Pramāṇa · Search · About · API

A History of Indian Medical Literature v1.c26.s3: Date of the Aṣṭāṅgasaṃgraha

A History of Indian Medical Literature v1.c26.s3

Language
English

Date of the Aṣṭāṅgasaṃgraha

v1.c26.s3.1

One of the most controversial issues concerning the date of the Saṃgraha, as well as that of the Hṛdaya, stems from some remarks of I-ching, a Chinese Buddhist monk, in the report of his travels in India and the Malay archipelago.²⁶⁹ After an exposition on the eight branches of Indian medicine, the author says: “These eight arts formerly existed in eight books, but lately a man epitomized them and made them into one bundle. All physicians in the five parts of India practise according to this book, and any physician who is well versed in it never fails to live by the official pay”.²⁷⁰

v1.c26.s3.2

These sentences enticed a host of scholars to speculate on the identity of the medical treatise I-ching had in mind.

v1.c26.s3.3

Before discussing this point, it may be useful to deal first with I-ching’s reliability in general and his knowledge of medicine in particular.

v1.c26.s3.4

Max Müller wrote in a letter to Takakusu²⁷¹ that we must not expect from I-ching any trustworthy information on the ancient literature of India;²⁷² he added that, though the works of Chinese pilgrims throw little light on the ancient literature, or even on what is called the Renaissance period up to 400 A.D., they have proved of great help to us in fixing the dates of Sanskrit writers whom they either knew personally or who had died not long before their times.²⁷³

v1.c26.s3.5

This estimate of I-ching’s dependability is rather encouraging.

v1.c26.s3.6

Medicine is a subject on which I-ching himself remarks that he studied it successfully, but, as it was not his proper vocation, he finally gave it up.²⁷⁴

v1.c26.s3.7

The extent and accuracy of I-ching’s knowledge of Indian medicine can be gauged and checked by the material on the subject in his travel report, and, as has often been overlooked, by a close study of his Chinese translation of the Suvarṇaprabhāsasūtra.²⁷⁵

v1.c26.s3.8

The travel report contains three chapters devoted to medicine. These are: chapters XXVII: on symptoms of bodily illness, XXVIII: rules on giving medicine, and XXIX: hurtful medical treatment must not be practised.

v1.c26.s3.9

A digression on these chapters is necessary in order to be able to form an opinion on I-ching’s sources and his way of handling Indian medical concepts.²⁷⁶

v1.c26.s3.10

The very first statements of chapter XXVII make one suspicious. The author says that the small meal in the morning should be taken in accordance with the condition of the four great elements of which one’s body consists. This group of four great elements (mahābhūtas) recurs again and again in the medical chapters of the report and in the Suvarṇaprabhāsasūtra translation. Chapter XXVIII even begins with the assertion that every living creature is subject either to the peaceful working or failure of the four great elements.

v1.c26.s3.11

The source of this view is, according to I-ching himself, a Sūtra on the art of medicine, preached by the Buddha, which mentions earth, water, fire and air as the four elements; the morbific factors associated with these elements are designated by Chinese characters corresponding to Sanskrit guru,²⁷⁷ śleṣman, pitta and vāta.²⁷⁸

v1.c26.s3.12

This emphasis on a doctrine making four great elements and their imbalance into basic concepts of medical theory disagrees completely with āyurveda,²⁷⁹ which is characterized by a theory acknowledging the three doṣas and their imbalance as basic, while five, not four, great elements are also given considerable scope. This is the more remarkable when it is taken into account that the tridoṣavāda was not unknown to I-ching.280

v1.c26.s3.13

The translation of the Suvarṇaprabhāsasūtra by I-ching confirms that he reinterpreted the central concepts of āyurveda under the influence of Buddhist thought and theories characteristic of Chinese medicine.

v1.c26.s3.14

The two sources at our disposal as testimonies of I-ching’s acquaintance with āyurveda thus show unambiguously that he was conversant with part of its theoretical structure and a large number of details concerning treatment, but failed to grasp the essential differences between Indian and Chinese medical thought. The travel report amply attests to his tendency of finding parallels and of his attempts at equating disagreeing concepts of both systems.

v1.c26.s3.15

The question to be tackled now is whether or not the evidence available suggests that I-ching was informed about Vāgbhaṭa and a work attributed to him.

v1.c26.s3.16

Several arguments tell against the assumption that he was. Characteristic of Aṣṭāṅgasaṃgraha and Aṣṭāṅgahṛdaya is a strong emphasis on the doctrine of the three doṣas and the seven elements of the body, while the five mahābhūtas retreat into the background in comparison with their role in Caraka- and Suśrutasaṃhitā. In contrast with this, I-ching lays stress on the crucial position of four mahābhūtas.

