Vagbhata's Astangahrdayasamhita
.jpg)
.jpg)
Vagbhata's Astangahrdayasamhita Aṣṭāṅgasaṅgraha (अष्टाङ्गसंग्रह) and the Aṣṭāṅgahṛdayasaṃhitā (अष्टाङ्गहृदयसंहिता)
THE FIRST FIVE CHAPTERS OF ITS TIBETAN VERSION
ABHANDLUNGEN
Ft'R DIE KUNDE DES MORGENLANDES
IM AUFTRAG DER DEUTSCHEN MORGENLANDISCHEN GESELLSCHAFT HERAUSGEGEBEN VON HERBERT FRANKE
XXXVII, 2
VAGBHATA'S
m
ASTANGAHRDAYASAMHITA
• • • o
THE FIRST FIVE CHAPTERS OF ITS TIBETAN VERSION
EDITED AND RENDERED INTO ENGLISH ALONG WITH THE ORIGINAL SANSKRIT
BY
CLAUS VOGEL
ACCOMPANIED BY A LITERARY INTRODUCTION
AND A RUNNING COMMENTARY
ON THE TIBETAN TRANSLATING-TECHNIQUE
DEUTSCHE MORGENLANDISCHE GESELLSCHAFT
KOMMISSIONSVERLAG FRANZ STEINER GMBH WIESBADEN 1965
AUe Eeohte vorbehalten Ohne ausdrttckliche Genehmigung des Verlages ist es nicht geatattet, das Werk oder einzelne Teile daraus nachzudrucken oderauf photomechanischem Wege (Pbotokopie, Mikrokopie usw.) zu verrtelfaltigen. © 1965 by Franz Steiner Verlag GnibH, Wies- baden. Als Habilitationsschrift auf Empfehlung der Philosophischen Eakultat der Philipps-TJniversitSt Marburg gedruckt mit Unterstfltzung der Deutschen Eor- schungsgemeinsobaft. Gesamtherstellung: Wiesbadener Graphische Betiiebe GmbH Printed in Germany
Contents
Preface VII
A. Introduction 1
I. Vagbhata's works in general and the question of their authenticity .. 1
1. Traditional views 1
2. Cordier's view 2
3. Jolly's view 3
4. Hilgenberg-Kirfel's view 4
II. Vagbhata's circumstances of life 7
5. His origin 7
6. His date 8
III. Vagbhata's Astangahrdayasamhita 10
i. Its commentaries and their authors 10
7. Synopsis 10
8. Arunadatta 12
9. Indii 13
10. Candranandana 15
11. Hemadri 16
12. Others 16
ii. Its later history 17
13. Excerpts and summaries 17
14. Manuscripts 17
15. Editions 18
iii. Its Tibetan version 18
16. Eeason why translated and canonized 18
17. The colophon 18
18. The translating team 19
19. Rin-chen-bzan-po 20
20. Xylographs used for this edition 21
a. Chone 21
b.Derge 22
c. Narthang 23
d. Peking 23
21. Their stemmatical relationship 24
22. Principles followed in this edition 33
23. Corruptions 33
24. Sanskrit variants to be inferred 34
VI Contents
25. Interpolations to be detected 35
20. Remarkable interpretations 35
27. Some aspects of the translating-technique 36
a. Circumlocution 36
b. Redundancy of speech 37
c. Fluctuation of terminology 38
d. Ambiguity of nomenclature 39
e. Incongruity of expression 39
f. Replacement of unsuited words 40
g. Verbalization 40
h. Handling of word-order 41
i. Sanskritisms 42
j. Faithfulness 43
B. First chapter 44
C. Second chapter 82
I). Third chapter 121
E. Fourth chapter 171
F. Fifth chapter 198
G. Bibliography, abbreviations, sigla 265
H. Appendix (to Introd. § 4) 281
Preface
After Beineich Latjfee's Beitrage zur Kenntnis der Tibetischen Medizin (Berlin, 1900) had been for decades the only comprehensive if preliminary work on the topic, the study of Lamaist healing-art has received of late a new and unexpected impulse from three publications, each of which is meritorious in its own individual way : Cyrill von Korvin-Keasinski's strictly scientific Tibetische Medizinphilosophie (Zurich, 1953), Theodoe Btjeang's mainly popular Tibetische Heilkunde (Zurich, 1957), and Ilza Veith's richly illustrated Medizin in Tibet (Leverkusen, 1960). What is still a desideratum— though it should properly be the starting-point of any such research— is a complete edition and translation of the rOyud bzi, the standard book of Tibetan medicine, which is supposed to have been adapted from a now lost Sanskrit original by the Kashmirian physician Candranandana about the middle of the 8th century A.D., and which is said to have been written by none other than Kumarajivaka, the famed contemporary of Buddha Sakyamuni 1 . The indispensable condition, however, of a correct understanding of the rGyud bzi is an intimate knowledge of Tibetan medical terminology, which in its turn can be acquired only by closely comparing an extant medical Sanskrit text of some length with its Tibetan counterpart. No work seems better suited for this pur- pose than Vagbhata's Astmgahrdayasamhita, the only representative description of Indian medicine incorporated into the Lamaist canon.
The plan to bring out a critical edition of the Tibetan Astmgahr-dayasamhita, a specimen of which— along with the original Sanskrit, a literal translation, and a running commentary on the translating- technique— is now placed before the learned public, was conceived in the winter of 1958—59, during a prolonged stay at the International Academy of Indian Culture in New Delhi, where the present writer made a complete transcript of the text from the Peking xylograph: a tedious job that was, however, well paid in the end since the Japanese photomechanical reprint, like the Narthang xylograph, turned out to be difficult to read in many places. It is intended to publish all 120 chapters in Sanskrit, Tibetan, and English and to prepare a trilingual glossary of the medical terminology that may serve, as it were, for a master-key to the locked treasures of Lamaist healing-art.
1 This is not to answer beforehand the question of its true provenance and authorship, on which now see Unkeio in Kobvtn-Kkasikski's Medizinphibsophie ■p.xvni sq.
VIII Preface
In concluding, the author wishes to express his sense of obligation to Prof. Dr. Wilhelm Rac for the indefatigable support given at all stages of this work; to Dr. Wolfgang Voigt of the State Library, Marburg, and the staff of the India Office Library, London, for the prompt services rendered in the procurement of urgently needed books; and, last but not least, to Mr. William Fielding Hatton for a styUstic check-up on the Introduction.
Marburg, June 25, 1963 Clatts Vogel
Introduction
1. In the medical literature of the Hindus, three voluminous works generally go by the name of Vagbhata 1 ; they are the Astangahrdaya-samhita [[[Ah]].] or "Collection of the Essence of the Octopartite (Science)," the Astahgasamgraha [As.] or "Compendium of the Octopartite (Sci- ence)," and the Rasaratnasamuccaya or "Jewel Accumulation of Metallic Preparations 2 ." While the genuineness of the Ah. and As. has never been seriously questioned, the Rasaratnasamuccaya is in some manuscripts attributed to one Nityanatha or Asviniktimara and at any rate disagrees so much in content and style from those two works that Vagbhata* s authorship appears rather dubious 3 .
In return for this, the Ah. and As. pose another problem, namely, whether they originate with one and the same author or with two differ- ent authors who are namesakes. For though the colophons of both works give Vagbhata, the son of Simhagupta, as the author, Indian commentators often refer to the former simply as Vagbhata, but to the latter as Vrddha-Vagbhata 4 , a term that is usually interpreted as "elder
1 Other forms are Babhata (Bengal) and Vahata (South India). The Tibetans have reproduced the name with Pha-gol or Pha-khol, which comes nearest to the Bengali spelling. For details, see notes on Opening Statement.
2 Vagbhata is further credited with an autocommentary on his Ah., which is extant only in Tibetan (cf. Hum, SPAW 1895 p. 270), as well as with the following works: Bahatanighantu, Bhavaprakaia, Padarthacandrika, Sastradarpana, Satasloki, Vagbhatiya, and Vamanakalpa (cf. Coedier, Vagbhata pp. 7 & 16). It is virtually certain, however, that the bulk of these works originate from different authors, the name Vagbhata being fairly common after all (of. Godb, Studies
Ip.l71*g.)-
3 Cf. COBDIEE, Vagbhata p. 7 sqq.; Jolly, Medicin p. 4. The authenticity has
recently been defended by Bhattachaeyya (ABOEI xxviii p. 121 sq.), whose argumentation centres around the fact that Candrata in his Yogaratnasamuceaya (gadasantivarga, fol. 4b) ascribes a passage from the Rasaratnasamuccaya (xxi 145—149) to one Rasa-Vagbhata. But how "the inclusion of the Rasa-Vagbhata along with Vagbhata . . . and Vrddha vahada . . . among the sources of Candrata is a most important discovery on the problem of the identity of the different Vagbhatas" in the way Bhattachaeyya has it remains a complete mystery. Any disinterested person would take it for evidence to the contrary, namely, that the three of them are only namesakes and have actually nothing to do with one another.
4 It should be observed, however, as Bhattachaeyya (ABORI xxviii p. 116 sqq.) rightly points out, that this policy was adopted only by later scholiasts, say from Hemadri onwards, and that Arunadatta, Indu, and Candranandana leave
1 Vogel, Vagbhata
2 Introduction
Vagbhata." This distinction between a junior and senior writer of the same name is at first sight corroborated by a passage from the Ah. (VI 40.79) which expressly states:
artawjavaidyakainaltodadkimantkanena
yet 'sUdigasamgrahamahamrtarasir aptah j
'tasntftd analpaphalam alpasamudyamanarri
prtfyartlutm dad uditam prthag em tantram //
The big nectar heap of the Astangasamgraha, which (was) obtained
by churning the big ocean of the octopartite medicine— from this
the present work (was) produced separately, for the joy of those who
show (but) little zeal, (being) of no little use.
2. Once the precedence of the As. seemed established, further osten- sible evidence was adduced in its favour. Cokdieb. 1 , who was the first to tackle the problem methodically, relied on the following five arguments . ( 1 ) The title Astangasamgraha is identical with the one quoted by ancient Arab sources 2 as Kitab Asankar or Astankar. (2) The As. is the only work in which the traditional division of medicine into eight parts is fully implemented. (3) The mixture in the As. of prose and verse reminds one of Caraka and Susruta, while the Ah. consists only of verses. (4) Four stanzas of the As. that do not occur in the Ah. have been included in
no doubt about the fact that they consider the authors of both works identical. A few examples will suffice to prove this.
Arunadatta, in commenting on the cursory treatment which honey is given in Ah. I 5. 51 sqq., remarks: tatha, hy ayam eva tantrakarah samgrahe madhuno bheddn akhyat "thus, for instance, the present writer in his Compendium specified various sorts of honey."
Indu, in pondering over the different effect on digestion which radish is accorded in As. I 17 ~ I p. 128b 3/4 and Ah. I 6. 102, professes: vrddhamulakasya tridosa- kartuh katukasya kaphakartrtve yad deary a-Vahatena maihuravipakitvam, kdra/ijam uktam tat svayani hrdayapathitasyaiva vrddhamulakasya katuvipdkitvam smrtam kim vanyai Jdmcid iti na jane "while (in the present case) of full-grown radish, (which is) productive of (all) three humours (and) pungent because of its productive- ness of phlegm, sweetness of digestion (has been) named as (its) action by Master Vahata, (in the analogous case) of full-grown radish mentioned in his Essence pungency of digestion (has been) recorded by him (as such); if (there is) anything else (to it) I (do) not know [ ?]."
Candranandana, in comparing the Ah. with the As., often employs phrases like tatha. ca samgrahe proktam dcdryeqa "and so (has been) taught by the Master in his Compendium" (Ah. I 5. 13) or ity uktam, samgrahe tantmkartrd "thus (has been) said by the Author in his Compendium" (Ah. I 20. 39).
1 Donnees p. ? As this pamphlet is not available to us, we must depend for what follows on Jolly's abstract (ZDMG liv p. 262).
2 Mainly the Fihrist (Flttgel, ZDMG xi p. 148 sqq.), Mas'udfs Golden Meadows, and Ibn Abi UseibiVs Fountains of Information (Dietz, Analecta I p. 117 sqq. ; Cceeton, JRAS vi p. 105 sqq.; Muller, ZDMG xxxiv p. 465 sqq.).
Introduction 3
the Madhavanidana, in the same way that Vrnda and Cakradatta have only used the As., not the Ah. (5) The As., despite its greater volume, makes no mention of quicksilver, while the Ah. does.
3. Discussing these arguments, Jolly 1 called them plausible but not decisive and raised the following objections: (1) The dubious Arabic term Astankar may apply to the As. as well as to the Ah. 2 . (2) The eight- fold division is observed in the Ah. too, though not as strictly as in the As. According to Arunadatta (on Ah. VI 1 init.), the first part is treated in the first five sections, whereas the other five parts are dealt with in the sixth section. And by the same scholiast (on Ah. VI 38 & 39 fin.), the chapters on poisons and elixirs are both called "treatises" (ta?itra), a term synonymous with "part" {anga). (3) It is true that a mixture of prose and verse is characteristic of the old sutra style ; but the possibility of a later imitation must be reckoned with, and the earliest datable documents of Indian medicine as preserved in the Bower Manuscript are versified almost throughout. (4) In view of the large number of mnemonic verses that are common to nearly all medical books, no undue emphasis should be placed on some special points of agreement. That both the As. and the Ah. contain rather old material appears from the many prescriptions they share, for instance, with the Bower Manuscript. (5) Quicksilver must have been introduced into Indian pharmacopoeia much earlier than is commonly conceded, because it occurs already in Susruta 3 .
Notwithstanding his criticism of Corbieb's views and his additional counter-argument that the Ah. had been circulated in countless manuscripts, expounded in numerous commentaries, and held in high esteem, while the As. remained, as it were, an unnoticed wall-flower, Jolly stuck to the priority claim of the As., and that mainly for three reasons 4 : (1) the above-quoted statement from the Ah., (2) the archaic mixture of prose and verse in the As., and (3) the Buddhistic tendencies in the
1 ZDMG liv p. 262 sq.
2 Jolly seems to have overlooked that the texts in question actually read kitab asdnkar [w. 11. asatar, astankar, asdtkar] al-jami' "the summarizing book Asankar," which indeed agrees with Astangasamgraha rather than with Astan- gahrdayasamhita (unless one considers the final r in asankar a relic of hrdaya). But even so, the argument is by no means conclusive, because it only proves that the As. is fairly old, a point that goes undisputed.
The JPihrist describes this work as a translation into Arabic by Ibn Duhn, who appears to have been director of the hospital of the Barmecides (jl. 752—803 A. D.). Cf. Fltjgbl, ZDMG xi pp. 149 & 151.
3 IV 25.39, V 3.14, VI 35.7. Jolly still considered the mention of, or silence on, quicksilver not unimportant for determining the chronology of the Ah. and As. We now know, however, that reference is made to quicksilver in at least two stanzas of the As. (VI 30 ~ III p. 226 a 1 ; VI 49 ~ III p. 446 a 2) which word for word agree with their counterparts in the Ah. (VI 25.61, VI 39.162).
4 Medicin p. 8.
4 Introduction
As., which, though still existent, were largely obscured in the Ah. He was followed without reserve by Hoernle 1 , who introduced the names Vagbhata I and Yagbhata II for the authors of the As. and the Ah. respectively, by Keith 2 and Wintermtz 3 , who included Jolly's reason- ing in their histories of Indian literature, and by Gode*, who recorded the progress of Vagbhata research until 1938. All these scholars contrib- uted substantially to impressing the antecedence idea on a wider public 5 . The issue seemed definitely settled.
i. However, in a learned introduction to their pioneering translation of the Ah. 4 , Hilgesbebg and Ktjrfel undertook to closely re-examine the entire problem and arrived at an altogether different conclusion. Because of the prime importance of these findings, their train of thought may here be reproduced at some length.
Starting with those four lines in the epilogue which supposedly prove the Ah. to be a condensed version of the As., it should be observed that they are quite incompatible with a statement in the prologue (1 1.4 sq.), which says:
tebhyo 'tiviprakirnebhyah prayali sarataroccayah //
kriyate ' stangahrdayairi natisaniksepavistaram j
(These) [i.e. the works of Agnivesa etc.] being too widely scattered,
there is (now) made from them, as a collection for the most part of
very essential (matter), the Astaiigahrdaya, without too much brevity
or prolixity.
If the Ah. were regarded as a sort of story within a story, then its head and tail pieces would not fit together. Another inconsistency lies in tbe fact that in the epilogue (VI 40.59 sqq.) Atreya's disciples Agnivesa etc. are assumed to have been present at the first recital of the Ah., whereas in the prologue nothing like that is intimated. Furthermore, the whole debate among Agnivesa and his fellow-pupils on the merits of distinguishing between wholesome and unwholesome, which actually only serves the purpose of motivating and praising the new book, would be much more in keeping with the supposed original (the As.), where it is missing though, than with its alleged epitome (the Ah.). Add to this further discrepancies : in v. 62 the name Punarvasu is abruptly substituted for Atreya without having been introduced so far, neither in the text nor in Arunadatta's commentary 7 ; and in v. 59 the sage Bheda is thought to be attendant, while in v. 87 the question is raised why his work is neglected. Such absurdities are apt to throw discredit upon the
1 Osteology p. 6 sq. 2 History p. 510. 3 Gescbichte III p. 549.
4 Studies I p. 171 sqq.
5 E.g., Mukhopabhyava, History III p. 790 sqq.; Kashtkae, ABORI xxxvii p. 338 sq. 6 p. xvii sqq.
7 It. may rightly be argued, however, that the identity of Itreya and Punarvasu, which is confirmed by Arunadatta (on Ah. VI 40.59), was so familiar to the Indians as not to require special mention.
Introduction 5
whole epilogue, and the suspicion suggests itself that we have before us an interpolation. Since the stanzas in question have been translated into Tibetan, however, they must have been added at a comparatively early stage.
Next there is the mixture of prose and verse, the dubious value of which for proving the As. prior to the Ah. Jolly himself had already conceded. When he nevertheless adhered to this argument, he did so because internal evidence seemed to speak in its favour. But the reasons he advanced do not stand a critical investigation. By comparing line for line the Cikitsasthanas and Kalpasthanas of both works with each other and then with the relevant passages of the Carakasamhita, [CaS.]
1 , Hilgenbbeg and Kikfel reached the following conclusions:
(1) In the metrical parts, the extra verses of the As. may be isolated almost every- where without any difficulty or violence to the wording; what is left behind is the text of the Ah. Hence the As. and the Ah. are in the ratio of two different recensions of the same text, one of which appears enlarged or abridged as against the other. (2) Both works show a considerable identity or resemblance of wording with the much earlier CaS., which applies to their common stock as well as to the additional verses of the As., but not to the prose portions. No matter whether Vagbhata drew on Caraka or whether both authors relied on the same source, it is not the presumedly older prose but the supposedly younger verses that form the loan-texts
2 . Now if verses that are found in one text (the Ah.) inde- pendently occur in another (the CaS.) in a similar connection, while yet a third (the As.) gives prose instead, then the verses cannot be a transposition from the prose, but the prose must be an adaptation from the verses ; and it is indeed possible in the prose portions of the As. to trace not only signs of the original metrical version such as unusual word-order and rhythm, but also vestiges of the gradual transition from verse to prose as preserved in the variants 3 . Thus Jolly's second argu- ment is not tenable either.
1 See Appendix.
2 The same situation could be demonstrated by means of the Garudapuranaj which almost literally contains the Nida.nastha.na of the Ah. (see KntFEL, Festgabe Garbe p. 102 sqq.), provided it can be definitely shown that both texts are not interdependent.
3 That the prosification of the Ah. was not the work of a single man but extended over a longer space of time also appears from the fact that Arunadatta (on Ah. 1 12.52 sqq.) cites 24 slokas from the As. which in the present text correspond only to prose (I 20 ~ I p. 149b 10 sqq.; ef. Cobdiee, JA ix 18 p. 152). It is interesting in this connection to learn of the existence of a Madhya- or "middle" Vagbhata and his Madhyasamhita. or "middle collection" (already see Coediee, Museon N.8. iv p. 334) which, though lost as such, is known from many quotations in NMcalaka- ra's Ratnaprabha, and Sivadattasena's Tattvabodha and, judging by these, repre- sents an intermediate stage in the course of prosification. A number of passages given by Niscalakara have been analysed by Bhattachaeyya, ABORI xxviii p. 113 sqq.
(J Introduction
Finally, as concerns the Buddhistic tendencies prevalent in the As., JorxY referred in the main to a prayer addressed to Buddha, which he thought was missing in the Ah., and which runs as follows (I 27 ~ I p. 2(?3a,5 aqq.):
o)ii namo hhagavate bhaisajyagurave vaiduryaprabharajaya tathagatayar-
hate mmyaksanibuddhaya /
Om! Reverence to the Victorious One, the Medicine Master, the
Cat's-eye-splendoured King, the Thus-gone One, the Saint, the Fully Enlightened One! It was already Cordier 1 who set him right by tracing this prayer in the Ah. too (I 18*18). On the other hand, it is possible in the As. to find, besides Mahayanie traits, typical Hinduistic features, such as relating the story of Virabhadra's creation with fever (III 1 init.), suggesting the presentation of a Vedic offering against phthisis (IV 7 fin.), and praising the Asvins by adducing all the Vedic and Brahmanic myths connected with them (VI 50 med.). So there can be no talk of a specifically Buddhistic character of the As.
After the theory of a senior and junior Vagbhata has been deprived of its basis, the question remains to be answered how else the term Vrddha-Vagbhata can be understood. Here, Hilgenberg and Kireel say, analogous cases will help us on. In his Catalogus Catalogorum
2 , Aufrecht lists, inter alia, a Vrddharyabhata besides an Aryabhata, the Vrddhayavanajataka of a Vrddhayavanacarya besides the Yavanajataka of a Yavanacarya, a Vi'ddhagargyasamhita, besides a Gargyasamhita, and a Viddhayogasataka besides a Yogasataka . Turning to indigenous authors , Bhavamisra records among his sources not only a Vagbhata and Vrddha-vugbhata, but also a Susruta and Vrddhasusruta
3 , and Todaramalla mentions, apart from these, an Atreya and Vrddhatreya, a Harita and Vrddhaharfta, and a Vrddhabhoja
4 . It stands to reason that in these instances (the number of which can probably be increased) the attribute vrddha signifies, not a senior writer or an older work as opposed to a junior writer or a younger work, but rather the author of an enlarged recension as against that of a shorter original or such books themselves. That this is the only interpretation possible for the aphorisms going by the name of Vrddha-Canakya has been irrefutably demonstrated by Kbessler 5 .
