down and toes up'). In Paiyal. 269 it has been translated with tiriccha 'oblique(ly), slanting'. Hemacandra (1 26 and II 92) derives it explicitly from tryasra (synonymous with tryasra) 'triangular'. A comparison with caturaśra(ka) 'a particular position of the hand, a particular posture in dancing' (from 'having or making four angles') would suggest for tamsa a meaning 'making three angles (on (the rest of) the body)'. It would subsequently have acquired a somewhat less technical sense, like 'slanting' or 'oblique'.
vāluhk T, 'a kind of cucumber'. Skt has masculine or neutre valurka (MW, s.v.), feminine vālunkt being found only in the lexicons. Pāiyal. 172 has valuñk aṁ and Mar. vālūka (quoted in Ed., p. 7). Feminine vāluikt is probably ad hoc, the speaker by mentioning this plant referring to a woman.
ruvvau is found only in Bh (Bhuvanapāla: rudyatām). The corruption of ruvvau into ruvvasu, found in the other MSS, probably rests on a confusion between the grammatical and the logical subject, the latter being a second person. In his Abh.(p. 6) Weber, to whom at the time the variant ruvvau was not available, explains ruvvasu as a deponens. In his editions this explanation turns up several times more. However, a closer look at the instances concerned shows that most if not all can one way or another be eliminated. It follows that as far as the Sattasaf

Page #214

201
is concerned deponentia do not exist in Pkt. It should be noted that Alsdorf (1937: 61-8) after studying those verbs in Apa. earlier identified as deponentia, reached a similar conclusion regarding that language.
An obvious case is tīrae. This is not, as Weber (Abh., p. 64) assumes, a deponens but a regular passive of the verb tara-, the usage of which is comparable to that of Skt śakyate. gasijjihii in 804, likewise, is a regular passive. Gathā 804 (only found in S) reads as follows:
vārijjartT navako- mā te amvupisão
mui tti ma putti aṁgane suvasu caído tti muhaṁ gasijjihii.
If someone tries to stop you, saying '(it is) full moon', you should (obey and not sleep in the courtyard, little girl, lest your face will be swallowed by Rāhu) who may think the (real) moon (only) a piśāca in the water (arvupisão).
A different case is pijjai in 678. Gathā 678 (only found in w and Y ; and in G and ) reads as follows:
pijjai kannańjalihim duddha jalas amilia
janar avamilia pi tujjha salāvam să vālā rāahaṁsi vva.
As far as the first half is concerned the passive pijjai is in perfect order. An anomaly arises only in connection with the second half. One may therefore consider the possibility that the Gathā is a patch-work verse assembled, with little regard for syntax, from two independently existing half-verses. Note in this connection the awkward repetition of the word (Saṁ)miliań. The fact that Hemacandra in IV 10 includes the passive stem pijja- among the Adeśas for pā- is no independent support for Weber's (and the traditional) interpretation for pijja- in this Gathā. The inclusion of pijja- in Hemacandra's sutra may in fact rest upon a misunderstanding of an instance like this one. Another instance of the 'deponens' pijja- quoted by Pischel (in his translation of Hem., p. 132), namely pijjanti in Śakuntala (ed. Pischel, p. 29, line 5) is in my opinion a regular passive.

Page #215

202
gammasu in 819 should be eliminated in favour of gamesu which has become available through Ma, Ti and Tp. ruvvasu as a corruption for ruvvau is not unique. Another instance is gammasu in 715 in T, for gammau in the other MSS available. It also explains gammihisi in 609, which against the evidence of all the available MSS should be emended into gammihii. A possible third instance is bhaņņihisi in : 906, which, alternatively, is a regular passive, meaning 'you will be told'.
The case of lagga-, pahuppa-, vacca- and dhakka- (Abh., p. 64) is different yet again. The final -kka in dhakka-, 'to cover', is not from -kya but a MIA verbal suffix (see *14). lagga-, 'to cling to', and vacca'to go', go back to *lagyati and *vrajyati (IV class), rather than being derived from the respective passives. lagga- can also have been based on the past participle lagga (lagna) (the origin of the voiceless cluster -CC-in vacca-, instead of expected -jj-, will be discussed elsewhere). Finally, the instance pahuppa- is, as has already been noted in *8, doubtful, the derivation still being far from certain.
11( 12 Ma, 13 Tp; 10 Bh, 11 R; K, V, B, Y, P, G, 5; 590 T; 322 s) Missing in W.
pāavadiassa paino didhamannudūmiāe vi
patthim putte samāruhaṁt ammi hāso ghariņte nikkaṁto.
pāapadiassa Ma (paa), Tp; R; K, M, B, Y, P; T; S -- patthiṁ paino Tp; puţthim R; K (oththo), V, B, Y, P; T; putte patthim S -- ruhaṁt am Ma, Oruhaît ainmmi Tp -- dadha Bh, R; K, V (datha), Y, P; S -- mamna Bho, manna Bh"; manu K -- dūmiai Y, dummiāe P; důmiāi S -- ghāso Ma -- harinte Ma; gharinti Bh; gharinta B, Y; S -- nikvasto K.
When their son climbed on her husband's back who had fallen at her feet, the wife could not help laughing, even though she was tormented by a fierce anger.
For the reasons to adopt pāavadiassa found in Bh rather than padiassa in the other MSS, see above, p. 164.

Page #216

203
paţthi, 'back', corresponds to Skt prstha. The final -T has been explained as arising from contamination with pitthT (Skt prsti) 'rib' (see Andersen, Pāli Glossary, s.v.). Ma, Ti and Tp throughout have pațghi, with -a- for -r-. Likewise T, except here (T also often has paţtha (+ prştha), for which, see * 35 (also W), 87 and 223). Bh varies: patthi here and in * 35, 216 and 223, but puţght in 313 and pittht in 87. The other MSS generally have puţthi (note, however, that S has paţthi here). The -u- for -r- is due to the influence of the preceding labial. It should be noted that pattht (or puţthi) is often found rhyming with pa- as well as with pu-: *11 paiņo patthin putte, 35 paţthin palfvesi, 87 patthipulauggamo, 216 patthie (c), pulaubbhee (d), 313 paţthin p(h)usasu, and 676 padara pattht.
For drdha Ma, Ti, Tp and T have throughout didha. The other MSS usually have dadha (exceptions: didha in *11 B, 108 K, Y, P and 276 K, S, (x)). For the principle underlying the decision to adopt the forms of the South-Indian recensions here, namely pațghĩ and didha respectively, see above, p. 165. .
For didhamannudūmiae vi, see also 74. Var. VII 8 and Hem. IV 23 mention dūme- (actually they have dūma-, but the grammars quote all verbs as a-stems irrespective of their actual conjugation) as an Adeśa for Skt du- 'to burn, to afflict'. The verb occurs frequently in M. Pkt and also occasionally in the Panhāvāgaraņāim, a relatively late Amg. text. The most common glosses found in the available commentaries and Chāyās are du-, duḥkhay-, vyathay-, today- and tāpay-. The instances of the word show that the affliction brought about is both mental and physical. For the former meaning, see, for instance, Gāthās 792 (dūmesi maņaṁ) and 937 (dūmes i maha hiaaṁ) and Panhāvāgaraṇāin 2, 17 (hiyayaman adūmakain). A clear case of the latter meaning is found in Panhāvāgaranāiỉ 1,30, where the expression pahāradūmaņa occurs in a list of mutilations, such as aţthibhasjana, nās ābheya and chaviccheyaņa. Note, furthermore, Setubandha XIII 96 (riusūladūmio). In the same text, in V 24 (anuaṁpādūmiņana) and A 8 (saravian ādūmiānana), it seems to me an 'to distort, to disfigure'. It is also possible that dūmia in the latter two cases is in fact a corruption for dūsia (Skt dūşita).
These two sets of meanings refute the derivation, first suggested by Weber (Ed., p. 8) and later elaborated upon by Schwarzschild (1962:

Page #217

204
70), of dūme- (Schwarzschild 'se désespérer') from Skt durmanas 'in low spirits'. This derivation assumes a development from 'to afflict mental pain' to 'to afflict physical pain', which is not very likely. Furthermore, formally (dummaņa + dummia + dūmia → dūme-) .it rests entirely on the variant dumme - occurring in some of the MSS of the Sattasaf. This dumme - almost certainly is an innovation, probably introduced only. after the link between dūme- and durmanas was made. Instead, Turner, CDIAL 6493, suggests to derive dūme- from a non-existent * dūman 'distress', a noun based on the stem du-. However, by doing so the problem of the derivation of the verb dūme- remains unsolved.
Beside dūme- 'to afflict' Hemacandra, in IV 24, mentions a verb dūmameaning 'to whitewash', which would as such be a variant of duma-, for which latter verb, see Gathā 747. dūma- 'to whitewash' may, however, well be the same verb as the one discussed above, its meaning having been arrived at ad hoc, in which its resemblance to duma- may also have played a role. The only instance of dūma- 'to whitewash' I have been been able to trace back 'is Kalpasūtra, p. 42, where it occurs in a standard description, also occurring in Nāyādhammakahāo 1, 1, 18, of a bed-chamber the outside walls of which are dūmiyaghattamatta. In his word-index Jacobi gives for dūmiya the gloss dhavalita. It should, however, be noted that dūmiya, which, as has been shown above, denotes, among other things, some more or less rough and painful treatment, occurs here side by side with the participles of two other verbs (ghrs - and mrj-), which denote more or less rough treatments as well.
12( 14 Ma, 15 Tp; 11 Bh, 12 R; K, B, P, 13 V, Y, 12 G, 5; 313 T; 472 S; 6 V)
saccam jāņai daţthum marau na tuman bhanissam
sarisammi janammi jujjae rão maranań pi silāhanijjam se.
In Ma lacuna between da(tthun) and (sa)risammi. du(thun) Ma; dadhdhun K -- rāom k -- marane tuma Ma -- silānagijjan Ma; salāhanijjaħ Bh, R; K, M, B, Y, P; T; S.
Daśarūpa, p. 29 (salao).

Page #218

205
I must say, she knows what looking is! Passion for a man who is one's equal is right (but not for a man like you). Let her die! I will not tell you (what she asked me to tell you). Even such a dreadful thing as death is to be welcomed by her (in this situation).
A messenger addresses her friend's lover. The girl has threatened to die if she does not meet him. For this threat, see also 178 (= 739), 588 and 604 (Pāda d in 604 reads marau na bhanissan). However, when the messenger sees that the lover is unworthy of her friend she decides not to intervene. In her opinion the girl is better off dead than alive. saccam in saccaṁ jāņai datthum is evidently meant cynically here.
In the source of Ma the words marau na seem to have been partly or wholly lost, as would appear from the fact that in the accompanying Chāyā these two-words have been skipped. mar ane looks like an ad hoc attempt to fill in this lacuna: 'at her death (I will tell you)'. : For silahan ijja, found in Ma and Tp (see also 114 Ma, Ti, Tp and
ī (s11°)), beside salo of the other MSS, see Pischel § 133. For the principle underlying the decision to adopt the form found in the South-Indian recension, see above, p.165.
U, B, Y, G, 13 5; 389 T; 139 s)
13( 19 Ma, 20 Tp; 12 Bh, 13 R; K, P, 14 Missing in W.
randh anak amman iunie muhamāruan pianto
mā jhūrasu rattapādalasuandhan dhūmāi sinf na pajjalai.
katmma Ma -- jūrasu Ma, Tp; Bho, R; Y, sūrasu B; jūrasu T; S -- pādali S -- suamdh1 Ma; suaṁdhin T -- pivasto Ma; pibato B -- vijjhāi Bh; dhūmăhi P; dhūmei T -- pajalai Ma.
There is nothing wrong with your cooking. Don't be annoyed. The only reason the fire doesn't start to burn but keeps smoking is that it wants to drink your breath which is as fragrant as the red pāțalā-flower!

Page #219

206
A woman, probably newly married, is teased with her clumsiness in lighting the kitchen-fire. A woman's clums iness in the kitchen is also the subject of the next Gāthā and of 732, 733, 734 and 735.
jhūrasu of Bha, K, V and P is preferred to jūrasu of the other MSS. Hemacandra mentions the latter verb in IV 131 as an Adeśa for khid- 'to be annoyed' (together with visūra-, which explains the variant sūrasu in B) and in IV 134 as an Adeśa for krudh- 'to be angry'. jhūra- is mentioned in IV 74 as an Adeśa for smr- 'to remember, to think of'; it would seem that Hemacandra thus tried to account for the meaning 'to worry' (see below). Turner derives jūra- from jvar- 'to be feverishly hot' (CDIAL 5259) and jhūra- from jr- 'to become old' (5409). He suggests that the initial jh- of jhūra- is due to the influence of jhi- from ķşi'to destroy'.
It is doubtful, though, that we have here to do with two different verbs. In the MSS they occur as variants. Admittedly, jhūra- is rare. Thus, in *41 jhūra- is restricted to K, in 570 to R and in 610 to y. In the other instances of the word, viz. 354, 454, 695 and 822, all available MSS have jūra-. Compare the situation in the Setubandha: in XIV 1 jūrai (v.1. jhūrai), in XI 3 jūrai (v.l. kuppai) and in XI 1 cittavia (i.e. cintavia; v.1. jūravia). The latter instance seems to undermine Hemacandra's distinction between jūra- 'khid-, krudh-' and jhūra- 'smr-'. As already said above his classification of jhūra- as an Adeśa for smr - seems an attempt to account for the meaning 'to worry', which as Setubandha XI 1 shows is equally well conveyed by jūra-. In this connection it is not without significance that jūra- does not have any NIA derivatives; see Turner, CDIAL 5259. Turner mentions only Prasun žora 'to be angry', which could, however, just as well go back to jhūra-. In contrast to jūra-, jhūra- has survived in most NIA languages; see Turner, CDIAL 5409, who quotes P. jhūrnā 'to repine', H. jhūrna 'to waste away, dry up'. OGuj. jhūrai 'to repent' and Guj. jhur vũ 'to languish'.
I should therefore like to suggest that Pkt jūra- is in fact a mere ghost-word. However, the reasons that may have been behind the supposed change of jhūra- into jūra- remain largely unclear. It may have been due to a simple orthographical problem, the subscribed ū having blurred the distinction between the (Jaina) Nāgarī characters for jh and j.

Page #220

207
In the commentaries on the Sattasaf and the Setubandha there is no evidence of any etymological speculation concerning the word, which might have resulted in a change in form. The usual glosses, khid-, krudh- and kup-, are translations rather than etymologies. For this it would seem that we have to turn to Skt which has jūr- 'to be angry, to be annoyed' (only once, namely in Bhattikāvya XI 8). This jūr- could be a loanword from Pkt, but in the light of what has been said above regarding Pkt jūra- it is more likely a loan-translation, showing that jhūra- was etymologically connected with jvar- 'to be feverishly hot'; compare jūr- with jūrni 'glowing fire', which according to the lexicons would also mean 'anger'.
suadht in Ma stands for suandhin (cf. T) or for suaħdhan, -I, -in and -aṁ in the Malayālam script being hardly distinguishable.
pādala, 'Stereospermum suaveolens'. For the variant pāçalt in S, see also 468 (Y, T), and 621 (Y). Compare Pkt halidd beside Skt haridrā in Gāthās 58 (Ma, Tp, Bh), 80 (Ma, Tp, Bh, R) and 246 (Ma, Bh, R). haliddt (see Var. V 24 and Hem. I 88) is ascertained for the archetype in these instances. Each time, though, a variant haliddā is available, undoubtedly introduced under the influence of Skt haridrā. The case of the variation of padala and padalt is different. It is not possible to say with certainty that pāçalā was secondarily introduced under the influence of Skt for pādali as Skt has pāțalt as well, even though it occurs only relati.vely late in that language.
:14( 15 Ma, 16 Tp; 382 Bh, 14 and 386 R; 14 K, P, 12 V, Y, 13 B, G, 14 €; 388 T; 203 S) Missing in W.
ghar inte mahānasakam- chikkan muhan hasijjai
malaggamas imailiena hatthena candāvatthaṁ gaan painā.
ghar inie Ma; na R (14, 386); B, Y; S -- kāma Ma -- mailienan K, maliena Y; mailena S -- cchakkam Ma, chittan Tp; chikvań R (386), chittań R (14); K (chimnnam), V, B, Y, P (ci"); cikkan T; chippan S -- hisijjai Y, hasijjayi P.
The husband laughed at his wife's face which, touched

Page #221

208
with her hand dirty with soot from working in the kitchen, had come to look like the moon.
mailia 'made dirty' is a past participle based on the adjective maila 'dirty', which is from malila (mala+ ila; see Turner, CDIAL 9904). For mas imailia, see also 70; mas imaila in 117.
chikka is the past participle of chiva- 'to touch'. For the suffix -kka, see Schwarzschild (1958).
Hem. IV 182 mentions chiva- among the Adeśas for sprś= 'to touch'. The derivation is uncertain. Turner, CDIAL 5055, suggests that it goes back to chup- (IE skup-) contaminated with kşip-. By way of alternative I should like to suggest that chiva- goes back to Skt kşip- 'to throw, to strike', which became khiva- 'to throw', on the one hand, and chiva'to touch', on the other. The meaning 'to touch' would have developed from that of 'to strike'. For the existence in Pkt of two alternative forms going back to one Skt word, each with a different meaning, compare khaņa 'moment' beside chana 'festival' (from 'right moment'), which both derive from Skt kşaņa (see *42).
The variant chitta shows the influence of Skt kşipta.
15( 17 Ma, 18 Tp; 13 Bh, 15 R; K, B, P, 174, 16 Y, 15 G, 5; 653 T; 611 S) Missing in W.
kiñ kin de padihāai padhamullaadohaliņi-
sahihi ii pucchiae muddhāe e navari daian gaā dittht.'
In Ma lacuna from the beginning to (padha )muo. kim ki y -- te Tp; R -- padibhāai Tp; padihāsai Bh, R; U, B, Y, padihā k, padiāsahi P; padibhāai T; padihāsai S -- sahihi R; K, Y, P; sahti S -- ia Tp; Bh, R; K, 4, B, Y, hia P; i T; ia s -- puchiyai Bh, pucchiai R; Y, puchiae K, y, puchiĪa B; puchiai S -- muddhāi S -- (lacuna)mullua Ma, padhamallaa Tp; padhamu(lacuna)haliņie Bho, padhamunnaya Bh; pamthamullaa K, padhamuggaa B, Y, pathamullaa P; padhamullia T; padhamuggaa S -- dohanie R; dohaninta y ; dohalinta S -- nava Ma, navara Tp; Bh, navaram R; naṁvar am K, navara , B, navaran P; T; ņavara S -- dimmi Ma; diththt K.