v1.c26.s3.17

I-ching’s travel report enumerates and characterizes the eight branches of āyurveda.281 The same subject is dealt with in the Suvarṇaprabhāsa translation. These two versions are largely in conformity with each other; some differences in detail are interesting without warranting the conclusion that they cannot stem from the same author, as J. Nobel suggested.282

v1.c26.s3.18

Remarkable about the order of presentation of the eight divisions is their conformity with the arrangement found in the Suśrutasaṃhitā (Sū.1.8), with only this dissimilarity that agadatantra precedes kaumārabhṛtya instead of being the next item. The order found in Aṣṭāṅgasaṃgraha (Sū.1.8cd) and Aṣṭāṅgahṛdaya (Sū.1.5cd) diverges considerably. Very strikingly, I-ching’s list begins in the travel report with an unusual equivalent of śalya (surgery); he defines this branch as the one dealing with sores, adding farther on that these are of two kinds, inward and outward.283 The Suvarṇaprabhāsa translation deals with śalya differently and uses a Chinese character meaning thorn, which better approximates Sanskrit śalya (all sorts of foreign bodies).284 The eighth item of the series, vājīkaraṇa, the branch dealing with aphrodisiacs, is incorrectly described in the travel report and said to consist of methods of invigorating the legs and the body.

v1.c26.s3.19

The evidence about I-ching’s knowledge concerning the eight aṅgas of āyurveda285 is obviously not in favour of his having been acquainted with Aṣṭāṅgasaṃgraha or Aṣṭāṅgahṛdaya.286

v1.c26.s3.20

Some details in his information about Indian medicine reinforce this impression.

v1.c26.s3.21

He relates, for example, that people in India do not eat any kind of onions287 and that these are not permitted as food (to Buddhists), except in case of illness.288 Onions (palāṇḍu) and garlic (laśuna) are, however, the subjects of elaborate kalpas in the Saṃgraha, while the Hṛdaya, which has no palāṇḍukalpa, does contain a laśunakalpa. Both Samgraha and Hṛdaya declare, however, that brāhmaṇas abstain from garlic (A.s.U.49.162 = A.h.U.39.112), which is not śasta (A.s.U.49.164). I-ching’s data on harītakī are also peculiar in mentioning the use of the bark instead of the fruit. 289

v1.c26.s3.22

All this negative evidence is not counterbalanced by a passage, found in the travel report, 290 declaring that, if a disease be not cured by abstaining from food for seven days, one should then seek help from Avalokiteśvara. Paying homage to Āryāvalokita and other figures from the Buddhist pantheon is recommended to patients suffering from fever in the Samgraha (Ci.2.144), but this does not make it certain, as sometimes assumed, 291 that I-ching knew the work, because the veneration of Avalokiteśvara is such a widespread phenomenon in Mahāyāna Buddhism.

v1.c26.s3.23

In short, the communication in the travel report about a generally known medical treatise amounts to no more than a rumour about a book never studied by I-ching if he meant an āyurvedic treatise. More probably, he got his knowledge of Indian medicine from one or more Buddhist texts dealing with the subject. 292

v1.c26.s3.24

The vagueness of I-ching’s communication was a moving force, actuating many scholars to invent ingenuous reasonings in their efforts to solve the problem and to refute competing theories of colleagues. Rather often, it was overlooked that I-ching was not well informed about Indian medical literature; otherwise, he would never have asserted that only lately 293 an epitome of the eight branches of medicine had been composed, and that, formerly, these were dealt with in separate textbooks. Dṛḍhabala’s revision of the Carakasaṃhitā and, probably, the new version of the Suśrutasaṃhitā with the Uttaratantra appended to it, were already available as complete textbooks when I-ching was in India. Suspicion is aroused too by his explicit statement that one single treatise was the basis for medical practice in the five parts of India.

v1.c26.s3.25

As already mentioned, all these considerations did not withhold speculations about the identity of the treatise mentioned.

v1.c26.s3.26

J. Takakusu 294 suggested that the Suśrutasaṃhitā might be meant and pointed to the very similar series of the eight branches in that treatise, when compared with I-ching’s enumeration. J. Jolly, 295 not wanting to exclude this option, collected some additional data in support 296 of the hypothesis that a treatise closely related to the Suśrutasaṃhitā might be the book in question; however, not feeling sure of this, he brought forward, as an alternative solution, that it could be some version of the Aṣṭāṅgasaṃgraha, still nearer to Suśruta than the one known to us.

v1.c26.s3.27

The idea that I-ching’s information on Indian medicine could, at least partly, ultimately derive from the Suśrutasaṃhitā, was brushed aside, too easily in my opinion, in later writings on the subject.

v1.c26.s3.28

A.F.R. Hoernle 297 strongly objected to the views of J. Jolly. He advanced that the Suśrutasaṃhitā could not possibly be implicated, since its great antiquity conflicts with I-ching’s reference to a recent book and also because its title cannot be connected with I-ching’s epitome. The similarities between the lists of the aṅgas of āyurveda were reasoned away with unacceptable arguments. Moreover, Hoernle was wrong in giving much weight to the age of the Suśrutasaṃhitā, because the later version is concerned here; this later version, not the original, more ancient saṃhitā, may well be described as comprising all eight branches of medicine.