Other evidence points in the same direction. To give only one example, the Ah. (II l.S sq.) says that, healthy conditions granted, a woman who has completed her 16th and a man who has completed his 20th year will produce a sturdy child, while younger parents will beget a sick, short- lived, hapless creature at the most. This statement, which is in perfect
1 JA is 18 p. 168. 2 s. w.
3 Cf. Aufrecht, Cat. Bodl. viiip. 311b.
- Cf. Webee, Verzeiohniss I p. 289 sq. 5 Stimmen p. 38.
Introduction 7
harmony with Indian, circumstances, has been amplified in the As. (II 1 init.) to the effect that a 21-year-old man shall marry an approximately 12-year-old girl, but will father with her a sturdy child only when he is 25 and she 16. The As. here makes a positive concession to the Hindu point of view that marital cohabitation should take place immediately after the first menses (which start at the age of 12 ; cf. Ah. II 1.7) at the risk of otherwise committing a serious offence 1 . If the Ah. really were an abstract of the As., then it would be quite inconceivable why the epit- omizer should have swerved from common opinion in so vital a matter. Even though the As. has lost all of its originality and much of its importance by what has been said hitherto, it cannot be entirely dis- missed as second-hand. On the contrary, the scholiastic exhaustiveness and pedantic consequence with which every subject is treated make it a valuable supplement to the much condensed and often enigmatic Ah.
5. About Vagbhata' s life 2 nothing else is known than what he himself has handed down in a short autobiographic note (As. VI 50 fin.) :
bhi§agvaro Vagbhata ity abhun me pitamaho namadharo °smi yasya / suto 'bhavat tasya ca Simhaguptas tasyapy aham Sindhusu labdhajanma // samadhigamya guror Avalokitat gurutarac ca pituh pratibham maya / subahubhesajasastravilocanat suvihito 'ngavibhagavinirnayah //
My paternal grandfather, whose namesake I am, was the eminent physician Vagbhata; his son was Simhagupta, and his (son) again (am) I. Among the people of Sind I was born.
Having obtained my knowledge from the venerable Avalokita and my (even) more venerable father, whose eye (represents) medical science in a very high degree 3 , (there was) well made by me (this)
1 Cf. Jolly, Becht p. 55 sqq.
2 Tradition now takes him for Dhanvantari, the physician of the gods and founder of Ayurveda, identifying him with one of the fourteen gems obtained at the churn- ing of the ocean; then for the great sage of the present Kaliyuga as opposed to Atri (Caraka) and Susruta, who are related to the past Krta- and Dvaparayugas respectively (Har. VI 2 fin.); then for an incarnation of Buddha; and then again for a voluptuous brahmin given to all sorts of revelries and lost in love with a low- caste woman. There is also a story current among the learned pundits of South India that Vagbhata, formerly a brahmin, was persuaded by a Buddhist priest to adopt his religion, which he embraced in the latter part of his life. Cf. Gode, Studies I pp. 174 & 178.
3 Indu explains: adesdrthaparijiiaTiahetutvad "on account of its being the cause of the full perception of the entire object." Htlgenberq & Kike-el, who render this line "nach Prufung vieler Medizinbucher" (p. xlvii), seem to read vUokanal instead of vihcanat.
8 Introduction
complete exposition 1 , arranged according to the (eight) parts (of
medicine).
That Vagbhata, as stated in this note, hails from North India 2 also appears from the fact that he reckons seeing a Dravida or Andhra among the bad omens (Ah. 116.49; As. II 12 ~ I p. 367M3), and that he regards the earth as chiefly occupied by the Himalaya and Vindhya mountains, taking herbs which grow on the former for soma-like and wholesome, but such as come from the latter for fire-like and unwhole- some (Ah. V 6.29; As. V 8 fin.).
6. Vagbhata' s date, however, once the two-author theory is discarded, may be fixed with considerable accuracy 3 . For the well-known Chinese pilgrim I-ching, who stayed in India from 672 till about 688 A.D., after enumerating in his book of travels "the eight sections of medical science," proceeds as follows 4 :
1 vinirttaya, by Indu glossed as viniscaya, is given in the dictionaries only as denoting "complete settlement or decision, certainty, a settled rule" (MW p. 970) or the like, which does not make sense here. The above translation rests upon BhP. VI 2.20 where the corresponding vinirifi is used in the meaning of "to explain fully" (Sanyal):
ta emm. suvinir^Xya dharmam bhagavatam nrpa /
tarn yamyapaMn nirmucya vipram mrtyor amumucan jj
These [i.e., Visnu's emissaries], having thus well and fully explained the divine
law [i.e., the Bhagavata creed], O king, (and) having freed him [i.e., Ajamila]
from Yama's fetters, released the brahmin from death.
s In the colophon of Ja(i)jjata's Susrutatlka, Vagbhata is given the appellation Mahajahnupati "'Lord of Mahajahnu." Bhattacharyya (ABORI xxviii p. 122) identifies Mahajahnu with the present Majhanda (spelt Mahajanda by Preedy, SRBG N.S. xvii p. 673, and Manjhand on Bartholomew's map), a small township and subdivision of the Karachi district, situated about 50 miles north of Hyderabad on the west bank of the River Indus. It is not altogether impossible that this was Vagbhata's domicile.
3 Till now it varied between such extremes as the 2nd cent. B.C. (Kttnte, Introd. p. 15) and the late 12th cent. A.D. (Cobdier, Vagbhata p. 3; refuted by Jolly, ZTXMG liv p. 262), though usually scholars placed the "elder" Vagbhata in the early 7th and the "younger" one in the 8th cent. A.D. Cordier founded his chronology on a spurious passage from Kalhana's Rajatarangini, which is missing in Stein's edition and runs as follows: [[Simhaguptasutah paramabauidJw Vagbha- tdcaryah Kdsmlratmrapati-Jayasimhasya prajapalanasamaye varttamana asit "the prominent Buddhist Vaghhatacarya, the son of Simhagupta, lived under the reign of the Kashmirian king Jayasimha [1196—1218 A.D.]." When faced with Jolly's retort, he strove to make good ( JA ix 1 7 p. 1 83 sq. ) by advancing an equally untenable theory based on a rather fantastic narrative in Merutunga's Prabandhacintamani (V20), which assigns Vagbhata to the court of King Bhoja of Dhara, (r. 1018—60). It is well known, however, that this quasi-historical work stops at no anachronism and hence commands little if any authority (Winternitz, Geschichte II p. 332).
- Kan-hai chi-kuei nei-fa-chuan, ch. 27 ~ Taisho Tripitaka, vol. 54, p. 223, col. 3. The transliteration is owed to Prof. Wolf Haenisch, Marburg. The English rendering is that of Takakusu (Record p. 128).
Introduction 9
ssu chih pa shu hsien wei pa pu. chin-jih y% jin liieh wei i chia. wu-tien chih ti hsien hsi ch'iu hsiu. tan ling chieh cM wu pu shih-lu. These eight arts formerly existed in eight books, but lately 1 a man epitomized them and made them into one bundle. All physicians in the five parts of India practise according to this booh, and any physician who is well versed in it never fails to live by the official pay. Though neither name nor title are given, there can be little doubt but that this passage refers to Vagbhata, the author of the Ah., who must consequently have flourished near the middle of the 7th century 2 .
1 Mukhofadhyaya (History III p. 794), while rightly correlating "lately" (chin- jih) with "formerly" (hsien), goes so far as to maintain that by "lately" any period subsequent to "formerly" can be understood. But neither the English nor the Chinese admits of such an interpretation, which would amount to a meaning "later, after- wards" untenable for chin-jih. (By courtesy of Prof. Alfred Hoitmann, Berlin.)
2 Some dissentient views must here be touched upon. Takakusu suggests in his additional notes (p. 222) that "this epitomizer may be Susruta, who calls himself a disciple of Dhanvantari, one of the Nine Gems in the Court of Vikramaditya." It need hardly be mentioned that Susruta's relation with Dhanvantari (Susr. 1 1.12) and Dhanvantari' s relation with Vikramaditya (Pseudo-Kalidasa, Jyotir-vidabharana XXII 10) are of a purely legendary nature and of virtually no chron- ological value (Wintbbnitz, Geschichte III p. 42 sq.), and that Susruta must have lived far earlier (Jolly, Medicin p. 9 sq.). In making his suggestion, Takakusit apparently relied on the fact that I-ching recorded the eight branches of medicine in very much the same way that Susruta did, with only nos 5 & 6 (pediatries and toxicology) given in reverse order (Jolly, ZDMG lvi p. 566 and JRAS 1907 p. 173 sq.). On the other hand, Susruta did not epitomize those eight branches at all. This notwithstanding, Jolly saw unfit to rule him out altogether, on the ground that I-ching might have read Susruta's introduction and "looked upon his work as a recent compilation, because it purports to be an extract in eight parts . . . from an earlier work in 100,000 verses." If such were really the case, however, then I-ching would indeed be guilty, as Hoernle (JRAS 1907 p. 414 sq.) pointed out, not only of lightly brushing aside Indian tradition, which regards Susruta's work as one of great antiquity (and on which I-ching had to rely in his day), but also of grossly misunderstanding Susruta's introduction, which puts matters into quite a different perspective, the relevant passage (Susr. 1 1.6) reading as follows: iha khalv ayurvedam namopangam atharvavedasydnutpddyaiva prajah slokasa- tasahasram adhydyasahasram ea hrtavan Svayambhuh j tato 'Ipayustvam alpame- dhastvam ccUokya naranam bhuyo 'stadha pranttamn //
Now then, the so-called Ayurveda (is) a subsidiary part of the Atharvaveda. Before even creating man, the Self-existent One [i.e., Brahman] composed (it) in a hundred thousand stanzas and athousand chapters. Afterwards, considering man's shortness of life and narrowness of intellect, he recast (it) into eight divisions. Bhattacharyya's opinion (ABORI xxviii p. 127) that I-ching referred to an- other medical compendium existent at his time (such as Ravigupta's Siddhasara or Acyuta's Ayurvedasara), apart from being purely hypothetical and rather far- fetched, is equally unsatisfactory; for a book that once was so popular as to have been read all over India is not very likely afterwards to have fallen into complete oblivion. Besides, Vagbhata does not mention any such work among his sources, an omission that would be hard to explain in the case of so renowned a predecessor.
10 Introduction
I>lung's account fully agrees with the fact that the Persian, physician. C A1I ibn Sahl Rabban at-Tabari, wl]LO included "a survey of the Indian system of medicine" in his Firdaus al-Hikma or "Paradise of Wisdom" (dated 849/50 A.D.), names Garak, Susrud, the Astanqahradi, and the Nidan as his sources 1 . If the Ah. was so famous in Persia by that time as to be put on a par with Caraka and Susruta, it must have been written during the 7th century at the latest 2 .
T. As is also manifest from these testimonies, Vagbhata enjoyed early and widespread recognition both inside and outside India. Hence it is no wonder that numerous commentaries on the Ah. (as against only three on the As. a ), many of them unedited so far, can be traced in manu- sript. catalogues, publishers' lists, etc. The following have come to our notice (those fully or partly printed are marked by an asterisk) 4 :
1 VII 4.1 (p. 557 Siddiqi; p. 1110 Siggel). Cf. Meyebhof, ZDMG Ixxxv p. 63 sq. and Isis xvi p. 12; Mcllee, JRAS 1932 p. 791 sqq.
- Disregarding this impregnable testimony, which could have easily saved him from a serious misconception, Bhattachabyya (ABOM xxviii p. 122 sqq.) assigns our author to the 9th century and identifies him with Vagbhata, the father of TTsata (who wrote the CiMtsakalika). After some chronological speculations on the iconolatry of Pamasavari (As. IV 2 fin.) and the twelve-armed Avalokitesvara (Ah. VI 5.50 ~ As. VI 8 fin.), he refers for the terminus a quo to a Vagbhata quotation in Niscalakara's Ratnaprabha, the first line of which is lost (A fol. 1 1 7 a) :
- //
[bojdhicaryavatdroktam hatnaiokadininditam j
aturam sravayed dhiman bodhayec ca muhur muhuh jj
That said in the Bodhicaryavatara, by which lust, grief, etc. have been censured,
a wise (physician) shall iet his patient hear and understand time and again. Since Santideva's Bodhicaryavatara dates from the middle of the 7th century, so he argues, its recognition as a scriptural text places Vagbhata not before 800 A.D. However, the present stanza is contained neither in the Ah. nor in the As. and, like many similar quotations in Ni^calakara's seholion, apparently originates from the secondary Madhyasanihita (see above, p. 5 n. 3). What is more, it occurs in only one of the three manuscripts utilized by Bh., which does not contribute exactly to its credibility. If anything at all, this sioka simply proves the Madhya-samhita to be later than 800 A.D. Bh. labours under the misconception that the so-called Vrddha-, Madhya-, and Svalpasamhitas are three separate works of the same author rather than three different recensions of the same text. The revisor of the Madhyasamhita may well have been a namesake or even descendant of the great Vagbhata, who, judging hy his own words (As. VI 50 fin.), belonged to an old family of physicians. There is no need to go into Bh.'s views on the terminus ante quern since naturally they do not conflict with the present investigation.
3 Viz., *Indu , s Sasilekha, the *Arthaprakaiika (in Hindi), and a fragmentary Vyakhya (in Canarese script) on which see KtnppuswAMi, Descriptive Catalogue xxiii no 13071.
4 Cf. Cobdieb, Vagbhata pp. 6 & 14 sqq. and Museon N.S. iv p. 335j Gode, Studies I p. 181 sqq. ; HmaENBEBG-KntPEL, Transl. p. xxiii; Mooss, Ed. II p. v.
Introduction 11
(a) Commentaries with author and title indicated :
(1) *Atridevagupta's Vidyotini (in Hindi);
(2) *Arunadatta's Sarvangasundari or Sundara ;
(3) ASadhara's Astangahrdayoddyota or Uddyota;
(4) *Indu's Sasilekha;
(5) Udayadityabhatta or Suvarnapurandhara's Hrdayaprabodhika ;
(6) *Uppotta Kannan's Bhaskara (in Malayalam) ;
(7) Kahnaprabhu's Nidanacintamani (on Nidanasthana) ;
(8) *Candracandana or Candranandana's Padartha(guna)candrik5
or Gunapatha (also in Tibetan) ;
(9) Damodara's Saiiketamanjari;
(10) *Paramesvara's Vakyapradipika;
(11) (Bhatta-)Narahari or Nrsimhakavi's VSgbhatakhandanaman- dana;
(12) Ravi's Vyakhyasara;
(13) Vagbhata's Astahgahrdayanamavaiduryakabhasya (also in Tibetan) ;
(14) Vasudeva's Anvayamala ;
(15) Visvesvarapandita's Vijneyarthaprakasika, ;
(16) Sankarasarman's Lalita;
(17) *Sivadasasena's Tattvabodha (on Uttarasthana) ;
(18) *Sivasarman's Sivadipika (in Hindi);
(19) *Sridasapandita°s Hrdaya(pra)bodhika ;
(20) Hatakanka(ga)°s Astangahrdayadipika 1 ;
(21) *Hemadrf s Ayurvedarasayana or Dinacaryaprakarana.
(b) Commentaries with author only indicated :
(22) Todaramalla (on Nidanasthana) ;
(23) Ramanatha ;
(24) Ramanujacarya (in Telugu).
(e) Commentaries with title only indicated:
(25) Andhratika (in Telugu) ;
(26) *Kairali (on Uttarasthana) ;
(27) P&thya 2 ;
(28) Balaprabodhika ;
(29) Brhatpathya ;
(30) Brhadvyakhyasara ;
(31) Vyakhya 2 ;
(32) Htdayabodhika s ;
(33) Hrdya.
1 Hatakanka's Sarvangasundari recorded by Bubnell (Index I p. 65b) looks suspicious. There may be a confusion -with Arunadatta's commentary at the bottom of it. Of. Cordieb, Vagbhata p. 6.
2 Behind these titles several commentaries may be concealed.
8 This is perhaps identical with [[UdaySdityabhatta's] or with gridasapandita's Hrdayaprabodhika.
12 Introduction
Among these commentaries, there are four that have reached some fame : Arunadatta's Sarvangasundarl, Indu's Sasiilekha, Candranandana's Padarthacandrika, and Hemadri's Ayurvedaras&yana. A few facts may here be given on their authors.
8. Arunadatta was the son of Mrgankadatta
1 and probably hailed from North India
2 . Besides the Sarvangasundarl
3 , he appears to have written a commentary on Susruta
4 and, as will presently be seen, on the As.
His term of life may be fixed indirectly with some accuracy. For Dalliana, while elaborating on the term aksivairdgya (Susr. V 1.30), records the following alternative :
rupagrahane 'lasatvam iti Gayi j vigatarage aksinl bhavata iti Sam-
graharunau, /
Inactiveness in the perception of colours : thus Gayin ; the eyes become
discoloured : thus the Samgraha and Aruna. Since the parallel passage of the Ah. (I 7.15), like Susruta, has nominal aksivairagya, which Arunadatta glosses by nominal aksiviraktata, the second reference can only be to the As. (I 8 ~ I p. 65a 17/18) and to a lost commentary on it by Arunadatta; for there alone it is said that at the sight of poisoned food cakorasyaksiyi virajyete "the eyes of the Greek partridge [Perdixrufa] become discoloured," a verbal phrase which may well have been explained as indicated. As Dalhana is assigned, by general consent 5 , to the 12th century 6 , Arunadatta must have lived
1 Thus the colophons. He himself gives his own name as Aruna and his father's as Mrganka (1 1 init.). These short forms are doubtless due to the metre.
a Under the lemma oiananga "extravaginal congress" (Ah. I 7.71), he observes: d<iJ;sit}atya hi mukkena kurvanti "for Southerners perform (coitus) through the mouth." In Bkattachakyya's opinion (IHQ xxiii p. 132), which there is no reason to dispute, only a Northerner could have ventured such a sweeping assertion. At any rate, Arunadatta's remark runs diametrically counter to Vatsyayana's doctrine (p. 130), repeated by most subsequent authorities (cf. Schmidt, Beitrage p. 233 sqq.), that oral intercourse was typical of the inhabitants of the Punjab. As a Northerner or even a Punjabi he probably took offence and committed a pious fraud, putting the blame for this disdainful practice on some far-away people. VatsySyana, on the other hand, was less sensitive about the matter, although he too must have been a Northerner, referring as he does to the southern custom of marrying one's maternal girl-cousin as something unusual (p. 207 sq. ; cf. Schmidt, ZDMGlviip. 706).
Das Gupta (IC iii p. 159 sq.) reckons the Sarvangasundarl among the medical works of Bengal, unfortunately without offering any proof to that effect.
s Simply styled Sundara by Paeadkak Shastki in the N.S.P. edition of 1939.
- Cf. Auebecht, Cat. Cat. I p. 30 a.
5 Cf. Jolly, Medicin p. 10; Hoernle, Osteology p. 16: Bhattachabyya, IHQ xsiii p. 132.
8 He is quoted by Hemadri (according to Cordiek, Donnees p. 3), the archivist and chief minister of the kings Mahadeva (r. 1260—71) and Ramacandra (r. 1271 —1309) of Devagiri (see below, § 11), and in his turn quotes Cakrapanidatta
Introduction 13
prior to him 1 . Whether he is identical with the lexicographer of the same name first mentioned in Vardhamana's autoeommentary on Ganar. II 77 (dated 1140) cannot be ascertained 2 .
9. Indu is still a largely unknown quantity as far as the circumstances of his life are concerned. Judging by the fact that he expressly defines Andhra and Dravida as the names of two southern peoples or kingdoms 3 and repeatedly mentions Kashmirian terms for particular plants 4 , he is likely to have been a Northerner and a native of Kashmir 5 .
Tradition
(on Susr. VI 49.19), whose father Narayana, or (according to a variant reading, on which see Bhattachabyya, IHQ sxiii p. 134 sq.) who himself, was the head-cook and minister of King Nayapala (r. 1038—55) of Bengal (see colophon).
1 So far it was customary to determine Arunadatta's date according to Hoesnle's method (Osteology p. 17), which is as follows.Vijayaraksita and Srlkanthadatta (on Nid. LIX 29) controvert a certain doctrine of Arunadatta (on Ah. VI 12.1) regarding the position of the so-called "first" (prathama) membrane of the eye, which, in keeping with common opinion, they equate to the "innermost" (sarva- bhyantara) membrane, while he takes it for the "outer" (bahya). Since these scholiasts cite Gunakara (on Nid. V 31), who commented on Nagarjuna's Yogaratnamala in 1240 (Aufbecht, Cat. Cat. I p. 155b), and in their turn are cited by Vacaspati (Atank. introd. v. 8), whose father Pramoda was the chief court-physician of Hammira Mahammada (ibid. v. 2 sqq.), that is, Amir Muizzuddm Muhammad or Muhammad of Ghur (d. 1205), they must have flourished about 1240, and Arunadat-ta must consequently have lived prior to them, probably around 1220.
This ingenious calculation was dealt a serious blow by Bhattachabyya (IHQ xxiii p. 130 sqq.), who pointed out that Indu's Sasilekha (on As. VI 15 init.) and the anonymous Kairali (on Ah. VI 12.1) controvert the same doctrine, again without mentioning any name, and that evidently all three scholiasts just refer to a dissentient view common in their day. Further, the identity of the medical author Gunakara with a Jaina scholar of the same name, who commented upon a work on magic rites, incantations, and sorcery, should by no means be taken for granted.
What, incidentally, is meant by those "membranes" (patala), of which four are distinguished in Indian ophthalmology, cannot be ascertained for the time being. Magnus' view (Augenheilkunde p. 37) that they represent retina, choroid, sclera, and mucosa is categorically rejected by Esseb (Ophthalmologic p. 14). As for the above "outer" membrane, Hobrnlb (Osteology p. 17) considers it to be the cornea plus aqueous humour.
^ Cf. Gods, Studies I p. 185. s On As. II 12 ~ I p. 368 a 20/21.
4 E.g. in his notes on the vegetable section (iakavarga) Ah. I 6.73—116 and As. 1 7 ~ I pp. 50-54.
5 Here a conjecture made by Bhattachabyya (ABOBI xxviii p. 118) deserves special attention. In commenting on Vagbhata's view that even a brahmin may take garlic if it is through the milk or curd of a cow that has been fed garlic after a three days' fast (As. VI 49 ~ III p. 423a 4 sq.; cf. Bower Ms. I 34), Indu remarks: ity deary usya desasiddhdh kdSmirakah j vayam etan na mdtnah "thus the master's Kashmirian local authorities; we (do) not know this (practice)." Stumbling at the collocation desasiddhdh M&mlrahah "Kashmirian local authorities," Bhatta- OHABYYA proposes to punctuate not after but before MSmirakah, so that the passage
1 4 Introduction
makes him a pupil of Vagbhata 1 , and his referring to the author of the All. simply as "Master" {acarya}* seems to point in the same direction 3 . What is more, a contemporary of Vagbhata by the name of Indu, or rather Indukara, is quite familiar to us; he is the father of Madhavakara, the renowned author of the Madhavanidana, which can hardly have been written later than the 7th century, since in 849/50 it is already quoted by a Persian physician as an authority equal to Caraka, Susruta, and the Ah. *. It must, be stressed, however, that this identification is by no means conclusive, based as it is on very shaky evidence 5 .