Page #222

209
Pāda a quoted in Hem. III 80 (kin ki te padihāi).
When her friends asked the young wife who was pregnant for the very first time and experienced longings (common to pregnant women): 'whatever do you like?', her eyes fell on her husband and on nothing else!
The word dohala (Skt dohada) refers to a woman's cravings for unusual kinds of food during her pregnancy. The woman in this Gāthā, who is pregnant for the first time, apparently mistakes the nature of her cravings, thinking they are a desire for her husband.
The Skt enclitic pronoun te 'you' is found as de and occasionally after vowels as e. The instances of de and e are the following. 547 kim de (S; te in all the other MSS), 609 muhan de (Bh, , Y, T, S; te Ma, Ti, Tp, R). In these two instances te in the South Indian MSS, which often have -ind - for -it- (677 saņdāva T, Ma, Tp), may be hypercorrect. Other instances are: 554 tam pi de vahuan (Ma, TP, X, Y, T, S; te W; he Ti), 895 atthi de kajja (Ti, Tp, T), 896 kkhu de (Ti, Tp, T; me Ma), 946 je de eaṁ (Ti, Tp, T), 948 jaha de pia (T; te Bh, R, S). Instances of e are: 683 ma e maņņu (Ma, TP, R, T), 734 kaha e (R; de T), 737 vie (R; de s), 908 aha e icchai (Ma, Ti, Tp; de T), 925 kaha nu e 'māņo (Ti, Tp; de T).
The same development can be seen in the pronouns deņa 'by him' (tena) and e 'they' (te). For deņa, see *29 (T), 164 (Tp, T), 276 (R), 438 (S), 627 (1), 651 (T) and 905 (T). For e, see 732 chivasti e jala (Ma, R; de Tp; te T, S) and 733 saṁdhei e jāla (R; te T).
de for te is the result of the lax pronunciation which goes with the often unemphatic use of the pronoun, which is especially the case with their enclitic forms. This phenomenon has been discussed by Turner (1928). As such the development of te into de has parallels in, for instance, vi (api) and dāva (tāvat). The instances of e could be due to accidents in the transmission of the text, the original initial consonant, d- or t-, having been deleted unconsciously as it occurred between vowels On the other hand, the loss of the initial consonant could be the result of the same phenomenon discussed above. As in the case of de there are several parallels, namely a (ca), i (vi; see *17)

Page #223

210
ira (kira; see *36), ā (tā(t)), uņa (puna) and u (tu). For e, the personal pronoun of the second person, generally a variant de is available (throughout in T and once in S) except in 683, in which case e was misunderstood (see above, p. 163). This marks de as the innovation. The nature of the innovation is in itself peculiar in that it is another typical Pkt word.
For the possible use of de, the personal pronoun of the second person singular, as a kind of particle, see *16.
padihāai (TP, T) corresponds to Skt pratibhāti. pad ihāsai (pratibhāsate) found in the other MSS is most likely originally nothing but a clerical error, the -s- having been intruded from the following word sahThi. It was retained as it did in no way disturb the original sense.
For ii of the text the available MSS read i(y)a. Var. I 14 and Hem. I 90 (see also I 42) mention this ia as a substitute for Skt iti. Its occurrence would be restricted to the beginning of a word (pādādau; Var.) or to the beginning of a sentence (vāk yādau; Hem.). As such it differs from the other substitute for iti, namely (t)ti, which functions as an enclitic.
Pischel ($ 116) looks upon ia as a direct continuation of Latin ita. However, a comparison with AMg. shows that the situation is somewhat different. In AMg. one finds beside iya a form ii, which goes back directly to Skt iti. In North-Indian, and especially in Jaina MSS, the syllables -i- and -ya- are frequently interchanged, which is due to the fact that they are very similar in (the North-Indian) pronunciation. This explains iya beside ii. It should be noted that Pischel (loc. cit) leaves open another possibility, namely that ii in Amg. stands for original iya. It is very doubtful, though, if a copyist would accept such an awkward form as ii unless it was already there. Compare the case of the ending -(y)a in kodi(y)a, for koqti, which latter form was apparently considered awkward, discussed in * 2.
The occurrence in M. Pkt, throughout, of i(y)a would only show that the MSS of most of the earlier M. Pkt texts, go back to North-Indian MSS. This is certainly the case with the MSS of the Sattasaf. For the occurrence of i(y)a in such later texts as the Kams avaho (see I 55) the grammars may be held responsible. In the text I have everywhere adopted ii against the unanimous evidence of the MSS in favour of ia. .

Page #224

211
Vararuci and Hemacandra make a clear distinction between navara(n) 'no other, nothing else' (Var. IX 7 and Hem. II 187) and navari(in) 'immediately after' (Var. IX 8 and Hem. II 188). Hemacandra in his own commentary on II 188 adds, however, that according to 'some' navara (in) and navari (n) are synonyms.
In the Setubandha the two meanings are neatly sorted out between the respective words. Note that once, though, in IV 78, navara 'no other' has a variant navari. In the MSS of the Sattasaf the situation is less regular. The words occur nine times in all (I leave out the instances in 515 and 527, where they are found only in Bh and R as variants for nāma and taha vi respectively of the other MSS), each time meaning 'no other, nothing else'. The majority of the MSS have navara(in), pavari () occurring as an isolated variant:* 15 navari Y, *34 Bha, 172 V, Y, S, (x), 248 S, 303 Bho, S, 485 R (navari), Ma (navarin), 614 y. Gathas 875 and 953 are found only in T, which has navari.
Thus in the Sattasaí navara(m) and navarim) are merely variants of the MSS. It should be noted that the occurrence of navari (n) beside navara(m), though each time restricted to only one or two MSS, is too persistent to allow to consider the word as a mere scribal error. What we see is in fact the substitution of navara(m) by navari(n) or vice versa. As far as the Sattasas and the Setubandha are concerned the variation is significantly restricted to the instances where the word means 'no other, nothing else'. A confusion of the two respective contexts is in fact unlikely to have taken place, navari 'immediately after' apparently occurring only in the phrase navari a (for instances, see Setubandha, word-index, s.v.). As a result navari a was taken as one entity; see päiyal. 17: navariya...sahasatti. In the grammars navari is recognized as an independent word, meaning 'immediately after'. From these points of view navara(mn) is more likely the innovation here than zavari(i).
The situation may also be approached from an entirely different angle, namely by putting the question whether we have here to do with the substitution by an independently existing synonym or, alternatively, vith a change or emendation of the original word. Evidence bearing on this problem, again, points to the secondary nature of navara(in), the independent existence of a word navara being highly uncertain. Thus,

Page #225

212
in Pkt adverbs ending in a bare -a are virtually non-existent. The supposed instances in Gathās 113, 153 and 256 are in reality first members of a compound (navara should be kept apart from such words as jaha and taha, which belong to a different category. Moreover, in jaha and taha the alternation is not between -a and -am as in navara(m), but between -a and -a).
The evidence given above points to the conclusion that originally there existed only one word, navari or navarin, meaning both 'immediately after' and 'no other, nothing else'. The variants pavara and navaram may have been introduced on the basis of etymological speculation. An 'etymology behind this may have been the Skt phrase na param, which was in fact suggested by P. Goldschmidt (as quoted by Pische. S 189) as the possible source for navaram). Goldschmidt derived navari (m) from another such combination, namely na pare. I should like to suggest that navari(in) derives from na and vari(i), the latter an apocopated form of avari which goes back to Skt upari (see Pischel § 123; for vari, see the word varilla 'overcoat' quoted there). This navarim) may be assumed to have the following meanings: 'no higher, no further, not afterwards' out of which the meanings 'no other, nothing else' as well : as 'immediately after' (from 'and ((c)a) not (long) afterwards') could have developed.
The word navara(m) thus formed may, partly or wholly, have replaced navari(m) in the MSS of Pkt texts. On the basis of its occurrence in these texts it was recognized as an independent word by the grammarians. This, in turn, may have had a further 'normalizing' effect on the texts. Thus, we see a highly regular situation in the MSS of the Setubandha. In the MSS of the Sattasaf navari(m) occasionally appears to have escaped normalization. These instances have in the present edition been treated as retentions.
padhamullaa 'very first' consists of pachama and -ullaa. The suffix denotes not only possession but also intensity, as here (see Pischel $ 595 and Schwarzschild, 1962: 518). For padhamullaa, see also 187 and 218. For padhamuggaa (prathamodgata) of B, Y and S, see ibidem, MSSV , Y, P and S.

Page #226

213
Y, 16 G, 5; 261 T; 356 S)
16( 18 Ma, 19 Tp; 14 Bh, 16 R; K, W, B, P, 15 Missing in W.
amaamaa gaanasehara chikko jehi piaamo
raaņTmuhatilaa cada de chivasu mamaṁ pi teh is cia karehi.
amia K, B, P; T -- gagana K, P -- seara TD -- cchanda Ma -- he TP; B; T -- chivasu Ma; civasu Ta -- chitto Tp; Bh, R; K (chinno), , B, Y, P; S; cikko T -- jehi Bh; K -- pio Bho, pi pio Bho -- manh Ma -- tehi ccia K, teh in vvi P; tehi via S -- karehim K, B, P; T; S; (BH, R, W, Y).
0, moon, consisting of nectar, crest-jewel of the sky, tilaka on the face of the night, will you please touch me with those same beams with which you touched my lover!
According to Hem. II 196 the particle de is used in addressing a person and in a friendly request, which would correspond to its occurrence before vocatives (see, for instance, *20) and before imperatives (as here.) respectively (see also Dundas, 1981: 205).
It is tempting to connect this de with the enclitic form de of the personal pronoun of the second person, discussed in *15. The latter is used in this text specifically for the genitive-dative. It should be noted, though, that the enclitics te and me (Skt) are basically kinds of stem-forms unmarked for case. This explains their use, occasionally, for other cases than the genitive-dative as well. For the use of these enclitics, see AIG III S 235. .:. For the verb chiva- 'to touch' and the past participle chikka, see*14.
, 17 G, E; 263 T; 367 S)
17( 20 Ma, 21 Tp; 15 Bh, 17 R; K, B, Y, P, 18 Missing in W. .
ehejja so pauttho ii kassa i phalai mano-
ahaan kuppejja so vi aņunejja rahāna māla piaamammi.
ehijja so pauttho Ma, Tp; R, ehai so pauttho Bh; ehijja so pauttho K, P,

Page #227

Saham pi Tp
ehii so vi pauttho, B, ehai so upaduttho y; ehijja so pauttho T; kuppajja Ma, Tp; kupijjha Bha; kupejja K anupajja Tp; anigijjha Bha
so pi K
ia Ma, Tp; Bh, R; K, , B,
Y, P; T; S halai Ma
kie vi Ma, Tp; kassa vi Bh, K, V, B, Y, P; T; S manomuhaŋa Ma
piaamanmT Ma.
--
--
--
214
--
May he come from abroad, may I be angry at him, and may he try to please me. (Perhaps) for some (lucky) person this string of wishes with regard to the dearest one is fulfilled.
--
--
The optative ehejja, of which no other instances are known, is peculiar in that it is formed on the future stem ehi- instead of, as is usual, on the present stem e- 'to come' (see Pischel § 529). It should be noted, though, that if the optative had been formed on the stem e-, the verb itself would have become virtually unrecognizable, completely disappearing in the suffix. The context definitely requires an optative. Therefore the future ehii in ehii so vi pauttho, found in Bh,, B and Y (for ehai in Bh and Y, see Berger, 1955: 79, note 6), for the unusual optative ehejja, is almost certainly secondary.
A different problem concerns the spelling ehijja (for ehejja; cf. kuppejja and anunejja). This spelling most likely goes back to the MS of the archetype (see above, p. 172). However, it is not clear why in, for instance, Ma and Tp the -e- in this word was not restored for the i-, as it supposedly was in the other optatives in this Gatha.
For kuppajja in Ma and Tp (for kuppejja) and anumajja in Tp (for anunejja), see 1.
For ii 'thus', see *15.
kassa i is based on R. For the principle underlying the decision to adopt this reading found in R, instead of kassa vi found in the other MSS, see above, p. 163. For other instances of i for vi, see 24 do i (T), 149 jTviasesă i, 255 do i (Bh, R), 304 vamke i (R), 534 mudhahiai (Ma) and 904 do i (T). Te vi in Ma and Tp (Madhavayajvamiśra: kasyaścid dhanyayas; the Chaya in Ma has kasyapi) seems the innovation, The possibility of a reverse change of kTe into kassa is less likely in the present context.

Page #228

215
The variant halai in Ma for phalai is most probably due to confusion of the Malayālam characters for ph and h which are very similar.
18(16 Bh, 18 R; K, B, Y, P, 19 y, 18 G, 5; 81 s) Missing in Ma, Ti, Tp, T and W.
duggaakudumvaal/itthi das iosaraṁtasalile-
kaha nu mae dhoieņa sodhavvā na uaha runnaṁ va padaena.
kudarba R; kudanva v , kudumba B, P, kudduña y -- aftht Bh, R; iththt K, Y, addht , aththt B, ittht P; aththt S -- dhovieņa Bh; toiena P -- iti sothaccā K -- ořs ar ata k -- runa K, B, tuidam P; ruņa S -- cca K.
"How can I, who have been washed, bear the al/i)tthi of a poor family.' Look, it seems as if the cloth cries, water dripping from the seam.
The meaning of kudumvaaţght or kudumvaittht is unclear to me. The variants seem related through the spelling vayaţth. The commentaries gloss aftht or ittht with krsti, which Weber translated with 'pulling'. It is not clear to me, though, what might be meant by this 'pulling of (or by) a poor family'. It should be noted that the commentaries assume that the initial plosive of krsti as the second member of a compound is lost. krsţi is then not the only possibility. itth1 may stand for, for instance, drsti 'glance' or vişti '(forced) service, drudgery', and at tht for, for instance, patthí 'back'. . The occurrence of the ka-suffix twice in the second line alone, namely in das jā (Skt daśikā, from daśā) and in paqaa (Skt pataka (L), from pata) is significant. It is also possible to recognize the suffix in kudumvaal/ittht, which may be divided into kuțumbaka and tth1 or al/i)ţtht. The meaning remains, however, just as unclear as with the possibilities mentioned above. Besides, it raises another problem, namely what implications, if any, the occurrence three times in one Gātha of the ka-suffix could have here.
For dho(v)a- 'to wash' two explanations are available. Pischel $ 482

Page #229

216
suggests it is from dhu(v)a- 'to shake, to wash', which became secondarily inflected according to the I present class. Turner, CDIAL 6886, suggests, instead, that it goes back to *dhauvati, replacing dhāvati after dhauta.
For uaha 'look', see *4.
19( 21 Ma, 22 Tp; 17 Bh, 19 R; K, U, B, Y, P, G, 5; 498 S) Missing in T and W.
kos aṁvakisalavannaa hiaecchia gharam vac-
taŋŋaa uņņāmiehi kaŋgehi camāņa dhavalattana pāva.
kos umbha Ma, Tp; kosambi B, kos arħbo P (BH, R, V, Y) -- kisalaa 'ų , B, r (°laya), P; S -- Carinaya Bha (varnaya Bh); vanga y -- tanaya Bha (tarinaya Bh); omitted in Ma -- ukvaņiehih R -- kimhnehi Bh; kannehin K, B, P; S (R, W, Y) -- ehiaaccia Ma, hiaacchia Tp; hiyaichiya Bh, R; hiaatthia K (oththo), B, Y (oththo), P, hiaichian ; hiaatthian S -- vacchamāŋa Tp.
Calf, coloured like the leaf-buds of the kośāmra-tree! (Before) you go, with your ears pricked up, to the house of one's heart's desire, become a (white) bull!
A man is called a 'young calf' when on his way to his mistress's house he is unable to hide his anticipatory excitement. The phrase dhavalattanań pāva contains a pun in this connection. It means 'become a bull (dhavala)' and 'become white (dhavala)', i.e. 'try not to show your agitation'.
kos ava would correspond to Skt kośāmra, a kind of mango-tree. For the red colour of the sprouts (navapallavam) of the mango-tree, see 586. kosumbha of Ma and Tp corresponds to Skt kausumbha, an adjective of kusumbha 'safflower'. The Chāyā in Ma has in fact kusumbha! A substantive kausumbha is mentioned only in the lexicons, meaning 'wild safflower' (see PW, s.v.). Then a problem arises in connection with kisala 'sprout, leaf-bud', for it is not the safflower's sprouts or leaf-buds which are red, but its flowers. The division of the .com

Page #230

217
pound into kosumbha 'safflower-coloured' and kisalavannaa 'coloured like a kisala' does not make sense either. Probably the word kosurħbha is the result of contamination of kos añva and kusumbha, which latter plant may have been more familiar from its other instances in the text; see 145, 392 and 546.
kisala 'leaf-bud' is rare. The usual form is kisalaa. Skt kisala is found only in the lexicons. The only other instance in Pkt is, according to PSM, found in the Kumār apālacaritra (no place), where it is evidently meant by its author Hemacandra as an exemplification of Sutra I 269 of his own grammar, in which kisala is explained from kisalaa.
tannaa 'calf' corresponds to Skt tarnaka, which contains the word taruna (see Mayrhofer, EWA I, p. 485 and Weber, Abh., p. 79, note 2). See Pāiyal. 235: tannao vaccho.
hiaacchiaṁ in Ma and Tp probably stands for hiaecchian, showing confusion of -e- and -a- before the palatal consonant -ch- (for this phenomenon, see the note to Gatha *1).
Hem. IV 225 mentions vacca- as an Adeśa for vraj- 'to go'. Its derivation will be discussed elsewhere. Hemacandra mentions vacca- a second time, in IV 192, among the Ādeśas for kārks- 'to desire'. It is not unlikely that this interpretation of vacca- was based on the occurrence of the word in the Gathā under discussion: 'longing (vaccamāņa) for the house desired in your heart'.
201 53 Ma, 55. Tp; 18 Bh, 20 R; K, B, Y, P, 21 y, 20 G, 5; 140 T; 174 S) Missing in w.
aliapasuttaa viņimi- gaind apar i urvanā pula
liaccha de suhaa majjha oāsam iasga na uņo cirāissa.
Yom
alie Ma -- pasutta Ma, Tp; P; T -- 'liacha R; K, olidaccha B -- desu suhaa majjha Tp; de suhaya ahma Bh, de suhaa amha R; desu haamajhū K, de suhaa majhjha B, dehi me suhaa majhjha y, desu haamajha P; suhaa maha desu T; de suhaa majja s -- uvās a Bh; osās am B, oăsa P -- pariurbana Ma, pariurbană Tp; pariunvană Y; (BH) -- angari Ma -- po puno , B, na puno Y, na vuno P -- cirāyisan P.

Page #231

218
Sāhityadarpaņa, p. 90.
(I see,) you pretend to be asleep, keeping your eyes closed. You, handsome man! I promise you -- look! your hair. stands on end as I kiss you on the cheek! -- at a (next) occasion I will not be late again.
The woman speaking is probably an abhisārikā meeting her lover. He feigns to be asleep in order to make clear to her that she has kept him waiting much too long.
The commentaries, and following them Weber, take de as the imperative of the verb de- 'to give' (Skt da-) and link it to oāsa, viz. de... :: oās añ. de, however, cannot possibly be the imperative of de-, which is dehi or desu (actually found in Tp, K, P and T and in y respectively). Most likely we have to do here with the particle de (de 'suhaa) discussed in *16. The problem then is how to fit in oasa (Gangadhara suggests to supply dehi: he subhaga mamāvakāśam dehiti śeşah; see Ed., p. 10). I should like to suggest to take it, despite the distance (see below), as an accusative of time to na uno cirāissań: 'during (or better: at) a (next) occasion I will not be late again'. For the accusative, which usually denotes the time during which, denoting the time at which as well, see Speyer $ 54. Admittedly a locative would be more regular. It is in fact tempting to suggest that, oasa is an old 'innovation', going back to the MS of the archetype, for oase, for which, among others, the superficial resemblance of de...oās ar to the idiom avakaś aṁ dā- may be held responsible.
Behind all this is the failure to recognize the intentionally interrupted sentences in the Gathā and to appreciate its implications. The sentences are interrupted because the woman is speaking and kissing at the same time. Another clear example where the particular circumstances under which the words are spoken have left their traces on the text itself is found in Gathā 160, which should read:
acchodiavatthaddhan- cintesi thaŋabharāā-
tapatthie martharaṁ pie vacca siassa majjhassa vi na bhagań.