v1.c26.s3.29

Unfortunately, Hoernle’s criticisms were thought to be convincing by a number of later scholars. 298

v1.c26.s3.30

Hoernle tried to defend as a much better solution that I-ching had the Aṣṭāṅgasaṃgraha in mind. His arguments were that its title agrees with an epitome of the eight branches, that it was a recent work in the latter part of the seventh century, the period of I-ching’s stay in India, and that it was current during that period. He ruled out the Hṛdaya as a treatise with a much less suggestive title, and, also, because of his opinion that the Hṛdaya cannot be placed earlier than the eighth century. The considerable disagreements between the names of the eight aṅgas, their contents and their order, as found in I-ching’s work on the one hand and both Saṃgraha and Hṛdaya on the other, are in this way completely played down and the Saṃgraha is allotted a much more prominent place than it ever had. The thesis that I-ching alludes to the Saṃgraha was repeated in Hoernle’s work on Indian osteology. 299

v1.c26.s3.31

J. Jolly 300 wrote a short article in reply to Hoernle’s earliest claim 301 that I-ching refers to the Aṣṭāṅgasaṃgraha. He brought forward that he was not prepared to question the possibility or even plausibility of the proposed identification and emphasized that he had suggested much the same thing himself. In his view, however, it was still undecided whether the author of the Saṃgraha or Suśruta had the better claim, while the author of the Hṛdaya and the numerous writers of compendia only known from quotations should neither be lost sight of. Jolly went on with some remarks on the list of the eight aṅgas of āyurveda in I-ching’s work. He argued, quite sensibly, that the list is more in conformity with the Suśrutasaṃhitā than with Saṃgraha or Hṛdaya. A strong point mentioned in support is the noteworthy detail that I-ching uses the Sanskrit term agada, as Suśruta does, instead of giving a Chinese equivalent or paraphrase of agadatantra; this means that his source, whatever it may have been, had this term, and not daṃṣṭrā, employed in Saṃgraha and Hṛdaya.

v1.c26.s3.32

Jolly’s concise, but very clear, exposition shows unmistakably that he had a preference for regarding the Suśrutasaṃhitā or a work related to it as the treatise mentioned by I-ching, although he ended his article cautiously by saying that it does not seem to be sufficiently established that the Aṣṭāṅgasaṃgraha is the anonymous textbook referred to by the Chinese monk.

v1.c26.s3.33

Hoernle’s conviction, expressed as an authoritative conclusion, led to a rather general consensus that Vāgbhaṭa is the author of the book mentioned by I-ching, 302 without critically assessing the arguments, which, in my view, have no cogency at all.

v1.c26.s3.34

Differences of opinion arose concerning the issue whether the Saṃgraha or the Hṛdaya, both usually regarded as works of Vāgbhaṭa, should be regarded as the winning card. These discussions on the most likely candidate could originate because of the weakness of the arguments determining Hoernle’s choice of the Saṃgraha. First, his late dating of the Hṛdaya is wrong; second, the title of that work, which is often seen as an epitome of the Saṃgraha, is not less apt; third, evidence that the Saṃgraha ever enjoyed popularity is completely absent. Jolly’s perception of the situation testifies to more perspicacity, but got less attention, possibly due to the, undeserved, reputation of Hoernle.

v1.c26.s3.35

Arguments invalidating Hoernle’s reasoning about the Saṃgraha as the best choice were adduced by many scholars, but the majority of them did not question the major point and kept to Vāgbhaṭa as the author alluded to.

v1.c26.s3.36

P.V. Sharma³⁰³ exposed in detail his reasons for regarding the Hṛdaya as the recent treatise mentioned by I-ching. He thinks that I-ching may refer to the Saṃgraha once, namely in the passage where he says that the old translators declare that, if a disease be not cured by abstaining from food for seven days, one should then seek help from Avalokiteśvara.³⁰⁴ This interpretation becomes untenable when one considers the sentences that follow, where I-ching points to the want of knowledge concerning the science of medicine on the part of the old translators. P.V. Sharma adduces as an argument for I-ching’s acquaintance with the Hṛdaya his assertion that onions (palāṇḍu) are not eaten in India.³⁰⁵ Another piece of evidence consists, in his view, of the pills prepared from equal quantities of harītakī, dried ginger and sugar, described as being of great benefit.³⁰⁶ He refers to prescriptions found in the Hṛdaya, but these do not mention the equal quantities of all three substances.³⁰⁷ As a general point, P.V. Sharma brings forward that the Hṛdaya, being in verse, soon became popular.

v1.c26.s3.37

G. Mukhopādhyāya’s objections to Hoernle’s views,³⁰⁸ partly sound, partly based on the idea that the Saṃgraha is very old, resulted in the conclusion that I-ching may have had the Hṛdaya in mind, but that it is impossible to say whether his remarks may not appropriately refer to other authors whose works are lost.

v1.c26.s3.38

I-ching’s remarks also occasioned some interpretations of a quite different kind than those from the side of Hoernle.