Another way of fixing Indu's date also leads to nothing definite. Among the medical writers he mentions by name is the Caraka scholiast rMuittara Hari(s)candra*, whom Maheivara 7 states to be the court physician of King Sahasahka. Wilson 8 suggests the possibility of Sahasanka being a title of Sricandradeva, who (according to a copper- plate inscription from Nidigal 9 ) founded towards the end of the 11th
would ran as follows: "Thus the master's local authorities; we Kashmirians (do) not know this (practice)." Attractive though his suggestion is, it must be left alone for the time being. Its tenability hinges on the open question of whether or not Kashmirian brahmins were allowed to take garlic under the above circumstances. BCJnXER, while dealing with the peculiarities of Kashmirian brahmins (Report p. 1ft sqq.), does not mention anything like that; and of little avail is a reference in Rajat. I 342 to King Gopaditya (r. 370-310 B.C.), "who," it is said, "after having banished garlic eaters to Bhukslravatika, directed brahmins devoid of ordinances of their own to Khasata" (bhukslravdpikdyam yo nirvasya lasunaiinah / khdsatuyam vyadhdd vipran nijdc&ravivarjit&n) .
1 T. Rudbapabasava, in his preface to the Trickur edition of the As. (p. iv), gives the following "meditation stanza" (dhya7iasloka), which he says is "universally known" (lokaprasiddha) :
lamhaimairukalapam ambujambhacchayadyutim vaidyakan antevasina Indit-Jajja{a-muMian adhyapayantam soda / aijulphamalakaTtcukamitadaralaksyopavitojjvalam katfthastkagarusdram aujitadrsam dhyaye drdham Vdgbhatam //
1 steadily meditate on Vagbhata: the tassel of his beard-hair dangling (and) the brightness of his complexion resembling a lotus; always instructing his medical pupils Indu, Jajjata, et al. ; the splendour of his sacred thread being (but) slightly visible, distinguished (as he is) by a spotless coat reaching down to bis ankles; aloe-sap being in his throat (and) his eyes bedaubed.
On the coat (kuncuka), see Bhttshatj, Costumes p. 20.
2 On As. II 4 — I p. 304b 13 sqq. et passim. The reference is to Ah. II 1.94.
3 Thus Gode, Studies I p. 162 sq. But Bhattachabyya (ABORI xxviii p. 117 xq.) objects that in eommentatorial literature the word acdrya, if used without preceding wad or asmad, refers to the author of the text which is explained rather than to the teacher of the scholiast who explains it. 4 See above, § 6.
5 Bhattachabyya (IHQ xxiii p. 139) refutes it on the ground that Indukara, like Madhavakara, belonged to Eastern India.
6 On As. Ill 2~II p. 12a 14.
7 Visvaprakasa, introd. v. 5, reproduced by Webee, Verzeichniss II p. 261.
8 Works V p. 215 sq. » Cf. Colebbooke, Essays II *p. 286 (~ 2 p. 253).
Introduction 15
century the ruling dynasty of Kanauj, which realm he acquired "by his own strength" ; but this is, as he himself puts it, nothing else than "a mere etymological speculation" on the phrase "by his own strength," Sahasanka being a possessive compound formed of sahasa "strength" and anha "mark."
Besides his commentary on the Ah., styled Sasilekha, Indu also wrote a commentary on the As., which bears the same title and follows the wording of the former wherever Ah. and As. agree with each other. As this work is frequently mentioned as "Indumati" in Niscalakara's Ratnaprabha, which was written between 1110 and 1120 1 , Indu must have nourished in the 10th century at the very latest. He often adduces anonymous definitions of a pharmaeopoeial nature 2 and evidently commands a thorough knowledge of the medical convertible terminology 8 ; so he may be identical, for all we know, with the author of a medical glossary, also called Indu, who is very frequently cited by Ksirasvamin in his commentary on the Amarakosa 4 . Since this commentary dates from the second half of the 11th century 5 , the glossarist Indu must have lived prior to the year 1050 6 .
10. Candranandana (Tib. Zla-ba-la dga-ba), the son of Ratinandana (Tib. Chags-pa-la dga-ba), hailed from Kashmir'. His terminus ad quern is established by the fact that the Padarthacandrika (Tib. Thsig-gi don-gyi zla-zer) was translated into Tibetan by Rin-chen-bzan-po between the years 1013 and 1055 8 . He may or may not be identical with the physician Candrabhinandana (Tib. Zla-ba-la mhon-dga), who is
1 Cf. Bhattaobabyya, ABORI xxviii p. 118 & IHQ xxiii p. 129 sqq. His chro- nology depends on the reign of Ramapala, which, he fixes at 1078—1120 (IHQ iii p. 583 sq.), not without being contradicted (Majumdab, HB I p. 180 sq.). But objec- tions are raised only against his evidence, not against his result, at which others arrived independently of him, and which may now be regarded as common opinion.
2 E.g. on Ah. 1 3.30—32 & 48. 8 See his notes on Ah. I 6.94 sqq.
- Especially in his notes on the tree and herb section {vanausadhivarga, H 4).
5 It often refers to Bhoja, the author of a Sabdanuiiasana, who (according to Ganar. I 2 schol.) is identical with Bhoja, the king of Malwa {r. 1018—60; cf. OHI 3 p. 204) and author of the Sarasvatlkanthabharana, and is repeatedly quoted by Vardhamana, the author-cum-scholiast of the Ganaratnamahodadhi, which was written in 1140. Cf. Attfbecht, ZDMG xxviii p. 104. The identity of the grammar- ian and the writer on poetics is questioned by Zachariae, GGA 1885 p. 377.
6 The suggestion that the vyakhyakara Indu and the naighantuka Indu may be one and the same person was first made by Das Gupta (IC iii p. 154). It was repeat- ed, not without reserve, by Gode (Studies I p. 159 sq.).
7 See Tibetan colophon. Htjth (SPAW 1895 p. 270), evidently depending on the Mongolian, Manchurian, and Chinese indexes, wrongly transcribed the name as Candrananda. He was followed by Uhkbig (KoEVXN-KBAsrNSKi, Medizinphilo- sophie p. xix), through Coediee (JA ix 17 p. 185; BEEEO iii p. 614) had corrected this mistake long before.
8 See below, § 19. According to Cobdieb, Catalogue iii p. 472, Candranandana was a contemporary of King Abhimanyu II of Kashmir (r. 958 — 972).
16 Introduction
said to have assisted "Vairocana in putting the Four Tantras (rGyud bzi) into Tibetan at the time of King Khri-sron-lde(u)-btsan 1 , or with the medical lexicographer Candranandana, who is frequently cited by the Aniara scholiast Ksirasvfimin in the late 11th century 2 . On at least two occasions, his views seem to have been refuted by Arunadatta, though no name is given in either case 3 .
11. Hemadri is the only one among the scholiasts under discussion who poses no problem as to his identity and chronology. Preluding Ms Ayurvedarasayana, he introduces himself as the author of the Catur-vargacintamani (v. 2), a standard encyclopedia on ancient religious rites. For a proper performance of the vows etc. set forth in that work, it is necessary to enjoy good health, and the present commentary has been written with a view to facilitating the attainment of such good health ; it follows the doctrines of previous authorities on the subject, such as Caraka and Susruta (v. 3), without repeating what has been said by their commentators (v. 4). Hemadri was the son of Kamadeva, grandson of Vasudeva, and great-grandson of Vamana. Besides the Catnrvargacintamani and Ayurvedarasayana, he wrote several other works (Sraddhapaddhati, Hemadriprayoga, Nanasantayah, Tristhalivi- dhi) and commentaries (on Vopadeva's Muktaphala and Saunaka's Pravanakalpa)*; but he is different from Bhatta Hemadri, the son of Lsvarasuri and author of the Raghuvamsadarpana 5 . While in the Catur-vargaeuitamani (1 1.6 & 13) he describes himself as being in charge of the state records of King Mahadeva of Devagiri (r. 1260—71), from the Ayurvedarasayana (introd. v. 5 sq.) and a contemporary inscription 6 he appears to have been the archivist and chief minister of his successor Ramacandra (r. 1271 — 1309). So Hemadri flourished in the second half of the 13th century, and it is reasonable to assume that he composed the Caturvargaeintamani under the reign of Mahadeva and the Ayurveda-rasayana under that of [[Ramacandra]'.
12. There is one more Vagbhata scholiast who deserves special men- tion, although no manuscript of his commentary has yet been traced. He is the Jaina teacher Asadhara (the son of Sallaksana and father of Chahada), who lived about 1236 A.D. 8 and is said to have written, be- sides the Uddyota, 17 more books, not all of which, however, are likely to be his 9 . According to a valuable eulogy attached to his Dharmamrta,
1 Cf. Csoma, JASB iv p. 1. According to the Tang Annals (Btjshell's trans- lation, JRAS 1880 pp. 473 sq. & 506), Khri-sron-lde(u)-btsan reigned from 755 to 797 A.D. * On Ak. II 4.63 et passim.
3 On Ah. I 5.17 [15] & 1 19.87 [75]. The bracketed verse-numbers refer to Raj-vaibya's edition. * Cf. Aotrecht, Cat. Cat. I p. 768, II p. 185, III p. 52.
5 Cf. Gode, ABORI xiv p. 126 sqq. 6 Cf. Barnett, EI xiii pp. 202 & 205.
7 For the whole paragraph see Kane, History I p. 354 sqq., and Gode, Studies I p. 186 sqq. 8 His Trisasthismrtisastra dates from that year.
9 Cf. Peterson, Report II p. 86"; Aoteecht, Cat. Cat. I p. 54.
Introduction 17
two manuscripts of which are found in the Bhandarkar Oriental Research Institute, Poona, he hailed from Sapadalaksaya (the vicinity of Jaipur) and, as his country was overrun by a barbarian king, fled to Malava, taking refuge at Dhara, where he was enthusiastically received by Bilhana, the chief minister of King Vijayavarman 1 . Since Asadhara was apparently a senior contemporary of Hemadri, the discovery of his commentary would be a notable gain indeed, if only because of its antiquity.
As for the other scholiasts, Todaramalla was the Hindu financier of the Emperor Akbar (r. 1556— 1605) 2 , while Sivadasasena was the court physician of Barbak Shah of Bengal (fl. 16th cent.) 3 .
13. Of excerpts and summaries, there are found an anonymous Astangahrdayasamgraha embracing 700 granthas and Govindadeva's Astangahrdayasamhita or Yogalilavati comprising 300 slokas*. How popular the Ah. has remained up to the present day may be seen from the fact that as late as the year 1864 a certain Paramesvara of South India completed an abstract from it entitled Hrdayapriya, which closely follows the wording of the Ah.
14. Numerous manuscripts (as against but a few of the As. 5 ) are extant either of the whole or of parts of the Ah. and its commentaries 6 , some of them being fairly old. The earliest traceable so far was written by one Samalabhatta at Ahmadabad in samvat 1544 (i.e. 1486/87 AJD.) ;
i Cf. Gode, Studies I p. 182. 2 Cf. Irvine, IGI II p. 399.
3 Cf. Das Gotta, IC III p. 157. Lane-Poole, Dynasties p. 307, dates this ruler as early as 1459 A.D. See Godb, Studies I p. 123.
- Cf. Sastri, Catalogue xvi no 11034,- List MSS Benares p. 129.
5 G. S. Tabte consulted two MSS for his edition: one owned by G. S. Nirantar of Nasik (~ Bhandarkar, Lists I nos222— 27) and dated saka 1794 (i.e. 1872 A.D.), the other owned by B. S. Mate of Poona (Introd. p. 2). Cordeer's confrontation (JA is 18 p. 152 sq.) of Bhandarkar's MS. with Tarte's edition is completely illusory, because the former was actually used in the latter (JELelgenberg-Kire'el, Transl. p. xxii).
T. RudrapaRasava gives no information about his sources. He only states (Introd. p. v) that Indu's Sasllekha., given along with the text, was reconstructed during several years of extraordinary labour from highly dilapidated palm-leaf MSS in the possession of eight physicians and the royal library in Cochin.
R. S. Kinjawadekar based his text on Tarte's and Rtoraparasava's editions as well as on two more MSS : one procured by P. K. Gode for the Bhandarkar Oriental Research Institute in Poona, the other owned by G. S. Godbole of Bombay. Incidentally, as both these MSS only contain the As. proper, Indu's SaSilekha, also given in Kinjawadekar's edition, must be a reprint from Rudra- paRasava' s edition. — Finally, there is a fragmentary palm-leaf MS. in Canarese characters, recorded by Kupptjswami, Descriptive Catalogue xxiii no 13070.
6 Cf. Auprecht, Cat. Cat. I pp. 35 sq. & 773, II pp. 7 & 188, III p. 8; Kttpptt- swami, Descriptive Catalogue xxiii nos 13072—85; Rangacharya-Ktoptjswami, Triennial Catalogue I sqq. (v. indexes); Sastri, Catalogue xvi nos 11007—33.
2 Vogel, Vagbhata
18 Introduction.
two others are dated samvat 1576 (i.e. 1519/20 A.D.) and samvat 1735 (i.e. 1078/79 A.D.) 1 .
15. The t ditto princeps, accompanied by a modern commentary (entitled "Bhiiskara") and printed in Malayalam characters, is owed to Uitotta Kankak and appeared 1874—78 in Calicut. Since then countless other editions, with or without scholia, of the Ah. in its entirety as well as of individual sthanas have been published 2 , but nevertheless the text still awaits philological treatment. A first step in this direction was taken only recently by N. S. Mooss, who consulted no less than 11 M8S of the Ah. and 4 MSS of Indu's iSasilekha, noting every single variant in the apparatus criticus and thereby setting a standard for any future enterprise of this sort 3 . As in many similiar cases, however, it must be seriously doubted if the benefits derived from a critical edition of the Ah. would compensate for the pains taken in preparing it.
16. Mention was made above (§4) of Buddhistic tendencies existent in the Ah., and a magic formula directed to Buddha was adduced as evidence. Other characteristic examples are an allusion in 1 1.1 to the three moral poisons, which almost certainly shows Buddha to be "the unprecedented physician" in question, an enumeration in I 2.21 sq. of the ten commandments, and a reference in I 2.46 sq. to the application of awareness. This is not of course the place to decide whether Vagbhata himself was a Buddhist or whether he only had Buddhistic inclinations ; but any future research on the problem * will also have to take cognizance of such typically Hinduistic features as the three objectives of life (I 2.29 ; cf. 1 1.2) and the aversion to Buddhist sanctuaries (I 2.33 & 37). What- ever the final solution may be, there ean be no doubt but that it is those Buddhistic tendencies which led the lamas to translate, of all works, the Ah. into Tibetan and to incorporate it into the Tanjur 8 .
1 7. According to the colophon, which is reproduced below, the trans- lation was made by the Indian pundit Jarandhara and the Tibetan lama Rin-chen-hzan-po 6 :
1 Cf. Ktjnte, Pref. p. 1. 2 See Bibliography.
3 The present work is mainly concerned with establishing and interpreting the Tibetan version. Hence no attempt has been made to produce a critical edition of the original Sanskrit, which could be done only after the wording of those copies used by the Tibetans and all the scholiasts has been ascertained. In fact, the text given is merely that of A. M. Kunte's second Nirnaya Sagara Press edition (marked B for Bombay), with variant readings found in N. S. Mooss' Vaidya Sarathy Press edition (marked K for Kottayam) being listed in the foot-notes.
4 Useful preliminary work was already done by Coedibk, JA ix 18 p. 167 sqq.
5 The medical works contained in the Tanjur and the chronological questions connected with them were first discussed by Huth, SPAW 1895 pp. 269 sqq. & 283 sq., with additions in ZDMG xlix p. 280 sqq. A more detailed description was later given by Cokdieb, BEFEO iii p. 604 sqq.
6 (336a 2), ((335a 6)), [337a 6], 322b 3. For the meaning of the abbreviations and brackets, see below § 20.
Introduction 19
in Tibetan —
sman-pai bdag-po dGe-'dun-gsan-ba zes bya-bai bu /j 1 sman-pa chen-po
Pha-hkol 2 zes bya-bas yan-lag-hrgyad-pai snin-po mdzas -pa-las 3 rgyud
4 phyi-mai gnas[l]-te drug-poo // //
((7)) sman-dpyad yan-lag-brgyad-pai snin-po bsdus-pa zes bya(B)-ba
ji-sned-pa rdzogs-so // //
rgya-gar-gyi mkhan-po Dza-ran-dha-ra* dan /j s zu chen-gyi lo-tsa e -ba
dge-slon Rin-chen-bzan-pos 5 [bsgyur-cin] 7 zus-te gtan[337bl]-la
phab-pao* // //
in English —
Of the work Astasigahrdaya by the great physician named Vagbhata, a son of the master of physicians named Sanghaguhya 9 , (this) is the sixth (section), being the Uttarasthana.
The medical research-work entitled Astangahrdayasamhita is (here- with) entirely finished.
By India's Professor Jarandhara 10 and Revising Great Translator Monk Bin-chen-bzah-po it has been [translated], revised, and edited.
18. This translating team worked together on no less than nine canonical texts dealing with such diverse themes as hymnology, mystics, meditation, discipline, medicine, and chemistry. Besides the Astan- gahrdayasamhita, their list of publications includes :
(1) the Visesastavatlka or "Commentary on the Hymn on Distinction" by Prajnavarman ;
(2) the Devatisayastotratika or "Commentary on the Hymn on Pre- eminence to Gods" by the same author;
(3) the Tattvasarasamgraha or "Compendium on the Essence of Truth" by Dharmendra ;
(4) the Yogavataropadesa or "Guide to the Understanding of Medita- tion" by the same author;
(5) the Pratimoksabhasyasampramusitasmaranamatralekka or "Writing on Nothing but the Unlost Memory of the Commentary on the Pratimoksasiitra" by an anonymous author;
(6) the Padarthacandrikaprabhasa or "Moonshine Splendour of Word Meanings" by Candranandana (a commentary on the Ah.) 11 ;
1 N; no Sad CD; simple Sad P. s CD; gel NP.
3 NP; double Sad C; simple Sad D. 4 CD; Dza-randha-ra'N'P.
5 NP; simple Sad CD. 6 NP; tsthsa CD.
7 CD ; missing in NP. 8 NP; pa CD.
9 This is the exact retranslation of the Tibetan name. In the Sanskrit colophons of Ah. I, II, III & VI and As. I & VI, however, Vagbhata's father is called Simha- gupta, -which would be Sen-ge-sbas-pa in Tibetan. Interestingly enough, Aufbecht (Cat. Bodl. viii p. 303 a) records Samghagupta as the principal spelling of his MSS no 741— 42.
10 Elsewhere the spelling varies among Janardana, Jarandana, Jarandana, Jarandhara, and Jarandhara. u See above, §§ 7 & 10.
2*
20 Introduction
(7) the Dhupayogaratnamala or "Jewel Garland of Incense Prepara- tions" by Nagarjuna; and
(8) the Astapadikrtadhupayoga, a treatise on "Incense Preparations Made into Chess-board Squares" by the same author.
1J*. While Jarandhara is not otherwise known, Rin-chen-bzaii-po (or Ratnabhadra, as his name would be in Sanskrit) is one of the most outstanding figures of his time, bearing the honorary title of "great translator" {lo-tm-ba chen-po) and holding the notable record of 150-odd translations and revisions— not to mention his many original works, among them a Sanskrit-Tibetan glossary.
Descended from a Kashmirian family, he was born 958 A.D. in the West Tibetan province of m^a-ris. When he was 13 years old, he received his consecration at the hands of Abbot Ye-ses-bzah-po. Buddhism then being at a low in his native country, he paid three visits to adjacent Kashmir and there studied Tantrayana with 75 pundits; the most eminent among them were Sraddhakaravarman, Kamalagupta, and Jinamitra. At the age of 49, he joined the order of Bla-chen, which sought to remedy the abuses of Buddhism that had cropped up after its persecution by King Glan-dar-ma. Some time later, Grand Lama lHa-lde-btsan appointed him "head monk" {dbui mchod-gnas) and "diamond professor" (rdo-r]e-slob-dpo?i), presenting him with the estate of ier in the South- West Tibetan province of sPu-hrans. In this capacity, he erected numerous monasteries, temples, and symbols at Khra-tsa, Ron, and other places. An adherent of the Madhyamika school of philosophy founded by Nagarjuna, he also developed a brisk teaching activity, especially in the field of Prajnap&ramita and Anuttarayoga literature, and produced many learned disciples, among whom were some ten Sanskritists capable of doing translation work. Besides this, he rendered signal service to the reformation and propagation of Bud- dhism, known as the "later spread" (pkyi-dar) of the doctrine. At the age of 85, he met Atlsa, the famous patriarch of Magadha, who had come to mNa-ris on the repeated entreaties of King Od-lde. Being 26 years his senior, he did not at first bow to him in reverence ; but when, on a visit to the Golden Temple at mTho-ldin, he found him correctly describing the attributes of, and extemporaneously composing hymns on, every deity he saw, he paid due homage to him and entered into close relations with him. Among other things, he learned from him the method of "propitiation" (sgrub-pa), that is, how to coerce the Godhead through profound meditation. Because of his advanced years, he had to decline an offer to accompany the master on his trip to Central Tibet. He died, aged 97, at Khva-tse-vin-gir on the 17th of Magha 1055.
Thus far indigenous sources 1 , which differ only in small details. What is particularly interesting is Sum-pa-mkhan-po°s remark that
1 Sum-pa p. 182; g&m-nu pp. 68 sq., 249 sq., et passim.The entire material available on Rin-chert-bzan-po has been gathered and sifted by Tucci,Iiido-TibeticaII&IIIl.
Introduction 21
"at the age of 55, when the pundits Sraddhakaravarman, Karagupta [ ?], Buddhasrisanti, Buddhap&la, Kamalagupta, et al. were invited to Tibet, he copiously translated sutras (and) mantras such as Prajnaparamita, Yoga, and Saniaja (scriptures) 1 ." From this it would seem that Rin- chen-bzan-po started his translations at a comparatively late date, and that the Tibetan version of the Ah. too was made between the years 1013 and 1055 2 .