Page #232

219
The woman addressed has in vain tried to prevent her lover from leaving her. She is now running after him, being in turn followed by a friend who tries to stop her. This friend addresses her, apparently panting and out of breath. This seems to be the most plausible explanation for the chaotic word-order: acchodiavatthaddhant apatthie...pie is a locative absolute construction, interrupted by mamthar an, which is to be taken with vacca. In the second line the finite verb is standing at the beginning of the sentence, while the interrogative particle vi (Skt api) occurs practically at the end! The Gāthā should be translated as follows:
Now that your lover (pie), held back by you (acchodia), has been able to draw himself free (vyasta) and finally (adhvānte) has left (prasthita), go slowly! Are you then not afraid (cintesi...vi na) that your waist, from the burden of your heavy breasts, might break?
As in the case of the Gathū under discussion the state of the sentence was not recognized. The text was interpreted quite differently and in the course of the transmission of the text some changes were brought about in it to suit this interpretation. The result of these changes can be seen in Weber's Ed.
The commentaries link the word majjha with immediately following bāsam. It is doubtful if this still makes sense. I have, alternatively, taken it with. gad apar iurvaņā in the second line.
Skt punar, meaning 'on the other hand', becomes uņa (occasionally uņo) losing its initial consonant (for an explanation of this loss, see *15). When meaning 'again' the initial p- is usually retained (see puņo in 375), except in the phrase na uro (here and in 941; na una in 818), which apparently was looked upon as one entity, in which case the rules of internal sandhi came to be applied (compare na āņa-, for ņa jāņa-, discussed in *3).
The Gathā is found inscribed on the pedestal of a basrelief, probably from the eleventh century, found near Jabalpur, Madhya Pradesh (see Katare, 1952). The relief itself depicts a man lying on a couch in a kind of bower with a woman kissing his cheek. The inscription has been deciphered -- rather imperfectly as the accompanying ink-impression

Page #233

220
shows -- by Katare as follows:
aliapasuttaa viņimtlimai de mūha(a)mha uvās an gançapariumba(na) pulai anga na une cirāismam.
, 21 G, 5, T; 488 T;
.
21( 56 Ma, 57 Tp; 19 Bh, 21 R; K, B, Y, P, 22 175 S; 23 W) # begins here.
asamattamaño ana ccia volaiahalahalaas
vacca gharaṁ se sakouhallassa sa putti citte na laggihisi.
ciya Bh; via y; via S -- vaccha T; vauva S; gacha W -- harah S; cari W -- (lacuna se sa)kāuhallassa Ma -- bolia Tp; volavity)a Bh, R; K, M, P, bolăvia Y, B; T; volāvia S; W -- halah aliassa T -- laggihist Ma; ggilarihisi Bha, ligjihisi Bho, laggihasi R; W.
Who cares if you haven't finished applying your make-up. Go to his house while he is still eager! Daughter, after his passion is spent you won't stay in his mind anyhow!
For vacca-, 'to go', see *19.
For kouhalla, 'eagerness', see Pāiyal. 156 and Hem. I 117.
Hem. IV 162 mentions vola- among the Adeśas for gam- 'to go'. The instances in the SattasaT show that it means in particular 'to walk past, to disappear'. Turner, CDIAL 12167, suggests that vola- goes back to vyapacal-. Earlier Weber had suggested a derivation from vyapali- (Abh., p. 32, and, again, in Ed., p. 10). Indeed vola- has much in common with the verb 11-. Beside alliai (898, from a-17-) and samalliai (532) is found ahileñti (266, from abhi-11-), with which may be compared volei (711, 854). Beside volia (232; for vilia, from vi-11-, see 53) is found volīņa (57, etc.), which corresponds to Skt 1Tna. For the prefix vo-, from Skt vyapa- or vyava-, see Schwarzschild (1965: 350-4).
For the variant bola- found in the South Indian MSS for vola-, see above, p. 170 (volaia in Ma is an exception in this MS).
For the interpretation of vola(v)ia there seem to be two possibili

Page #234

221
ties. volāvia may be the past participle of the causative *vyapalāpayati (cf. vilāpayati mentioned in CDIAL 11906). This explanation of volāvia as a causative must have been current as is shown by the variant bolávia. It is not clear, though, why the -v-, which is otherwise always maintained in such causatives, was deleted in volāia. For that reason I have adopted, volāia of Ma and Tp, which would represent * (vo)lāyita (compare *apalāyayati beside apaltyate, quoted by Turner, CDIAL 447).
For halahalaa, 'agitation' or 'passion', see Pāiyal. 246 and Desin. VIII 74; see also Gathā. 780.
For the possible derivation of Pkt lagga- 'to cling to', see *10.
, 22 G, E, T; 649 T;
22( 61 Ma, 62 Tp; 20 Bh, 22 R; K, B, Y, P, 23 619 S) Missing in W.
Tarapanāviotthan vannaggh aauppamuhi-
aghadianās aṁ as agaanidālam ette pariumvi am bharimo.
panāmioţtham Bh (uţthan), R; K (ooththan), , B, Y, P; T; S -- as aṁhaa K, M, B, Y, as aghaqida P; as andata S -- vadālaṁ Bho, nalādam Bha, niladań R; nidāham W; lalādam S -- vannaghaa Ma, Tp; R; K (ghia), y (ghia), vannaraghaa , vannacchia P, vannagdhaa B; T; S -- tuppa Ma, litta (sec. m. tuppa) Tp; tuppa R; litta K (lipta), Y, P, luppa V, B; tuppa T; S -- muhti Bh; munie K, B, Y, P; T; S -- tii Bh; tta S. -- pariumban a Ma, Tp; pariumvium Bh, pariumbiya R; pariumv anam , ombana B, ombh anam Y; ombanan T.
I still think of how I kissed her while her face was covered with coloured tuppa or ghee: our lips carefully pursed to keep our noses from touching and our foreheads from meeting.
During their monthly period women smeared their faces with a kind of coloured ghee (Gangadhara: haridrāvarmapradhān an ghrtan varnaghstan; & and T: varnapradhānaṁ ghrtam kumkumaharidrāsādhitam; Ed. p. 10). probably to mark themselves as 'untouchables'. Contact with them during this time was, as it still is today, prohibited, a taboo which merely

Page #235

222
worked for men as an excitant. When nevertheless engaging in sexual activity, the men had to be careful not to get stained by the coloured ghee, which would betray their transgression (see also 529). This explains the complicated way of kissing described here. For other references to women in their period, see 191, 289, 480, 520, 529 and 530.
For the change of -m- into -v- as seen in paņāvia (Skt pranāmita), see Brough (1962: 88-9), who has clearly shown that it is conditioned. by the presence of another nasal in the word. Other instances in the Sattas af are *29 nisāvio Ma (nisāmio), 83 vane Ma (maņe) (vase Tp), 111 vane Ma, Bh (id.) (vase Tp), 154 vane Bh (id.), 332 panāvesi Ma, Ti, Tp, T, W (paņāmesi). See also vammaha (mammaha) in 327 (R) etc.
Through its occurrence in Ma and Tp, on the one hand, and in Bh, on the other, the variant vannaghaa(t)uppamuh Te can be traced back to the MS of the archetype. It is, however, to be abandoned in favour of vannagghaa(t)uppamuh Te found in R, V and P, as it gives an unmetrical text with u-u in the third Gaņa.
For the reasons to adopt vanna(9) ghaauppa found in Bh, rather than tuppa found in the other MSS, see above, p. 164. The compound, which is also found in 520, contains two synonyms for 'ghee'. The first is a commonly known Skt word, ghrta, which is also widely distributed in NIA (CDIAL 4501). This is followed by a regional word for 'ghee', tuppa (Pkt, Mar. and Guj.; see CDIAL 5864), which is not known from Skt. It is probably a loanword from Dravidian (see DED 2685; but Turner, loc. cit. suggests, instead, that Kannada tuppa is a loanword from Mar.). The compound occurs itself again in a compound (see also the instance in 520). Theoretically it is therefore possible to translate it as a Dvandva, 'coloured ghee and tuppa(-ghee)'. However, the resemblance of the part ghaa(t)uppa to sakhasippt (discussed in* 4) is too obvious to be ignored. This suggests for vannagghaa(t)uppa the translation 'coloured tuppa, i.e. ghee'.
tuppa as occurring in vann agghaa(t)uppamuh Te was traditionally misunderstood as meaning 'smeared with' (Bhuvanapala: upp anh snigdha (ISt., p. 36), Pāiyal. 233: makkhiyam tuppan, and Deśfn. V 22: siņiddha... tuppo); from there litta (Skt lipta) in K, Y and P and luppa (+ tuppa x litta) in v and B. The meaning 'smeared with', however, is, among others, conveyed by tuppalia, tuppavia (529) or tuppia (Vivāgasūyam

Page #236

1,2,14; 1,3,13; 1,9,6: nehatuppiyagatta). For tuppalaa (tuppa plus laa; for the suffix, see Pischel $ 595), see 289.
For pariumviah Ma, Tp. v. B, y and T have pariunv(/b) anamh. The latter would seem to be the innovation. For the use of the past participle as an action noun, see Renou § 153 c.
Hem. IV 74 includes bhara- among the Adeśas for smr- 'to remember'. Pischel 267 suggests that bhara- goes back to this smr-, in the following way: smarati mharaimbharai bharai. See also Mayrhofer, EWA II, p. 482.
23 655 T1, Tp; 21 Bh, 23 R; K, B, Y, P, 24, 23 G, E, π; 163 T; 200 S; 10 V)
anasali deftT gose vi on aamuhT
223
taha surae haris aviasiakavola aha se tti piarh saddahimo.
F
ana Ti; pasaydim R; aşasaahim K daft Ti, Tp; dett K, detti P; danti T; S ha (taha) Ti hasiaviasia Y gose aya Bha b gose ya Bh; gos ammi P unaya Bh; onaha Y, ona P -- muhim Ti aha sa cci (breaks off) Ti; aha sa tti Bh, aha si tti R; aha so tti K, aha sa tti, B; sahasatti S pi(y) 8h, R; Y na visasimo Tp, (missing in Ti); saihihmo Y, sahippo P.
--
--
--
--
--
When that woman even in the morning still hangs over him (on aamuhT) giving hundreds of instructions and (taha) having cheeks burning with excitement during the love-play, I do not think she is dear (to him).
--
Weber seems to have been quite at a loss what to do with the compound on aamuhT. In Abh., p. 81, he changed it, against the metre, into anonaamuhT, 'die...nicht das Antlitz neigt' (from shame). See also Retr., p. 355 and Ed., p. 11. I think, though, that on aamuhT should not be taken as a sign of shame here. It seems, instead, to indicate the position of the woman relative to the man: she is hanging over him or even sits on him (viparttasurata) all the while giving instructions how she wants. to have it. The same situation is understood in Gatha 299.

Page #237

224
aha se tti should be divided into aha sa and tti. Var. VI 24 and Hem. III 87 take aha as a substitute for asau. Like the latter it is used with reference to the masculine: *32 aha sambhāviamaggo, 300 aha...cando, and 908. aha icchai, as well as to the feminine: 613 aha...paai, and 717 aha suai. For the phrase aha sā, see also 57 and 318. In 418 aha so occurs. aha is Apa. in appearance; compare Apa. ehu for eşaḥ and eha for eşā.
For the occurrence of setti (TP, R, Y, P, T) beside sa tti, see also 123 mame tti (Ma, TP, R, T), 360 mae tti (T), 498 mahile tti (Ma, Tp) and 855 na ve tti (Ti, Tp). These instances may be explained by assuming the interference of the corresponding Skt phrases; e.g. seti, mameti, etc. More likely, however, the e for a is due to the palatalizing effect of tt on a preceding a; for another instance of this phenomenon, see Pkt metta beside Skt mātra. On Pkt metta, see Mayrhofer (1955: 224) who refers to an article by Schneider (1954: 576-583) in which the latter shows the existence in the Pāli canon of a form khettiya beside Skt ksatriya.
As such these instances may be compared to 88 se ccia (y) beside sa ccia or to 69 pāniggahana ccia (K, S; for gahaņe ccia), 129 ekkaggāma ccia (Ma, Tp, K, M, Y, P, T), 130 sāloa ccia (Ti, Tp (pr. m.), W, Y, P, T, S), 290 hiaa ccea (Ma, Ti, Tp, K, , Y, S), 383 hattha ccia . (Ma, Ti, TP, T, S), 612 hiaa ccia (Ma,Ti,Tp,Bh,x,y) and 654 uara ccia (Ma, Ti, Tp). They may also be compared to those instances in which o before the enclitic va is replaced by a; e.g. *29 padaha va (Ma, Tp), 74 mutthia va (y, P, T) and 253 murava vva (Tp). In all these cases it concerns a short vowel of which the particular colouring may already have been less distinct than that of the corresponding long one. This may also explain the replacement of o by the more neutral vowel a in 489 ettia ccea (Ma, Ti, Tp, v), 619 magga ccia (Ma, Ti, Tp) and 767 sea ccia (Ma, TP, R).
A problem arises in connection with those instances in which a before vva is replaced by e; e.g.* 32 devae vva (Tp, v), 75 bhatthe wa (Ma), 77 sāņusaene va (v), 91 dāse vva (Ti, Tp), 120 pade vva (Ti, Tp, T), 183 sake vva (Ma, Tp, T), 217 pańcālie vva (T), 417 jāe vva (Ti, Tp), 485 cittālihie wa (Ma, Ti, TP, Y, T), 496 male wa (Ma, Ti Tp), 575 mehale wa (Ti) and 654 paohare wa (Ma, Ti, Tp). In the

Page #238

225
following cases the e replaces an o: 253 murave va (Ma) and 714 asarane vva (Ma; asarone vva Ti). Evidently we have to assume intermediate* murava va and *asar ana wa here, of which the former is actually found in Tp (see above).
The status of these instances in the MSS is uncertain. The occurrence in 217 of pascālie vva in T (beside vāullaaṁ va in the other MSS; the compound in which the word occurs is a neutre Dvandva, for which see *37) should make us aware of the possibility that the instances of -e vva for -a va, which are mainly restricted to the South-Indian MSS, are the result of an independent development in these MSS. This would explain murave vva in 253 in Ma beside murava va in Tp; the latter can hardly be considered 'an emendation of the former, the most likely emendation being muravo va. On the other hand, instances are also found in y, which is not known to have any South-Indian source. In fact, the occurrence of sāņusaene vva in u in 77, a Gathā which was most probably added later to the text in the North-Indian branch, suggests that the development could also have taken place in the North-Indian MSS independently.
A difficulty arises when one tries to identify the cause or causes underlying the replacement of a before va by e. This replacement cannot be explained with reference to the effect caused by vva, which would have a labializing effect. It is possible, though, that the e reflects a kind of hypercorrect or exaggeratedly fronted pronunciation of the a in order to prevent any labializing effect caused by following wa. In any case it cannot be explained with reference to the pronunciation found in South India of a after certain consonants as e (see *1). This would not account for the instances in v or for e in, for instance, jāe va, in which it occurs after a vowel.
A second possibility is, as in the case of setti discussed above, to assume the interference of the corresponding Skt phrases; e.g. māleva, sānus ayeneva, etc. (compare s odaradt via (avatarattva) and rodidi via (roditfva) occurring in Bhāsa's Cārudatta in 51, 2 and 66, 6 respectively; quoted by Printz, 1921: 9). This, however, does not explain dāse va and asar ane wa which correspond to Skt dāsā iva and ašarana iva respectively. It probably goes too far to assume Skt dāseva and aśaraneva, representing cases of 'double' sandhi (for this

Page #239

226
phenomenon, see Renou $ 41 b). If indeed in the origin of male wa Skt is involved it would not do either to explain dāse va (beside dāsa vya) as formed analogically after māle vva (beside māla vva) since this analogy does not exist in Skt: dāsā iva and māleva.
In the edition such instances as se ccia, pāņiggahaņa ccia, padaha vva and ettia ccea are considered as retentions. The reason for doing so is that, whether one starts from Pkt or Skt, the cause for emending them into sa ccia, pāniggahane ccia, padaho va and ettio ccea respectively presents itself easily. The instance uara ccia in 654 is uncertain as it is restricted to the MSS Ma, Ti and Tp which also have pacchai (for pecchai; see *1) and hiaacchian (for hiaecchian; see *19).
The case of such instances as se tti is slightly different, because of Skt seti. In this connection it is to be noted that most of the instances concerned are restricted to the MSS of the South-Indian recensions. An exception is setti which is found in the MSS of the NorthIndian branch as well. At the same time this case differs from the others in that se could somehow be fitted in as the pronoun se 'to him'; e.g. 'we do not believe that she (aha) who... is dear to him (se)'. Admittedly, this involves ignoring tti.This is actually done in Bh, R and y (the latter two having se!) which for pian read pi(y)a. As such se tti could have escaped emendation more easily than the other instances given above. I feel therefore justified in taking these instances as retentions as well.
The case of māle va etc. constitutes a problem. In the first place it is almost certain that not all instances are original or go back to the MS of the archetype. In the second place the origin of the e beside a remains unclear. My decision to adopt these instances in the text is more or less arbitrary. It is done in order to emphasize the uncertainties concerning the forms, rather than because I believe they are original. An exception is made in the case of pacālie va in 217 (for obvious reasons) and for murave vva in 253. In the latter case murava vva of Tp will be adopted, thus making allowances for the possibility that murave va in Ma is the result of some independent development.
For datt in Ti, Tp and T for dett, see *1. The a in dantt in s should probably be explained from the influence of Skt dadatt. Note that MS S makes a strict distinction between a and e-stems (see*5).

Page #240

227
24 58 Ma, 59 Tp; 22 Bh, 24 R; K, B, Y, P, 25, 24 G, E, π; 632 T; 530 S) Missing in W.
piaviraho appiadamJTe tuman karijjasi tie namo ahiāte.
sapa ca garusi do i dukkhai
a K, B, Y, P; T; S
piya dannah (dasanah) Ma gurusi Tp; Bh; K (im), B, P do vi Ma, Tp; Bh, R; K, W, B, Y, P; S -- duḥkhāi Tp; dukhkhaim K, dukavāim B, duḥkhaim P; dukkhaim T; dukhkhaim S; (Bh, R, V, Y) jTa Ma; jTi Bh; ja w, jTa B; S kārijjast Ma, kariisšihi Tp; kǎrijjai y; kārijacasi S tia Ma; V, Y, tTa B; S (ahi missing) aTe Ma, ahijaтe Tp; K, B, Y, P, ǎhija e ; ahijāte T; S.
--
Being separated from what is dear and seeing what isn't are two causes of great grief. Honour to that noble birth which makes you bring about both!
--
--
Weber (Retr., p. 355) suggests that the man much against his liking. left his dear mistress (piaviraho) to show himself in front of his wife whom he does not love (appiadams an am); see also Bhuvanapala as quoted in ISt., p. 36. However, the causative karijjasi shows that the man brings grief not on himself but on someone else, in case his wife. She scolds him for his good upbringing which brought him to her, his heart, though, still being with his mistress. Thus her separation continues and above that he offers to her an unpleasant sight.
For the reasons to adopt do i found in T instead of do vi found in the other MSS, see above, p. 163. For other instances of i for vi, see *17. For do (Skt dvau) used with reference to the neutre, see Pischel § 436; see also Gathā *27.
For Pkt Shit (see also Pali abhijati in chalabhijati), beside Skt abhijāti, see Var. I 2 and Hem. I 44. Pischel § 77 and Geiger § 24 wish to derive ThiaT and abhijati directly from Skt abhijati. Geiger explains the lengthening of the initial syllable from the position of the accent. For the same accent shortening a preceding or following syllable, see also Pischel § 79 ff. However, for most of the other

Page #241

228
instances of either lengthening or shortening of vowels explained by Pischel and Geiger from the position of the accent, another and better explanation is available. I need not go into details here as several instances will be discussed in the course of this edition (e.g. alia in *27). The lengthening of the initial syllable in āhiat may have its origin in the use of the word in an exaggerated sense, abhijāti meaning - "high birth' and āhiat 'very high birth. Compare the -i- in English big which in exaggerations becomes lengthened. A similar explanation may be considered for Pkt sāmiddhi beside Skt samrddhi 'wealth' and for Pāli! ānubhava beside Skt anubhava 'power'. For other words of which the forms seem to have been influenced by emotional factors, see diara and viaņā discussed in *38.
In the text āhjāte of Bh and R (and Ma?) is preferred above āhijate of the other MSS as it is the form deviating most from Skt. Compare 265 with āhiaiṁ in R, K and and āhijāin in Ma, Ti, Bh, Y, T and S. Gathā 905 is found only in Ti, Tp and T which all have āhijāt. The case of the past participle ahiāa is strikingly different in that for this word no variant ahijāa is available (see *41 and 266)
; 205 T;
25( 59 Ma, 60 TP; 23 Bh, 25 R; K, B, Y, P, 264, 25 G, 5, 176 S) Missing in W.
ekko vi kālasāro kin uņa vāhāulian
na dei gantur paāhiņavalato loanajualan piaamāe.
kannasāro ta, kahņasāro Tb -- vadaṁto R; calato , B; valanto P; T; S -- puņa Bh -- bāhāulian Tp; vāhāulia K (pr. m.), oulia K (sec. m.) -- jaalan Ma.
For a possible translation, see below.
Weber (Abh., pp. 81-2) translates the Gathā as follows:
Ein Reh schon, das rechtshin flieht, lässt (den Wanderer) nicht weiter ziehen. Wie viel mehr dies von Thränen umf lorte (Ren-Augenpaar der Liebsten!