v1.c26.s3.39

D.Ch. Bhattacharyya,³⁰⁹ who places one single Vāgbhaṭa, author of Saṃgraha and Hṛdaya, in the ninth century, a long time after I-ching’s stay in India, suggested that the latter’s reference to an epitome of the eight branches of medicine might aim at any of the compendia which existed at the time and would have achieved notoriety and popularity. The availability of these compendia is not doubtful, since they were the sources of Vṛnda, Caṇḍraṭa, and other writers. Examples given by D.Ch. Bhattacharyya are Ravigupta’s Siddhasāra, Acyuta’s Āyurvedasāra, and the works of Bhadravarman and Bindusāra.

v1.c26.s3.40

Generally, this hypothesis has not sympathetically been considered, in spite of the fact that Jolly already remarked in passing that the works known from quotations only should not be lost sight of in attempts at identifying the treatise referred to by I-ching. Bhattacharyya’s reasonings did not gain much attention on account of the obvious flaws in determining Vāgbhaṭa’s date.

v1.c26.s3.41

C. Vogel³¹⁰ described Bhattacharyya’s opinion as purely hypothetical, rather far-fetched, and unsatisfactory. He argued that a book once so popular as to have been read all over India is not very likely afterwards to have fallen into complete oblivion. This is, however, not a strong point, first, because it gives too much credit to I-ching’s reliability, and, second, because many once popular medical works did get lost. Vogel’s second objection, namely that Vāgbhaṭa does not mention any such work among his sources, is even weaker with a view to the extreme paucity of the latter’s references to authors and works he borrowed from.

v1.c26.s3.42

J. Filliozat³¹¹ repeatedly expressed as his view that the treatise mentioned by I-ching cannot possibly be the Saṃgraha or the Hṛdaya, nor the Uttaratantra of the Su- *śrutasaṃhitā*.³¹² He argued that the Uttaratantra, being an appendix to the *Suśrutasaṃhitā* that deals with additional matter, does not answer to the description of a complete medical textbook. The *Saṃgraha* and *Hṛdaya* do not qualify due to their extent, because I-ching had a short work in mind. Filliozat concluded that one cannot prove that Vāgbhaṭa is the author meant; he considered any work that can be regarded as a summary of the eight branches of medicine to be eligible as one of the candidates.

v1.c26.s3.43

All these doubts are, according to J. Filliozat, cleared away when the *Yogaśataka* is taken into consideration. That work is the proper candidate, being a short and complete manual for practitioners, divided into parts corresponding to the eight branches, and very popular too in a period not far removed from the times of I-ching.³¹³

v1.c26.s3.44

J. Filliozat was less confident in a later publication, where he conceded that I-ching is mistaken about the recent date of the *Yogaśataka*, if that is the work meant. He mentioned as another difficulty the absence of the name of the author in I-ching's travel report, which requires an explanation in discussions about the *Yogaśataka* as a possible candidate, since this treatise is often attributed, particularly in Buddhist circles, to Nāgārjuna, a name well known to I-ching as a Chinese Buddhist.³¹⁴ Filliozat admitted that I-ching's information is often questionable. He concluded, finally, that a basis for reaching a definite answer is absent.³¹⁵

v1.c26.s3.45

Filliozat's conclusion was endorsed by H.H.M. Schmidt in his notice of the former's views.³¹⁶

v1.c26.s3.46

The harvest yielded by this long digression is obviously not rich, but important nevertheless. I-ching makes us acquainted with a rumour about an unidentifiable book he never set eyes on. Evidence that Vāgbhaṭa was its author is entirely lacking. The medical historian is faced with a situation resembling that vis-a-vis the stories about Kaniṣka and Caraka.

v1.c26.s3.47

Information on the works attributed to Vāgbhaṭa can be gleaned not only from Chinese, but also from Arabic sources.

v1.c26.s3.48

One of the earliest scholars to discuss the date of composition of the *Aṣṭāṅgasaṃgraha*, P. Cordier, called attention to the resemblance of this title to that of an Indian medical work known to early Arabic writers and referred to as *Kitāb asānkar al-jāmi‘*.³¹⁷ The translator of the *Asānkar* was Ibn Duhn, who also rendered another Sanskrit medical treatise into Arabic.³¹⁸

v1.c26.s3.49

A consensus on the interpretation of the full title, *Kitāb asānkar al-jāmi‘*, has not been reached so far. G. Flügel translated al-jāmi‘ as ‘the collector’ (der Sammler), A. Müller as ‘consisting of something abbreviated’ (bestehend in etwas Abgekürztem), C. Vogel as ‘summarizing’.