20. The Ah. belongs to the mDo or Sutra section of the Tanjur, four different xylographs of which are available, and occurs at the following places (the abbreviations and brackets are those used in the present edition) :
Ghone Tanjur (C), vol. he (119), fol. (44b 1— 336a3) 3
Derge Tanjur (D), vol. he (119), fol. ((44bl-335a7))
Narthang Tanjur (N), vol. se (118), fol. [15bl— 337a7]
Peking Tanjur (P), vol. se (118), fol. 15a5— 322b5
For establishing the Tibetan text, all these xylographs have been con- sulted, and a short account of them may be given here.
The Chone* Tanjur is extant but in one copy 5 , which was purchased for the Library of Congress, Washington, in 1928 by Rock e . It comprises
1 na-lnai dtts-su bod-dv, pafy-di-ta Sri-dha-lca- [read Srad-dhd-ka-ra-] varma dan Ka-ra-gup-ta dan BwJ-dha-dri-dan- [read dan-] ti dan Bud-dha-pa-la dan Ka-ma-la- gup-ta sogs spyan drafts thse der-phyogs [read -phyiri] dan yo-ga dan 'dus-pa sogs mdo snags man-du hsgyur-£in (line 11 sqq.).
2 Cordier. ( Vagbhata p. 11), referring to Fotoaux (Lalit. ed. II p. is), still held that the Tibetan translations of Sanskrit works, at least of those pertaining to medicine and magic, were made as late as the 12th or 13th century, at the time they were incorporated into the Tanjur. His view is no longer tenable.
3 In the provisional numbering of the Library of Congress, this is vol. 198.
4 Co-ne, less correctly spelt Co-ni, is a village (pop. 2000) and monastery in the A-mdo province of North-East Tibet or, according to another version, in the south- western part of the Kansu province of China. Situated not far from the upper course of the Hwang-ho, it once was the centre of all yellow-cap lamaseries in that region and later came to be a stronghold of the Yellow Church there, housing some 3800 monks under the rule of Yung-lo (1403—25). The wood-blocks of the first Kanjur and Tanjur edition, which along with many others were deposited in the huge archives, are said to have dated from the 15th or 16th century. The library and the printing press were completely destroyed, and the buildings badly damaged, when Mohammadan fanatics under the command of the four brothers Ma devastated the province in the early 1930s. The monastery itself was reconstructed in after years, but its treasures are lost for ever.
5 On the distribution of the various Kanjur and Tanjur editions among the libraries of the world, see Ch'en, HJAS ix p. 60 sqq.; Tucci, HJAS xii p. 480 sq.; Meisbzahl, PG viii 2 p. 27 ; Mauser., JOIB ix p. 99 sqq.
6 See his interesting report in NGM liv p. 569 sqq. , partly reprinted in The Illustrat- ed London News of September 1929. As the impression was struck in his presence, Rook could supply some valuable details on technical matters. Thus he was told
22 Introduction
209 volumes and is being catalogued by Mbisezahl. The colour of ink i.s black, the size of the folios e. 60.5 by 15.8 cm., the size of the printed surface 56.5—57.5 by 9 cm. 1 , and the number of hues per page 7. The text has been enclosed with a rectangular frame. The left-hand margin of the obverse bears, perpendicularly to the main body, the title of the subdivision, the number of the volume, and the number of the folio (e.g. gno-ba rig-pa. he. £e-h\a). The print is easily readable but contains, despite assertions to the contrary by the Chone lamas themselves 2 , not a few mistakes.
The Derge 3 Tanjur is available in several libraries, both Oriental and Occidental, and has been catalogued by Japanese scholars of the TShoku Imperial University, Sendai 4 . The copy used for the present edition is that of the former Prussian State Library, Berlin, now stored partly in the State Library, Marburg, and partly in the University Library, Tu- bingen. Details of its procurement could not be learned. It consists of 214 volumes, 14 of which got lost during World War II and were not recovered so far. As a microscopic analysis has shown, its paper was made from fibres of a genus of Thymelaeaceae, probably Daphne or Wikstroemia 5 . It is struck in black ink. The folios measure 62 by 11.5 cm. The printed surface is 51.5 cm. long and 7.5 cm. wide. The lines number
7 a page. The text is placed in a rectangular box with a 1.5 cm. margin on either side, that on the left giving the number of volume and folio (recto) and the title of the subdivision (verso) respectively, while that on the right remains blank. The print is perfectly legible and virtually free of carver's errors.
that a skilful lama needed 4 days to cut 1 block, and that 16 years were necessary to carve the blocks of the Tanjur alone, the wood used being walnut and a lighter species not identifiable. He also learned that 45 monks required 3 months to print the Kanjur and almost 6 months to strike the Tanjur. The paper was bought at Kungehang in Eastern Kansu, 11 days distant from Chone; it was very thin, and
8 sheets were pasted together to yield 1 leaf. The "wages" of the printers were 250 cash, i.e. 5c, per day, plus rations of barley flour, tea, and yak butter.
1 Working from a microfilm, we have to rely on the data kindly given us by Dr. Walter H. Maurek, Washington.
2 Rock, XGM liv p. 581.
s sDe-dge, also spelt bDe-dge and sDe-deg, is a remote place and principality in Eastern Tibet. It lies on the left bank of the Yang-tse-kiang in the Hsi-K'ang province. Of the several monasteries belonging to this principality, that of Derge itself is the most famous. It is occupied by followers of the Tantric rSfan-ma-pa sect, who disavow the reforms of Atisa and bTson-kha-pa. As late as the 19th century, the Yellow Church had not yet succeeded in getting any hold on the region. The Kanjur and Tanjur editions of this lamasery must therefore be considered heterodox. The Kanjur blocks of the first impression are said to have dated from the loth or 16th century.
4 Ui [et al.], Catalogue p. 181 sqq. The Ah. bears the serial number 4310.
5 Cf. Meisezahl, Libri x p. 292 sqq.
Introduction 23
The Narthang 1 Tanjur too is accessible in several research, centres. An abstract of its contents was given long ago by Gsoma 2 . The xylograph utilized in the present edition was acquired for the Prussian State Library, Berlin, in 1889 by the imperial minister to Peking, Herr von Bbandt. Like the Derge Tanjur, it was transferred to a salt-mine during World War II, where it sustained considerable damage through inade- quate keeping, and is now deposited in the State Library, Marburg, with 12 (out of 225) volumes still missing. Its paper exhibits a yellow- brown colour, apparently because it was impregnated with an arsenical lye against destruction by insects 3 , and consists of Thymelaeacea fibres, probably of the Daphne genus*. It is entirely done in black ink. Its measurements are 63.5 by 16.5 cm. for the folios and 58 by 8 cm. for the printed surface. Each page bears 7 fines. The text has been put into a rectangular case, with a 1.5 cm. margin on the left of the obverse giving the title of the main division and the number of volume and folio (e.g. mdo. se. bcu-drug). The print is to a large extent difficult to decipher on account of blurs 5 . Sometimes resorting to abbreviations 6 , the carvers have made comparatively few mistakes; in fact, Lattfeb 7 was told by several lamas that the Narthang edition was considered by far the best of them all from the point of view of textual criticism.
The Peking 8 Tanjur, of which apparently only a few impressions were ever struck, is within easy reach of scholars now that the Japanese have
x sNar-than is a small village in Southern Tibet. It is situated about six miles south-west of bKra-sis-lhun-po (Tashilhunpo) in the gTsan (Tsang) province, and is touched by the road leading from gZi-kha-rtse (Shigatse) to Sa-skya. The monas- tery is the personal property of the head-lama of Tashilhunpo, who is known as the Tashi Lama. It was founded in 1153 and saw many a great man. Here the reformer bTsoh-kha-pa (1357—1419) studied Asanga's Yogacara philosophy and the first Dalai Lama (1391—1475) began his clerical career at the age of 7. The wood-blocks of the Kanjur und Tanjur encyclopedias were prepared under Abbot Rig-pai-ral- gri in the years 1730 to 1742 (cf. Tucei, HJAS xii p. 479 sq.). The edition itself goes back to an order given by the sovereign Pho-lha-nas, also called Pho-lha- nas-mi-dban or simply Mi-dban. 2 AR xx p. 553 sqq.
3 Cf. Kirfel, LO xlv p. 5. 4 Cf. Meisezahl, PG viii 2 p. 18 sq.
5 The poor state of the Berlin copy gave rise to the opinion that the Narthang edition as such was inferior in get-up. Nothing could be farther from the truth than this. Laoter (Account p. 7) describes a perfectly clear Narthang Kanjur in the possession of the Newberry Library, Chicago. Obviously the legibility depends on the kind of paper used.
6 E.g. rjesu for rjes-su 1 1.31; yonsu for yons-sw I 2.12 & 3.55; sod for sogs I 3.19; grod for grogs I 3.22; mum for swum I 5.39. ' Loc. cit.
8 Peking, the capital of China, has been the place of issue of several Kanjur edi- tions conveniently named after the emperors on whose direction they were published. The earliest among these is the Yung-lo edition completed in 1410 (Pander, ZE xxi p. (201) ; cf. Laueer, JRAS 1914 p. 1129 sq.), which is said to be a true copy of the old Narthang edition (Ch'en, HJAS ix p. 58) released under Buyantu Khan (1312—20) ('Jigs-med p. 105), and which, incidentally, appears to have been the prototype of the Berlin manuscript (Lauter, JRAS 1914 p. 1129 sqq.). Next came a
24 Introduction
finished reproducing it, along with the Kanjur, by photolithography 1 . It is made up of 224 volumes. An inventory, prepared from the Paris block-print, is owed to Coedieb 2 . The copy consulted in editing the present text was procured in 1955 for the International Academy of Indian Culture. New Delhi, by Rag hit Vera. As is the case with all Peking xylographs, it is printed in red throughout 3 , the paper used being of a stiff, cardboard-like quality. The folios are 70 cm. in length and IS) em. in width, having 8 lines to the page. The 62 by 13.5 cm. frame encasing the text has a 2 cm. margin on either side: that on the left bears a Tibetan legend stating the title of the subdivision or, in the case of longer works, that of the work itself 1 (recto only), the number of volume and folio (recto and verso alike), and whether it is the obverse or reverse (e.g. ylag-brgyad-pa*. se. bcu-drug. gon and se. bcu-drug. og respectively) ; that on the right gives a Chinese translation of it (e.g. ti i-jm shih-pa. i-ching pa chih. shang [ksia]. shih I'm 6 ). The print makes easy reading but shows not a few mistakes; what is particularly interesting in this connection is a dittograph of 87 padas beginning in fol. 25a 7.
til. As concerns the interrelationship of these four xylographs, all that is essential has already been said by Laufee 7 , who took a lifelong
reprint of the Ymig-lo edition, made from 1602 to 1605 and called the Wan-li edition. Scarcely a century afterwards, between 1683 and 1700, another reprint was undertaken, known as the K r ang-hsi or first palace edition (Laufee,, BAIS 1909 p. 567 sgq.). Some forty years later, in 1737, a thoroughly revised and partly enlarged version of it came out, the Ch'ien-lung or second palace edition, which de- rives from the redaction of the fifth Dalai Lama (1617—82) (Citr. p. 44) and passes for the Peking edition par excellence. In contrast to this, the Tanjur seems to have been issued only once, and that probably after 1742; for details see below, p. 32 n. 2.
1 Tripitaka, vol. 46 sgg. The Ah. is found in vol. 141.
- Catalogue, pt. 2 sg. The Ah. is listed in III p. 470.
3 Printing a xylograph in red is considered to be 108 times as meritorious as printing it in black. Similarly, in copying manuscripts and xylographs, the merit of the scribe is raised to the third power of 108 by using various inks, to the fourth power by using silvery ink, and to the fifth power by using golden ink. The figure 108 is sacred; thus the Kanjur usually has 108 volumes, and the lamaist praying string 108 beads. Cf. Unkbig, Erdball II p. 330.
- which then lends its name to the whole volume; thus ylag-brgyad(-pa) occurs throughout the margins of vol. 118, although not only the Ah. but also the Yogasa- taka (fol. lal— 9b5), the Jivasutra (fol. 9b5— 13b4), the Avabhesajakalpa (fol. 13b4— 15ao), and the first hook of the Astangahrdayanamavaiduryakabhasya (fol. 322b5— 410a8) are contained in it.
5 Abbreviation for yan-lag-brgyad-pa, with subjoined I, a short form of yan-lag- brgyad-pai snin-po bsdus-pa also employed by Bu-ston in his reference to the Tibetan Ah. (cf. gSun-'bum, lHa-sa edition, vol. ya, fol. 18b 6). Contracted spellings like ylag are very popular in marginal legends, especially as far as the pagination is concerned. Cf. Meisezahl, Oriens xiii/xiv p. 285 sgg.
6 i.e. "volume 118, medical sutra Astanga, recto [verso], 16." (Transliteration and English rendering by courtesy of Prof. Aitoed Hobtmann, Berlin.)
7 Citr. p. 42 sgq.
Introduction 25
interest in the history and evolution of the Tibetan canon. Thus little remains but to repeat and modify his statements, which are chiefly based on the index volume of the Chone Tanjur, and to make additions wher- ever this seems desirable.
The codification of the Tibetan Tripitaka goes back to the early 14th century 1 when at the instance of the Rev. 'Jams-dbyans, the court chaplain of the Mongol emperor Buyantu Khan (1312—20), a great number of sacred texts still scattered among the country's various hermitages and lamaseries were assembled and incorporated into the old stock of scriptures already existent at the Narthang monastery. This collection, which was in manuscript form, was deposited in the 'Jam Temple and represents the prototype of all subsequent Kanjur and Tan- jur editions 2 .
Every available text having been utilized, it is by no means astonish- ing that the original Narthang copy included a considerable amount of duplicates. To eliminate these as well as to augment, revise, and classify the remaining vast material was the object of research done some years afterwards by Kun-dga-rdo-rje and Bu-ston on the Kanjur and Tanjur respectively 3 . When completed, their redaction, which was again a manuscript set, received an adequate place in the Golden Hall of 2a-lu, where it is still kept*, unless it has fallen a victim to recent developments. In view of its great antiquity and direct relation to the original, the knowledge that it is as good as lost to western scholars is particularly distressing.
Independent of Bu-ston's tradition, there seems to have existed another Tanjur recension, connected with the Central Tibetan district of Chongay fPhyon-rgyas) 5 , of which unfortunately no details are known. To this, Sans-rgyas-rgya-mthso (1620—1705), from e. 1645 ad- viser of the 5th Dalai Lama (1617—1682) and from 1679 regent of Lhasa, drew the attention of his spiritual master, who, in order to facili- tate his struggle against the heterodox red-caps, promptly undertook to prepare a new edition of it. Though doubtless inferior to Bu-ston's version as far as authenticity is concerned, it had a lasting effect on all later editions. Not only was it made the basis of the Peking Tanjur published under the Emperor Ch'ien-lung; it was also amalgamated with
1 Not the first part of the 13th century as claimed by Ch'en, HJAS is p. 53. Cf. Tucci, HJAS xii p. 477.
2 Cf. Ttjcci, Scrolls I p. 107 sq.
s Lattfeb (Citr. p. 53) assigns the Tanjur redaction to the reign of Buyantu Khan (1312—20); but Tucoi points out that Bu-ston (1290—1364) -was still too young at that time to command the experience necessary for such a difficult task (Scrolls I p. 108), and that he first came to iSa-lu in 1320, under the local rule of Grags-pa-rgyal-mthsan (Indo-Tibetica iv 1 p. 79 sq.), while the revision was made even later, under the auspices of the latter's son Kun-dga-don-grub (Scrolls I p. 258). * Cf. Ttjcci, HJAS sii p. 478 sq.
8 For the spelling see Richardson, CAJ viii p. 73.
26 Introduction
Bu-st oil's, revision into the Derge Tanjur of 1733—72 and, by the con- current aid of the original, into the Narthang Tanjur of 1742.
Some extra words must be said about the Chone Tanjur issued in 1772. Lx\ufee 1 , who had no copy of it at his disposal then, took it for a sister contamination of the Derge Tanjur. However, as a mere glance at the apparatus crilicns of the present edition will show, the harmony of the two is so far-reaching as to virtually preclude this possibility. Irrespec- tive of the almost identical pagination, C agrees with D and differs from X, P, or NP in no less than 504 out of 572 instances (misprints reckoned). For countercheck, here is a list of those 68 cases in which C deviates from D :
1.19 yan-dag-par sbyor-bas for yaii-dag sbyor-bas; 1.3S zug-rnu 'byutt-bai dpyad for zug-rnu dbyun-bai dpyad; 1.46 ma-nad snad dan kha-nad for rna-nad sna dan kka-nad; 2. 2 nya-glo-dha for nya-gro-dha;
2. 3 lud-pai hid for lud-pai hi;
2.24 phan-da-gdags-pa for phan-gdags-pa;
2.25 rgyvd for rgud ;
2.28 yoiis-m mgu-ba for yons mgu-ba;
2.30 dbu-mii lam for dbu-mai lam;
2.31 brus for khrus ;
2.41 gtsan-gzan for gcan-gzan;
2.46 Aid mthsan-du for nin mthsan-du;
3. 7 thsa for thse;
3.14 'difov'de;
3.15 slos for spos;
3.17 dgun-sman for dgun-smad;
3.19 pyi for pkyi;
3.21 skom-du sbyor-ba for skom-du sbyar-ba;
3.24 stsogs-sin for sgrogs-sin;
3.27 ni-mai od-zes span for ni-mai od-zer span;
3.30 bu-rum for bu-ram ;
3.35 'bob for 'dab;
3.37 bud-med bees gzugs for bud-med bcos gzugs;
3.39 rdzi-thsar for rdzi-char;
3.41 bde-rtsa for pad-rtsa;
3.44 mi-yi stobs for me-yi stobs;
3.48 sbyor skyo-me for sbyar skyo-ma (2 errors) ;
3.49 'phrul-du for 'phral-du; 3.53 ga-bur for ga-pur;
3.53 'phyens phren for 'phyans phren ;
1 Citr. p. 54.
Introduction 27
3.54 nid for ni;
4. 6 'byun-po for 'byun-no;
4. 6 mas-gtor-ba for mas-gton-ba; 4.10 mi mo for mi mo;
4.13 de-las zas for de-to zas;
4.22 bud-med gzon bston for bud-med gzon bsten ;
4.23 g>sod! dpyad for grso dpyad ;
4.27 srin-lcyi for srid-kyi ;
4.28 <fots dws for dws ;
4.29 sad da?i for sa dan ;
4.30 bsku-zin for &s&w-ze«i;
4.31 des «i mi for des ni me ;
4.32 chags-grugs for chag-grugs ;
4.33 wes zi for Eer zi ; 4.36 brtag-te for brtags-te; 4fin.yan-lag-pa for yan-lag-brgyad-pa;
5. 2 'gyur-pa for gyur-pa ;
5. 3 'bras-chen for 'iras-cfom;
5. 4 'byur-ba for 'gyur-ba;
5. 8 'dris for 'ires;
5. 8 rgya-mthsar bah for rgya-mthsor bab ;
5.24 i/a for j/aw ; 5.28 se£ for sa£;
5.32 mudgai srin for mudgai sran;
5.41 khu-bal for khu-ba ;
5.43 lan-thsa for lan-thsva ;
5.48 ZaZ cfter for pAaZ cAer;
5.56 s« °6n mkhris-par on for sa °6n mkhregs-par on ;
5.56 srid-buioT srin-bu;
5.57 rZ*gr ring's for rZigr Hwas ; 5.59 dti-6a for Mw-6a;
5.62 raft for ra?i ;
5.63 mdog 'gyur khon yuns-Sin for mdog 'gyur khon yans-sin; 5.71 srid-bu fox srin-bu;
5.76 srid for snn;
5.81 on-nad for or-WGwZ;
5.82 rigs rim ji-bzin for rag's rim ji-bzin.
Most of these deviations at once appear to be simple blunders that may be classified as follows :
(1) Dittographs:4.28;
(2) Superfluous letters: 1.46, 2.3, 2.24, 3.54, 4.13, 4.23, 4.29, 4.32, 5.2, 5.41;
28 Introduction
(3) Omitted vowel-signs, letters, or words: 3.7, 4.36, 4/m., 5.8, 5.24,
5.43, 5.58 b;
(4) Confused or miscarved letters: 1.38, 2.2, 2.31, 2.41, 2.46, 3.14, 3.15, 3.19, 3.24, 3.35, 3.39, 3.41, 3.53c, 4.6 bd, 4.27, 4.33, 5.4, 5.8, 5.48, 5.56c, 5.59, 5.62, 5.71, 5.76, 5.81;
(5) Levelled' vowels or consonants: 2.30, 3.17, 3.21, 3.27, 3.30, 3.37,
3.44. 3.48 (twice), 3.49, 3.53d, 4.10, 4.22, 4.31, 5.32, (5.56b), 5.57, 5.63, 5.82.
There remain only six items that cannot be explained as mechanical errors; they concern the insertion of two hyperinetric syllables (1.19, 2.28), the raalcorreetion of two rare words into commoner ones (2.25, 5.3), and finally the elimination of two evident misprints (4.30, 5.28). None of these variants cany enough weight, however, to justify the as- sumption of an independent Chone redaction; while the first four show every mark of an arbitrary change by either the scribe or the carver 2 , the last two may well prove non-existent in the end : be it that the missing vowel-signs broke off from the original wood-blocks or that they did not properly come out in the present copy. Going by the material ex- amined, the Chone Tanjur must be regarded as a direct offspring of the Derge Tanjur".
If this result is correct, it must be possible to find a number of errors that have been passed from the Derge into the Chone xylograph. As it is, however, sometimes difficult in the present state of Tibetology to know corrupt from variant readings, we have given in the following list all cases that appear suspicious either by themselves or from the context :
1 .28 mjr im for sgrin [dahsa] ; 2init. dpyad for spyad [car yd']; 2. 8 brtan for brten [silayet] ;
2. 9 thsul for ihsil [medas] ;
2.14 drags-pas for drags-par [vi°] ;
2.37 rifts for rin [ciram] ;
3. 3 las for lam [mdrga] ; 3. 6 stem for sten [tola];
1 The term "levelling" has tentatively been chosen for the frequent mistake of replacing a given vowel or consonant with a preceding or following one.
2 In making a printing-block, the text is first written on Chinese tissue paper and then pasted face-down, or traced in mirror- writing, on the block; after that, the wood in between the letters is carved off.
3 There is yet another piece of evidence that may be adduced in favour of this conclusion. According to an oral communication by Dr. Eiohaed O. Meisezahl, Beuel, the Chone Tanjur does not embrace a single work that is not found in the Derge Tanjur, while the Derge Tanjur contains quite a number of works that are missing in the Chone Tanjur.