Page #242

229
The first line would refer to a specific bad omen, which, however, does not seem to be otherwise known. Admittedly, I have not been able to check śrngāratilaka 4, 5 (Gieldemeister), which according to Weber would contain 'eine gewisse Analogie zum Inhalt des Verses' (Retr., p. 356). In the second line a pun would be made on the compound kālasāra, which beside 'antilope' also means 'eye' ('having a black -- kāla -- pupil -- sāra'); and vāhāulian meaning 'troubled by a hunter' (vyādha), on the one hand, and 'troubled by tears' (bāşpa), on the other. For this interpretation Weber relied in the first place on Kulanātha's commentary: yātrāsamaye eko'pi krsnamrgaḥ pradakşiņas calan vyādhenānudruto 'pyaśāk un atvena gamanavighnań karoti; vāşpākulitań, pakşe vyādhākulitań (Abh., p. 82). Apparently the other commentaries concur. The same interpretation is found in Madhavayajvamiśra's commentary found in MS Ma.
However, a closer look at the vocabulary of especially the first line might suggest quite a different idea. At least two words can be positively identified as technical terms relating to boardgames, such as Indian chess, backgammon or Caturanga. However, in the end there remains for me a problem to combine these terms meaningfully in order to make identification possible.
Thus, kālasāra (Skt kālas āra) is the common term for a black (kala) piece (śāra) in the Indian boardgames. The other term is paāhiņa (pradakşiņa) in paāhiņavalaṁto; it refers to the 'clockwise' movement of certain pieces (Thieme, 1962: 205).
In this connection also gam- (in gaṁtun) and val- (in valaito) are interesting. It is possible that the pair refers to particular but opposite movements. Compare Thieme's (212-3) analysis of the terms gama and cara, occurring in Balabhārata II 5, 14: 'moving clockwise' and 'moving counterclockwise', or vice versa. If val- (or perhaps we should read cal- as found in U and B, as paleographically the distinction is almost insignificant) is indeed a technical term for a movement in a specific direction, then the addition of paāniņa 'clockwise' would be redundant. Alternatively, it is possible to take gam- in the meaning 'to escape', which would in any case much better suit the content of the present Gatha:
If only one black piece, (by) making a clockwise movement,

Page #243

230
can prevent (a piece of the other colour) to escape (from the place it occupies), how much more (can prevent me to leave home) the two tear-filled eyes of my beloved.
The second line is a pun on kālasāra, which stands for Skt kālaśāra 'black piece in a boardgame' as well as for Skt kālas āra 'eye'.
The identification of the game remains uncertain. Backgammon can be ruled out positively. In this game the clockwise movement belongs to the white pieces. Furthermore it needs two pieces to obstruct the movement of another piece. This seems to limit the possibilities to either Indian chess or Catur anga. I would like to leave it to a specialist to decide which of the two may have been intended.
For the occurrence of vāha in Ma, which for Skt bāşpa usually has bāna, see above, p. 170.
26( 62 Ma, 63 Tp; 24 Bh, 26 R; K, B, Y, P, 27 W, 26 G, 5, 1; 324 s) Missing in T and W.
na kunata ccia maŋŋun sungaiavās apar imusa-
ņisāsu sahasuttadaraviuddhānań ņaveaņa jai si jānanti.
kunanti Ma, Tp; R; Y, kunamto K, U, B, P; S -- a Ma; via ; cvia S maîtu Ma, Tp; māṇaḥ Bh, R; K, W, B, Y, P; S -- pāšutta Bh, suhasutta R; K (suhasuh asutta), B, Y, P, sahasultati suhasutta si- ora (dara) Ma -- viddhana Ma, vibuddhānań Tp; viuddhāna Bha, vibuddhāņa Bho, vivuddhāna R; K, , B, Y, P; S; (R, W, Y) -- sunnāia R; sunnayia P -- pāsa Ma, Tp; Bh, R; K, M, Y, P; S -- parimas ana Bha (musaņa Bh"), par imhasana R; par imūs ana K, o musa B, mans ana Y, mūs ana p -- jai vi Ma, Tp -- jānaṁtt Ma, jāņañtï Tp; jānantu ccia Bho jānaṁtt R; Y.
I would not be so furious if you only realized how much pain is suffered by those who after having fallen asleep together (with their husbands ) wake up a little in the middle of the night to discover that the bed beside them has been vacated.

Page #244

231
As far as the participles kuņamta (ccia) and jāņaộto are concerned the text is based on MS Bh. For my interpretation of the text, see below. For these two participles two sets of variants are available, namely kun anti (ccia) and jānaṁtt (Ma, Tp, R and y), on the one hand, and kuņaito (ccia) and jānamto, on the other. On the basis of the results of the stemma the former set would have to be adopted. Then the situation is to be understood as follows. For some unspecified reason the woman is angry at her husband. She is warned not to behave in this way or otherwise one night she will find him gone. The speaker is most probably a sympathetic friend. It is not likely that the words are spoken by the husband bimself as he will hardly be aware of, or sympathetic with, the hurt feelings of deserted women as referred to here.
Weber in his editions reads kuņaṁto (ccia) and jāņaṁto. He analyzes the situation as follows (Abh., p. 82). This time it is woman who is speaking, to her husband. The latter is apparently in the habit of leaving the bed in order to meet his mistress. During this particular night his wife woke up a little and as it were instinctively felt if he was still lying beside her. The husband took this as a sign of mistrust and reacted angrily. To this his wife replied that he would not be angry if he knew how often she had come to grief in this way.
In my opinion each set of variants represents a different answer to the problem raised by the reading kuņaṁta ccia...jāņaṁto found in Bh, which consequently has been adopted. kuņamta in kuņaṁta ccia stands for kuņastā, the nominative of the feminine. For another instance of a present participle of the feminine ending in -a, see 136, in which the MS of the archetype read homtā (Ti, Tp and Bh; later changed into homtT). The situation is to be understood as follows. A woman is speaking to her husband, who the night before sneaked out of the bed in order to meet his mistress. In the morning he returns home as if nothing has happened, showing no regret or shame, which infuriates his wife. With feigned ignorance the husband asks her the reason for her anger.
The text in Bh is unclear in two respects. In the first place kunata in kuņaṁta ccia is ambiguous as it can also be taken as a masculine, e.g. kunaṁto (for the type of 'sandhi' assumed here, see* 23) This interpretation of kunaṁta may have been a real alternative, the more so because instances of the participle of the feminine ending in

Page #245

232
- are extremely rare. In the second place the subject of the main clause ('I'), which is different from that of the conditional clause ('you'), is not specifically indicated by, for instance, a personal pronoun to distinguish it from that of the conditional clause. This does not mean that the presence of a personal pronoun in the main clause is obligatory (compare Gatha 216: ga pattiat jai...vähavind bhijjamta), but in this particular case its absence combined with the ambiguity of kunafta is awkward.
The two sets of variants would seem to have resulted from attempts to make sense again of the unclear text of Bh. The two sets have in common that the subject of the main clause is the same as that of the conditional clause. In the one case the ambiguity of kunamta (ccia) was resolved in favour of kunato, thus bringing it in agreement with janamto. In the other case the rare feminine gender marker - was replaced by -1, the more common one. Compare the fate of homtā, cited above, which was changed into homtt. In order to suit the new context. janahto was changed into janantT.
If my interpretation of the text is correct we should prefer mannu 'anger, fury' (Ma, Tp) to the variant manah 'sulking' (all the other MSS), because a sulking woman will never give reasons for her behaviour. Part of mana is precisely the refusal to talk. The spelling mahtu found in Ma and Tp, with -mt- for -nn-, a confusion which is typical of Jaina Nagart, is explained by the fact that the word goes back to the common North-Indian source of Ma and Tp, which, though not necessarily identical to the MS of the archetype, was at least very close to it (see above, pp. 40-1).
For the peculiar use of the present participles kupata and janahto found in this Gatha, see Hem. III 180.
jai vi janamt 'even if you would know' found in Ma and Tp does not make sense. The position of si, the second person singular of as- 'to be', found in Bh, R, etc. before the word to which it belongs, viz. janamto, is peculiar. It is likely that si functions as a personal pronoun here (see Alsdorf, 1935-37: 326-7). In any case in the conditional clause no finite form of the verb 'to be' seems required. Compare the following Apa. verse quoted in Hem. IV 351:

Page #246

bhalla hua ju maria bahiņi mahara kamtu lajji(/e)jjamtu vayamsiahu jai bhagga gharu etu.
233
It is good, sister, that my lover was killed. For he would embarrass me in front of my friends if he came home beaten.
For the reasons to adopt viuddhaṇa (m) found in Bh rather than vivuddhana(m) found in the other MSS, and vasa found in B rather than pāsa in the other MSS, see above, p. 164.
sahasutta is found in Ma, Tp and . The remaining MSS have suhasutta, which probably is a mere corruption: gisasu sahasutta.
For viuddhanam, with -anamh as found in Tp, see p. 169.
sungaia 'vacated' is from sunga 'empty' (Sunya). For the type of formation, compare uttigaia (from uttrna) found in 170 (read: jhamjhavauttinaiagharavivarapalottasaliladharahim as in T).
parimusana (masana) is an action noun based on parimu (/a)sa-, from Skt parimrs- 'to touch, to discover'.
27 63 Ma, 64 Tp; 25 Bh, 27 R; K, B, Y, P, 28 , 27 G, E, π; 529 T; 325 S) Missing in W.
panaakuviana doŋha vi aliapasuttana maŋaillana giccalaniruddhaŋTs&- sadingaannana ko mallo.
--
pana
Y
ti Bha (vi Bhb)
kuvidana Tp; kuviagam K, kuviona y dohna Bhb (dunha Bha); dorahah K, doraha B; doŋa T pasuttana K magaillana Ma; maŋaillagarh K, B, maŋaijjāņa P; mānaillanam T; S; (Bh, R, V, Y) -- bhinnaannana Ma; dinnakannaṇa K (nam), Y. P, disaanana B; dingakannana S.
--
--
--
donna Ma, Tp;
Dasarupa, p. 184, Sahityadarpana, p. 81.
(I wonder) who will be the strongest of these two who, angry though loving, pretend to be asleep, are full of pride and lie still, listening carefully to their

Page #247

234
suppressed breathing?
For donha in the MSS of the North-Indian recensions those of the SouthIndian recensions have, here as well as in 162 and 627, donna; see also 878, a Gāthā found only in T. According to Pischel § 436 the genitive ending -nha() is typical of M. (cf. Var. VI 59 and Hem. III 123) as well as of Amg., while - nain) would be typical of ś. and Mg. Here they are MSS variants and the question is which was actually found in the MS of the archetype.
donna in the South-Indian MSS may be explained as a clerical error for donha. Compare kanna in *25 in To beside kahna in to, or saanna in 624 in Ma. Note in this connection that in Tp hņ is invariably preceded by the circle which indicates the doubling of the following consonant. Alternatively, it could be a saurasenism. Note in this connection the occasional retention of the intervocalic -t- as -d- typical of that dialect; e.g. bhūdań in *8, kuvidāņa here and padavie in 107 (but restricted to Tp and T). Another possible Saurasenism in the South-Indian MSS is khu (see *39). On the other hand, it is equally possible that donha in the MSS of the North-Indian recensions represents an instance of the influence of AMG., of which already other possible instances have been discerned (see *1 and *2).
It is obvious that these arguments do not lead to any decision as to which form should be adopted in the text. In this connection the derivations of the endings may be discussed. Irrespective of the questior whether the one is a saurasenism and the other an Ardha-Magadhism, it would appear that the ending -nha(n) is the more difficult one and should therefore be adopted.
The starting-point of the ending -nnath) may be found in, for instance, caturnām, şannām and possibly also trīņām. The derivation of -nha() is much less obvious. Pischel (loc. cit.) suggests that -nha(m) in donha (ñ) goes back to*dvausnām, itself a contaminated form of *donaṁ and *dosam. As an alternative I should like to suggest to derive -nha() from the feminine, as in catasrnām or tisrnām. It has to be assumed then that these numerals, as a special class of words, have lost their penultimate syllable. *catasņām and "tisnām would have developed into caunha() (the -u- from cau-) and tinha(in) respectively,

Page #248

235
and from there the ending -nha(m) would have been transferred to the other numerals. Regarding the use of the feminine of the number 'two' with reference to the masculine and the neutre as well it should be noted that in Pkt with the numerals above 'one' the gender-distinction appears to have been lost; see Pischel, loc. cit. Note do i dukkhãi, with do from Skt dvau, in *24.
alia 'false' (cf. Pāli alika) corresponds to Skt alska (see Pāiyal. 53 and Hem. I 101). Pischel $ 80 and Geiger § 23 derive ali(k)a directly from Skt alTka. They explain the shortening of the -T- in alia from the position of the accent on the first syllable. Evidence for the workings of such an accent is slender (see *24). Obviously we have here to do with two different suffixes, -ika in ali(k)a and -Tka in altka.
mānailla 'sulking' consists of mānaka and -illa. -ka in mānaka may be svārthe. It is also possible that mānaka means 'sulking' by itself (AIG II 2, § 362 d), in which case -illa has merely an emphatic function (compare -ullaa in padhamullaa, discussed in *15).
For the meaning of malla, 'the strongest, the winner', compare H. an. 2, 505 (balin) and Trik. 3,3,405 and Med. 1,44 (baltyaṁs) (quoted in PW, s.v.).
, 28 G, 5, T; 173 T;
28( 64 Ma, 65 Tp; 26 Bh, 28 R; K, B, Y, P, 29 490 S) Missing in W.
.
navalaapaharań age romańcadarndarat
jahin jahin mahai dearo daun tahin tahi dtsai vahūe.
kuvalaya Bh; navalai K, 4, B, Y, P; T; S -- hārań (paharan) B jaha jahim Ma -- devaro Ma; deuro Bh, devaro R; K, Y, P; T; S dandarat K, dastur aimya Y -- vaehūe Ma.
-- --
Wherever on her body her brother-in-law takes pleasure in beating her with a young creeper, there appears (her defence in the form of) a line of sticks formed by the thrill of her hairs.

Page #249

236
The relation between a woman and her husband's younger brother (devara) is characterized by flirtatious behaviour, in which often the woman herself takes the initiative (see Gāthā 456 quoted below). Equally frequently the devara goes too far and the woman has great trouble in keeping him in check (see *38 and 59). In 456 the woman provokes the devara into beating her with a young creeper. Gathā 456 reads as follows (text according to R):
dhāvai pur ao pase- navalayajuttaa ean
su bhamai ditthīpahammi saṁěhāi. de devara paharasu varāin.
She runs in front of you, wanders on both sides and stands in your view as you hold a young creeper in your hands. Go ahead, devara, use it on the poor thing!
From Gathā 862 it would appear that such beatings were part of the love-play between husband and wife. According to Ti and Tp this Gathā should read:
navalaapaharottattha- jaṁ ajja vi juvaijano
e tam ki pi haliaputt Te ghare ghare sikkhju mahai.
What even now every young woman in every house takes pleasure in learning, that thing the farmer's daughter does only when she is frightened of being beaten with a young creeper.
By provoking the devara into beating her with a young creeper she gives him all kinds of ideas. Furthermore, the way she defends herself can only encourage him to go on and take even greater liberties with her. The risen hairs on her body, which look like even so many sticks, are signs of her enjoyment of the situation. For hairs raised on a woman's body from pleasure functioning as a kind of protection, see also Gātha 96. Compare also the Skt expression romāñcakañcuka in Sprüche 3274.

Page #250

237
pavalaā is from Skt navalatā. The variant navalai is from navalaya, with -i- for -ya-.
Pkt pahara 'beating' corresponds to Skt prahāra; see Var. I 10 and Hem. I 68. Most likely they represent different formations from the verb prahr-, Skt prahāra being formed through Vrddhi and Pkt pahara through Guna. Pkt also has pahāra, for which, see *8.
maha-, mentioned in Hem. IV 192 among the Ādeśas for kānks- 'to wish', is from Skt mah- 'to rejoice, to like to'. It is constructed with an infinitive (139 and 862) as well as with an accusative (510).
For deara each time a variant devara is available. The latter seems to have been introduced secondarily under the influence of Skt devara (see above, p. 163). Here dearo in Tp, and B, devaro in Ma, R, K, Y, P, T and S, and deuro in Bh, 456 devara in R, deura in Bh, 690 deara in U, Y and S, devara in R, 771 deara in S, devara in R, 872 dearo in Tp (Ma has only'de), devaro in T, and 963 deara in W. Beside deara a form diara occurs, for which, see *38.
29( 65 Ma, 66 Tp; 27 Bh, 29 R; K, B, Y, P, 30 , 29 G, 5, ; 272 T; 368 .S) Missing in W.
ajja mae deņa viņā ahinavamehāna rao
aṇuhūasuhāi sambhar atte pisavio vajjhapadaha va.
tena Ma, Tp; Bh, R; K, 4, B, Y, P; S -- anumhūa k; anubhūda T -- sambhavaṁtie Ma; saħbharanfte K, satmaraṁtte , sammmar at Te B, sarambhat Te P -- mehāna k, meghāņa B -- ravo Ma, Tp; Bh, R, K,
U, B, Y, P; S -- ņisāmio Tp; Bh, R; K, B, Y, P, nisamio y ; qisāmio T; S -- vajja Tp; R; vajhjha K, vajja , Y, P;S -- padahu Bh, padaho R; W, B, Y, P, paduho K; padaho T; S.
Today thinking over the pleasures I had without him the thunder of the first rainclouds sounds like the execution-drum.
Usually in the absence of the husband thunder produces anxiety about his timely return home. In this Gāthā the thunder produces fear of

Page #251

238
punishment, as the comparison of it to the sound of the execution-ut um suggests. Therefore deņa viņā is to be linked to following anuht&suhāi. However, the position of deņa viņā after ajja mae and before anuhūasuhai is ambiguous, which seems intentional.
For the reasons to adopt deņa and rao, both found in T, rather than tena and ravo found in the other MSS, see above, p. 164. For other instances of dena, see *15.
For sabhara-'to remember and for nisavio beside nisāmio, see *22.
vajjhapadaha 'execution-drum'; compare Skt vadhyapataha and Pali vajjhapațahabheri. The variant vajja may be due to a misreading of the Jaina Nagart ligature for -jjh- ( 5 ) as -jj- ( 5 ). As Tp has vajja, vajjha in Ma is most likely an emendation.
30( 66 Ma, 67 Tp; 28 Bh, 30 R; K, B, Y, P, 31 W, 30 G, 5, ; 543 s) Missing in T and W.
nikkia jāābh Trua gāmo gāmaniņandana
duddarhs ana nimyakidaasariccha tujjha kae taha vi tanuāi.
nikkiva Ma, Tp; Bh, R; U, B, P, nikkava Y; nikkiva S -- jāhabh Trua P -- niddass ana Ma -- nibva Ma, nimba Tp; gimva K; (Bh, R, V, Y) -- kīdayas ar icha Bh; kīdas āriccha K (°icha), , B, Y, P; S -- gāmiņi Ma; Bh, R; K, ggamani v; gāmini S -- kujja Ma; tujhjha K -- tave Ma.
Afraid of your wife you are cruel, hiding yourself like a worm in a (bitter) neem-fruit. You are a true war-leader! Still the whole village grows thin because of you.
Thieme (1937: 117) has suggested that the word gāmaŋt (Skt graman occurring in the Sattas at is used in its Vedic sense of 'leader of an army (grāma)'. This in a way unexpected sense would explain why the word is marked as 'Dest'; see Pāiyal. 104 gāmant bhojo daina vady and Desin. II 89: gāmapahāne gāmaņi gānaudaggāmagohagohā ya. The function of the gāmani is to protect the village (gāma or, as in *33 and 630, a pallt). His heroism and constant desire for battle stand in the way

Page #252

239
of being a good husband or lover (*33) and are a worry to his relatives (630). He is cruel, always from home, preferring a live of hardship above a life of leisure at home. The man addressed is afraid of his wife (jāābh Trua) which makes him cruel (nikkia), too, namely in the eyes of the women he avoids (duddaṁs ana), as he apparently finds the pleasures they promise not worth a domestic quarrel. This has given him the reputation of being like a worm in a bitter neem-fruit (see also 878). With such a gāmaņi-like man about the village would be expected to prosper but instead its women grow thin being unable to meet him.
For the reasons to adopt nikkia found only in k instead of pikkiva found in the other MSS, see above, p. 164.
The word n andana 'son' in gāman in and ana would specifically denote 'someone belonging to the caste or profession of...'. Note Alsdorf's discussion of putra in the compound mantriputra (1950: 358-360). Beside naṁdana,, in this text occur (p)utta (haliautta in 84 and gāmaņiutta in *33) and sua (gahavaisua in 106); for the feminine are found (p)uttī, suā and dhūā.
Y, 198 P, 188 G,
31( 68 Ma, 69 Tp; 29 Bh, 31 R; 191 K, 190 W, 1.92 375, ; 389 T; 326 S) Missing in B, W and X.
neurakodivilaggań hiaa mānapauttham
cihurań daiassa pāavadiassa ummo asti ccia kahei.
--
aiura Y, nevura P; nepura T -- ciuran Ma; K (ciuran), U, Y, P; S daissa Bho, paiņo Bh-- pāapadiassa Ma, Tp; pāyalaggassa R; pāapadiassa K, M, Y, P; T; S -- māņapautthe Ma; padatthamāņa R; pautthamāṇah K, M, Y (padatthao), P; T; S -- umoyati Bha, omo asti Bh -- ccaya Bha (cciya Bh); via K, via y ; s.
Daśarūpa, p. 187 (māņ apauttham!).
By freeing the hair of her husband who had fallen at her
feet, from the clasp of her anklet, she indicated that • pride still kept her heart far away.