v1.c26.s3.50

P. Cordier³¹⁹ brought forward that *asānkar* seems to correspond to *Aṣṭāṅgasaṃgraha*, but the supposed antiquity of that work made him hesitant about the correctness of his interpretation. J. Jolly³²⁰ disagreed with him, arguing that *asānkar* may just as well designate the *Hṛdaya*. The fact that *al-jāmi‘* belongs to the title was probably overlooked by J. Jolly, as justly noticed by C. Vogel,³²¹ who expressed as his opinion that the full Arabic title is in conformity with *Aṣṭāṅgasaṃgraha* rather than with *Aṣṭāṅgahṛdayasaṃhitā*, unless one regards the final r in *Asānkar* a relic of *Hṛdaya*. In my own view, it cannot so easily be excluded that the term al-jāmi‘ is employed in order to express that the work translated was a summary (hṛdaya). One should also take into consideration that, preceding the Asānkar, the treatises of Caraka and Suśruta are mentioned in the Arabic sources, which might well be a testimony that they were acquainted with the triad now called Bṛhattrayī.

v1.c26.s3.51

The conclusion must therefore be that it cannot be proved that the Aṣṭāṅgasaṃgraha was known to the Arabs in the period of Ibn Duhn’s activity as a translator, i.e., about A.D. 800.

v1.c26.s3.52

Quotations from the Aṣṭāṅgasaṃgraha constitute much more solid evidence for its terminus ante quem.

v1.c26.s3.53

A discussion got going by J. Filliozat has to be dealt with first. This scholar³²² ventured to assert that Vāgbhaṭa reproduces verses found in the Yogaśataka.

v1.c26.s3.54

The facts in support of J. Filliozat’s thesis are slender. The Yogaśataka, which does not refer one single verse to a source, contains three verses found in both Saṃgraha and Hṛdaya: Yogaśataka 72 = A.s.U.8.21 = A.h.U.5.18; 73 = A.s.U.8.22–23 = A.h.U.5.19; 74 = A.s.U.8.24–27 = A.h.U.5.20. Three more verses occur in the Hṛdaya only: 22 = A.h.Ci.14.38; 80 = A.h.U.39.159; 90 = A.h.U.40.49. All these six stanzas have no close parallels in Caraka- or Suśrutasaṃhitā, which means that the Yogaśataka borrowed them or that they derive from the Yogaśataka, unless an unknown source is invoked.

v1.c26.s3.55

Evidence on the sources of the Yogaśataka would be illuminating,³²³ but most of the numerous commentaries, which might give some insight, have not yet been edited. The commentary of Pūrṇasena, which has been edited, is of no help in this regard.

v1.c26.s3.56

A curious fact, noticed by H.H.M. Schmidt,³²⁴ may, at first sight, be interpreted as giving support to the hypothesis that the Yogaśataka was indeed one of the sources of the Aṣṭāṅgasaṃgraha and Hṛdaya. One verse of the Yogaśataka (72), forming part of Saṃgraha and Hṛdaya, is said to be from the Yogaśataka in the Yogaratnākara (425). This stanza describes a formula, called māheśvaradhūpa in the Yogaratnākara. The seemingly obvious conclusion that the formula derives ultimately from the Yogaśataka is, nevertheless, without any firm basis, for the Yogaratnākara quotes, for example, the recipe of mahāpaiśācaghṛta (426) as coming from the Yogatarangiṇī, while it is already found in the Carakasaṃhitā (Ci.9.45–48).

v1.c26.s3.57

A similar state of affairs is met with elsewhere. Niścalakara remarks that a particular prescription (Cakradatta, gulma 40) comes from the Yogaśata, where it is found indeed (22), but it forms part of the Aṣṭāṅgahṛdaya as well. Śivadāsasena, also a commentator on the Cakradatta, and, moreover, one who bases his work on that of Niścala, attributes the recipe to Vāgbhaṭa, which proves that he regarded the Hṛdaya as the more original source.

v1.c26.s3.58

Though conclusive evidence on the sources of the Yogaśataka has thus not been obtained, it yet appears to be more probable that its author borrowed from the Hṛdaya than that the latter took part of its material from the Yogaśataka.

v1.c26.s3.59

The uncertain date of the Yogaśataka, however, makes that work unsuitable as a cue for establishing the date of Saṃgraha or Hṛdaya.

v1.c26.s3.60

One of the earliest works quoting the Aṣṭāṅgasaṃgraha may be the Mādhavanidāna, which incorporated a number of verses found in the Saṃgraha and absent from the Hṛdaya. Since Mādhava does not refer to his sources by name, we cannot exclude with certainty the possibility that he borrowed from some other, unknown, treatise, also made of by the author of the Saṃgraha.

v1.c26.s3.61

The chapter on fevers of the Mādhavanidāna contains four stanzas (2.40–43) which, in a slightly different order and with a number of variants, form part of the Saṃgraha (Ni.2.96cd–97ab, 95cd–96ab, 100cd–102). The not inconsiderable differences in wording throw some doubt on the Saṃgraha being their source. This doubt is deepened by the fact that the cluster is preceded by an ardhaśloka (2.39cd) and followed by a verse (2.44) from an unidentified source. In addition, the commentator Vijayarakṣita says in the Madhukośa that the first verse of the short series (2.40) is taken from a treatise that goes against the (traditional) grain (pratilomatantra); it is improbable that he had the Saṃgraha in mind, which was well known to him and quoted in other contexts as Vṛddhavāgbhaṭa.