Introduction 29
3. 7 thse for che [agrya]; 3.15 bdug for bdugs ; 3.18a stsogs for sogs [cita]; 3.18o ?ms for nod [roga];
3.34b Wa» dan for <&wi M<m [sKstfa] ;
3.34c khan-pa for kha-ba [hima];
3.35 hal-lha-ra for Tca-lha-ra or lealha-ra [kahlara] ;
3.43 min for smin [pa&a] ;
3.46 tea %i for Ina-yi or tern [jacmca ];
3.53 ur-sin for w-sir [imra];
3.58 thog-mai zag bdun for *Aogr ?n{Aai zagr Mimi [a/i^adKsapfa/ia] ;
4. 6 bcin for spy in [avagaha] ;
4. 9 frjag' for bhag [dharana] ; 4.19 frros for zos [bhuktva] ; 4.27 gzu for 6sm [pacaraa] ; 4.29 dro for gro [godhuma] ; 4.32 dti^s for d/itgr [visa] ;
5. 3 gan-zin for gan-zig [yena] ;
5. 8 6a6for 5 ia6[°sra,];
5.38 rma for rmya [glapita] ;
5.57 shran for sjfcrare [^wZma] ;
5.60 shran skye for dfcra skyed [kesya] ;
5.63 t/aws for yan [laghu] ;
5.66 2/t'w for min [na].
Even if some of these turn out to be genuine variants, they cannot possibly have entered the Chone text otherwise than through the Derge edition.
What has hitherto been said on the transmission of the Tanjur may be illustrated by the pedigree overleaf.
Thus the interrelationship of the several block-prints would be satis- factorily established, were it not for the above-mentioned dittograph in the Peking edition, which requires further investigation. For, strange as it may seem, this dittograph (marked P 2 in contrast to P x ) is not a mere repetition but rather a different recension of what precedes. The matter stands shortly as follows. In 87 dittographed padas there are found 46 variant readings (errors excluded). In no case does P 2 go with P x alone against the other xylographs. In only 7 cases it agrees with NP against CD; but in as many as 21 cases it concurs with CD against KfP. In further 9 cases it deviates from all block-prints. The remaining instances show in 4 cases accord with N, in 1 case accord with DN, and in 4 cases accord with CDN.
Considering their length (for P 2 see cut), it is obvious that both P a and P 2 make almost precisely one folio, though they do not agree with any of the extant folios. Hence it may be assumed that in preparing the Peking edition a leaf of the lost exemplar (that is, the 5th Dalai Lama's
3G
Introduction
1312-20
1320-64
1645-82
1733-72
1742
1742-96
1772
edition) was inadvertently placed among the leaves of the revised text, from where it got unnoticed into the press copy. The question remains which one is the primary version.
The best way of tackling this problem is to examine those readings which Pj and P 2 share with none of the other versions. The following list gives the relevant material and its philological evaluation :
Introduction
31
Three pages from Peking Tanjur (fol. 25a, 25b, 26a) stowing dittograph, with irrelevant passages effaced
32 Introduction
(a) Peculiar readings in P x
5.43b dan for yin: change of words effacing period;
5.43 e rise for thsc [agre] : change of words restoring original ;
5.43d bcad-na for bcas-na [pidita] : change of words distorting sense ;
5.44a 'bus for dbus [jantu }: change of spelling constituting lectio faci- lior\
5.44b shags-pa for {s}pags-pa [samhara]; change of spelling constituting lectio facilior;
5.45 m/iar-ba for mhar-bas [°madhuryaih]: neglect of case-ending ob- scuring syntactical connection ;
5.46b skye for shjes: change of tenses not affecting sense, perhaps only haplography;
5.46b rem for rim [hramat]: error;
5.46d dro for drod [usqa] : interchange of cognates, perhaps only error.
(b) Peculiar readings in P 2
5.30b bskyed for skyed [Jcrt]: change of tenses constituting lectio diffici- lior, thus also in 5.33 bd, 5.37, 5.41b;
5.30c yid-ga for yi-ga: error;
5.34 bsri for sri: change of tenses constituting lectio difficilior;
5. 38 brdol for rdol: change of tenses constituting lectio difficilior;
5.40 sua for rna [karna] : change of spelling constituting lectio diffici- lior, perhaps arbitrary change of words ;
5.41b phyur-ba for chur-ba [kilata]: interchange of synonyms;
5.42b rmad for s-mad [nindita]: change of spelling constituting lectio difficilior.
Applying to this material the rules of textual criticism, especially that on "trivialization 1 ," it must be concluded that P a was made from P 2 , though not without having been heavily edited.
There is one interesting aspect to this situation. If P 2 belongs to the 5th Dalai Lama's edition, then any reading in discord with it (errors and individual changes excluded) must of necessity originate from Bu-ston's recension, which represents the only alternative branch of tradition. Now, as P x is independent of Bu-ston's recension but shares with it, on the strength of the above reasoning, no less than 21 variants, we are given no choice but to suppose that P x (and hence also P) is directly related to N 2 . Such a connection, which, by the way, would easily account
1 Cf. Maas, Criticism p. 13.
2 The Peking Tanjur cannot, therefore, have been published "in the 2nd year of Yung-eheng (1724)," as Cafmr is inclined to believe (HJAS ix p. 58 sq.), but must have been prepared after the completion in 1742 of the Narthang Tanjur, that is, under the rule of the Emperor Ch'ien-lung, which is also Lauter's opinion (Citr. p. 54). It is, consequently, a pendant not of the first but of the second palace edition.
Introduction. 33
for the many other cases of agreement between N and P x , is quite in tune with the eminent position assigned to the ISTarthang xylograph by
Tibetan scholars.
22. From our inquiry into the stemmatical relationship of the extant Tanjur editions it appears that, to use the terminology coined by textual critics, three of the witnesses in question (D, N, and P) have been con- taminated, while the fourth (0) depends exclusively on a surviving exem- plar (D) and must be ehminated as being worthless qua witness. That leaves us with no objective principle to follow in estabMshing the earliest get-at-able version, except where N goes against all other xylographs and hence is likely to reproduce either the original or Bu-ston's edition (peculiar errors and arbitrary changes barred) 2 . We have therefore adopt- ed, as a rule, those readings which are closest to the Sanskrit, which are best in form, style, and matter, or which make it easiest to see how the variants arose. Where two lections of equal value are available, we have given preference to that found in N as the most eminent witness.
23. On the other hand, we have deliberately abstained from altering, however slightly, the transmitted text, though its corruption is sometimes quite obvious, as may be seen from the following instances :
1 init. °sam-hi-tta for °sam-hi-ta l°samhita"\ ;
2. 2 c bslca dan thsa dan kha-ba-dag j rise 'jam for bsha dan thsa dan
kha-ba dan / rise 'jam [mrdvagram kasayakatutiktakam] ; 2.22 a de-ltar sdig-pai las-mams bcu for de-ltar sdig-pai las mam bcu
[papam karmeti dasadha] ; 2.39cd chan 'ihson-ba dan sbyor-ba-dag / sbyin dan len-la spyad mi bya for
chan 'thson-ba dan sbyor-ba dan / sbyin dan len-la spyad mi bya
\madyavikrayasamdhanadandda,nani nacaret] ; 3.38 d 'dod-pai rgyun for 'dod-pai rgyud [kamatantra]; 3.52c dan-bai dri-med chu zes-pa for nan-pai dri-med chu zes-pa [sue*
hamsodakam ndma] ; 5.34a skran-nad for skran-nad [sopha]; 5.45 c bsil dan mnar-bas for bsil dag mnar-bas [saityaprasadamadhuryaih].
Of these instances the sixth is particularly interesting inasmuch as CD have substituted dans-pai for dan-bai. Evidently the redactors did not recognize the corruption and, instead of eliminating it, perpetrated a malcorrection.
1 including mistakes like la for nal [Arama] 2.36, po for dbart-po [indriya] 4.25, Tier zin for ner zi [upajama] 4.33, gan for gar [ghana] 5.6, and ro-mi bsal-iifi for ro-ma bsil-zin \vrsyam himarri] 5.19.
2 Obsolete spellings also indicate an early stage of tradition. But as these occur only sporadically in NP and would disturb the otherwise uniform orthography of the Tibetan text, they have been relegated to the apparatus criticus. Such cases are, e.g., rgyun-tu for rgyun-du (1.1), brtan-cin for brtan-zin (2.26), dus-du for dus-su (3.17), mnar-ste for mnar-te (3.56), and gian-dag-du for gian-dag-tu. (3.57).
3 Vogel, Vagbliata
34 Introduction
2-1. More frequent than such cases where the Tibetan text may be corrected after the Sanskrit are those in which a Sanskrit variant can be inferred from the Tibetan. Here are some examples:
2. 1 smtha [bde-gnas] for svastha ;
2.10 trddho jlrnas ca [rgas-sin 'khogs-pa] for vrddho 'jtrril ca ;
2.44 madyadisaktim [chan sogs chags] for madyatisaktim;
3.29 b mbahuwri K [chu mart, bsres-la] for subahu van B; 3.29 e iom \kha bskams] for iopha;
3.39 jaJardratalavrntani [ta-lai bsil-yab spos-chus btab] for jalardras Uilarrntani ;
5.30 sltake visame jvare [gran-bai rims dan mi snoms rims] for sitake visamajvare ;
5.42/3 iksoli mro . . . rasah K [bu-ram sin khu 'khrii] for ikso raso . . .
sarah B :
5.52 meda [tJisil] for incha ;
5.55 ivagdosahrt [pags-pai Ties sel} for tvagdosakrt ; 5.70 sosaiopJio [Ins skem skran-or] for sophasoso° ; 5.77 hlnda [sdon-bu] for kanda.
Elsewhere the conjectural reading is too remote from the transmitted text to be generally acceptable 1 , or a satisfactory solution cannot be reached at all 2 .
Two further cases deserve special attention as they throw some light on the technique of revision adopted by the Tibetans. In 5.32 NP have klta-ra sa, which implies a variant sitapala "sugar and meat," while CD give m-Ma-ra, which agrees with the transmitted sitopala "sugar- crystal, sugar." Similarly, in 5.55 NP read bad-kan bskyed, which pre- supposes a variant kaphahrn na "not eliminative of phlegm," whereas CD offer bad-kan sel, which corresponds to the transmitted Jcaphakrn na "not generative of phlegm." No matter whether NP or CD furnish the primary version, neither change could possibly have been made without the help of a Sanskrit copy. This is particularly noteworthy in view of the fact that other redactionai changes leave a different impression. To cite but one characteristic instance, in 4.1 and (directly connected with it) in 4.17 NP write mchi-ma "tear" for the original airu and baspa, while CD, irrespective of the basic text, put nichil-ma "spittle" in its place, thereby affecting the very substance of the passage 3 .
- E.g. amayan [nod] for asrayau (5.58), pitta [mkhris-pd] for ku?tha (5.60), sukra or blja [khu-chu] for pitta (5.61), khedita [sun] for virikta (5.66).
2 E.g. arucau: kha-zas ten (5.30), vdaram, vi§amajvaram: Ito-nad Wan rims-nad (5.57), amJapdkarasam: zu rjes skyur-zin cun-zad 'khrv, (5.63), mrdu: sar-pa (5.66).
3 Similar distortions are rdzas ni khu-ba CD for rdzas-kyi zu-ba NP [wpako dravyasya] (1.17), og dkri NP for mgo dkri CD [rnaidi] (2.32), nal-log NP for nal Uog CD [svapnadhyayana] (2.42), lud-pa CD for lud nal NP l&ramasvasa] (4.1), rmen-bu CD for rma-f)bu NP [vrayajantv,] (5.59).
Introduction 35
25. In the same manner that variants are deducible from the Tibetan, interpolations may be uncovered with its aid. Generally speaking, any line or verse omitted by the translators is subject to suspicion, regardless of whether or not additional evidence shows it to be spurious. Thus 3.1 cd, 4.3, and 4.4cd do not occur in K either, whereas 2.41 cd, 4.20ab, 4.22cd, 5.36cd, 5.72cd, and 5.73ab appear in all the copies consulted. Things are different only with 3.35a, which is found in CD but wanting in NP ; here we probably have before us an oversight on the part of jNTP, though nothing definite can be told because of the intricate sentence-con- struction. It ought, however, to be borne in mind that the Tibetan reflects not the original but only the earliest reachable form of the text, and that inclusion in the Tanjur alone is no sufficient proof for genuineness. This applies, for instance, to the line on belching (1 4.9), which has no counterpart in the preceding argument, to the section on urine (I 5.80 sq.), which is not mentioned in the immediately following summary, and to the epilogue (VI 40.59 sqq.), which Htlgenberg & Iyirfel have demonstrated to be a counterfeit on internal grounds x .
Lacunae, on the other hand, could not be traced so far. There are a number of surplus lines in the edited portion, to be sure, but these result either from paraphrasing an unusual term (as 3.30 e) or from completing an ellipsis (as 5.55b), or else bear the stamp of an insertion (as 2.12ed). The reverse phenomenon that two stanzas have been condensed into one is also met with (cf. 1.37, 4.7 sq.), but cannot satisfactorily be ac- counted for.
26. When it comes to estabhshing the sense of a word or group of words, or determining the construction of a sentence, susceptible of more than one possibility, it is only natural that the Tibetans do not always agree with the opinions advanced in Axunadatta's Sarvanga- sundarl or Indu's Sasilekha, and followed in our translation of the San- skrit text. It is, however, noteworthy that they do not as a rule concur with Candranandana's Padarfchacandrika. either, in spite of this commen- tary itself being part of the Tanjur 2 . All such discrepancies will be stated at their appropriate places, though, and mention shall be made here only of some particularly remarkable interpretations, eccentric or otherwise.
In 2.3, for example, dantamamsa "tooth-flesh" [i.e. gums] is regarded as a copulative instead of determinative compound and is translated by so dan mil "teeth and gums," perhaps with reference to the parallel passage As. I 3.16, which has both danta and dantamamsa. In 2.7 and 5.17 just the opposite has happened: visamurcchd "poison and stupor" and amapinasa "rawness [i.e. indigestion] and catarrh" are considered determinative rather than copulative compounds (as is in the last case
1 See above, § 4.
2 Not until our translation was almost complete did we procure a copy of Can- dranandana's work.
3*
36 Introduction
also done by Candranandana) and are rendered dug-gis brgyal "fallen in a toxic stupor" and chain sar "raw catarrh" respectively, though. Vag- bhata does not describe these diseases anywhere. In 3.33 sitala "cold" is given the unusual though possible sense of dkar-ba "white," probably on account of the preceding sasita "sugared" having read sasita "cooled" in the translators' copy. In 5.67, lasty, grahani "dysentery" is changed into its antonym rtug skem "dry stool" [i.e. costiveness], maybe by association with the cognate grahin, which can also mean "constipating." In 3.24, on the other hand, the compound manikuttimakanti is ex- plained by all scholiasts as a possessive dependent : manayo vajramara- katadayas tatkrtani kuttimani taih kantir yesam tani "those whose splendour (is produced) by tessellated pavements made of jewels (such as) diamonds, emeralds, etc." (Indu's paraphrase) ; the Tibetans, however, resolve it as a possessive descriptive or rather appositional possessive : nor-bu bcag 'drai mdans Idan-pa "showing a splendour like (that of) tesseral jewels," thereby avoiding the somewhat fantastic idea of a forest ground laid with gems. In 3.37 saudha generally passes for a substantive synonymous with harmya "mansion," whereas the trans- lators understand it in the adjective sense of rdo-thal byugs "coated with mortar," thus saving common people from having to sleep on, of all places, a palaee-roof garden. In 3.41 the phrase mrnalavalayah kdntdh protphullalcamalojivalali "beautiful women possessed of lotus-fibre brace- lets (and) radiant with full-blown lotuses" has been turned chuh-ma gdu-bu pad-rtsa 'dra / padma rgyas-pa Ua-bur mdzes "women beautiful as full-blown lotuses in their lotus-fibre-like bracelets," with the simile of the next line having been extended, against tradition but nevertheless convincingly, to the whole stanza. In 5.11 pracya, avanti, and aparanta are interpreted to signify, not the peoples of Gaur, Malwa, and the Kon- kan, as the commentaries have it, but rather the lands inhabited by these peoples, which is equally correct and, in view of what follows, even more satisfactory (the Tibetan text itself being somewhat at variance with the original Sanskrit). In 5.81, finally, laghu is taken, not for a predicate noun ("light"), as the scholiasts suggest, but for an adverb ("quickly"), which, judging from its position, indeed seems to be its true function in the present context.
27. Turning in conclusion to the translating-technique of the Tibet- ans, a minute description has been given for each individual stanza, so that only the most salient features need be outlined in the present context. Among these is prominent the roundabout fashion of rendering certain terms, especially medical, for which no proper equivalents are at hand 1 . Two groups may be distinguished:
1 Sanskrit words, unless already forming part of the Tibetan language, are used only as a last resort. Those met with in the present text and not yet listed in the dictionaries include pa-ta-la [pStala] 3.32, kalha-ra with v. 1. kal-lha-ra [kalhara] 3.35, mudga [mwdga] 3.51, 4.29, 5.32, sa-hya [sahya] 5.11, pa-ri-ya-tra [pariyatra] 5.12, and Jsanta-ra with v. 1. ka-ta-ra [Mntara] 5.46.
Introduction 35
25. In the same manner that variants are deducible from the Tibetan, interpolations may be uncovered with its aid. Generally speaking, any line or verse omitted by the translators is subject to suspicion, regardless of whether or not additional evidence shows it to be spurious. Thus 3.1 cd, 4.3, and 4.4cd do not occur in K either, whereas 2.41 cd, 4.20ab, 4.22cd, 5.36cd, 5.72cd, and 5.73ab appear in all the copies consulted. Things are different only with 3.35a, which is found in CD but wanting in NP ; here we probably have before us an oversight on the part of jNTP, though nothing definite can be told because of the intricate sentence-con- struction. It ought, however, to be borne in mind that the Tibetan reflects not the original but only the earliest reachable form of the text, and that inclusion in the Tanjur alone is no sufficient proof for genuineness. This applies, for instance, to the line on belching (1 4.9), which has no counterpart in the preceding argument, to the section on urine (I 5.80 sq.), which is not mentioned in the immediately following summary, and to the epilogue (VI 40.59 sqq.), which Htlgenberg & Iyirfel have demonstrated to be a counterfeit on internal grounds x .
Lacunae, on the other hand, could not be traced so far. There are a number of surplus lines in the edited portion, to be sure, but these result either from paraphrasing an unusual term (as 3.30 e) or from completing an ellipsis (as 5.55b), or else bear the stamp of an insertion (as 2.12ed). The reverse phenomenon that two stanzas have been condensed into one is also met with (cf. 1.37, 4.7 sq.), but cannot satisfactorily be ac- counted for.
26. When it comes to estabhshing the sense of a word or group of words, or determining the construction of a sentence, susceptible of more than one possibility, it is only natural that the Tibetans do not always agree with the opinions advanced in Axunadatta's Sarvanga- sundarl or Indu's Sasilekha, and followed in our translation of the San- skrit text. It is, however, noteworthy that they do not as a rule concur with Candranandana's Padarfchacandrika. either, in spite of this commen- tary itself being part of the Tanjur 2 . All such discrepancies will be stated at their appropriate places, though, and mention shall be made here only of some particularly remarkable interpretations, eccentric or otherwise.
In 2.3, for example, dantamamsa "tooth-flesh" [i.e. gums] is regarded as a copulative instead of determinative compound and is translated by so dan mil "teeth and gums," perhaps with reference to the parallel passage As. I 3.16, which has both danta and dantamamsa. In 2.7 and 5.17 just the opposite has happened: visamurcchd "poison and stupor" and amapinasa "rawness [i.e. indigestion] and catarrh" are considered determinative rather than copulative compounds (as is in the last case
1 See above, § 4.
2 Not until our translation was almost complete did we procure a copy of Can- dranandana's work.
3*
38 Introduction
Above all belongs here the additional translation of Sanskrit pre- fixes already implied by the Tibetan simplexes, e.g. yons-su (tyrtag for parikseia '"one shall examine" (1.21), ner bsten for upasaya "reaction" (1.22), rjes (b)rtse for anurakta "beloved" (1.28), mam 'gyur for vikrii "alteration" (1.39). and Jcun-tu spyod byed-pa for xamacaran "practising" (2.47). (3) Hendiadyses such as bsku-byug "besmearing & anointing" for abhyanga "inunction" (2.7, 2.9, 4.6), dril-pfajis "rolling & wiping" for udmriana "massage" (2.14), mig nad gyan-pa "disease & itching [i.e. morbid itching] of the eyes" for aksikaiidu "itching of the eyes" (4.18), chag(s)-grugs "breaking & crumbling" for bJianga "fracture" (4.32), and bkres -Tien nad-pa "pained & diseased [i.e. painfully diseas- ed] with hunger" for ksudhatura "pained with hunger" (5.66).
This last mode of expression is employed frequently in rendering action- nouns.
It is well known that, as a basis for the translating committees to rely on, special vocabularies were prepared in which every Sanskrit word had got, as far as possible, only one Tibetan equivalent in order to secure a maximum degree of uniformity. Nevertheless the present text shows a certain fluctuation of terminology not readily under- standable from the necessity of suiting the metre. If this reflects an early stage of translating-technique or some kind of poetical license or just the natural unsteadiness of phrase to be expected in all such works, ancient or modern, there is no way of telling. A few examples will suffice to illustrate the point :
aiisara "tropical diarrhea" —
(1) thsad(-pai) nad "heat disease" 1.40, 5.24 (cf. thsad 5.52);
(2) thsad(-pas) 'khru "heat(-caused) diarrhea" 2.17, 5.13. aria* "hemorrhoids" ~
(1) gzail-'brum "anus pocks" 1.40, 1.42, 5.13 (cf. 'brum 5.59);
(2) gzah-nad "anus disease" 5.25. graham, °gada, °dosa "dysentery" — -
(1) pho-bai drod chuh "poor heat of the stomach" 5.14;
(2) 'kkru-bai nad "dysenteric disease" 5.31 ;
(3) pho-nad "stomach disease" 5.34, 5.70;
(4) rtug skem "dry excrements" 5.67 (see § 26). dipana "digestive" ~
(1) drod {b)skyed "producing (digestive) heat" 2.15, 5.19, et passim;
(2) drod che "rich in (digestive) heat" 5.25;
(3) drod 'bar "(digestive) heat being kindled" 5.62. hrdya "cardiac" ~
(1) yi-gar on "meeting one's appetite" 5.1 ;
(2) yid{-du) on "suiting one's mind" 4.30, 5.48, 68, 76, 79;
(3) silin-la phan "being wholesome for one's heart" 5.71 (cf. ahrdya - sniii gnod 5.26).