Page #253

240
For the position of this Gātha here as 31, see above, p. 161.
mānapauttha which is found in Ma and Tp, on the one hand, and in Bh, on the other, goes back to the MS of the archetype. Bhuvanapāla (ISt., p. 40) explains it as a BahuvrThi-compound in which the order of its members is reversed, meaning '(the heart) from which the sulking mood has disappeared' (for instances in Skt of such compounds, see Gonda, 1968: 246). This interpretation seems to have been generally accepted as appears from the variant pautthamāna. In my opinion, however, we should accept the compound as it stands and take it as a Tatpuruşa, meaning '(the heart) (still) away or distant because of the 'sulking mood' (compare pauttha, literally '(physically) abroad', in the expression ekkaggāma ccia pauttho in 129). My interpretation agrees with the fact that the woman is silent using gestures instead of words, which is an indication that her sulking mood is not over at all.
Var. II 4 and Hem. I 186 derive Pkt cihura by a special rule from Skt cikura 'hair on the head'. Skt cikura would regularly have become ciura, which is actually found in this and the other instances in which the word occurs as a variant; see, for instance, 273 in which cihura in Ma, Ti, Tp, Bh, R, V, P and S and ciura in Y and T. The latter is obviously an 'emendation' for cihura which does not have a clear derivation from Skt (Skt cikhura is found only in the lexicons):
For the reasons to adopt pāavadiassa as found in Bh instead of opadiassa found in the other MSS, see above, p. 164.
Y, 194.G, 38
5, TT;
32( 69 Ma, 70 Tp; 30 Bh, 32 R; 197 K, 1961, 198 322 T; 420 S) Missing in B, P, W and X. .
sā tai sahatthadinnan uvvasi anaaraharade-
ajja vi o suhaa gadhar ahi am pi vae vva nomālia vahai.
tue Tp; tań R; tui K, Y; tue T -- suhattha K -- dinnaa Ma; dinha K -- ajjha Bh; aja K -- u Bh; re K, y, Y; S -- gamdharaia Bh; andharahian K -- uccasia -- ghara Ma, Tp; Bh, R; W, ghara missing in Y; ghara T; S -- devaa wa Ma; Bh, R; devau vau K, devada va Y; devaa vya T; S -- omāliyah Bh, nomāņi aħ R; (devau va) u mālian K; omālian S.

Page #254

241
Pāda d quoted in Hem. I 38 (omālayan vahai).
Still she wears the navamālikā-flower, o handsome man, you gave her personally, though it has lost all fragrance, like a house-god (in a house) in a deserted town.
For the present position of the Gathā as 32, see above, p. 161.
For tai, the instrumental and locative (once, in 338) of the personal pronoun of the second person singular, practically each time a variant tui is available. Its long form tae has a variant tue. The MS of the archetype. seems in most cases to have had tai and tae, ascertained through its occurrence in Ma, Ti and Tp, on the one hand, and Bh and R, on the other. Admittedly Ti and Tp often have tui.(T always has tue, for which, see below. S and the Vulgata MSS vary.) tai is comparable to the nominative tan, tui to the other nominative tumam and the genitives tuha and tujjha. These three forms, which together are much more numerous than tan, may have suggested the idea that tui was more correct than tai. The instances of tai and tae are the following:
* 34 tai ti, TP, R, S, tae , B, tuha Bh, tue K, Y, tu(yavva) P, tue T, tui W; 137 tae Ti, Tp, Bh, R, K, U, tue Y, P, T, S;
194 tai Ma, Bh, W, S, tan R, tue TP, T, tui K, Y; 212 tai Ti, Tp, . Bh, R, K, M, S, tuha Y, ita P, tue T; 215 tai Ti, Tp, Bh, R, K,
V, Y, 'tuhi P, tue T, tuha S; 352 tae Bh, R, K, U, Y, tai S, tue T; 437 tai Ma, Ti, TP, U, Y, S, tue T; 478 tae Ma, Bh, R, tue Tp, (Ti lacuna), 4, Y, S; 667 tae , tue Y, T; 700 tai Ma, Bh, K, P, tuha R, tui Y, tue T, S; 811 tai S; 826 tue T; 860 tue W (tumae T); 876 tue Ti, TP, T, W; 912 tae Ma, tue Ti, Tp, T; 919 tai Tp (tui pr. m.), ruņai Ti, tue T.
The occurrence of tai and tae is entirely regulated by the metre. It is possible to explain the final -i in tai as a graphical representation of -ě. However, it is uncertain if and how far the writing of i for ě was carried out in the MS of the archetype, especially for -ě in an open syllable. Alternatively, tai may have been an independent word in Pkt existing side by side with tae. Its form, with -i after short

Page #255

242
a, became only common in Apa. For another Apa. word already occurring in Pkt, see aha discussed in *23. In fact, the occasional occurrence of the variant tae or tue (i.e. tаě, tuě) for tai would betray attempts to remove this Apa. form from the text.
For an explanation of the origin of the variant aħdharahia found in K, see above, p. 165.
For the particle o, see Var. IX 4 (o sūcanāpaścăttāpavikalpeşu) and Hem. II 203 (o sūcanāpaścăttāpe).
of the verb udvas- Skt has only the causative, i.e. udvāsay-, meaning 'to remove, to destroy' (cf. uvvāsia 'removed' in Gatha 106). The participle uvvasia, from uvvasa-, seems to me an 'deserted, fallen in disuse' (from there possibly also 'dilapidated'; cf. uvvasia in 233, quoted in the note on *7); compare (late) Skt udvasa 'uninhabited, empty' (RajataranginT), 'solitude' (satruhjaya-Māhātmya; MW).
For the reasons to adopt hara found in k instead of ghara found in the other MSS, see above, p. 164.
For devaeva (Tp, ) beside devaa va of the other MSS, see *23.
nomalia corresponds according to Hem. I 170 to Skt navamalika 'Jasminum sambac'. Var. I 7 has nomallia, i.e. Skt navamallika 'id'. The variant omālian (Bh, K, S) undoubtedly is a corruption for nomāliam, which may be due to preceding o suhaa. As becomes clear from the quotation of part of the last pada of this Gātha in I 38 Hemacandra used a MS of the Sattas at which contained this corruption. His subsequent dealings with the word omāliañ are illustrative of his method. The context suggests that it denotes flowers offered to a deity, which, moreover, had lost their fragrance. In II 92 Hemacandra ad hoc produced a Skt word avamālya, which would be synonymous with 'nirmālya 'the remains of an offering (of flowers) to a deity' (1 38).
, P, 31 G, E, T; 653 s)
33( 70 Ma, 71 Tp; 31 Bh, 33 R; 31 K, B, Y, 32 Missing in T and W.
pahar avanamaggavisame gāmaniuttassa ure
jāā icchena lahai se niddah pallt uņa se suhah suai.
papahara Ma -- viņa (vana) R
-- maga P; magva S -- kicchena Ma;

Page #256

243
Bh, R; K, W, B, Y, P; S -- mahai Bh (lahai Bh") -- niddhah P -- gomini Ma; gāmini Bh; K, Y; S -- outtasya B, P -- vallt P -- una sa K, Y, ua na sa P -- suhai (suh ari suai) P -- suvai Ma, Tp; S.
The war-leader's wife sleeps with difficulty, lying against his chest which is rough with scars from wounds of (sword-) strokes. His village, though, sleeps soundly.
For pahara 'stroke', see * 28, and for gamant 'war-leader' and the use of the word (p)utta, see *30.
For the reasons to adopt icchena found in Tp instead of kicchena found in the other MSS, see above, p. 165.
For the reasons to adopt suai instead of suvai, see above, p. 167.
34( 601 Ti, Tp; 32 Bh, 34 R; 32 K, B, Y, 334, 31 P, 32 G, 5, TT; 518 T; 533 S; 95 W)
aha saṁbhāviamaggo enhiñ hiae ann am
suhaa tai ccea navari nivvūdho ann am vae loassa.
sabbhāvia Tp; savbhāvi(y)a Bh, R; sambhovia K; sabbhāvia T; sabhbhāvia S; savbhāvia W - tuha Bh; tue K, Y, tae , B, tu(ya?) P; tue T; tui W - - vyea Ma; cceva R; jecca K, jevva Y, (tuya)vva P; vvea S; ccia W -- nivara Ma, qavara Tp; Bho, R; K (Oran), , B, Y (ora), P; T; S; W -- nivvodho Ma -- ihn im S -- iae annan Ma; ann an hi(y)ae Bh, R; hiae ahņań Ta -- second annan missing in Ma -- vai R; K, W, B, Y, P; S.
Handsome man, that (old) esteemed course is no longer accomplished by anyone else but you. Nowadays there is one thing in a person's heart and another in his speech.
The '(old) esteemed course' refers to speaking the truth. The Gatha is addressed to a man who accidentally addressed his wife with his mistress's name (gotraskhalana).
For aha, see * 23; for tai, * 32; and for pavari (Bha), * 15.

Page #257

244
35( 71 Ma, 72 Tp; 33 Bh, 35 R; 33 K, B, Y, P, 34 W, 33 G, 5, 7; 527 'T; 309 S; 36 W)
unha viņTsasaṁto hiaan palTviañ aņu-
alingasi kisa maṁ parāhuttin. saeņa patthiṁ palTvesi.
In Ma lacuna from para (huttin) onwards. uhņā vinto Ma; unhāi nisas arto Bh, R; K (unhāin), W, B, Y, P; T; ungāi pts as asto S; unga vinisas arto W -- sa(y) an addhe kisa me par āhuttin Bh" ("hutti Bh"), R; kТsa mae par ammuh Te saanaddhe K (htee), Y (ohta), P (sahan ao ), ktsa maha par aħmuh Ta saan addhe , B; kisa maha par arhk khuhĩa saanaddhe S; kfsa hma par armuh Te sayanaddhe W -- palTviurh BhD (°iyan Bha); U, palTviūņa K, palTviŭņa B, Y, paliliūņam P; palTvia T; palTvium S; paliyā W -- puţthi R; K (oththo), W, B, Y, P; paţthan T; puţthin S; petthan W -- palTvesu Y; valāvesi To, palāvesi To; peleppayasi W.
Why do you, sighing hot sighs, embrace me from behind? (First) you burnt my heart (and now) you burn my back with your repentance
The word aŋusaa (Skt anuśaya) 'regret, repentance' provides a pun, as etymologically it denotes 'lying close'.
For hutta 'directed towards' (Hem. II 158), see also 145 and 896 (parāhutta), 373 (antohuttan) and 696 (gharahuttamaŋena). The derivation suggested by Weber from bhūta (Retr., p. 410; cf. Hem. II 99: hutta from bhūta) fails to convince, from the point of the meaning as well as from the fact that hutta for bhūta does not actually occur.
For pattht 'back' beside Skt prstha, see *11.
unhan viņisas ato is based on Ma, Tp and W. unhāi in the variant unhāi nis as aṁto is problematic. It would through ellipsis stand for 'hot sighs'. However, such ellipsis is, as far as I know, not very common in Skt, which even tends to the opposite; note such expressions as tapas tap-. Possibly it originates from a corruption, like the one found in Ma, uhņā viņts as aṁto, which was subsequently corrected into unhāi nis asamto.

Page #258

245
The available variants seem to indicate that the peculiar position in the present text of kisa after, instead of what might be expected before, alimgasi was misunderstood. kisa was subsequently linked to palīvesi, as a result of which the word alimgasi lost its proper context, and the accusative mam made no longer any sense. alingas i was replaced by saan addhe, and the accusative maṁ by a genitive; e.g. me in Bh and R (side by side with accusative parāhuttim!). In this new sentence, kisa ...paltvesi, palTvia makes no sense. The word palTvia, which beside in Ma, Tp and I (and W) is also found in Bho, is therefore most likely a retention, even though in the adopted text the absolutive palTvium (Bho, v and S). would also fit well. The second line consists of two independent sentences, namely hiaan paliviań, 'my heart is already burnt (by you)', and anus aeņa paţthi palTvesi, '(now) you burn my back with your repentance.'
, 193
Y, 199 P, 189 G, 39 g,
36( 73 Ma, 74 Tp; 34 Bh, 36 R; 192 K, 191 ; 346 T; 410 S) Missing in B, W and x.
:
tujjhangarāasese- sā ira golātūhe
ņa sāmalt taha kharena somālā nhãā jau ūkas.eņa.
tumhangaasesena K, tujhao Y, P -- rāga S -- sāmila Ma -- teha Ma -- Śarena k -- somārā K, , Y, saumāra P; somārā S -- suiram Bh, R; să kira K, M, Y, P; T (kila); S -- godā K, goda S -- rūhe Ma, titthe Tp; rūhe R; K, lacuna in P; tîre T -- jambū Ma, Tp; K (mbū), B, P; T; S; (Bh, R, , Y).
They say that that dark and (taha) soft-skinned woman bathed herself in the Godāvart at the (same) bathing-place with the biting lotion of the jambū-treebark left over from your bath.
For the present position of the Gathā as 36, see above, p. 161.
janvūkasāa refers to the astringent juice extracted from the bark of the jambu-tree. Because of its astringent nature it must have hurt the woman who is expressly said to have a tender skin (somāla). The juice

Page #259

246
is used for a number of things, e.g. as a black dye (see Watts III, pp. 284-6). It is therefore foolish of the woman to apply it to her already dark skin (sāmall), the more so as for women a light-coloured skin is favoured, which they try to bring about by using a lotion of turmeric (see 58, 80 and 246). The woman is apparently desparately longing to be touched by the man (cf. *16 in which the woman begs to be touched by the same moonbeams that touch her lover). The Gathā has a counterpart in 246 in which a man is drinking the water of the river mixed with bitter turmeric used by his beloved as a bathing-lotion. In this case, too, it probably concerns a bathing-lotion floating on the water after the man's bath. In any case it is unlikely that he left it behind in some jar on the bank of the river for anybody else to use it. It would seem to appear from this Gāthā that men tried to enhance the blackness of their skins. Note in this connection the following Apá. verse found in Hem. IV 330, which indeed seems to comment favourably on the contrast of the lightness of the woman's skin and the blackness of that of the man:
dhollā sāmala dhan(i)a campavanpT nāi suvannareha kasavatgai dingT.
The dholla is dark, his wife campaka-coloured. It looks like a streak of gold on a (black) touch-stone!
For tūha 'bathing-place' beside Skt tirtha (Pkt tittha), see Mayrhofer, EWA II, p. 507, and Turner, CDIAL 5903. tūha is mentioned in Hem. I 104 and II 72. The variant rūhe in Ma, R and K probably rests on a misreading of the Jaina Nagarī symbol for tu as rū (note Weber's hesitation regarding the reading of Bh, which could be tūhe or rühe; ISt., p. 41). The variant titthe in Tp possibly is an emendation for this ruhe in the source of Tp. See also 58: rūha in R and Ma and tittha in Tp.
For the variant teha in Ma, for taha, see *1.
Other instances of ira (Ma and Tp; cf. suiram in Bh and R) beside kira are found in 499 tassa (vi?) ira in Ma (cf. Hem. II 186: tassa ira, which probably is a quotation) against tassa kira in Tp, V, Y, .T and W

Page #260

247
and tasseva in R and S, and 773 ko ira in S against ko kira in R. The loss of the initial k-, which is restricted to its occurrence after vowels (contrast kallar kira in *50), may be explained from the less emphatic or tense pronunciation of this particle (see also the note on *15). The variant kira would seem to have been introduced secondarily in the MSS individually or in some earlier stage in the transmission of certain branches. Pkt (k)ira corresponds to Skt kila (Var. IX 5 and Hem. II 186; the latter mentions ira as well). For the meaning 'it is reported or said', see Emeneau (1969: 43-54). .
For somāla 'soft, tender', corresponding to Skt sukumāra, see Var. II 30 and Hem. I 171, 254; see also Turner, CDIAL 13448. The variant somāra, which is not taught by the Pkt grammarians, is an innovation based on Skt sukumāra.
The word gola is glossed with godāvart 'the Godavart river' (Paiyal. 132 and Desin. II 104). The relation between this golā and Skt godāvarT is as yet unclear; see Lüders (1942: 34).
W, 34 G, E, T; 304 T; 341 s)
37( 35 Bh, 37 R; 34 K, B, Y, P, 35 Missing in Ma, Ti, Tp and W.
tuha virahe ciraāraa rairahasiharadhaena va
tissā nivad aitavāhamailena muheņa chāhi ccia na patta.
tuva B -- viraye P -- vi(y)alanta Bh, R; nivadaita k -- bāha B; T -- mainleņa K; saileņa s2 -- ravi Bh, R -- dhaveņa Bho; dhvaeņa k -- wa P; S -- chai Bh; chāyi P; chāah S -- cvia S.
As long as you haven't returned, slow traveller, her face spoiled by falling tears, will not regain its natural lustre like the chariot of the sun, a (mountain)peak or a banner, which never enter into the shade.
Weber translates rairahas iharadhaena with 'the banner on top of the chariot of the sun'. This translation does not do full justice to the word sihara (Skt śikhara). Instead I should like to suggest to analyze the compound as a Dvandva enumerating three objects which never enter

Page #261

248
the shade, namely the chariot of the sun, a mountainpeak and a banner.' For a Dvandva in the singular, see AIG II 1, $$ 68-70, which, however mentions only instances consisting of two items; but see vāyvagnivipram in Mānava-Dharmaśāstra IV 48. Note in rairahas iharadhaa the significant order in the enumeration of the items, from something high in the sky to something comparatively low to the ground, which is a feature of Dvandva-compounds (AIG II 1, $ 71). Another instance in the Sattasaf of this type of Dvandva is probably alihiabhittivāullia in 217. This Gatha reads as follows (vāullian as found in Ti and Tp):
tan mittam kāavvaṁ ālihiabhittivāul-
jam kira vasanammi desaālammi lian va na par ammuham thai.
That man is to be made a friend, who is known in times of misfortune, when and where he is needed, not to stand with his face averted, like something drawn, like an (immovable) wall or like a puppet '(? which cannot move away by itself).
For Pkt chāht beside chãa (Skt chāyā), see Var. II 18, where the intrusion of the -h- is noted, and V 24, where the commentator Bhāmaha mentions chāht as showing the substitution of the feminine gender marker -1 for -a (cf. Hem. III 34). Hemacandra in I 129 makes an interesting observation regarding the occurrence of chāht; when it occurs it would always me an 'shadow' or 'shade' and never 'beauty' or 'lustre' (akāntau). The few times chāht occurs in the Setubandha it indeed means 'shade' or 'shadow' (see X 37, 47, 48, 54 and 55). In the Sattasaf it occurs four times, namely here, in *41, 49 and 136. Here and in 49 and 136 (for the instance in*41, see below) it means indeed 'shade' or 'shadow'; in the Gāthā under discussion, however, only secondarily, by way of a pun. This shows that the distinction made between Pkt chăā and chaht is spurious. It is possible that Hemacandra overlooked this instance or that in the MS of the Sattas at he had before him the text read chaā, as is the case in Bh (chãi, i.e. chaya) and s (chẵan).
The origin of the form chāht beside chaa (chāyā) is uncertain. Turner, CDIAL 4027, suggests that it is due to contamination of chaa with a(b)ha or p(r)a(b)ha 'light'. Ghosal (1959: 256- 259) suggests that the.-h- is