v1.c26.s3.62

The chapter on sluggishness of the digestive fire and related disorders contains a group of verses (6.10–13ab) which, though their order differs slightly, are completely identical with stanzas found in the Saṃgraha (Sū.11.31, 33, 32, 36ab). The commentators Vijayarakṣita and Vācaspati are silent on their origin, which urges caution, but it cannot be ruled out that they are taken from the Saṃgraha.

v1.c26.s3.63

A third small group of one verse and a half, found in the chapter on kuṣṭha (49.40–41ab), and said to be from the Aṣṭāṅgasaṃgraha in Jādavjī Tricumjī’s edition, is not peculiar to that text, forming part of the Hṛdaya as well.

v1.c26.s3.64

The description of the children’s disease pārīgarbhika (68.10–11) is identical with that found in the Saṃgraha (U.2.97–98), while it is absent from the Hṛdaya. Mādhava may have taken the verses from the Saṃgraha, though the commentators are silent and a common source cannot be excluded.

v1.c26.s3.65

This evidence shows that the Mādhavanidāna may quote the Saṃgraha. The fact that Mādhava used the Hṛdaya rather extensively in the composition of his work suggests that he was acquainted with the Saṃgraha as well. The alternative, i.e., the supposition that the Saṃgraha was not well known or even unknown in Mādhava’s time, is improbable.

v1.c26.s3.66

Treatises which are later than the Mādhavanidāna rather often borrow, usually anonymously, from the Aṣṭāṅgasaṃgraha. Examples are the Siddhayoga³²⁵ and the Cakradatta.³²⁶

v1.c26.s3.67

A group of commentators quote from or refer to Saṃgraha and Hṛdaya, while the members of another group distinguish a Vṛddhavāgbhaṭa and Vāgbhaṭa. Aruṇadatta,³²⁷ Candranandana,³²⁸ Hemādri³²⁹ and Indu³³⁰ belong to the first group; Narahari, Niścala and Śivadāsasena belong to the second group.³³¹ This material from the commentarial literature does not give new clues to the date of the Saṃgraha.

v1.c26.s3.68

The external evidence, taken together, shows that the Saṃgraha is probably earlier than the Mādhavanidāna and certainly preceded the Siddhayoga.

v1.c26.s3.69

One testimony remains to be discussed. Śivadāsasena states in his commentary on the Cakradatta³³² that a particular verse of the Hṛdaya (U.22.41cd) is inspired by Jejjaṭa. The presence of a corresponding passage in prose in the Saṃgraha (U.26.31) might be regarded as an indication that Jejjaṭa antedates this work too, if Śivadāsasena is to be trusted.

v1.c26.s3.70

The anteriority of Jejjaṭa, who lived in the seventh or eighth century, would imply that the Saṃgraha cannot be earlier than the seventh century and may have to be assigned to its latter half or the beginning of the eighth century.

v1.c26.s3.71

Quotations from Vāgbhaṭa in non-medical works do not provide data helpful in narrowing down the limits of his period of activity.333

v1.c26.s3.72

The terminus post quem of the Aṣṭāṅgasaṃgraha is elucidated by its sources. The names of the authorities mentioned in it have therefore to be examined first.

v1.c26.s3.73

The Carakasaṃhitā is one of the works to which the Saṃgraha is heavily indebted. The chapters of both begin in a similar way, but the second half of this opening differs. The Carakasaṃhitā has: thus spoke the venerable Ātreya, whereas the Saṃgraha says: thus spoke Ātreya and the other great sages (maharṣi). The names of these great sages, to whom Indra revealed the āyurveda, are enumerated in the story about the descent of āyurveda to the world of human beings (Sū.1.4–13ab). The list begins with Punarvasu (= Ātreya) and Dhanvantari, who are known as the representatives of kāyacikitsā and śalya, and proceeds with the names of specialists in other branches of āyurveda.

v1.c26.s3.74

Ātreya is not only mentioned at the beginning of each chapter and in the list of sages referred to. His name appears among a series of deities and sages to be honoured during a treatment with clysters (Sū.28.34). He refutes the opinion held by women that, in the seventh month of pregnancy, due to the coming into being of the hair of the head in the foetus, vidāha arises in a pregnant woman. Ātreya declares that, due to the pressure exerted by the foetus, the doṣas move about and reach the cardiac region; this is the cause of the vidāha that occurs; from this vidāha itching arises, which causes in its turn kikkisa (striae) (Śā.3.3).334 Khaṇḍakāpya’s view on the diet in the eighth month of pregnancy is rebutted by Ātreya (Śā.3.5).335 Interesting is a group of verses in the chapter on rasāyana (U.49.262–263), which contain a mantra, to be recited when guggulu is used; this mantra, said to derive from the muni Ātreya, is addressed to Puṇḍarīkākṣa, 336 Purandara (= Indra), and other deities.337

v1.c26.s3.75

Atri is mentioned once (U.50.209) and may be the same as Ātreya in this case.