Introduction. 39
The last instance deserves special notice inasmuch as it presupposes a differentiation of meaning not immediately deducible either from the context or from the scholia. In 5.1, it is true, Indu glosses hrdayasya priyam "dear to the heart" and Candranandana remarks hrdayaya hitam hrdyam hrdayasya va priyam "hrdya (means) good for the heart or dear to the heart," whereas Arunadatta states hrdayaya hitam na tu hrdayasya priyam iti hrdyam iti vyahhyeyam "hrdya (is) to be explained as wholesome for the heart, but not as dear to the heart 1 ." On the other cases, however, Indu does not comment at all and Aruriadatta only at times {hrdayaya [a]hita "[not] good for the heart" 5.26, 48, 68), while Candranandana interprets hrdayaya kita "good for the heart" in 5.68, hrdayaya priya "dear to the heart" in 5.71, and hrdayapriya "not dear to the heart" in 5.26, which is just the opposite of how the Tibetans have understood these passages.
In the same way that one and the same Sanskrit term is rendered by different Tibetan terms, one and the same Tibetan term may correspond to different Sanskrit terms, thus bringing about a certain ambiguity of nomenclature. The majority of these cases are doubtless due to a lack of proper equivalents, as can be seen from the following examples : chan "spirits" -—
1) madya "spirits" 2.44 et passim;
2) surd "arrack" 3.12, 4.21;
3) arista "liqueur" 5.70 (cf. 3.22, 45). chan dans "clear spirits" —
1) acchasurd "barm" 3.12;
2) varunl "toddy" 5.68. to sbos "inflated belly" —
1) adhmana "inflation" 2.18, 5.17;
2) anaha "constipation" 5.25. bras "pimple" ~
1) hatha "urticaria" 4.18;
2) parlsarpa "erysipelas" 5.38. sa bkra "motley flesh" ~
1) svitra "white leprosy" 1.43, 5.81 ;
2) Tcotha "urticaria" 5.59.
Seeking a way out of the difficulty, the translators merely substituted general Iterms for specific ones.
On much the same level are a number of other cases in which, for no apparent reason, the Tibetan is slightly out of keeping with the original, implying either (1) more, (2) less, or (3) something else than is expressed
1 It may be noted en passant that according to Panini's Grammar the taddhita suffix ya means "wholesome" in general (tasmai hitam V 1.5), but "dear" after hrdaya (hrdayasya priyah IV 4.95). The problem with which the scholiasts struggle in the present context seems to be whether rule V 1.5 is annulled or supplemented by rule IV 4.95.
40 Introduction
by the Sanskrit. Though no serious distortions result, the matter seems important enough to justify a few examples for each category :
(1) iariracinta "care of the body" — - lus-lcyi bya-ba "affairs of the body" 2.1 ; tufa or anila "wind" - {b)ser-bu "fresh breeze" 2.40, 3.23, 3.55 ; hrllasa "palpitation of the heart" ~ ro-stod mi bde "indisposed upper part of the body" 4.18; lutaditantu "webs of spiders etc." — srog- ckags gdug dan-ba s{ts)ogs "webs etc. of poisonous insects" 5.8; hastisodhcma "purgative of the bladder" ~ gcin-nad sel "removing urinary diseases" 5.16.
(2) damstrin "tusked or fanged animal" -~ sbrul gdug "venomous snake" 2.41; snigdha "fat" ~ mum-bag "slightly fat" 2.11, 3.26; guhyavedana "pubic pain" ~ pho-mthsan na "penile disease" 4.20; dravadravya "liquid substance" ~ chu sna-ihsogs{-pa)-la s{ts)ogs-pa "all sorts of water etc." 5 init.; grakin "injurious" —- mi bzod-pa "unbearable" 5.7.
(3) garbhave&man "inside room" ~ khan-pa nis-rim byas "ground floor" 3.16; drava "liquid" ~ sla-ba "thin" 3.28; ksata "rupture" ~ mthson- rmas [v.l. -smas] "knife(-inflicted) wound" 4.32; lehhana "stimula- tive" — • nod sbyoii "curing diseases" 5.71 ; Mama "weariness" ~ rmya "exhaustion" 5.79.
An interesting counterpart of the last category form those compara- tively rare cases in which the translators, having regard to the changed circumstances, replace obviously unsuited words by substantially differ- ent ones. In 2.23, e.g., devout people are required to treat a cow [go) with deference; cow worship, however, being unknown in Tibet, dge-slon "monk" has been written instead of go. In 2.38 decent folks are warned against coming near a cremation ground (smasana) ; but since bodies are either buried, embalmed, burned, or cut in pieces and fed to animals by the Tibetans 1 , dtir-khrod "funeral place" has been substituted for smasana. In 3.48, lastly, health -conscious persons are advised to stay on sheltered roof-gardens free from vapour (vaspa) during the rainy season; moisture issuing from the ground, however, being more likely to freeze than evaporate in the rough climate of Tibet, ba-mo "hoar- frost" has taken the place of vaspa.
Another trait, one that is more or less contingent on the nature of the Tibetan language, is the practice of verbalizing the nominal style of the original Sanskrit. This policy often goes beyond the customary substi- tuting of full verbs for predicate nouns whose copulas are missing, as may be seen from the following confrontation (4.11 sq.) :
in Sanskrit —
sosangasadabddhiryasammohabhramahrdgada,^, // trsnaya nigrahat tatra sitah, sarvo vidhir hitah j angabhangarucigUnikarsyaiulahhramah ksudhah //
1 Cf. Koepken, Religion II p. 322 sq.
Introduotioa 41
Xerostomia, flaccidity of limbs, deafness, stupor, giddiness, and heart disease (result) from the restraint of thirst. In this case every cold application (is) -wholesome. Backing in the limbs, anorexia, lassitude, emaciation, stitches, and giddiness (result from the restraint) of hunger.
in Tibetan —
skom-pa bsgags-pas kha skams-iin /
yan-lag mi bde rna mi gsan //
mgo 'khor myos-sin snin-nad 'byun /
der bsil cho-ga ihams-cad phan /
bkres-pas lus iig yi-ga 'chus /
lus snoms nam chun gzer mgo 'khor \\
Through restrained thirst one's mouth is dry, one's limbs are not well, one's ears are not quick, one's head spins, one is stuporous, and heart diseases arise. In this case every cold application is wholesome. By hunger one's body is ruined, one's appetite disturbed, one's body lazy, one's strength poor, one feels pain, (and) one's head spins.
Then there is the peculiar manner of handling the original word- order, which, on the ground of idiom and metre, often runs directly counter to the fixed rules of Tibetan syntax. Two opposite courses are taken by the Lamaist translators in tackling this problem. Either the sentence-construction is preserved and the word- (or pada-) order changed or, less frequently but all the more remarkably, the word-order is retained and the sentence-construction altered, irrespective of any distortion that may result. Here is one such case by way of example (3.42):
in Sanskrit —
adanaglanavapusam agnih sanno 'pi sidati / varsasu dosaik //
The (gastric) fire of those whose body has been emaciated by (the period, of) absorption, though (being already) weak, is (further) weakened during the rains by the humours . . .
in Tibetan —
mthu 'phrogs-pas ni na-bai Iua / dman-par-las kyan dman gyur-te / dbyar ni gnod-pas gnod 'gyur-te / //
The body (which is already) sick by the absorption of strength, after having become even weaker than weak, gets (further) affected by the humours in the rainy season . . .
42 Introduction
Last but not least a phenomenon belongs here that is somewhat alien from the nature of Tibetan and, therefore, may well be characterized as a sort of Sanskritism. It concerns the proleptic use of the governing verb and its resumption by an auxiliary as met with, for instance, in the following stanza (5.68) :
in Sanskrit —
pei/am nosnopacarerjia na viriktaksudhaturaih j imtyarthatlksnamrdvalpasambharam kalusam na ca 1 1
(Alcohol is) to be drunk neither by one who is engaged in warming activities, nor by those who have been purged and who are pained with hunger, nor (when it is) exceedingly fierce and mild and made of inferior material, nor (when it is) turbid.
in Tibetan —
thsa-bai spyod-par Idem mi btw'i / sun-cm bkres fieri nad-pas min / iin-tu mo min sar-pa dun j legs-par ma bslans rnog-can min //
(Alcohol) shall not be drunk by one who is engaged in warming activ- ities, nor by one who is tired and pained & diseased with hunger, nor (when it is) exceedingly fierce, nor (when it is) fresh, not properly
prepared, (and) turbid.
Syntactically, the appended clause may be taken for an ellipsis to be completed by repeating the principal verb, say like this: Uun-bar bya-ba min. Judging from similar cases such as 2.37 sq. and 5.32, it would seem that this construction is preferred in negative sentences. But sometimes it occurs in affirmative sentences as well, so in 1.3:
in Sanskrit —
Brahma smrtvayuso vedam Prajapatim ajigrahat /
so 'svinau tau Sahasraksam so 'triputradikan mwnm \\
Brahman, having recalled medical science, taught (it) to Prajapati; he, to the two Asvins; they, to the Thousand-eyed One; he, to the sages Atriputra etc.
in Tibetan —
Thsans-pas thse-yi rig-byed dran j sKye-dgui-bdag-la bSad-pa yin j de-yis Tha-skar des brGya-byin / des rGyun-ses stsogs dran-sron-lao //
Introduction 43
Brahman, having recalled medical science, told (it) to frajapati; he, to (the two) Asvinl(putras); they, to the Hundred-powered One; lie, to the sages Atri(putra) etc.
Here the auxiliary is even missing, and its office is taken by the final o,
Time and again Tibetan versions have been praised for their almost proverbial faithfulness, especially when compared to their Chinese counterparts, and repeated attempts have been made even to recon- struct the wording of lost originals with their help. Justified though such a policy may seem in the case of prose compositions, it is certainly out of the question as far as metrical works are concerned, Here the fixed number of syllables and lines allotted to each line and stanza 1 calls for many changes, both slight and drastic, not to speak of those numerous instances where the text is altered for no apparent reason whatsoever. In fact, as the following analysis of the first five chapters of Vagbhata's Ast anganrday asamkita will show, hardly a single stanza can be traced that is exactly alike in Sanskrit and in Tibetan.
1 On Tibetan metrics see Beosh, Beitrage p. 53 &$,
First Chapter
Opening Statement
In Tibetan —
[15b 1] 15a5 yan-lag-brgyad-pai snin-po bsdus-pa slob-dpon Pha-gol-
gyis mdzad-pa bzags-so // // 1
(44 b 1 ) ( (44 b 1 )) rgya-gar skad-du j & astam *-ga-hri-da-ya z -sam-hi-tta na-
ma / bod skad-du / yan-lag-brgyad-pai snin-po bsdus-pa its bya-ba j bcom[2]-
Idan-'das de-bzin-gsegs-pasman-gyi bhbai-durya od-kyirgyal-po-h phyag*-
'thsal-lo jj
1 N ; double sad P ; passage missing in CD. a NP ; asp-am CD .
3 CD insert a simple sad here. * N adds 'phyag.
In English —
The Collected Essence of the Octopartite (Science), composed by Master
Vagbhata, is contained (in what follows).
In Sanskrit (it is entitled) Astangahrdayasamhita mama; in Tibetan,
Yan-lag-brgyad-pai snin-po bsdus-pa zes bya-ba.
Reverence to the Victorious One, the Thus-gone One, the Medicine
Master, the Cat's-eye-splendoured King!
Remarks
In their opening statement, the translators announce the new work and its author (this part is wanting in CD), record its full title in Sanskrit and Tibetan, and pay homage to Buddha as the lord of medicine (on this aspect see Waddell, Buddhism p. 353 sg.). Some points of interest may here be raised.
The Sanskrit title, to begin with, admits of two interpretations, both equally common in the existent literature, according as its initial member (astdnga) is considered a determinative or a possessive compound; in the first case it would read, "collection of the essence of the eight parts (of medicine)," in the second, "collection of the essence of the eight-membered (science)." The Tibetan title, on the other hand, bears only one interpretation, namely, "collected essence of the eight-membered (science)," the words yan-lag-brgyad-pa having a possessive meaning because of the suffix pa and ranking equal with terms like yi-ge-drug-pa "six-letter (sequence)" for the famous formula om ma-®i pa-dme Mm or sum-cu-pa "thirty(-letter sequence)" for the Tibetan alphabet (cf. Jaschke, Grammar p. 33). Besides, it should be noted that the final member, hrdayasamhita "collection of essence," has been denominalized into snin-po bsdus-pa "collected essence." This phenomenon, which occurs very frequently, is explained by the fact that in Tibetan verbal nouns and adjectives cannot be substantivated to such a degree as to assume nominal construction.
First Chapter 45
The author's name, Vagbhata (i.e. "slave of speech"), has been translated instead of transliterated, which is in keeping with common Tibetan usage. Its present equivalent, Pha-gol, is an inaccurate form of Pha-hhol (Mongolian Ebige-yin boyol, i.e. "slave of one's father"), which recurs in the colophon but has there been amended by CD. While the second members are thus brought into full harmony with each other, the first components remain at a sharp variance. It ought, however, to be noted that instead of Vagbhata the Prabandhac. V20 (p. 314 DinInatha, p. 199 Tawney) twice gives Bdhada, whereas Bengali and South Indian manuscripts often read Babhata and Vdhata respectively. This leads one to assume that Vag- bhata is nothing but a subsequent Sanskritization of an original Prakrit name, and that the translators had such a vernacular spelling before them (cf. Hindi & Bengali bap, Kashmiri bab, Nepali bd, "father").— In Svapnac. II 160, our author is called V&caspati (i.e. "lord of speech"), which presupposes a variant Vdgbhatta; but that seems to be an etymological pun rather than a serious attempt at explain- ing the formation of the word.
The salutation, lastly, reminds one of a passage in 1 18.18 which forms part of a spell (mantra) and runs as follows:
om namo bhagavate bhai$ajyagurave vaiduryaprabhardjaya 1 tathdgatdydrhate
samyaksambuddhdya /
Om! Reverence to the Victorious One, the Medicine Master, the Cat's-eye 2 -
splendoured King, the Thus-gone One, the Saint, the Fully Enlightened One ! Interestingly enough, the Tibetans have retained here the original Sanskrit, transliterating it in the manner of a mystic formula (dhdranl) : s
om (3) na-mo 6Aa-g , a-6a*((3))-*e / bhai-sa-dzya^-gu-ru bai-dHrya e -pra-bha-ra-dzd-
ya I ta-thd-ga-td-ya' 3 f arha s 4e / sam-tnyak 9 -sam 10 -btiddha-ya / This is not, however, to say that both salutations are directly related to each other; on the contrary, they doubtless derive from a current Buddhistic pattern.
As concerns the variant readings, CD have resolved the sandhi of a?tam-ga into asta-am-ga, while N has inserted an erroneous 'phyag after phyag. It may also be observed that all xylographs print sam-hi-tta, which appears to be an old corruption of sam-hi-ta.
1 B ; vaiduryaprabhdrajdya K.
2 For this identification of vaidurya see Finot, Lapidaires p. xlv sqq.
3 (83b2), ((83b2)), [58b 7], 57b 1. * CDN; m P. 5 CDF; jye N. « DNP ; durya, C. 7 CD ; ta-yd NP. 8 or a-rha CDN; ar-ha P.
9 P; sam-myag CD; sammyak N. 10 P; sam CDN.
II
Salutatory Stanza (1) Sanskrit Text ragadirogan satatdnusaktan asesakayaprasrtan aAesan / autsukyamoharatidan jaghana yo 'purvavaidyaya namo 'stu tasmai J I
Reverence be (paid) to him, the unprecedented physician, who destroyed all diseases— (such as) lust etc.— perpetually clinging to (and) spreading over all bodies [the whole body], causing desire, ignorance, and ill-will.
46 First Chapter
Tibetan Version
, dod(2)'7-chags-la ((2)) sogs ma-lus-pa-yi nad /
rgyun-du 1 'brel-bas Iws kun ma-lus khyab /
'dod dart- ijti-niug kkro 9 -ba sel-ba-yi /
sman-pa snoa-mcd de-la phyag[3]-'thsal-lo //
1 CD ; tu NP. 2 CDP ; dro N.
Reverence to him, the unprecedented physician, removing all diseases— (such as) lust etc.— perpetually clinging to and spreading over each and every body [the whole and entire body], (namely), desire, ignorance, (and) hatred.
Remarks
The salutatory stanza, composed in the Upajati metre, begins the Sanskrit text of the Ah. Like the final portion of the opening statement, it is addressed to Buddha in his capacity as lord of medicine; for though no name is mentioned and the terminology is slightly different, an allusion appears to be made here to the three moral poisons (visa, dug) in Buddhist philosophy. These are known as lust {raga, 'dod-chags), hatred (dvesa, ie-sdan), and ignorance (moha, gti-mug) and are symbolized by a dove, serpent, and pig respectively in the nave of the wheel of life (for details see Waddell, Buddhism p. 105 sqq.).
As regards the Tibetan version of this stanza, the following may be noted (the lemmata are arranged according to the Sanskrit text) :
'dod-chags, the common equivalent of raga "lust," is, properly speaking, a hendi- adys signifying "love and desire," that is, "amorous desire."
sogs is given here by all block-prints alike. As will later be seen, CD often write stsogs instead, both forms being approximately in the ratio of three to seven. Though nothing definite can be told yet, stsogs seems to be an orthographic peculi- arity rather than an antiquated spelling. For rgyun-du XT have an obsolete rgyun-tw.
asesakdyaprasrta, "spreading over all bodies" [or "spreading over the whole body," as this compound may alternatively be resolved according to Arunadatta] has been translated tautologically by lus kun ma-lus khyab "spreading over each and every body" [or "spreading over the whole and entire body"]. Both aiesa and ma-lus literally mean "without exception, without remainder."
asesa, separated from its governing noun by two intervening appositions, has been transferred to its proper place before roga, the rhetorical figure of hyperbaton being incompatible with Tibetan syntax. For khro-ba IS has a misearved dro-ba.
After autsukyamoMrati the root-suffix da "causing" has been omitted, which makes it necessary to interpret the remaining words as a specification of ragadi. This would also accord with the fact that raga and autsukya have been rendered by similar terms, namely, 'dod-chags and 'dod.
The relative clause jaghana yah "who destroyed" has been expressed by a parti- cipial clause, with the original perfect having been replaced by a present: sd-ba "removing."
The predicate astu "be (paid)" has been dropped and its office taken by the final o of phyag-'thsal-lo.
tasmai, lastly, has been interchanged with namas on grounds of syntax.
First Chapter 47
II
Introductory Line Sanskrit Text
athata ayuslcamiyam 1 adhyayam vyahhyasyamali / iti ha smakur Atre- yadayo maharsayah J I
1 K adds nama.
Now we shall set forth the_ chapter concerning the wish for long life. Thus spoke the great sages Atreya etc .
Tibetan Version
de-nas thse rin-bar 'dod-pai leu bsad8-par byao / rOyun-ses-kyi-bu-Ia (3)
sogs-pa dran-sron chen-po((3))-rnams-kyis 'di-ltar bsad-do //
Now will be set forth the chapter on wishing life to be long. Thus was spoken by the great sages Atreya etc.
Remarks
ayuskamiya "concerning the wish for long life" has been turned thse rin-bar 'dod-pai "on wishing life to be long." Putting nominal phrases verbal and express- ing suffixes of appurtenance by objective genitives is typical of the Tibetan trans- lating-technique.
vyahhyasyamah "we shall set forth" has been rendered Mad-par byao "will be set forth," with the agent left unnamed, so that the passive construction in English comes closest to the original.
iti "thus" usually refers not to what follows but to what precedes. Arunadatta, however, takes it here to stand for evam, which has both meanings. Hence its translation into Tibetan by 'di-ltar.
Atreya has, in accordance with Mvy. 3461, been Tibetanized as rCfyun-des-hyi-bu "son of the always knowing one." While bu clearly represents the patronymic suffix eya, the correlation between rOyun-des and Atri remains obscure, the latter being usually etymologized as the "devourer" (attri). There is a remote possibility that Atri has been associated with Agni, who is known in Vedic literature both as "omnivorous" (viivad RV. VIII 44.26 etc.) and as "omniscient" {viivavedas RV. I 128.8 etc.).
II. 2
Sanskrit Text
ayuh kamayamanena dharmarthasukhasadhanam j
ayurvedopadesesu vidheyah paramadarah //
By him who wishes a long life leading to virtue, wealth, and happiness, the utmost attention (is) to be paid to the precepts of medical science.
Tibetan Version
thse ni rin-bar 'dod-pa-yis / chos dan nor dan bde-ba sgrub 1 j ihse-yi rig-byed luh bsad\4P\-pa / rab-tu gus-par 15b 1 bya-bar gyis // 1 NP; bsgrub CD.
48 First Chapter
By Mm who wishes life to be long (the mind) shall be made deeply intent on what is set forth (in) the precepts (of) medical science, (which) leads to virtue, wealth, and happiness.
Remarks
In contrast to the salutatory stanza, which alludes to the Buddhist roots of vice, the present stanza refers, if covertly, to the Hindu aims of life. But this seems to be a commonplace of Indian medicine rather than a clue to the author's faith. Caraka already observes that "health is the supreme foundation of virtue, wealth, love, and release" {dJm.rmdrtJmMrrianioksa'^dm arogyam mvlam uttamam 1 1.15), whereas the As. speaks of "diseases causing the obstruction of virtue, wealth, love, and release" (dliarmdrthakdniamoksdvdm vighnakaribhir amayaih 1 1 init.).
Going into particulars, ayuh kamayamdna "he who wishes a long life" has been translated thse ni rin-bar 'dod-pa "he who wishes life to be long."
dharmdrthasukhasddhana "leading to virtue, wealth, and happiness" has been connected not with ayus "long life," to which it belongs grammatically, but with dyurvedopades'a "precepts of medical science." This change of construction became necessary once the original word-order was retained.
For sgrub GD have substituted the corresponding future stem, bsgrub, which comes to the same.
dyurvedopadesa "precepts of medical science" has been paraphrased as thse-yi rig-byed lun Mad-pa "what is set forth (in) the precepts (of) medical science."
vidheyah paramudarah "utmost attention (is) to be paid" has been verbalized into rab-tu gus-par bya-bar gyis "shall be made deeply intent."