Page #262

249
a glide. A fanciful derivation is found in Pischel § 255. Whatever its origin, the form is in any case also found in some NIA languages.
38( 74 Ma, 75 Tp; 36 Bh, 38 R; 35 K, B, Y, P, 36 V, 35 G, 5, TT; 394 T; 526 S) Missing in W.
diarassa asuddhamaņas - diahaṁ kahei rāmā-
sa kulavahū piaakuddaliniai nulaggasomitticariai.
daar aṁsa Ma, devaraassa Tp -- addhasuddhamanas sa Ma, asuddhaamanassa Tp; asuddao k, suddao y -- kulabahū R; kulavahuā k, kulabahuāY, P- nani Ma; niya Bh; nia K, W, B, Y, P -- kuddi Bha (odda Bh"); kudu K, P, kui B; kuddu S -- lihi(y)āin Bh; K (°āmin), B, P; lihiaāin T; lihiāin S; (R, X, Y) -- somitta Bh; saumitti k -- cariyāin Bh; cariānis K, cariāim B, P; T; S; (R, W, Y).
The good wife takes all day to tell her young brother-in-law, who has improper intentions, about the adventures of Sumitra's son, the devoted brother of Rāma, though these were drawn on the walls of their very own house.
The term kulavahū denotes a wife who by her behaviour avoids any harm or embarrassment to her husband's family. In this Gathā she tactfully deals with her husband's younger brother, devara, who misbehaves towards her. She does not expose his behaviour to the other members of the household which would only result in a quarrel between him and her husband and finally in a rift in the family (see Gathā 59). Instead she tells him edifying tales about the ideal devara Lakşmaņa. Thus she makes him understand she knows what he is up to. However, these tales were depicted on the walls of their own house, so the devara will already have been quite familiar with them. This and the fact that she takes the whole day for it should indicate to him that she is not at all averse from his company.
diara 'brother-in-law', husband's younger brother' goes back to Skt devara (see Var. I 34 and Hem. I 146). Otherwise Pkt has deara (see * 28). It is possible to recognize in this diara beside deara an early instance

Page #263

250
of the development found in Bengali, of -e- in the neighbourhood of an hiatus changing into -i-. As an example Bloch (1934: 44) quotes siult from Skt śephalika. Turner, on the other hand, suggests that the special development found in diara goes back to the use of the word expressing a certain degree of intimacy (CDIAL 6546). The pair diara-deara has an exact parallel in vianā (64) beside veanā (e.g. * 26) (both from vedanā), of which the former could then be taken to denote a 'terrible pain' against the latter which would simply denote 'pain'.
39( 75 Ma, 76 Tp; 37 Bh, 39 R; 36 K, B, Y, P, 37 V, 85 PP, 36 G, 5, ; 404 T; 572 S) In P this Gātha occurs as 36 and a second time after 85, itself numbered 85. Missing in W.
.
caccaraghariņt piadań- asaT saajjhia dug
saņā a tarunt pautthavaia a gaā a na hu khandi an silan..
cattara Bh; K, W, B, cattaa y, po, a P -- dassană Ma; daśaņā P, dasana po -- e Ma; vi S -- pautthapajā Ma, Tp; pautthadaiyā Bha (vaiya Bh); odaia K, Y, P, o taia PD, paia , B; T -- e Ma; a missing in P, pB -- saaṁdhia Tp; saajjiā R; K, V, B, Y (onjiā), P, sahadhyā P; saaccia T; saajhjhia S -- duggaā e Ma; a duggaa , B -- na kkhu Ma, na khu Tp; na ya Bh, puna R; ņu hu B. -- khandia B, kharidi am missing in P -- sīlai Ma.
A beautiful housewife living on a busy road, a young woman whose husband is abroad, an unchaste neighbour who is poor, (who believes that) propriety is not lost?
Hem. II 12 equates caccara with Skt catvara 'crossroad' (i.e. where four roads meet). The etymology of caccara is as yet unclear. Schwarzschild (1956: 112) suggests that caccara derives from cattara (+catvara) which was contaminated with the verb car-, as in sacara 'road'. I should like to suggest that Pkt (and Pāli) caccara is from*carcara 'a place where there is a constant coming and going', from there 'crossroad' (as a place where traffic accumulates). Note that in the Sattasat caccara each time indeed in particular denotes a place where a lot of people

Page #264

251
pass (see *47). catvara is merely a convenient Skt gloss. Its Pkt derivative, cattara, has in some MSS been entered into the text as the more familiar formation, for original caccara.
For the reasons to adopt pautthavai ā (Bho, s) instead of paiā, see above, p. 164.
Pāiyal. 236 mentions the word saajjha, of which saajjhia is an extended formation, as a synonym of samosia 'living together' (sam-a-vas-). Deś in. VIII 10 translates saijjha (for sayajjha) with padivesia 'neighbour', which as a gloss is also found in most of the commentaries on the Sattasat; see, for instance, Kulanātha, quoted in Abh., p. 86, and pādivesa in 335 in W in the Pkt text. In this connection it should be noted that Deśfn. VIII 49 translates samosia (see above) with paçivesi. This meaning 'neighbour' returns in the derivatives of the word in Mar.; e.g. sejt 'neighbour'. Turner, CDIAL 13077, derives saajjha from * sagrhya which would mean 'living side by side' (cf. Vedic sakşit
living together or side by side'). . The variant saajjiā may be due to a misreading of the Jaina Nāgari ligature for -jjh- as -jj-; see the note to * 29. saandhia in Tp is the Pkt form of sagandhika 'relative' with which Madhavayajvamiśra translates saajjhiā. See also T, quoted in Retr., p. 428: (ta?) sagandhika prātivesika, and. V, quoted in Ed., p. 124: sam) gandha. For the variant saandhia, for saajjhia, see also 335 (Tp dhia), 848 (Ti, Tp) and 849 (Ti, Tp, I).
In Pkt. the Skt particle khalu was first shortened into khu, which is found after Anusvāra (see, for instance, *6). After vowels its pronunciation in M. Pkt was further weakened, according to the rules of internal sandhi, into hu. This development of khalu into khu and hu should probably be explained from the unstressed or unemphatic nature of the particle (see *15). Note in this connection that khalu is never found in the emphatic position at the beginning of the sentence. On the other hand, it is often found in fixed idioms, such as na hu (here, 48, 227, 279, 478, 666), mā hu (521, 607, 909; in 909 kiń nu khu in Ti and Tp) and taha vi hu (229, 275 (?), 470, 647, 741), of which the first is known from Skt, whereas the latter two are not. For taha vi hu a variant taha vi a tathāpi ca) is found occasionally in the MSS of the South-Indian recensions (275 Ma, Ti, Tp, 470 T,

Page #265

252
647 Tp; in the last instance taha vi a khu in Ma). Otherwise hu occurs freely (203, 212, 230, 242, 265, 401, 425, 498, 775, 889, 891, 896, 903, 926). For the meaning of (k)hu, see Emeneau (1969: 240-68).
For hu in the Jaina and the Vulgata MSS, Ma, Ti and Tp have throughout a variant khu (note one typical reversal, in 478, with hu in Ma, Ti and Tp, and khu in v and ku(?) in R and y). khu is also occasionally found in T and, where available, in W. S often has kkhu (against the metre). Khu in these MS would seem to be secondary. As such it may be an indication of the influence of Saurasent on these MSS, which dialect has either khu or kkhu but never hu (see Pischel & 94) (for other possible traces of this influence on the South Indian MSS. see *27).
40( 38 Bh, 40 R; 37 K, B, 38 V, V, P, 37 G, 405, IT; 560 S) Missing in Ma, Ti, Tp, T and W.
.
tālūrabh amādanakhudi- daravuddavuddanivvud-
akesaro girişate pūrena damahuaro hirai kalavo.
khuddiya Bha, bhamaņas ankhudia R; bh amāulakhudia K, M, B, Y, P ! (okhadia); S -- naie K, Y; naya , nata B, ņaf P; nata S. -- daravuddapun ibudda Bho, vudduvuddun ibudda Bh", "buddabuddan ibbudda R;
cuduvudun ivudu K, vuddaghudanighuda W, dakhuduchuuņichuda B, ovuduvudunivvudu Y, budubuduni budu P; ovuddhavuddhanivuddha S -- h Trae S -- kalambo K, B, P; kaavo S (Bh, R, W, Y).
The Kadamba ( -flower) is dragged along the stream of the mountain-river, its filament broken by the violent movement of the whirling water; the bee (sitting on it) first sank a little, then sank and finally went under completely.
For tālura 'whirlpool', see Pāiyal 129: tālāro āvatto, and Desin. V 21 (Hemacandra's own commentary): tālāro avartaśceti kecit.
For bhamāçana 'tossing around', see Hem. III 151, IV 30 and 161, Pischel S 554 and, finally, Turner, CDIAL 9648. Bloch (1934: 224) connects the suffix -ad- with the nominal suffix -- which is unknown in

Page #266

253
Pkt but characteristic of Apa. (see Alsdorf 1937: 18-9). The readings found in R (bhamanasan(khudia)) and the Vulgata MSS (bhamāula(khudia)) seem to represent respective attempts to remove this Apa. formation, which is foreign to Pkt proper, from the text. A similar phenomenon can be seen in 92.
The fact that bhamādana is a relatively modern formation and the fact that the Gathā was not present in the South-Indian recensions, suggest that the Gathā was not included in the archetype but was added to the text only in a later stage. A similar conclusion on similar grounds can be reached regarding 77 and 92 (see also above, p. 162).
In Hem. IV 116 khuda- is enumerated among the Adeśas for tud- 'to break'; for the participle khudia, see Hem. I 53. In the SattasaT otherwise only ukkhudia 'broken open' occurs (331). For khuda-, see also Turner, CDIAL 3892.
For vudda- 'to sink' (or 'sunk' the past participle being identical to the stem), see Hem. IV 101. The verb is amply attested in NIA (see Turner, CDIAL 9272). vudda- does not occur in the Setubandha or the Gaud avaho. The verb is, however, found in Apa. (see Hem. IV 415 and 423); nir- in nivvudda-, seems here to intensify, as in Skt nirmajj
kalava "Nauclea cadamba' corresponds to Skt kadamba (Paiyal. 255, Var: II 12 and Hem. I 30, 222). As Tamil katampu shows, the - 1 - in Pkt is more original than the -d- in Skt (see Mayrhofer, EWA I, p. 144).
,
;
41( 76 Ma, 77 Tp; 39 Bh, 41 R; 38 K, B, P, 39 X, Y, 38 G, 41 384 T; 293 S) Missing in W.
ahiãamāṇino dugniavandhavāņa jhūrai
gaassa chāhim paissa rakkhaîtT ghariņT vihavena ertāna.
In Ma lacuna from ahiūamān(iņo) to (ghari)nT. ahiai Tp; ihiiyāa Bha ahiyāi Bho, ahiyāni R; ahiāi , B -- dugāassa P -- chāań Tp; 'Bh, R; K, Y, P, chāahi U, chaam T; S -- payassa Bh; piassa K, Y, P -- rakhkhanty K -- bandhavāna Tp; B, P, vandhavāṇaḥ K; baṁdhavāna T; (Bh, R, X, Y) -- kuppai Tp; jūrai Bh, R; V, Y, kujhjhai B, sūrai P; jūrai T; jūrai S -- vibhaveņa Tp -- ertānā Ma; istānañ Bha (ento Bh), R; pattāņa K, Y (°āna), P, ettāna , B;

Page #267

ehtagam T; ettaga S.
254
Not wishing to rob her husband who was poor but had high notions of his birth, of this delusion, the house-wife was angry at her own relatives when they arrived with pomp and splendour.
I suggest to take chaht in chahim...rakkhamt in the sense of 'delusion' (i.e. about his own greatness). Compare Skt chaya 'hallucination' in Susruta-Samhita 1.114, 13, 15 (PW: 'Schattenbild, Hallucination"). For the meaning given to rakkha-, compare the expression papaan...rakkhanto in Gatha 122, "protecting", 1.e. "not wishing to hurt" the affection (his first wife has for him)'.
chahi is restricted to and B, the other MSS having chaam. There are reasons to assume that chahim is actually a retention.. It is in any case difficult to see why, if the text originally had chaam, the latter was replaced by the more difficult form chahim (no clear explanation of the form chaht beside chaa is available; see *37). There was no necessity to do so from the point of the meaning of the word either. chahi occurred side by side with chaa.
If chahT is indeed original here, one has to explain what in this particular instance could have been behind its change into chaa. The fact that this change took place independently in the South as well as the North-Indian branch points to an external factor. In this connection the traditional interpretations of the word should be considered. (Unfortunately in Madhavayajvamisra's commentary in MS Ma nothing specific is found in relation to chahim.) Bhuvanapala (ISt., p. 42) apparently takes the word in the sense of 'complexion': vailaks (y)ena ma malinimanam Systu iti nijabahdhavebhyo as@yati, in order to prevent (the husband) from acquiring a dark complexion from embarrassment she shows anger to her own relatives'. In S (Retr., p. 360) and in E and m (Ed., p. 15) the word is glossed with mahatmyam; in G (Retr.. p. 360) with mahattvam. 'Greatness' can here only denote the image. which the husband thinks he projects. As such the meaning of these glosses comes very near to that of 'delusion' suggested above. The gloss kirti 'fame' for chaa found in DesTn. III 34 may well have been based

Page #268

255
on these, taken literally. On the other hand, these interpretations all seem to proceed from the meaning 'beauty' or 'lustre'. In *37 I have referred to Hem. I 129 where it is said that where the word chaht occurs it invariably means 'shade' or 'shadow'. Undoubtedly this restriction was initially based on the observation of the occurrence of the word in the texts. It is not unlikely, though, that it subsequently acquired the character of a prescriptive rule. This may have been responsible for the change of chāht into chaa in the Gathā under discussion. In any case the rule did not prevent the substitution of chāā as the easier form, for chāht, the more difficult one.
For jhūra- 'to be angry at' and the variant jūra-, see *13.
The variant pattāna(m) for entāṇa rests on a misreading of the Nagart character e- for pa-. Regarding the variant entāṇā found in Ma it should be noted that in this MS a vowel at the end of the Gatha is occasionally lengthened; e.g. piaamammt in * 17.
,
;
421 78 Ma, 79 Tp; 40 Bh, 42 R; 39 K, B, 40 V, Y, P, 39 G, 42 449 S) Missing in T and W.
sahtne vi piaame duggaapautthavai am
patte vi chane na madio appa saajjhiam santhavaṁtte.
sahine Ma -- pi P -- pate Ma -- cchane Ma; khane S -- duggai R; dugāa P; ciraara S -- pautthapaiaṁ Ma, Tp -- saacchiaaṁ Ma; sajjhian R; saajjian K, V, B, Y, P; saajhjhia S.
Though her dearest one is at her disposal (and) the festivalday has arrived she has not adorned herself, (ostensibly) in an attempt to cheer up her neighbour who is poor and whose husband is abroad.
The woman would not adorn herself out of sympathy with her neighbour, who is poor, whose husband is still abroad and who consequently cannot afford and is not allowed to adorn herself. However, so much sympathy between neighbours is suspect. A comparison, with, for instance, 848 suggests that the woman may in fact have an affair with her neighbour's

Page #269

256
husband and does not adorn herself because he has not yet returned home. The Gatha reads as follows:
ruai ruaṁtte mae piavirahe māmi saaj-
ohidine ganai jhijjai ahaṁ va jhiae neho ccia apuvvo.
She cries as I cry, she counts off the days and grows thin like I do. Aunt, the affection of the the neighbour (for me) during my dear one's absence is unprecedented.
chaņa 'festival' (see Pāiyal. 248 and Hem. II 20) goes back to Skt kşana '(right) moment', which otherwise became khana, meaning 'moment'; see *14 and Turner, CDIAL 3642.
For the reasons to adopt pautthavaiam as found in Bh, R, etc. in- . stead of paiah found in Ma and Tp, see above, p. 164.
For saajjhia 'neighbour', see *39.
5, it;
43( 77 Ma, 78 Tp; 41 Bh, 43 R; 40 K, B, 41 W, Y, P, 40 G, 43 305 T; 437 s) Missing in W.
tujjha vasai tti hiaa tuha virahe kisiaiỉ
imehi dittho tuman ti acchti ti tte angāi vi piāi..
tujha K, tujhjha , B, P; S -- vasahi Bh, R; vasihi S -- diththo K -- acch Thi Ma; achThin Bh, achihim R; Y, achTin K, acchTim B, P; T; S; (v) -- kisiāati tti Tp; kisiāati R; kisiāņi tik, kisiāi ti , B, Y, P; kasiãi ti T; kisiāi ti S -- tena Bh; tfa S -- asgāin K -- pūāi Ma; piāin K, B, P; T; S; (Bh, R, W, Y).
Her heart is dear (to you) because it is your house, her eyes because through these she looks at you, (and) her limbs (I suppose) because they have become thin during your absence.
The variant kisiāaiti (R, Tp) is due to the spelling kisiāyah ti for

Page #270

257
kisiāin ti. kisiāati is interpreted as krśāyante (R). In Tp the quotative particle which was thus lost is added again: kisiāati tti (cf. & and : krs Tbhavanttti).
,
;
44( 79 Ma, 80 Tp; 42 Bh, 44 R; 41 K, B, 42 U, Y, P, 41 G, 44 557 S) Missing in T and W.
sabbhāvanehamaie anahiae una hiaan
ratte rajjijjai tti juttam in a ja dijjai tam jano hasai.
sabbhāvangeha Ma; sabhbhāva k -- bharie R; W, B, mahie P-- rattijjai Ma, raccijjai R; rattijjai y -- annahiae y -- in Ma lacuna for hasai.
You may be full of true love (but, mind you) falling in love with a loving man, that is (a) good (combination). But at this people laugh, namely when the heart is given to a man who is completely heartless.
Kulanātha (Abh., p. 87) and Bhuvanapāla (ISt., p. 42) translate the word maie with maye. maia would be an extended formation: * mayika. It may be asked, though, if maia does not go back to madita (as in ummaia, from unmadita, in 498 etc.) 'drunk' or 'full with' (from saturated!; cf. matta in 66). For the translation this hardly makes any difference. nehamaia occurs also in 450.
Weber, following the commentaries, takes sabbhāvaņehamaie as a locative singular masculine, linking it to following ratte. Alternatively, it could be taken as a vocative of the feminine, in the same way as vahuvihavilās ar asie in 477 (not so interpreted by Weber and the commentaries). This interpretation may in fact be preferred. Note in this connection the absence of a connective particle between omaie and ratte. Furthermore, maie as a locative linked to ratte, seems to disturb the clear parallelism between ratte rajjijjai, on the one hand, and anahiae una hiaas dijjai, on the other.
For the variant raccijjai in R, see Hem. IV 422 (23) (Apa.) and Turner, CDIAL. 10583.