v1.c26.s3.76

Punarvasu, i.e., Punarvasu Ātreya, is presented as the authority to whom Agniveśa addresses himself for instruction (Ka.5.57). He is referred to as munivṛṣabha in the verses that follow.

v1.c26.s3.77

Finally, Punarvasu is said to be the only authority to distinguish eight instead of seven vegas of poison (U.40.28).

v1.c26.s3.78

Caraka is mentioned a number of times. Three āsthāpana enemas are sufficient in order to subdue the doṣas in cases of saṃnipāta according to Caraka; a larger number is not necessary at all, because there does not exist a fourth doṣa (Sū.28.54). Interestingly, the Hṛdaya has only the first two pādas in common (Sū.19.59cd); it proceeds with a verse (Sū.19.60), which, though having the same content as the next four pādas of the Saṃgraha, does not refer to Caraka, but, instead, to other physicians, who are, moreover, represented as having a divergent opinion, disagreeing with that of the author.338

v1.c26.s3.79

As far as I know, the *Carakasaṃhitā* never refers to the theory that some authorities acknowledge a fourth doṣa, which raises the problem of the *Saṃgraha*’s source. The second part of the statement may be directed against Suśruta, who says (Sū.35.6) that a basti is beneficial in disorders caused by vāta, pitta, kapha, blood, in cases of saṃsarga, and in saṃnipāta.

v1.c26.s3.80

The second quotation from Caraka (Śā.10.31) deals with the essential character of a riṣṭa as a certain sign of approaching death; those without an accurate knowledge of the matter make mistakes in not recognizing a riṣṭa or seeing one where none is present. The same verse forms part of the *Hṛdaya* (Śā.5.2), where it is not ascribed to Caraka. The stanza is absent from the *Carakasaṃhitā*, as already suggested by the use of riṣṭa instead of ariṣṭa, but has a parallel there, though quite differently worded (I.2.6). Again, the source of the verse remains unknown.

v1.c26.s3.81

The paribhāṣā section at the end of the Kalpasthāna refers to Caraka several times (25, 27, 61, 70).

v1.c26.s3.82

Next to the references to Ātreya and Caraka, those to Agniveśa have to be dealt with.

v1.c26.s3.83

Agniveśa is one of those who composed a medical treatise (Sū.1.11). He is mentioned as such again and belongs to a series of deities and sages to be honoured during a treatment with clysters (Sū.28.34).

v1.c26.s3.84

Agniveśa requests Punarvasu to explain to him which fruits may profitably be used in clysters (Ka.5.57). The stanzas giving Punarvasu’s opinion on the subject agree literally with some found in the Siddhisthāna of the *Carakasaṃhitā*,³³⁹ where not Agniveśa, but a group of other sages approach the son of Atri (= Ātreya Punarvasu) with an identical question. Some more verses from the same chapters are clearly related to each other as well.³⁴⁰

v1.c26.s3.85

Agniveśa’s theory on the pathophysiological processes operative in a saṃtata fever is described (Ni.2.62–64a), preceding the rival theory of Hārīta. These verses are noteworthy in not presenting Ātreya as the authority, but his pupil, and contrasting him with Hārīta.

v1.c26.s3.86

The material collected so far demonstrates that the *Saṃgraha* is later than Dṛḍhabala’s version of the *Carakasaṃhitā*, in spite of the absence of his name. Additional support is provided by the numerous verses from Dṛḍhabala’s contributions which are identical with or closely related to verses occurring in the *Saṃgraha*. The Kalpasthāna of the *Saṃgraha* illustrates this, for it contains, when compared with the *Hṛdaya*, a large number of additional verses, literally borrowed from the Kalpasthāna of the *Carakasaṃhitā*, which derives from Dṛḍhabala.

v1.c26.s3.87

The evidence notwithstanding, some scholars disagree with this point of view.

v1.c26.s3.88

A.F.R. Hoernle was convinced that Dṛḍhabala is posterior to the period of composition of the *Saṃgraha*. He claimed³⁴¹ that Dṛḍhabala obtained his total of ninety-six eye diseases by adopting the number of ninety-four of the *Saṃgraha* and adding the two new diseases of Mādhava. This quite erroneous interpretation has been discussed in the section on Dṛḍhabala. P.V. Sharma³⁴² showed that Dṛḍhabala simply accepted Karāla’s number of eye diseases.³⁴³ The other reasons that induced Hoernle to assigning Dṛḍhabala to a period even later than Mādhava are equally wrong.

v1.c26.s3.89

S. Dasgupta³⁴⁴ accepted Hoernle’s statement that Dṛḍhabala was acquainted with the Saṃgraha and, for that reason, he reproduced part of Hoernle’s references.

v1.c26.s3.90

P.V. Sharma³⁴⁵ expressed as his opinion that the Saṃgraha dates from shortly after Dṛḍhabala or from about the period of his activity, which explains that Kapilabala, Dṛḍhabala’s father, is mentioned in it, not Dṛḍhabala himself. He asserts that the Saṃgraha does not quote from Dṛḍhabala’s parts of the Carakasaṃhitā, because it dates from a period in which this version of the saṃhitā was not yet well known. An additional argument adduced by P.V. Sharma is the concise treatment of pañcakarman in the Saṃgraha, while this subject is dealt with much more elaborately by Dṛḍhabala.