11.3
Sanskrit Text
Brahma smrtvayuso vedam Prajapatim ajigrahat j
so 'svinau tau Sahasraksam so 'triputradikan munin \\
Brahman, having recalled medical science, taught (it) to Prajapati; he, to the two Asvins; they, to the Thousand-eyed One; he, to the sages Atriputra etc. ;
Tibetan Version
Thsans-pas thse-yi rig-byed dran / sKye-dgui(4)-bdag-la Mad-pa yin / de-yis Tha-skar ((4)) des brOya-byin j des rGyun-ses stsogs 1 dran-sroh-lao //
1 CD; de-yis rGyun-ses NP.
Brahman, having recalled medical science, told (it) to Prajapati; he, to (the two) Asvini(putras); they, to the Hundred-powered One; he, to the sages Atri(putra) etc. ;
Remarks
The Tibetan translators follow the original Sanskrit word for word (with the exception that smrtva is put after ayuso vedam on grounds of syntax). Their proce- dure is all the more remarkable here as by doing so they get into conflict with
First Chapter 49
elementary grammar, which requires the verb always to stand, at the end of its clause or sentence. On this peculiar aspect of translational Tibetan see Introd. § 27.
ajigrahat "he made grasp, taught" has been rendered by bsad-pa yin "he told."
Asvinav, "the two Asvins" has been represented by Tha-skar, which properly signifies the goddess AsVim, the mutual wife of the two Asvins later considered to be their mother. As in the case of the following rOyun-ses (~ Skr. Atri; see note on v. 2), the name of the parent serves here as a substitute for that of the son.
Sahasraksa "the Thousand-eyed One," an epithet of Indra (who, endeavouring to seduce Ahalya, was cursed by her husband with a thousand female sex organs later changed into eyes; cf. MBh. XIII 34.27 sq.), has been replaced by brQya-byin "the Hundred-powered One" ( — Skr. Satakratu), another epithet of Indra meaning that he has sacrificed a hundred times (cf. MBh. IX 49.1 sqq.).
Instead of des rGyun-des stsogs MP simply read de-yis rGyun-ses, the diastole of des leading to the omission of stsogs.
II. 4
Sanskrit Text
te 'gnivesadikams te tu prthak tantrani tenire / tebhyo 'tiviprakirnebhyah prayah sarataroccayah //
they, to Agnive&a etc. But they composed (their) works separately. (These) being too widely scattered, there is (now) made [kriyate 5 a] from them, as a collection for the most part of very essential (matter),
Tibetan Version
de-yis Me-bzin- 'jug-la sogs 1 / de-mams-hyis 2 rgyvd so-sor [5] byas / sin-tu 'thor-ba 2 de-rnams-las / rab-gces 3 phal-cher btus(5)-pa ni \\
1 NP; stsogs CD. 2 NP; 'thob-pa CD. 3 NP; ces CD.
they, to Agnivesa etc. (But) they composed (their) works separately. (These) being too widely scattered, there was (then) made [byas 5 b] from them, as a collection for the most part (of) very essential (matter),
Remarks
The name Agniveda has been Tibetanized as Me-bivn,-$ug "he who enters like fire" (thus also Mvy. 3471). This translation presupposes for AgniveSa the etymology "he whose entrance is like that of fire" (as against PW I 34 "he who has a fire- temple"), by which a brahmin must be understood; cf. Vas. XI 13 [~KatbUp. I 1.7]: vativanarah pravUaty atithir brahmanyo grham \grhan\ "as fire enters a brahmin guest the house [the houses]."
The particle tu has been omitted for lack of space and the adverb prthak inter- changed with tantrani on grounds of syntax. For the same reason, tebhyah has been placed after ativiprakiri),ebhyah and prayah before uccayah
The word viprakirya (and its equivalent 'thor-ba) should be interpreted to mean, not "umfangreich" as Hilqenbebg & Kxbfel have it, but "scattered" as the com- mentators suggest; see Arunadatta's paraphrase viksiptebhya uccdvacoktarthata- yaivetascetaSca gatebhyo 'ta eva kascid evarthah kasmad em tantrdntaraj jnayate
4 Vogel, Vagbhata
50 First Chapter
"dispersed, gone hither and thither because of the diversity of subjects treated so that every subject is known from another -work."— For 'thor-ba CD have substituted 'ikob-pa, which occurs only as the future of thob-pa "to gain"; but this makes no sense in the present contest.
The comparative suratara is used here for the elative and, consequently, has been rendered by rab-gecs "very essential (matter)." CD give the short form rab-ces instead, which is not attested elsewhere.
1 1. 5
Sanskrit Text
kriyate ' tfangahrdayam natisamksepavistaram j
kaytdialagrahordhimigasalyadamstrajaravrsan \\
the Astangah; daya, without too niucli brevity or prolixity. Body, child, demon, upper part (of the body), dart, fang, old age, and potent man
Tibetan Version
iirt-tu bsdiis min rgyas min-par /
yan-lag-brgyad((o))-pai snin-po bya-s /
lus dan byis-pa gdon las-stod j
3 mthson dan mche-ba rgas ro-tsa //
the Astaiigahj'daya, neither too brief nor (too) prolix. Body, child, demon, upper part of the body, dart, fang, old age, (and) potency:
Remarks
kriyate "is made" has been placed at the end of the sentence and turned into the past tense: byas "was made."
natisamJcsepavistaram "without too much brevity or prolixity" has been inter- changed with astdngahrdayam and put verbally: iin-tu bsdus min rgyas min-par "neither too brief nor (too) prolix."
Mya "body," bala "child," etc. stand for the eight parts of Indian medicine usually known as general therapy, pediatrics, psychiatry, supraclavicular surgery, general surgery, toxicology, rejuvenation, and virilification. Bu-ston, who has quoted the Tibetan version of this and the following line in his Chos-'byun 1 , holds a different view, understanding the first lus as "womb," the second lus (in lus-stod) as "interior or trunk," and rgas ro-tsa as "senile lust." On this misinterpretation, as well as on the nomenclature and arrangement of the eight parts of Indian medi- cine, see Vogel, II J vi p. 290 sqq.— Since Vagbhata names these parts metonymi- cally after the objects with which they are concerned, it is unsatisfactory if not incorrect to render Mya by "Sonde," as Hilgenbeeg & Etrfel do; one should rather say "dart," the object of general surgery being that of removing darts and other foreign bodies.
vrsa "potent man" has been translated by the corresponding abstract noun ro-tsa "potency," spelt ro-rtsa in Bu-ston's quotation.
1 gSun-'bum, lHa-sa edition, vol. ya, fol. 18b 4—6. Cf. Obeemilleb, Transl. I p. 48.
First Chapter 51
II. 6
Sanskrit Text
astav angani tasyahus cikitsa, yesu samsrita / vayuh pittani kaphas ceti trayo dosah samasatah jj
they call its eight parts with which therapy is concerned. Wind, choler, and phlegm, (are), in short, the three humours.
Tibetan Version
gso-dpyad [6] gan-la gnas-pa-yi /
yan-lag brgyad-du de bsad-do j
rlun dan mkhris-pa bad-kan (6) yan j
nes-pa rnam-gsum mdor bsdus yin jj
these are said to be the eight parts with which therapy is concerned. Wind, choler, and phlegm are, in short, the three different humours.
Remarks
The Tibetan differs from the Sanskrit only in some minor points: astav, lias been put after angani; tasya, which refers to asfangahrdayam in 5a, has been given the role of summing up the preceding objects; the relative clause cikitsa yesu samsrita has been placed before its governing noun; iti has been left untranslated; and trayah "three" has been interchanged with dosah and qualified by the addition of mam "different."
A few words may be said on the term dosa and its pendant nes-pa. In non- technical language, they both mean "fault"; in medical literature, however, they denote a morbific agent not necessarily defective itself and hence conveniently called "humour." This differentiation, accepted by nearly all authorities on the subject, is vigorously denied by Muller (JMV xvii p. 76 sqq.), who wants to see the basic meaning also applied to medical texts.
11.7
Sanskrit Text
vikrtavikrta deham ghnanti te vartayanti ca j
te vyapino 'pi hrnnabhyor adhomadhyordhvasamiraydh jj
(According as they are) changed (or) unchanged (in their state), they (respectively) destroy and sustain the body. Though spreading every- where, they (are) seated (respectively) below, between, and above breast and navel.
Tibetan Version
de ni mam gyur 1 ((6)) ma gyur-pas / 4 lus ni 'joms dan gnas-pa yin j des khyab gyur kyan sriin-ga dan j [7] Ue-bai og bar sten-na gnas jj
1 CD; 'gyur NP.
52 First Chapter
According as they are changed (or) unchanged (in their state), they (re- spectively) destroy and sustain the body. Though spread everywhere, they arc seated (respectively) below, between, (and) above breast and navel.
Remarks
vikrtii "changed" has been metaphrased by rnam gyur (v. 1. rnam 'gyur; cf. Mvy. 2574 & 7315), with mam standing pleonastically for vi°, while avikrta "un- changed" has been translated simply by ma gyur-pa. The collocation vikrtdvikrtah may be considered to be either a copulative adjective compound (see the similar combinations hitahita 5.2 and varavara 5.20) or a case of effaced hiatus (~ vilertd avikrtdh; ef. Whitney, Skr. Gr. § 177 b).
te has been transferred to the beginning of the sentence on grounds of syntax.
vartayanti "let live, sustain" has been rendered by gnas-pa yin, which literally means "stay," but here seems to signify "stay up, sustain," in the same way that the corresponding noun gnas "place" often represents the Buddhist term nisraya '"support" (cf. Edgekton, Diet. p. 306 sg.). Such a transitive use of gnas-pa is not, however, recorded in the existent dictionaries.
vyajrin "spreading everywhere" has been turned lchyab gyur "spread every- where."
hrd and its equivalent stiin-ga should be understood here as "breast" rather than "heart," this contrasting better with the following nabhi "navel."
11.8
Sanskrit Text
vayohoratribhuktanain te 'ntamadhyadigah kramat /
fair bhaved vimmas tihsno mandas cagnih samaih samah jj
During life, day and night, and meals they (are) apparent towards the end, middle, and beginning successively. Through them, the (digestive) fire may become irregular, violent, and sluggish (respectively and, if they are) balanced, regular.
Tibetan Version
na-tfood nin-zag zas zos-la /
de ni mtha dbus sogs 1 (7) rtogs rim j
des ni 'ju-ba mi 5 moms mo /
chun* dan mnam-pas ((7)) mnam-par "gyur //
1 NP; stsogs CD. * CD ; chu NP.
During (any) period of life, day, (and) intake (of) food they are noticed towards the end, middle, (and) beginning (successively). Through them, digestion becomes irregular, violent, (and) sluggish respectively and, if they are balanced, regular.
Remarks
vayas Arunadatta and Indu explain as "man's life" (manusyayus, purusayus), while Candranandana takes it to mean any "stage of the body produced by time such as youth" (kdlakrta kurlrasyavastha yauvanadih; cf. Hemadri's gloss iarlra-
First Chapter 53
pari<Qamah). The Tibetans have adopted this second interpretation, reproducing vayas with na-thsod "period of life." Vagbhata distinguishes three such periods of life: youth, till 16; manhood, till 70; and senility, above 70 (Ah. II 3.105).
ahoratri "day and night" is used here in the technical sense of a span of 24 hours (like the Greek vvx&tffiegov). Hence its translation by the tautologic nin-iag instead of the customary nin-mthsan.
bJiukta, a perfect past participle turned action-noun, has been metaphrased by zos, to -which a cognate accusative has been added, by -way of specification.— The genitive, separated from its governing noun by the subject and represented in Tibetan by a dative of sphere, is best considered to be one of general concernment (cf. Whitney, Skr. Gr. § 300b).
ga "coming, appearing" has been rendered more freely by rtogs "are noticed."
rim, the pendant of kramdt "successively," poses a syntactical problem. Follow- ing rtogs, it cannot well be connected with the previous sentence, though that is what the Sanskrit would suggest. The easiest way out of this dilemma seems to be that of referring rim to the next sentence and understanding it in the sense of "respectively."
taih is used here elliptically for fair visamaih "through them (if they are) unbal- anced," as appears from the analogous samaih in pada d.
bhavet "may become" has been shifted to the end of the stanza and translated by 'gyur "becomes," such nuances of speech as the optative of softened statement being unknown in Tibetan.
For chun NP write ehu which, in view of the following dan, seems to be nothing but a haplography passed from N into P (cf. Introd. p. 33 n. 1).
agni "(digestive) fire" has been placed after taih and altered to 'ju-ba "digestion ."
II. 9
Sanskrit Text
Jcosthah hruro mrdur madhyo madhydh syat taih samair api /
sukrartavasthair 1 janmadau viseneva visaJcrimek //
- B; sukla K.
The belly may be hard, soft, (and) normal (respectively, that is), normal only when they are balanced. (As they are) seated in sperm and menstrual blood, (there are produced) by them {taih 10a] in the beginning of birth, just as by poison (in the case) of a poisonous insect,
Tibetan Version
de ni Ito-ba sra sni ran j
mnam-'pas kyan ni bar-mar 'gyur j
[16a 1] slcye-bai dan-por 1 hku Ichrag gnas j
dug-can srin-bui dug dan mthsuns //
1 CD; dban-po NP.
(Through) them, the belly becomes hard, soft, (and) normal (respectively, that is), normal only when they are balanced. As they are seated (in) sperm (and) menstrual blood in the beginning of birth, (there are produced by) them [de ni 10a], just as (by) poison (in the case) of a poisonous insect,
54 First Chapter
Remarks
at/at "may be" and iaih "through them" hare been transferred to their appro- priate places at the end and the head of the sentence and altered to 'gyur "becomes" and de. ni "as to them" respectively, while the second madhyah has been interchanged with fnmair api.
jamimdan "in the beginning of birth," glossed by the scholiasts as garbliMha- nakale "at the time of conception," has been referred not to what follows but to what precedes and, consequently, has been put before iukrartavastliaih. Instead of dah-por, the usual equivalent of adi, XP read dban-po, which is likely to be corrupt.
khrag must be understood here, like the original artava, in the specific sense of "menstrual blood." Impregnation, according to Indian physicians, is brought about by the union of sperm and menstrual blood, which are considered the male and female generative fluids.
vimkrimeh and its corresponding dug-can. srin-bui ought to be connected, not with rimya and dug as is done by Hilgenbeug & Kxrfel contrary to all commen- taries, but with a hypothetical prakrtih {ran-bain) to be inferred from the following prakrtayah (ran-bzin). In other words, the present genitive is again one of general concernment rather than of possession. It must be admitted that the Tibetan, taken by itself, points at first sight in a different direction, especially as visakrimeh has been placed before viseixeva; but then it should be borne in mind that elliptical clauses like this confront a Lamaist translator with serious difficulties and that, under such circumstances, keeping to the original as closely as possible is about the best solution to the problem.
II. 10
Sanskrit Text
taii ca tisrah prakrtayo Mnamadhyottamah prthah j
sarnadhatuh samastasu srestha nindya dvidosajah //
the three constitutions, each (being) weak, average, and strong. That in which the elements are balanced (is) the best of all; to be regarded as inferior (are) those originating (only) in two humours.
Tibetan Version
de ni ran ^[QJl-Min rnam-gsum-ste /
(45a 1) chun 'brin chen-po so-sor blta s j
ma-lus-pa-la Ichams mnam-pa /
((45a 1)) mchog yin nes 3 -pa gfiis 'byun smad //
1 CDN;mj»P. 2 NP,-ZtoCD. 3 CD; mnesKP.
the three different constitutions, each to be viewed as weak, average, (and) strong. That which is balanced as to its elements is the best of all ; those originating (only) in two humours are regarded as inferior.
Remarks
taih "by them" has again been altered to de ni "as to them," in the same way that the preceding vis&na "by poison" has been modified to dug "as to poison." This is indicative of a slightly different sentence-construction (not expressed in our English rendering), the verb to be supplied being something like sampadyante
First Chapter 55
"are produced" in Sanskrit (thus the commentators) but yin "are existent" in Tibetan ; for if -we understand the translation correctly, then ste in pada a serves as a continuative while (b)lta in pada b, together with so-sor, forms the equivalent of prthak.
tisrah has been interchanged with prakrtayah and qualified by the addition of main "different."
rin-bzin (for ran-bzin "constitution") in P is a mistake evidently caused by the preceding ni.— The three constitutions are the so-called wind, choler, and phlegm types of man, on which see Ah. II 3.84 sqq.
samastasu has been represented by a dative of sphere and transferred to the head of the sentence, for the obvious reason that it belongs to srestha as well as to nindydh.
samadhatu "whose elements are balanced" has been reproduced by hhams mnam-pa "balanced as to its elements." (dhatu is used here as a synonym of dosa.) Placing the final member of a bahuvrihi before the initial member and treating it as a modal accusative is one of the commonest ways of turning such compounds into Tibetan.
nindya "to be blamed, regarded as inferior" has been put at the end of the stanza and rendered simply by smad-pa "to blame, regard as inferior," the gerun- dive having been neglected without any palpable effect on the context.
For nes-pa NP read mnes-pa, which is unattested in the meaning of dosa.
II. 11
Sanskrit Text
tatra ruhso laghuh sitah hharah suksmas calo 'nilalj, / pittam sasnehatlksnosnam laghu visrani saram dravam If
Among the (humours), the wind (is) rough, light, cold, pungent, subtle, (and) volatile; the choler — slightly unctuous, violent, hot, light, musty, liquid, (and) flowing ;
Tibetan Version
de[2]-la rlun ni rtsub-cin yah /
gran-zin sra-la phra^zin gyo /
7 mkhris-pa snwm-bcas rno-zin dro j
yan-zin dri(2)-mnam 'khru-iin g&er //
l CD;'jpferoNP.
Among the (humours), the wind is rough, light, cold, pungent, subtle, and volatile; the choler — slightly unctuous, violent, hot, light, musty, liquid, and flowing;
Remarks
ruksa does not mean "dry" in this and similar contexts, as supposed by Jodly (Medicin p. 40) etal.; rather, it is synonymous with parusa "rough," as may be seen from its Tibetan pendant rtsub(-pa). On the physical properties of the humours (according to Caraka, Susruta, and Vagbhata) see VoaBL, PO xxiv pp. 31 & 35.
khara and sra{-ba), both of which literally mean "hard," are here used in the pregnant sense of "hard on the tongue, pungent," since the wind is taught by Susr. I 20.28 to be soft.
50 First Chapter
jihra occurs in NP as 'phra, a rare secondary spelling also recorded in the diction- aries.
anihi has been placed after tatra for syntactical reasons.
n-irn '"musty" has been paraphrased by dri-mnam "reeking of dirt."
dram "flowing" has been rendered by gser(-ba) which, as an equivalent of dram, has been noted already by Das (Diet. p. 1251), but only in the nominal sense of •'water, liquid" (cf. Ah. I 5.82 & 28.37). However, the corresponding adjective meaning is established for the Ah. beyond doubt by such occurrences as I 9.6, 12.11, 16.25, 17.18 & 19.
II. 12
Sanskrit Text
xnigdhah iito gurur mandah slaksno mrtsnah sthirah kaphali / mnisargah samnipatas ca taddvitriksayakopatah //
the phlegm— unctuous, cold, heavy, sluggish, soft, slimy, (and) solid. Combination and junction (result) from the diminution or ebullition of two and three of the (humours respectively).
Tibetan Version
bad-kan snum bsil Ici-ba dan j
rtul-zin 'jam brtan 'byar-bag-can j
((2)) de gnis gsum zad 1 [3] 'khrugs -pa-las j
Idan-pa 8 dan ni 'dus-pa yin jj
i N adds la.
the phlegm— unctuous, cold, heavy, sluggish, soft, solid, (and) a trifle sticky. From the diminution (or) ebullition (of) two (and) three (of) the
(humours respectively) result combination and conjunction.
Remarks
Maks)}a and mrtsna are hardly separable from each other without difficulty. In PO xxiv p. 35, we had understood them to mean "slimy" (~ picchila) and "soft" (~ mrdu), tacitly equating daksna with 'byar-bag-can "a trifle sticky" and mrtsna with 'jam "soft" (which, in itself, is quite possible). On second thought, however, it seems more likely that slaksna corresponds to 'jam and mrtsna to 'byar-bag-can, particularly since Arunadatta and Candranandana (whose commentaries we had not at our disposal when writing the above article) explain slaksna with aparusa "not rough" and mrtsna with mrdyamano 'nguligraJii picchilagwfyayuktas' cakacakd- yamdnah [v. 1. kacakacayamdnah] "sticking to the fingers when squeezed, endowed with a slimy quality, glimmering 1 ."
In other regards, kapha has been transferred to the beginning of the stanza and pada c interchanged with pada d on grounds of syntax. After zad N adds a super- numerary la probably miscarved for the following ' in 'khrugs.
1 Doubtless to be derived from cak [kac] "to shine."
Sanskrit Text
rasasrnmamsamedosthimajjaSukrani 1 dhatavah [ sapta dusyah mala mutraiakrtsvedadayo 'pi ca // 1 B; "majjasukldni K.
First Chapter 57
Chyle, blood, flesh, fat, bones, marrow, and sperm (are) the seven ele- ments ; (they are) liable to be spoilt (by the humours). The secretions (are) urine, feces, sweat, etc. ; and (they are liable to be spoilt by them) too.
Tibetan Version
lus-zuns dans-ma khrag dan sa / tksil dan rus rhan khu-chu-ste / (3) gnod bya bdun yin dri-ma gcin j bsan-ba rhul-la sogs 1 -pa yin If 1 NP; stsogs CD.
The elements are chyle, blood, flesh, fat, bones, marrow, (and) sperm; they are liable to be spoilt (by the humours and are) seven (in number). The secretions are urine, feces, sweat, etc.
Remarks
dans-ma, as the pendant of rasa "chyle" is spelt throughout the Ah., is not in this form given by the lexicographers ; they write dvans-ma instead.
khu-chu, the equivalent of sukra "sperm," is a tautology literally to be turned, "sperm & semen."
dhatu "element" has been transferred to the head of the stanza and etymologized as lus-zuns "body-hold." Though in keeping with the definition offered by the scholiasts (dariradharai/dd dhatavah), this etymology does not conform to Nir. I 20 where dhatu is explained more correctly as a derivative of dhd "to put" (dhatur dadhdteh). The usual Tibetan correspondent is lehams.
sapta is treated by the commentators as an attribute of dhatavah, while dusyah is taken for a short sentence of its own (the neglected sandhi indicating a period after dusyah). The Tibetans do not follow their example; they rather make both sapta and dusya specifications of dhatu, connecting them to what precedes in reverse order by means of ste (the annunciatory continuative called dam-boa by native grammarians; see Bacot on Thon-mi 1 13).
mala and dri-ma properly signify "dirt, impurity." As in the case of dosa and nes-pa, however, their original meaning has become somewhat obliterated in medical usage. Both terms now denote the waste products or "secretions" of the elements, which are respectively phlegm, choler, dirt in the apertures, sweat, nails & hair, fat of eyes, skin & feces, and vital essence (Ah. 113.63 sq.; cf. Jolly, Medicin p. 43).