Page #271

258
ana- in anahiae is a double a-privans, an- + a (cf. CPD I, p. 142"), meaning '(completely) without' (cf. Pāli anabhāva 'non-existence, annihilation'). See in this text also anahontaa (212), anarāma aṁ (so read; 325), anarasia (555) and ana(horana) (774). See also Hem. II 190 and Pischel in his translation of Hem., p. 92, where earlier litarature on ana- is quoted.
45( 25 Ma, 26 Tp; 43 Bh, 45 R; 42 K, B, 43 X, Y, P, 42 G, 45 Į, T; 30 T; 600 S; 26 W)
ārambhamtassa dhuan ta maraŋam aņārasbhe
lacchi maranai va hoi purisassa vi hoi laccht una na hoi.
ār arbhatassa Ma -- dhura Ma, Tp; dhuvan Bho; dhur an T; dhar an W -- lacht -- ca R -- purusassa Tp; R -- maranam ar an ār abhe Ma -- hovi P -- lacht K.
A man who shows initiative is certain of either wealth or death. Even when he lacks all initiative (that same) death is there but there will be no wealth.
dhuan of the North-Indian branch is preferred to dhuram of the SouthIndian branch. dhur am would go back to Skt dhurā 'yoke, burden'. It is doubtful, though, if dhurā provides a suitable object to arambha- 'to undertake, to take the initiative' or if arambha- is a suitable verb to dhurā (cf. Pāli vahanto...dhuram; PTSD, p. 342"). Originally dhura may well be nothing more than a corruption for dhuam. Note the frequent occurrence of -r- in this and the next line.
For purisa 'man', see Var. I 23 and Hem. I 111. Ma, Ti, Tp and W have in all instances purusa (purisa here in Ma is the sole exception in this MS), a form which is also occasionally found in some of the other MSS, as here in R, in 196 in R and S, and in 380 in 8. purusa is most likely an innovation in these MSS introduced under the influence of Skt.

Page #272

259
46( 82 Ma, 83 Tp; 44 Bh, 46 R; 43 K, B, pa, 44 , 37 Y, P, 43 G, 46 5, T; 579 T; 342 S) In P as 37 and as 43 (pa), in the latter case with Chāyā but without commentary. Missing in W.
virahānalo sahijjai ekkaggāmapavāso
āsāvandhena vallahajanassa mãe maraŋam visesei.
saijjai y -- āsībaṁdhena Ma, Tp; P, po; T; (Bh, R, W, Y) -- dullaha R; vallabha K, vallaa P; pa -- ikkagjāma Bh -- u mãi Bha, māmi Bho.
The fire of separation from a beloved person is endured by the band of hope. (But) a man's absence from home while he is still in the same village, mother, is worse than death.
Garrez (1872: 213) compares the first line with Sakuntala, verse 91 (ed. Böhtlingk): garuan pi virahadukkham āsābandho sahābedi. For ās āvaṁdha earlier Weber (Abh., p. 88) referred to Meghadūta 11 (ed. Hultzsch).
For the present interpretation of ekkaggāmapavāsa, see Garrez, loc. cit. Compare ekkaggāma ccia pauttho in 129.
47( 83 Ma, 84 Tp; 45 Bh, 47 R; 193 K, 192 , 194 Y, 200 P, 190 G, 425 T; 343. S: 21 W) in k the commentary is also found between 43 (*46) and 44 (+48), numbered '44'(!). Missing in B, E, and x.
ajjaħ cea pauttho racchāmuhadeulacac-
ajja cia sunnai jāāi carāi amh aṁ ca hiaai.
ajja ccea Tp; ajja ceva R; ajja ccea K; T; ajja cia W -- a cia Ma, ajja ccia Tp; ajja cia K, ajaṁ cia P; ajja cchia T; ajja cvia S -- sunnaāim R; K (suttanaain), W, Y, P (suņamā in); S (sungain); W -- jāyām in R; jāāin K, P; T; S (Bh, X, Y, W) -- ratthā k -- devaulacaccarãi Ma, Tp; caccaradeulaimh Bha, deulacaccarain Bho, ocaccarāin R; X, Y, cattarāi K, P; devulacaccarāi T; cattaradeuldim s - - ahma ca Ma, Tp; amh an va Bh -- hiaāin K, P; T; S; (Bh, R, W, Y) --

Page #273

260
racchāmuhañ devaulla cattarai amha ca gihaai sungai W.
Today he went abroad and today everything has become empty: the streets, the temples, the crossroads and our hearts.
For the present position of the Gāthā as 47, see above, p. 161.
For the phrase ajja cea pauttho, see also 58.
deula (see also 109) is from devakula, with loss of the syllable -va-; see Hem. I 172. Compare rāula from rājakula, quoted by Bhāmaha, ad Var. IV 1. For deula, see also Turner, CDIAL 6524.
For caccara, see *39.
, 44 G, 47
,
;
48( 602 Ti, Tp; 46 Bh, 48 R; 44 K, B, Y, P, 45 125 S) Missing in T and W.
akkhudai pia hiae ditthe sarisammi gune
ann an mahildan am ramastassa sarisammi gune aTs arte.
na kkhudai Tp; akhudai Bho, avadai R; akkhadai K (akhkho), \, B, Y, P (akhkho); S -- mahilājana Tp; B -- maratassa Ti -- diththe K - - sarimmi Ti -- sarimmasammi Ti; sarise va S. -- afsate Ti; ad Tsamte K, Y, P, aisante y.
A man's dear mistress inevitably comes back to his mind when enjoying other women, namely when he sees a virtue equal to hers as well as when he doesn't!
akkhudai has been based on Bh and on Ti and Tp (for Tp na kkhudai, see below). akkhuda- seems a variant of āhuda- the past participle of which, āhudia, is mentioned in Deśfn. I 69 among the synonyms for padia 'fallen'. A NIA derivative is S. ākhuraņu 'to trip' (see Turner, CDIAL 1028: from *ākşuțati, or 1036: * ākhuţati, Pkt āhuda- 'to fall', *ākhoțayati, Pkt āhoda- 'to strike, to beat' (from 'to cause to fall')). The variant akkhadai shows the influence of askhalati; cf. the gloss ăskhalati given by PTtāmbara and Gangādhara (Retr., p. 360). na kkhudai

Page #274

261
found in Tp seems due to the connection made with khuda- 'to break' (see *40). The Chāyā in Ti reads na trušiti, and Madhavayajvamiśra has khuhai (khudai?) trutati vicchidyate (passive!), smaryata iti yāvat.
The emphasis in the first line is on the verb, which is found at the head of the sentence (see 8).
ats aṁte, from a + disante, shows a rule of internal sandhi (see also 224 and 388; cf. aittha in 414, 526 and 748). This seems to have to do with the relatively frequent occurrence of this combination, as a result of which it has come to be treated as one word. Compare such phrases as na ittha (502, v.1.) or na āṇasti (*2), for which a similar explanation should be given..
, 45 G, 48
5, T;
49( 603 Ti, Tp; 47 Bh, 49 R; 45 K, B, Y, P, 46 533 T; 682 S) Missing in W.
naipūrasacchahe jov- aniattāsu a rat-
vanammi diahesu niccapahiesu su putti kim daddhamānena.
;
sachahe K, B, P, satthahet; sariche T; sachahe S -- jauvvanammi Ti; jovvaemmi K -- divasesu piccapahiesu R; niccapahiesu diahesu K (diahaṁsu), Y (dihaesu), P (pahiesa), aipavasiesu dianesu , B; saiposiesu diahesu T; aipavasiesu diahesu S -- aniattāsu vi R; K, Y,.P; T;. anivuttāsu vi S -- rāsu V , rattin T -- puttali T -- dhaddha Tp; d'addha Bh, R; daththa K, dagha v', dattha ?, didhdha B, ddaththa Y, lacuna P.
What use is (this) cursed sulking, little girl, when youth (passes quickly) like the stream of a river, when the days are constantly on the move and the nights do not return:
For sacchahe 'similar to', see 5.
In MIA the initial d- of the Skt verbs dah- 'to burn' and dam) - 'to bite' becomes - Pāli invariably has dahati, Skt dāha occurs as dāha as well as as dāha. The past participle invariably occurs as daddha. Thus either the initial plosive or the medial cluster is retroflex, but

Page #275

262
not both at the same time (on the medial -ddh-, see below). Likewise . Pali has dasati, sandāsa and daṁsa, and daţtha and dāțhā. In the Sattasas one finds dakka 'bitten' in 237 (Ti, Tp, Bh, R, K,Y; dakka in
W, P, T, S, W) and 532 (Ma, Ti, Tp, Bh, R, W, S; dakka in Y, T, 54, W), dattha in 636 (Ma, TP, W, Y, T; çakka in R (Bh not available)), dādha 'fang' in 402 ( U, Y, S) and in 761 (Bh, R, T, S), and dasana 'tooth' in 522 (Ti, TP, BH, K, T, S, nahara in R, W, Y, P) (note that Var. II 35 explicitly mentions dasaņa!). In 105 dajjhihisi is found (but bhaṁjihisi in Ma and Tp), in 321 Jahai (dahai in T), in 373 qajjhai, in 401 dajjhasi (su), in 758 dahiūņa, in 193 qāna (Ma, Tp, Bh, V, Y; daha R, K, S), in 726 dahaņa (R; dahaņa s), and in 799 dahaņa (R). The past participle of daha- 'to burn' occurs frequently, in several cases, as here, meaning 'cursed'. Both daddha and daddha are found, which are variants. Thus, the MSS of the South-Indian recensions, Ma, Ti, Tp and T, where available, have qaddha throughout. The other MSS have daddha as well as daddha:
*49 gaddha Ti, Tp, P, T, S, daddha Bh, R, K, W, B, Y, 134 daddha Bh, K, , P, daddha R, Y, 149 daddha Tp (lacuna in Ma), Bh, K, V, S, daddha R, Y, P, 229 daddha Ma, Tp, Bh, k, , (utthe, but uprobably for da-), daddha R, 502 daddha Ma, Ti, Tp, V, T, daddha Bh, R, S, 511 daddha , daddha S (dutthu Bhi, ucha R), 558 daddha
U, S, daddha R, Y, 602 daddha Ma, Ti, Tp. V, T, S, 710 daddha S, daddha Y, 714 daddha Ma, Bh, T, S, daddha Tp(!), R, V, W, maddha Ti(!), 781 daddha R, daddha S.
This state of affairs partly coincides with the observation made by Hemacandra in I 217: - would be optional in daśana and dasta, and in dagdha and dāha, but obligatory in dahati and daśati. Note that Var. II 35 mentions only dasana.
In the Gathā under discussion the two branches are diametrically opposed, Ti and Tp having daddha and Bh and R daddha. It is almost certain that the latter, here as well as in all its other instances, is secondary, introduced in order to remove caddha from the text which with its many retroflexes must have appeared a linguistic monstrosity.
A different question concerns the origin of the initial d- in MIA

Page #276

263
Jaha- and casa-. H. Smith, quoted by Bloch (1934: 58) suggests that the
- represents a case of anticipatory assimilation to the medial -ddhand - th- in the respective past participles. On the other hand, we may have to do here with early instances of the variation occasionally seen in NIA between an initial d- and 4-. See H. dtthna/dtthna (CDIAL 6518) and P. dambhņā/ Jambhņā (CDIAL 6121). As far as NIA is concerned an explanation for the variation is probably that at the beginning of a word the phonemic distinction between a d- and a d- is less clear. The variation may also reflect a difference in the speech-situation, varying between quiet and un involved and emphatic and involved. Such a distinction seems interesting when applied to MIA daha- and dasa-, which both denote more or less violent or painful actions.
A second problem is the origin of the medial -ddh- in caddha. In fact the problem of its origin is largely ignored (not discussed by, for instance, Geiger, in § 42, 3, and Pischel, in § 222). AIG I S 236 suggests that Pāli daddha is from *dazgha, i.e. as if the -h in dahbelongs to the old palatal series. However, *dazdha should then have become *dādha (cf. 1Tdha from *lizdha). In fact, beside daddha a form *daddha seems to have existed. Thus, side by side with dārhna 'to burn' and dādha 'to blaze' or 'fire' H. has dadhna 'to burn' and dadhāna 'to blaze furiously'. P. dambhnā (da) 'to burn' (see above) supposes *daddha, which after laddha : lambha, has developed a stem dambha-. In my opinion the existence of daddha beside *daddha should be compared to Pali kathati 'to boil' beside Skt kvath- and Pkt kutthasu 'boil!' (for Pāli kațhati, see von Hinüber, 1979). The fact that it concerns two semantically closely related words can hardly be accidental. Probably the cerebralization of the medial dental(s), which appears to be optional, is contextual, signifying, for instance, the speaker's fear or horror of the actions denoted by the verbs.
50( 84 Ma, 85 Tp; 48 Bh, 50 R; 46 K, B, Y, P, 47 U, 46 G, 49 }, ; 160 T; 336 S) Missing in W.
kallan kira kharahiao taha vaddha bhaavai pise
pavas ihii pio tti suvvai janammi jaha se kallam cia na hoi.

Page #277

264
kallin S -- khala T -- hiyau Bh -- pavas ihai Bh, pavisahai R; pavasai K, pavasei Y, pavasehai p -- piu tti Bh -- Sungai K, u', sunai P -- manammi Ma -- vaddaha Ma; vaththa K, V, Y, vatha P; vadhdha S -- jahi Bha (jaha Bho) -- mey - kalan Ma; kallaa P; kallin S -- cia missing in P.
Pāda a quoted in Hem. II 186.
People say that at daybreak a lover will leave cold-heartedly. Venerable Night, grow so long that for him daybreak will never come!
For kalla 'dawn, next morning (or 'yesterday'; cf. Pāiyal. 211: hijjo (hyas) kallam), from Skt kālya or kalya, see Turner, CDIAL 3104.
For kira beside Skt kila, see *36.

Page #278


Page #279


Page #280

Index of words discussed in the notes
aisaṁta
akkhuda
aṇa
anusaa.
attā
alia
asoa
aha
44
uva(ha)
uvvasia
48
35
8
27
7
23
Thi(j) T 24
i (v.l. vi) 17
ia (v.1. of via)
1
ia (v.1. of ii) 15 ii (v.1. ia) 15 icchena (v.1. kicchega) ira (v.1. kira) 36 iva (v.1. of via) 1 ua(ha) 4
-utta 33
4
32
48
e (pronoun)
ehejja
0 32
omalia
kalamva
kalla 50
kicchena 33
kira 36
kisala 19
kudumvaa(/i)tthī (?)
kuravaa 8
kulavahū 38
kouhalla 21
kosamva
19
17
32
40
15
18
267
33

Page #281

268
khu 39 khuda- 40 gāmant 30 golā 36 ghaa(t)uppa 22 caccara 39 cattara 39 cihura 31 chana 42 chāht 37, 41 chikka 14 chiva- 14 jūra- (v.1. of jhūra-) 13 jhūra- (v.1. jūra-) 13 daddha 49 na āņa- 2 ņa uno 20 -naħdana 30 navari(n) (v.l. navara(ih)) 15 navara() (v.l. of navari(m)) pavalaa 28 . na hu 39 . nivvudda- 40 nisāvia (v.1. nisāmia) 29 nomālia 32 nohalia 6 tamsa 10 tai 32 tae 32 taŋŋaa 19 tatt 2 taha vi hu 39 tālūra 40 tui 32 tue 32 tuppa 22
15

Page #282

tuha 36
daddha
dadhal
11
dasia
18
diara
38
digha. 11
dūme
11
49
de (pronoun)
de (particle)
deara
deula
dega
devara
donga (m)
dogha (m)
dohala
dho (v)a
patthT 11
padaal
18
28
47
15, 29
28
putth
purisa
27
15
27.
18
15
padhamullaa 15
panavia (v.1. pagamia) 22
parähutta 35
parimusaga
26
16.
pahara 28
pahuppa- 7
paada (v.1. of paua)
padala (11)
13
piva (v.1. of via) 1
11
45
purusa 45
bola (v.1. of vola-)
bhamadana 40
bhara- 22
bhisinT 4
'maia 44
2
21
269

Page #283

mailia
majjhaara
mar agaa
malla
maha
reha
lagga
ladahavila
lua 8
vacca
14
27
28
4
māŋailla 27
mā hu
39
miva (v.1. of via) 1
rakkha
41
4
21
19
vajjhapadaha vannagghaa(t)uppa
valaa 4
vola- 21
volaia 21
sarkhasippT
sambhara
sacchaha 5
saajjia 39
sal hagijja
sippi 4
silahagijja
sungaia
somala
hu 39
-hutta
2
vālumkī 10
via (v.1. fa, iva, piva, miva, viva)
viaga 38
vilaa
7
viva (v.1. of via) 1
vudda- 40 vesavila 7
26
36
7
35
29
4
29
22
12
121
270

Page #284

271
Index of grammatical points discussed
-a
-accia + -eccia Accusative Anusvāra
Apabhras a
: -a tti + -etti
-a va * -e va bhūa
in the South Indian MSS ending of the instrumental and genitive of feminine I-stems, e.g. kodta feminine gender-marker, e.g. erisā e.g. kuņamtā e.g. seccia denoting the time at which, e.g. oāsaḥ added after the endings of the plural ending in -i and -a, e.g. -āim, -ehiń and -āņa
p. 168 ff. early Apa. forms in Pkt, e.g. aha
23 e.g. tai
32 Apa. features in Gathās added later in the North-Indian branch, e.g. bhamādana 40, p. 162 e.g. se tti
23 e.g. devae va
23, 32 functioning as a tense-marker, e.g. mand anañ bhūam position of the particle of comparison non-existent in Pkt consisting of more than two members, in the neutre singular, e.g. rairahasiharadhaa in the South Indian MSS e.g. pāņiggahana ccia development in Pkt to a-stems position at the beginning of the sentence 7 retention in a foreign loan-word, e.g. maragaa position of ii, in contrast to that of (t)ti, which is an enclitic possible emphatic function, e.g. māņailla (cf. -ullaa)
Comparison Deponentia Dvandva
-eccia + -a ccia e-stems Finite verb -9
ii (Skt iti)
-illa

Page #285

272
-jjh-/-jj
Karmadhāraya
-kka Lengthening
numbers
confusion, e.g. vajjha/vajja e.g. saajjhia/saajjiā
bilingual', e.g. sankhasippi e.g. (vannag) ghaa(t)uppa verbal suffix of initial syllable, e.g. āhi(j)at in the neighbourhood of other nasals above 'one', loss of gender distinction e.g. sea ccia e.g. padaha va e.g. murave va functioning as an action noun in conditional clauses internal sandhi in compounds influence of Skt on the spelling influence on the MSS broken-up functioning as a personal pronoun
23
-occia + -a ccia -Ova + -a vva -Ova + -evva Past participle Present participle Sandhi Sanskrit Saurasen 1 Sentence
23 22
26
p. 164 p. 163 27, 39 20
si
<6
taha...sariccha -ullaa -V
emphatic, e.g. padhamullaa secondary insertion or deletion in -ua- (or -uva-)
p. 165 ff.