The words api ca, which according to the scholiasts stand elliptically for te 'pi ca du§yah (Indu's paraphrase), have been omitted by the Tibetans, apparently for lack of space. They have not been translated by Hxlgenbekg & Kibeel either.
II. 14
Sanskrit Text
vrddhih samanaih sarvesam viparltair viparyayah [
rasah svadvamlalava^atiktosanakasdyakah //
(There will be) an increase of all (humours, elements, and secretions) through homogeneous (substances) ; through opposite ones, the opposite. The six [sad 15a] flavours (are) sweet, sour, salt, bitter, pungent, and astringent ;
58 First Chapter
Tibetan Version
thams-cad mnam-pas 'phel 'gyur-zin j
[4] 16a 1 de-las ((3)) bzlog-pas gnod-par 'gyur /
ro ni mmr skyur lan-thsva dan /
ihaa dail kha dan bska-ba-ste //
AH (humours, elements, and secretions) will be increased by homogeneous (substances) and spoilt by those different from them. The flavours are sweet, sour, salt, bitter, pungent, and astringent;
Remarks
In the first sentence, the nominal diction has been abandoned and the wording modified. In the second, the terms tikta "bitter" and usaya [rare for usqa ~ hatu- {ka)] "pungent" have been translated by thsa(-ba) and kha(-ba) respectively, which is just the opposite of what should be expected on the strength of the dictionaries (including Jlvy. 1901 sq.). Glancing over the first book of the Ah. 1 , however, we have found 4 other eases (1.15, 3.4, 5.24, 5.50) in which thsa(-ba) corresponds to tikta and no less than 10 other cases (3.4, 5.24, 6.79, 6.108, 9.21, 9.29, 12.52, 18.21, 18.35, 19.59) in which !:ha{-ba) corresponds to katu(ka). Under these circumstances, we dare not assume a change of word -order in the present stanza.
1 With the aid of a handwritten Tibetan-Sanskrit glossary kindly placed at our disposal by Dr. Lokbsh Chandra, New Delhi. This glossary proves invaluable on many occasions for tracing parallel passages.
II. 15
Sanskrit Text
sad dravyam asritas te tu 1 yathapiirvam balavahah /
tatradya marutarri ghnanti trayas tiktadayah kapham //
x B;caK.
inherent in a substance, they (are) generative of strength in (descending) order; the first three of them destroy the wind, the bitter and following ones— the phlegm,
Tibetan Version
rdzas drug-la gnas (4) de-rnams 1 kyan / go-rim* b£in-du stobs chen yin / de-la 3 2 dan-poi gsum-po ni / rlun 'joms thsa sogs* [5] bad-kan sel \\
1 NP; dag CD. 2 NP; rims CD. 3 NP; las CD. 4 NP; thsva stsogs CD.
inherent in six substances, they are also great in strength in (descend- ing) order; the first three of them destroy the wind, the bitter and following ones remove the phlegm,
First Chapter 59
Remarks
sad has been regarded by the scholiasts as an attribute of rasa (14c), in obvious analogy to sapta and dk&tu (13bc), though an enjambment like this is none too frequent in Sanskrit. In Tibetan, it has been connected -with dravya, -which shows that the translators read saddravyam (in one word) rather than sad dravyam (in two words), taking the phrase for a dvigu.
For de-mams CD have substituted the commoner de-dag; for go-rim, the alter- native go-rims. The enclitic kyan would fit the text of K (ca) better than that of B (tu).
balavaha "generative of strength" has been rendered by stobs chen "great in strength," which corresponds to maJiabala in Sanskrit (see Mvy. 3343). Here again it would be quite possible to assume a variant reading in the translators' copy.
For de-la CD give de-las, which occurs only at times in the sense of tatra.
ghnanti has been left in its original position after the first object (marutam), a new verb (sel) having been added to each of the other two objects (kapham 15d and pittam 16b).
trayah has been transferred to its appropriate place after adyah, while tikta has again been represented by thsa (spelt thsva in CD).
II. 16
Sanskrit Text
kasayatiktamadhurah pittam anye tu kurvate j
samanam kopanam svasthahitam dravyam iti tridha //
the astringent, bitter, and sweet ones— the oholer, the other ones pro- duce (them). Sedative, irritative, (and) good for normal (humours, elements, and secretions) : thus a substance (is) threefold (in its effect) ;
Tibetan Version
bska dan Jcha dan ((4)) mnar-ba ni J
mkhris sel gian-dag byed 1 -pa yin j
de-ltar rdzas ni rnam-pa gsum /
zi (5) 'khrugs tha-mal gnas-la 3 phan //
1 MP; skyed CD.
the astringent, bitter, and sweet ones remove the choler, the other ones produce (them). Thus a substance (is) threefold (in its effect) : sedative, irritative, (and) good for (humours, elements, and secretions) that are in a normal state ;
Remarks
tikta "bitter" has here been represented by its usual equivalent kha(-ba). Con- trast w. 14 & 15.
j-1 The words anye tu hurvate and gian-dag byed-pa yin "the other ones produce (them)" must be interpreted to mean, as the commentators put it, that the bitter, pungent, and astringent flavours produce wind, the sweet, sour, and salt ones— phlegm, and the sour, salt, and pungent ones— choler. In CD byed-pa has been replaced with the synonymous skyed-pa.
60 First Chapter
svastka "normal" has been paraphrased by tha-mal gnas, which is best turned "being in a normal state." Strictly speaking, tha-mal(-pa) alone -would have done in Tibetan ; for gnas (~ stTm) has been added merely in an effort to make the translation more literal.
The clause dravyam iti tridha has been shifted to the beginning of the sentence, ■while the particle iti has nevertheless been rendered by de-ltar, which refers as a rule not to what follows but to what precedes. It is doubtful, though, if the Tibetans really had any such relation in their mind.
II. 17
Sanskrit Text
umasilagunotkarmt tatra viryani dvidha smrtarn j tridha vipako dravyasya svadvamlahatuhattnakah [j
according to the prevalence of the qualities hot or cold, the power in it (is) taught (to be) twofold, (namely heating or cooling) ; threefold (is) the digestion of a substance, (namely) of a sweet, sour, or pungent na- ture;
Tibetan Version
de-la thsa bsil phul-byun-bas j
nus-pa mam-pa griis-su bsad [
[6] rdzm-kyi zu 1 -ba rnam-gsum-ste j
mnar slcyur thsa-bai bdag-nid-can //
1 KP;nikhuCI).
according as (the qualities) hot (or) cold are prevalent in it, (its) power is said to be twofold, (namely heating or cooling); the digestion of a substance is threefold, namely of a sweet, sour, (or) pungent nature ;
Bemarks
■usjfasitaguyx>tkarsdt "according to the prevalence of the qualities hot or cold" has been denominalized into thsa bsil phul-byun-bas "according as (the qualities) hot (or) cold are prevalent," with gwna "quality" having been omitted.
tatra "in it," which belongs to usyasltagu'witkarsat as well as to mryam, has been connected only with the former, the construction and xoivov being impracti- cable in Tibetan.
The words tridha vipako dravyasya have been inverted for syntactical reasons. Instead of rdzas-kgi su-ba CD read rdzas ni khu-ba, which appears to be corrupt.
II. IS
Sanskrit Text
gurv,mandaki7nasnigdha^lak§nasdndramrdusthirdh / gunah sasiiksmavisada vimsatih saviparyayah //
heavy, sluggish, cold, unctuous, soft, viscid, pliant, and solid as well as subtle and dry (are) the qualities (of a substance: they are) twenty together with their opposites.
First Chapter 61
Tibetan Version
((5)) yon-tan lei rtul bsil dan snum j
'jam dan bska4:-ba mnen (6) dan brtan j
■phra dan bcas-sin skam 1 -bag-can j
de bzlog-pa dan rii-su yin //
1 NP;&sfoxCD.
the qualities (of a substance are) heavy, sluggish, cold, unctuous, soft, viscid, pliant, slightly subtle, and a trifle dry: these, together with their opposites, are twenty.
Remarks
Except for the inevitable transposition of gu%a and vimsati, the Tibetan version. is a faithful reproduction of the original Sanskrit— so faithful, in fact, that the merely connective office of sa° in sasuksmamiada seems to have escaped the trans- lators' attention; for phra dan bcas-Sin sham-bag-can cannot 'well be interpreted otherwise than indicated.
The opposite qualities not mentioned here are specified by Arunadatta and Candranandana as laghi^kanos^ruksakharadravakathinasarasthiilapiaAildh "light, violent, hot, rough, harsh, mobile, hard, liquid, coarse, and slimy." The confron- tation of vi&ada and picchila is noteworthy in that it presupposes a meaning "dry" or the like not attested for vidada but corroborated by the Tibetan sham (which CD have malcorrected into bska "astringent"). For vUada^skam, also see I 9.7.
II. 19
Sanskrit Text
kalarthakar-manam yoga 1 hlnamiihyatimatrahah 1 /
samyagyogas ca vijneyo rogarogyaihaharanam jj
1 B ; yogo and °kah K.
The weak, wrong, and excessive connections (of a humour) with season, object, and action and (its) proper connection (with these are) to be known as the sole cause (s) of illness and health (respectively).
Tibetan Version
dus don las-hyi sbyor[l]-ba-rnams / dman dan log^pa Uiag-pa dan / yan-dag 2 sbyor-bas nod dan ni / nad-med 5 rgyu ((6)) gcig yin zes bya //
1 NP ; par CD. 2 C adds par .
According as the connections (of a humour) with season, object, (and) action are connected weakly, wrongly, excessively, and properly, they are said to be the sole causes of illness and health.
Remarks
It must be observed at the outset that the present sioka is not readily under- standable by itself, neither in Sanskrit nor in Tibetan. Judging from Arunadatta' s
62 First Chapter
and Candranandana's comments as well as from a more detailed account in I 12.34 sqq. (which see for further information), the attributes hlna, mithya, atimatra, and samyaTic arc meant to qualify Mia, artha, and karman rather than yoga. Such a phenomenon is known to classical scholars as a metathesis qmlitatum.
Mnamiihyaiimatraka and samyagyoga, which are respectively subordinate and co-ordinate to yoga (the singular in K is not confirmed by the Tibetan), have been combined with it into a gerundial clause, with the pertinent changes limited to a minimum degree.
For log-pa CD have substituted log-par, which seems to be influenced by the adverbial mithya but does not harmonize with the sentence-construction.
After yaii-dag C has added a s\ipernumerary though grammatically correct
-par,
vijneya, "to be known" has been transferred to the end of the stanza and rendered by sen bya, a phrase usually corresponding to nama, namocyate, or the like. Its being used as a verb of declaration, with a supine (yin) depending on it, is very strange. Since vijneya would ordinarily be ses bya in Tibetan, the present zes bya bears every mark of an old corruption.
11.20
Sanskrit Text
rogwi tu domvaisamyam dosasamyam arogata / nijagantuvibhagena tatra roga dvidha smrtah 1 / tesam kayamanobhedad adhisthanam apt dvidha // 2
1 B; matah K.
2 We have retained this peculiar grouping of lines in order to avoid any discrep- ancy in numbering between the Nirnaya Sagara Press edition and ours.
Illness (is) disharmony of the humours; harmony of the humours (is) health. Of these (two conditions), the diseases (are) taught (to be) twofold because of (their) division into endogenous and accidental ones ; on account of the distinction between body and mind, their seat too (is) twofold [on account of their classification into (such of the) body and (such of the) mind, (their) seat too (is) twofold].
Tibetan Version
nad tii nes-pa ma-snoms yin /
ues(7)-pa' snom 1 -pa iwd-med-nid* j
de-la ?iad ni ran-bzin gnas j
glo*-bur clut-yis g¥iis\\6hl]-su bsad j
de-mams lus sems dbye-ba-yis j
6 gnas kyaii rnatn-pa gnis yin-no //
1 M> ; moms CD. 2 CD ; yin NP. 3 CDN ; bio P.
Illness is disharmony (of) the humours; harmony (of) the humours (is) health. Of these (two conditions), the diseases are said to be two(fold) because of (their) division into endogenous (and) accidental ones; on account of the distinction between body (and.) mind, they are twofold
First Chapter 63
with regard to (their) seat too [on account of their classification into (such of the) body (and such of the) mind, (their) seat too is twofold].
Remarks
The chiastic word-order in the first and second padas has been retained by the translators.
snom-pa has been altered to snoms-pa in CD, evidently for uniformity's sake.
nad-med-nid, as given in CD, is the precise correspondent of arogatd "state of non-illness, health." The reading nad-med yin in NP is a redactional change obviously prompted by stylistic considerations.
tatra, which Arunadatta and Candranandana refer to the preceding roga and arogatd, has been placed at the head of the sentence.
glo-bur is written blo-bur in P, both spellings being equally current.
dvidha "twofold" has been rendered simply by gnis "two" (instead of the usual rnam-lpa] gnis), a brachylogy doubtless caused by lack of space.
For smrtah "taught" K prints matah "held." The Tibetan bead "are said" does not clearly show which reading the translators had before them.
The last sentence admits of two slightly different interpretations, depending on whether tesam is referred to Myamanobhedad or adhisthdnani. The Tibetan is of no avail here, being ambiguous itself.
II. 21
Sanskrit Text
rajas tamas ca manaso dvau ca dosav udahrtau / darsanasparsanaprasnaih parlksetatha 1 rogi'nam //
1 B ; parik§eta ca K.
Passion and delusion (are) taught (to be) the two mental disorders. By inspection, palpation, and interrogation (the physician) shall examine a patient ;
Tibetan Version
'dod-chags gti({l))-mug hyan yid-hyi j
nes-pa gnis-su bstan-pa yin /
(45b 1) nad-pa blta dan reg-pa dan j
dri-ba^-yis ni yons-su brtag 2 //
1 NP ; dris-pa CD. 2 CD ; rtag NP.
Passion and delusion are taught to be the two mental disorders . . .
Remarks
The sentence starting with nad-pa in 21 c and reaching up to 'thob-pa yin in 22 b,
though reproducing the original word for word, differs sharply from it in structure,
and besides is obviously corrupt. As it stands in the block-prints, the Tibetan can
only be rendered—
"with regard to a patient, there are: examination by inspection, palpation, and interrogation, portent of the cause of disease, symptom, reaction, (and) course"
which does not make any sense. One should rather expect it to read —
64 First Chapter
"with regard to a patient, there is an examination by inspection, palpation,
and interrogation ; with regard to a disease, there are cause, portent, symptom,
reaction, {and) course" with pada 22a changed into nad ni gii-ma dan-poi thsul. But this is still somewhat unsatisfactory. Perhaps one should also change (b)rtag into rtog and translate as follows —
"a patient is examined by inspection, palpation, and interrogation; a disease is
(diagnosed by) cause, portent, symptom, reaction, (and) course" with the predicate left in its mid-sentence position and taken up at the end by an auxiliary (see Introd. § 27), which would come closest to the basic text.
The variant dris-pa found in CD is merely a substitute spelling of dri-ba.
The particle atlta, replaced by ca in K, has been omitted in Tibetan; it seems to be expletive anyhow.
1 1. 22
Sanskrit Text
rogani nidanapragrupalaksanopasayaptibhih / hhumidehaprabhedena desam ahur iha dvidha \\
a disease (he shall diagnose) by cause, portent, symptom, reaction (to certain articles of food or medicine), and course. On account of the dis- tinction between land and body, (scholars) say the (meaning of) region
(is) twofold here, (namely tract of land and part of body).
Tibetan Version
nad-hyi gzi-yi dan{[T]}-poi thsul /
[2] mthsan-nid ner 1 bsten 'thob-pa yin /
sa daii Iv^-kyi bye-brag -gis j
'di-la yul-sa rnarn-ghis Mad Jj
1 CD'P;ne-barN.
... On account of the distinction between land and body, the (meaning of) region is said to be twofold, (namely tract of land and part of body).
Remarks
'Tier has been resolved by N into tie-bar, which does not agree with the metre.
deia "region" has been translated by yul-sa, a tautological compound properly meaning "place & region."
ahuh "(scholars) say" has been shifted to the end of the sentence and rendered by the impersonal bsad "are said" as a matter of course.
iha "here" refers to the science of medicine; it has been put right after °prabhe- dena.
II. 23
Sanskrit Text
jangalam vatabhuyistham anupam 1 tu Icapholbanam / sadharartam samamalam tridha bhudeiam ddiiet //
1 B; anupam K.
First Chapter 65
Jungle (is) full of wind, swamp (is) rich in phlegm, (and) ordinary land (is) possessed of balanced humours: (thus) one may define a tract of land as threefold.
Tibetan Version
((45b 1)) skam 1 -sa phal-cher rlun bskyed 2 -cin j rlan-can bad-kan 'phel-ba yin j (2) 8 cha-mnam nes-pa mnam-pa-ste j sa-yi phyogs ni [3] gsum-du biad //
1 CDP ; shams N. 2 NP ; skyed CD.
Jungle chiefly produces wind, swamp increases phlegm, (and) ordinary land is indifferent to the humours : (thus) a tract of land is said to be three(fold).
Remarks
The first half of the sentence has been paraphrased rather than metaphrased. But if the meaning has been grasped correctly remains doubtful. From the explana- tion given by the commentators it would seem that the terms jafigaia, anupa, and sadhara'Qa stand metonymically for the plants, birds, human beings, wild animals etc. indigenous to these regions (jatausadhik7iagapuritsamrgddayah Arunadatta & Candranandana) .
skams-sa in N is obviously a miswriting.
bskyed is, strictly speaking, the perfect of skyed-pa "to produce." in NP, however, it is used also for the present (as often elsewhere). CD have skyed in virtually all the cases.
mala serves here, according to the scholiasts, as a substitute for dosa; hence its translation by nes-pa.
tridha "threefold" has been interchanged with bhudeia and reproduced simply by gsum "three" (see v. 20). So we should perhaps understand the final clause as follows: "(thus) the tracts of land are said to be three (in number)."
II. 24
Sanskrit Text
ksan&dir vyadhyavastha ca halo bhe§ajayogakrt /
Sodhanam iamanam ceti samdsdd a-usadham dvidha jf
Moment etc. and state of disease (make up) the right time deterrruning the (suitableness of the) administration of a remedy. Purgative and sedative : thus, in short, a medicine (is) twofold.
Tibetan Version
skad-cig-la sogs nad-kyi skabs / dus-su sman ni sbyor-bar byed / mdor-na sman ni rnam-gnis-te j sbyan-ba dan ni ((2)) zi-bao //
Moment etc. (and) state of disease let (the physician) administer a remedy at the right time. A medicine, in short, is twofold : purgative and sedative.
5 Vogel, Vagbhata
60
First Chapter
Remarks
The terms k?a^adi "moment etc." and vyadhyawsika "state of disease" represent the idea of time in its astronomical and medical aspects. According to Arunadatta and Candranandana, the units of time not specified here comprise lava, truti, muhurta ("hour"), yama ("night-watch"), ahoratra (day & night), paksa (half- month), mam, (month), rtu (season), ayana (half-year), and samvatsara (year) 1 , while the states of disease are characterized as sama (immature), nirama (mature), rnrdu (mild), madhya (moderate), and tiksffa (severe).
The nominative kalo lias been reproduced in Tibetan by the terminative dus-su, which makes it almost certain that the basic reading was kale; hence the difference in interpretation. It is equally certain, however, that this reading is corrupt inasmuch as a dual predicate (bhesajayogakrtav) would then be required, which is incompatible with the metre.
The third and fourth padas have been transposed, with the annunciatory te taking the place of iti.
1 Indian writers differ considerably on the division of time, especially as regards the smaller units up to a muhurta (~ 48 minutes). Cf. Cqlebrooke, Essays I p. 540 sqq. This applies also to medical authors. Hemadri, for example, while comment- ing on the present stanza, takes ksai^a to mean aksinimesa (twinkling of an eye) and adi to stand for kd^hakaM(J>hagay7iddikdmuhurtaya7nahoratrapaksa7nasartvayana- varsai^i, equating —
15
30
17
20
'>.
aksinimesa
kastha
kala
bhaga
nadika
3 3 / t muhurta
yama
ahoratra
paksa
masa
rtu
ayana
8
15
2
2
3
2
to
to
to
to
to
to
to
to
to
to
to
to
kastha
kala
bhaga
nadika
muhurta
yama
ahoratra
paksa
masa
rtu
ayana
varsa
Sulruta, on the other hand, who is the with the problem extensively (I 6.4 sqq.
15 aksinimesa
30 kastha
20Vio kal5
30 muhurta
15 ahoratra
2 paksa
2 masa
3 rtu
2 ayana
5 samvatsara
only physician of the classical triad to deal ), assigns—
to
to
to
to
to
to
to
to
to
to
1
1
1
1
i
I
1
i
i
1
kastha
kala
muhurta
ahoratra
paksa
masa
rtu
ayana
samvatsara
yuga
The terms lava and trufi recorded by Arunadatta and Candranandana occur in neither system.
First Chapter 67
II. 25
Sanskrit Text
sarirajanam dosanam kramena par am ausadham J bastir virelco vamanam tatha tailam ghrtam madhu //
The best medicine for somatogenic diseases — according to the (above) order (of humours) — (is) a clyster, cathartic, (and) vomitive; in the same way, sesame-oil, ghee, (and) honey.
Tibetan Version
16b 1 lus-las byun-bai nad-rnams-kyi j rim-gyis sman-gyi (3) dam-pa ni / mas-btan 1 bkru-sman slon[4:]-sman dan / de-bzin til-mar mar sbran-rtsi //
1 NP; gton CD.
The best medicine for somatogenic diseases— according to the (above) order (of humours) — (is) a clyster, cathartic, (and) vomitive; and — in the same way— sesame-oil, ghee, (and) honey.
Remarks
The words Sariraja dosa must here be understood, both as "bodily humour" and as "somatogenic disease" : in their primary meaning "bodily humour" they depend on hramenjM and refer to the sequence vayu — pitta— kapha in v. 6 ; in their secondary meaning "somatogenic disease" they belong to param ausadham and contrast with manodosa in v. 26. The Tibetans proceed eclectically, adopting the latter connota- tion but retaining the former construction.
The terms basti "clyster," vireka "cathartic," and vamana "vomitive" have been paraphrased by mas-btan "moving-domrward (drug)," bkru-sman "washing-off drug," and slon-sman "thrusting-out drug" respectively. Instead of mas-btan, CD read mas-gton throughout.
1 1. 26—27
Sanskrit Text
dhldhairyatmadivijnanam manodosausadham param j bhisag dravyany upasthata rogl padacatustayam // cikitsitasya nirdi§tam pratyekam tac caturgunam / dahlias tirthattas'astrartho drstakarmd Sucir bhisak //
Intelligence, firmness, and knowledge of one's self etc. (are) the best medicine for mental disorders. Physician, (