Page #286

273
Bibliography
AKM
Abbreviations
Abhandlungen für die Kunde des Morgenlandes BB Bezzenberger's Beiträge zur Kunde der Indogermanischen Sprachen BSOAS · Bulletin of the School of Oriental and African Studies IHQ The Indian Historical Quarterly IJJ Indo-Iranian Journal
Indian Linguistics ISt Indische Studien JA Journal.Asiatique JAOS Journal of the American Oriental Society JOIB Journal of the Oriental Institute of Baroda KZ (Kuhn's) Zeitschrift für vergleichende Sprachforschung STII Studien zur Indologie und Iranistik ZDMG Zeitschrift der Deutschen Morgenländischen Gesellschaft
Alsdorf, L., 1935-37: 'The Vasudevahindi, a Specimen of Archaic Jaina
Māhārās tri', BSOAS VIII, pp. 319-333 (Kleine Schriften, pp. 56-70) Alsdorf, L., 1937: Apabhraía-Studien, AKM XXII 2, Leipzig Alsdorf, L., 1950: "Pancatañtra-Miszellen', ZDMG 100, pp. 356-361
(Kleine Schriften, pp. 586-91) Alsdorf, L., 1974: Kleine Schriften, ed. by A. Wezler, Glasenapp
Stiftung Band 10, Wiesbaden Andersen, D., 1901: A Pāli Glossary, repr. Delhi, 1979 Berger, H., 1955: Zwei Probleme der mittelindischen Lautlehre, München The Bhattikāvyam of Bhatti with the Commentary (Jayamangala) of Jaya
mangala, ed. by V. N. S. Joshi and K. V. Joshi, Bombay, 1900 Bloch, J., 1934: L'Indo-aryen. Du Veda aux temps modernes, Paris Bloomfield, M., 1899: The Atharvaveda, Encyclopedia of Indo-Aryan
Research II 1 B, Strassburg Brough, J., 1962: The Gāndhārī Dharmapada, London Burrow, T. and Emeneau, M. B., 1961: A Dravidian Etymological Dictionary,
repr. of the 1st ed., Oxford, 1966 Candralekhā: Rudradāsa's Candralekhā, ed. by A. N. Upadhye, Bharatiya
Vidya Series No. 6, 2nd ed., Bombay, 1967 The Dasarūpa, or Hindu Canon of Dramaturgy, by Dhananjaya; with the

Page #287

274
Exposition of Dhanika, the Avaloka, ed. by F. E. Hall, Bibliotheca
Indica, New Series Nos 12, 24 and 82, Calcutta, 1865
The Desinamamala of Hemacandra, ed. by R. Pischel, Bombay Sanskrit Series No XVII, 2nd ed., Bombay, 1938
The Dhvanyaloka of Anandavardhanacharya, with the Commentary of Abhinavaguptacharya, ed. by Pandit Durgaprasad and K. P. Parab, Kavyamala 25, Bombay, 1891
Dundas, P., 1981: 'Some linguistic Notes on the SattasaT', JOIB XXX, pp. 205-13
Emeneau, M. B., 1968: 'Prakrit etymological Notes', IL 29, pp. 30-33 Emeneau, M. B., 1969: 'Sanskrit Syntactic Particles kila, khalu,
nunam', IIJ XI, pp. 241-68
Fick, R., 1897: Die sociale Gliederung im nordöstlichen Indien zu
Buddha's Zeit. Mit besonderer Berücksichtigung der Kastenfrage.
Vornehmlich auf Grund der Jātaka dargestellt, Kiel, repr., Graz, 1974 Gairola, Ch. K., 1956: 'Die Anfänge der Satavahana-Dynastie in Indien',
ZDMG 106, pp. 155-165
Garrez, G., 1872: Review of Ueber das Saptasatakam des Hala. Ein Beitrag zur Kenntniss des Prakrits, ed. by A. Weber, AKM V 3, Leipzig, 1870, in JA, pp. 197-220
The Gaudavaho, a Prakrit historical Poem by Vakpati, 1st ed. by Shankar Pandurang Pandit, 2nd ed. by Narayan Bapuji Utgikar, Bombay Sanskrit and Prakrit Series No. XXXIV, Poona, 1927
Geiger, W., 1916: Pali Literatur und Sprache, Grundriss der Indo
Arischen Philologie und Altertumskunde, 1. Band, 7. Heft, Stassburg Ghosal, S. N., 1959: 'The Euphonic Glide H in Prakrit', JOIB IX, pp.
256-259
Godavarman, L., 1935: 'The Change of a to e in the Indo-Aryan Loan
Words of Malayalam', BSOAS VIII, pp. 559-562
Gonda, J., 1952: Remarques sur la Place du Verbe dans la Phrase active et
moyenne en Langue Sanscrite, Utrecht.
Gonda, J., 1968: 'Abbreviated and inverted nominal Compounds in Sanskrit', in Pratidanam, Indian, Iranian and Indo-European Studies Presented to F. B. J. Kuiper on His Sixtieth Birthday, ed. by J. C. Heesterman et al., Janua linguarum. Studia Memoriae Nicolai van Wijk dedicata Series Major 34, the Hague, pp. 221-246 (Selected Studies 111

Page #288

275
43-68) Gonda, J., 1975: Selected Studies, in five volumes, Leiden Handiqui, K. K., 1976: Pravarasena's Setubandha, translated into English
with extracts from unpublished commentaries, critical notes and an
introduction, Prakrit Text Series No. 20, Ahmedabad von Hinüber, 0., 1979: 'Pāli kathati: ein Beitrag zur Überlieferungs
geschichte des Theravāda-Kanons', IIJ 21, pp. 21-26 Jacobi, H., 1886: 'Unregelmässige Passiva im Prākrit', KZ 28, pp. 249
256 (Kleine Schriften I, pp. 110-117) Jacobi, H., 1898: 'Der Akzent im Mittelindischen', KZ 35, pp. 563-578
(Kleine Schriften I, pp. 73-88) Jacobi, H., 1970: Kleine Schriften, ed. by B. Kölver, in two volumes,
Glasenapp-Stiftung Band 4, 1-2, Wiesbaden Janert, K. L., 1965: An Annotated Bibliography of the Catalogues of
Indian Manuscripts, Part I, Verzeichniss der orientalischen Hand
schriften in Deutschland, Supplementband 1, Wiesbaden : The Kalpasūtra of Bhadrabāhu, ed. by H. Jacobi, AKM VII 1, Leipzig,
1879, repr. Nendeln, 1966 Kas avaho: Rāma Pānivāda's Kais avaho, ed. by A. N. Upadhye, 1940
2nd issue Delhi, 1966 Karpūramañjart: Rajasekhara's Karpūramañjart, ed. by S. Konow and Ch. R.
Lanman, Harvard Oriental Series Vol. IV, 1901, repr. Delhi, 1963 Katare, s. L., 1952: 'An inscribed Sculpture inspired by Hāla's Gatha
saptaśati', IHQ XXVIII, pp. 379-385 Kāvyādarśa: Dangin's Poetik (kāvjādarśa), ed. by O. Böhtlingk, Leipzig,
1890 Kavyaprakāśa of Mammaţa, with the Bālabodhint of the late V. R.
jhalakikar, ed. by R. D. Kamarkar, 6th ed., Poona, 1950 Keith, A. B., 1928: A History of Sanskrit Literature, Oxford Krishnamachariar, M., 1970: History of Classical Sanskrit Literature,
1st repr., Delhi Der Kumārapālapratibodha. Ein Beitrag zur Kenntniss des Apabhr ansa und
der Erzählungs-Literatur der Jainas, ed. by L. Alsdorf, Alt- und
Neu-Indische Studien 2, Hamburg, 1928 Lienhard, S., 1975: 'Sur la Structure poétique des Therather Tgāthā',
JA CCLXIII, pp. 375-396 Lévi, s., 1902: 'Sur quelques Termes employés dans les Inscriptions des
Kşatrapas', JA XIX, pp. 94-125

Page #289

276
Lévi, S., 1963: Le Théatre Indien, 2nd ed., Paris Lüders, H., 1940: Philologica Indica, Göttingen Lüders, H., 1942: 'Von indischen Tieren', ZDMG 96, pp. 23-81
(Kleine Schriften, pp. 490-548) Lüders, H., 1973: Kleine Schriften, ed. by 0. von Hinüber, Glasen app
Stiftung Band 7. Wiesbaden Mānava Dharma-śāstra, the Code of Manu, ed. by J. Jolly, London, 1887 Mayrhofer, M., 1955: 'Das Problem indo-germanischer Altertümlichkeiten
im Mittelindischen', in Studia Indologica. Festschrift für W. Kirfel,
ed. by 0. Spies, Bonn, pp. 219-241 Mayrhofer, M., 1956-1980: Kurzgefasstes etymologisches Wörtenbuch
des Altindischen, in four volumes, Heidelberg Meghaduta: Kalidasa's Meghaduta, ed. by E. Hultzsch, Prize Publications
Fund Vol. III, London, 1911. Mirashi, V. V., 1947: 'the Date of the Gathāsaptaśatt', IHQ XXIII, pp.
300-310 Monier-Williams, M., 1970: A Sanskrit-English Dictionary, New Edition,
Oxford Nāyādhammakahão, in Anga Suttani, ed. by Muni Nathamala, Vol. III, pp.
1-391, Ladnun, 1973 Nitti-Dolci, L., 1938: Les Grammairiens Prakrits, Paris The PaiyalachchT Nāmamālā, a Prakrit Kosha, by Dhanapāla, ed. by G.
Bühler, BB 4, 1878, pp. 70-166 Panhāvāgaraņāin, in Anga Suttāni, ed. by Muni Nathamala, Vol. III,
pp. 635-713, Ladnun, 1973 Pargiter, F. E., 1913: The Purāna Text of the Dynasties of the Kali Age,
Oxford Pischel, R., 1900: Grammatik der Prakrit-Sprachen, Grundriss der Indo
Arischen Philologie und Altertumskunde, 1. Band, 8. Heft, Strassburg,
repr. Hildesheim, 1973 The Prakrta-Prakasa or the Prākrt Grammar of Vararuchi, with the Com
mentary (Manoramā) of Bhāmaha, ed. by E. B. Cowell, 3rd ed. Calcutta,
1962 Prakrtaprakāśa of Vararuci with the Commentary of Rāmapāņivāda, ed. by
C. Kunhan Raja and K. Ramachandra Sarma, The Adyar Library Series No. 54, Madras, 1946

Page #290

277
Prākrtasarvasva: Mārkandeya's Prākrta-Sarvasva, ed. by K. Ch. Acharya,
Prakrit Text Society No. XI, Pkt Grammar Series No. 1, Ahmedabad, 1968 Printz, W., 1921: Bhāsa's Prākrit, Frankfurt Raghavan, V., 1968-... : New Catalogus Catalogorum. An alphabetical Register
of Sanskrit and allied Works and Authors, Madras Rama Rao, M., 1971: Studies in the early History of Andhradeśa, Madras Renou, L., 1930: Grammaire Sanscrite, Paris, 2nd ed., 1968 Renou, L. and Filliozat, J., 1947-1953: L'Inde Classique, in two volumes,
Paris-Hanoi Rhys Davids, T. W. and William Stede, 1921-1925: The Pali Text Society's
Pali-English Dictionary, repr., London, 1972 The Sahityadarpaņa or Mirror of Composition, A Treatise on Literary
Criticism, by Visvanatha Kavirāja, ed. by E. Roer, Bibliotheca Indica
Vol. X, Calcutta, 1851 Śakuntala: Kalidasa's Śakuntala, ed. by 0. Böhtlingk, Bonn, 1846 Sakuntala: Kalidasa's Sakuntala. The Bengal1 Recension, ed. by R. Pischel, • Kiel, 1877 Sandesara, B. J., 1972: 'Ājada's Commentary on the Gathās aptaśatt of
Hāla', JOIB 22, pp. 45-47 Ueber das Saptaśatakam des Hāla. Ein Beitrag zur Kenntniss
des Prāçit, ed. by A. Weber, AKM V 3, Leipzig, 1870, repr. Nendeln,
1966 Das Saptaśatakam des Hāla, ed. by A. Weber, AKM VII 4, Leipzig, 1881,
repr. Nendeln, 1966 Saptaśatīsāra with Bhāvad Tpikā of Vema Bhūpāla, along with the
Chappannaya-Gāhão (text and Chāyā), ed. by A. N. Upadhye, Shivaji
University Sanskrit and Prakrit Series Vol. III, Kolhapur, 1970 Schneider, U., 1954: 'Acht Etymologien aus dem Aggañña-Sutta', in
Asiatica. Festschrift Friedrich Weller, ed. by J. Schubert and
U. Schneider, Leipzig, pp. 575-583 Schubring, W., 1955: 'Jinasena, Mallinātha, Kalidāsa', ZDMG 105, pp.
331-337 (Kleine Schriften, pp. 314-320) Schubring, W., 1977: Kleine Schriften, ed. by K. Bruhn, Glasenapp
Stiftung Band 13, Wiesbaden Schwarzschild, L. A., 1956: 'Some Forms of the Absolutive in Middle
Indo-Aryan', JAOS 76, pp. 111-115

Page #291

278
Schwarzschild, L. A., 1958: 'Prakrit Thakka, "Tired", IL 19, Sir Ralph
Turner Jubilee Volume I, pp. 311-318
Schwarzschild, L. A., 1962: 'Ghummira, Gholira "agité, branlant",
JA CCL, pp. 65-75
Schwarzschild, L. A., 1962:
"First", "Second" and "Third" in Middle
Indo-Aryan', JAOS 82, pp. 517-522 Setubandha: Ravanavaha oder Setubandha, ed. by S. and P. Goldschmidt,
in two volumes, Strassburg-London, 1880 and 1884
Speijer, J. S., 1886: Sanskrit Syntax, Leiden, repr. Delhi, 1973 Sprüche: Böhtlingk, 0., 1863-1864: Indische Sprüche. Sanskrit und
Deutsch, in two volumes, S Petersburg
Sriramamurti, P., 1972: Contributions of Andhra to Sanskrit Literature, Andhra University Series No. 105, Waltair Śrigaraprakasa: Maharaja Bhojaraja's Sringara Prakasha, ed. by G. R.
Josyer, in four volumes, Mysore, 1955-n.d
Sternbach, L., 1974: Subhasita, gnomic and didactic Literature, part of Vol. IV of A History of Indian Literature, ed. by J. Gonda, Wiesbaden The Subhasitaratnakoşa compiled by Vidyakara, ed. by D. D. Kosambi and
V. V. Gokhale, Harvard Oriental Series 42, Massachusetts, 1957 Thieme, P., 1937: 'Uber einige persische Wörter im Sanskrit', ZDMG 91,
pp. 88-146 (Kleine Schriften I, pp. 278-336)
Thieme, P., 1938: Der Fremdling im Rigveda. Eine Studie über die
Bedeutung der Worte ari, arya, aryaman und arya, AKM XXIII 2, Leipzig, repr. Nendeln, 1966.
Thieme, P., 1962: 'Chess and Backgammon (Tric-trac) in Sanskrit Literature', in Indological Studies in Honor of W. Norman Brown, ed. by E. Bender, Amarican Oriental Series Vol. 47, New Haven, pp. 204-216 (Kleine Schriften II, pp. 413-425)
Thieme, P., 1971: Kleine Schriften, ed. by G. Buddruss, in two volumes,
Glasenapp-Stiftung Band 5, 1-2, Wiesbaden
Thieme, P., 1975: 'Kranich und Reiher im Sanskrit', StII 1, pp. 3-36 Tieken, H., 1978: 'A formal Type of Arrangement in the Vulgata of the Gathasaptasatт of Hala', Stil 4, pp. 111-130
Trenckner, V., et al., 1924-: A Critical Pali Dictionary, Copenhagen. Tripathт, Ch., 1977: 'Katalogisierungstradition der Jainas', ZDMG, Supplement III 2, XIX. Deutschen Orientalistentag, Vorträge,

Page #292

279
pp. 992-1001 Turner, R. L., 1928: 'Romani les and Sanskrit tasya', BSOAS V, pp.
43-51 (Collected Papers, pp. 310-318) Turner, R. L., 1966: A Comparative Dictionary of the Indo-Aryan
Languages, London Turner, R. L., 1975: Collected Papers, ed. by J. Brough, London Vajjalagga: Jayavallabha's Vajjālaggam with the Sanskrit Commentary
of Ratnadeva, ed. by M. V. Patwardhan, Prakrit Text Society Series
No. 14, Ahmedabad, 1969 Vivāgasūyam: in Arga-Suttāni, ed. by Muni Nathamala, Vol. III, pp.
715-813, Ladnun, 1973 Wackernagel, J. and, Debrunner, A., 1957-: Altindische Grammatik,
Göttingen Warder, A. K., 1974: Indian Kāvya Literature II: Origins and Formation
of the classical Kavya, Delhi Watts, G., 1972: A Dictionary of the Economic Products of India, in six
volumes, 2nd repr., Delhi Weber, A., 1872: 'Ueber das Saptaśatakam' ZDMG 15, pp. 735-745. Weber, A., 1883: 'Ueber Bhuvanapāla's Commentar zu Hala's Saptaśatakam',
ISt. XVI, Leipzig, pp. 1-204, repr. Hildesheim, 1973.

Page #293


Page #294

281
Nederlandse samenvatting
Het voorliggende werk is een voorstudie tot een kritische tekstuitgave van Hala's Sattasat, een anthologie, daterend uit het begin van onze jaartelling, van 700 Prākrit verzen erotisch van inhoud. De eerste editie van deze tekst, van A. Weber, dateert uit het einde van de vorige eeuw. Weber beschikte over 18 handschriften (HSS). De tekst bleek overgeleverd in zes verschillende recensies. Voor de huidige tekstuitgave wordt gebruik gemaakt van Webers materiaal, uitgebreid met drie tot nu toe onuitgegeven HSS uit Zuid-India. Deze laatste HSS bevatten de tekst van een zevende. recensie.
In deel I van dit boek wordt verslag gedaan van een onderzoek naar de onderlinge afhankelijkheid van de verschillende HSS (Stemma). De belangrijkste conclusie is dat de HSS zijn onderverdeeld in een Zuidindische en een Noord indische tak. Beide gaan uiteindelijk terug op één enkel Noordindisch HS, dat echter zelf verloren is gegaan. In de : tekstuitgave wordt getracht de tekst van dit ene origineel te reconstrueren. Een andere conclusie van dit onderzoek is dat de tekst van de afzonderlijke verzen van het verloren origineel in vele opzichten het best bewaard is gebleven in de Zuidindische tak, maar de volgorde van de verzen in de Noordindische tak. A
Deel II bevat een reconstructie van de tekst van de eerste 50 verzen van het bovengenoemde origineel. Hieraan vooraf gaan twee korte hoofdstukken. Het eerste hoofdstuk handelt over de vraag voor welk publiek de tekst was geschreven en/of verzameld. Gezien de superieure en patroniserende kijk die blijkt uit de verzen, op het leven op het platteland en op dat van de 'werkende klasse' moet worden aangenomen dat de tekst zich richtte tot een publiek van rijke stedelingen en hovelingen. Verder wordt gespeculeerd over de vraag waarom de dichter of dichters een Middelindisch dialect gebruikten in plaats van het 'hoof se' Sanskrit. Een mogelijk antwoord hierop is dat men zo poogde de woorden van de bewoners van het platteland, in wier mond de verzen zijn gelegd, een boers tintje te geven.
Het tweede hoofdstuk handelt over de gevolgde tekstkritische methode. Sommige principes vloeien voort uit de conclusies van het stemma. In een aantal gevallen echter werkt het stemma niet, met name daar waar

Page #295

282
het variaties betreft in de spelling van de woorden. Op dit punt is de invloed van het Sanskrit in elk stadium van de handschriftoverlevering aantoonbaar. Dit leidt tot het principe dat in geval er twee spellingsvarianten beschikbaar zijn, die waarvan de vorm het meest afwijkt van het Sanskrit, in de tekst moet worden opgenomen, ongeacht de positie in het stemma van het HS waarin deze wordt gevonden.
In het derde hoofdstuk volgt de eigenlijke tekstuitgave (verzen 1-50), met vertaling en noten. De vertalingen wijken in vele gevallen af van die van Weber. De verschillen worden voor een groot deel bepaald door een strikt volgen van de Prākrit tekst zelf, terwijl Weber zich juist vaak baseerde op de interpretaties zoals gevonden in de mee-overgeleverde commentaren. Deze commentaren blijken echter daartoe te onbetrouwbaar, zoals meer dan eens in dit proefschrift in de noten bij de vertalingen wordt aangetoond. In deze zelfde noten wordt verder ingegaan op, onder andere, problemen betreffende de reconstructie van de tekst, de oorsprong van de belangrijkste varianten en de betekenis van de woorden.

Page #296

283
Curriculum vitae
Geboren op 1 april 1952 te Lijnden, behaalde de auteur van dit proefschrift in 1970 het einddiploma Gymnasium Alpha aan het Bonaventura Lyceum te Leiden. Vanaf 1970 studeerde hij Indo-iraanse talen en culturen te Leiden. In 1973 behaalde hij het kandidaatsexamen; in 1975 het doctoraalexamen met als hoofdvak Sanskrit en Middelindisch en als bijvak Tamil. Sinds 1 januari 1976 is hij werkzaam als wetenschappelijk medewerker voor het Sanskrit en het Middelindisch aan de Rijksuniversiteit te Leiden. Naast zijn onderwijs en onderzoek houdt hij zich bezig met het vertalen van Tamil -, Sanskrit- and Prākritliteratuur in het Nederlands.

Page #297


Page #298
on कमीटनामधारणपतिनापटीयमाना तापदायमनिपातितानिपतोमेरठा धमधामगरलंयरोवनियतिईआणि द्वारगदानमसंपदाधिकमासखण्डा