v1.c26.s3.91

Unfortunately, P.V. Sharma does not set forth where all the material of the Saṃgraha dealing with subjects treated by Dṛḍhabala may come from. He also fails to explain why Dṛḍhabala’s name does not appear in the Hṛdaya. P.V. Sharma’s identification of the Kapilabala, quoted in the Saṃgraha (Sū.20.17), as Dṛḍhabala’s father, is no more than a guess and therefore unsuited for chronological purposes.

v1.c26.s3.92

One, rarely or never noticed, aspect of the relationship between Saṃgraha and Carakasaṃhitā has still to be mentioned. Material found in prose in one of the chapters of the Carakasaṃhitā appears in a versified form in the Saṃgraha.³⁴⁶

v1.c26.s3.93

The relationship between Saṃgraha and Suśrutasaṃhitā is more difficult to determine.³⁴⁷ In general, the Saṃgraha deals much more freely with material also occurring in the Suśrutasaṃhitā than with that found in the Carakasaṃhitā. The former is less often quoted literally, while very numerous verses of Saṃgraha and Carakasaṃhitā are identical. In spite of this striking difference, verses common to Suśrutasaṃhitā and Saṃgraha and verses closely related are not are at all.³⁴⁸

v1.c26.s3.94

P.V. Sharma³⁴⁹ is of the opinion that the revised and completed version of the Suśrutasaṃhitā as now known to us was not yet available to the author of the Saṃgraha, but he fails to give this claim a solid basis.

v1.c26.s3.95

A.F.R.Hoernle³⁵⁰ expressed as his opinion that this relationship is less certain than that with the Carakasaṃhitā because the Suśrutasaṃhitā is never actually quoted, but that, still, numerous indications point to a decided posteriority of the Saṃgraha to its revised and completed version, suggested, for example, by the treatment of the eye diseases, which follows the outline as found in Suśruta’s Uttaratantra. He also points to the division of the Saṃgraha into the same sections as the Suśrutasaṃhitā. Hoernle’s additional remark that the Indian tradition makes in the same direction cannot be accepted as a valid argument. His assertion that the Suśrutasaṃhitā is never actually quoted is far from the truth, but the presence of the same sections in Suśruta’s work on the one hand and Saṃgraha and Hṛdaya on the other, is a valid point, worthy of consideration.

v1.c26.s3.96

An examination of the actual state of affairs may elucidate this issue.

v1.c26.s3.97

Suśruta is mentioned several times in the Saṃgraha. The opening chapter (Sū.1.11) refers to him as one of the authors of a tantra, which is a straightforward indication that some version of the Suśrutasaṃhitā was known to the author of the Saṃgraha.³⁵¹

v1.c26.s3.98

Suśruta is one of the sages and deities to be paid homage to in the preparation of the sarvārthasiddhāñjana (Sū.8.59).³⁵²

v1.c26.s3.99

His opinion on the taste of pitta and kapha, when in a vidagdha state, is quoted (Sū.20.18),³⁵³ after Kapilabala’s view on the tastes of doṣas which have increased. Suśruta’s opinion as referred to is in conformity with the Suśrutasaṃhitā (Sū.21.11 and 15).

v1.c26.s3.100

The doṣopakramaṇīya chapter (Sū.21) discusses the views of a series of authorities on the order in which increased doṣas have to be controlled. The section referring to these very diverse opinions, absent from the Hṛdaya, mentions two authorities by name, Parāśara and Suśruta. The latter teaches that there is no rule valid for all diseases. He prescribes that in fever and diarrhoea pitta should be subdued first, kapha next, and vāta last (Sū.21.22),³⁵⁴ for reasons given in the next three stanzas.

v1.c26.s3.101

Suśruta’s point of view regarding ariṣṭas is quoted in the chapter of the Śārīrasthāna where Caraka and Kṛṣṇātreya are cited on the same subject. Suśruta asserts in a verse, borrowed from the Suśrutasaṃhitā³⁵⁵ indeed (A.s.Śā.10.32cd–33ab = Su.Sū.28.5), that the effect of an ariṣṭa can be counteracted by brāhmaṇas without blemish by means of rasāyana, tapas, etc.

v1.c26.s3.102

Finally, Suśruta is referred to in the paribhāṣā section at the end of the Kalpasthāna.³⁵⁶

v1.c26.s3.103

Two figures to be dealt with in the same context as Suśruta are Dhanvantari and Kāśirāja.

v1.c26.s3.104

Dhanvantari appears in the list of sages to whom Indra revealed the āyurveda (Sū. 1.6).³⁵⁷ He is, together with Suśruta, among the deities and sages to be honoured in the ritual accompanying the preparation of the sarvārthasiddhāñjana (Sū.8.59).³⁵⁸ Dhanvantari is also the authority who recommends the use of balātaila (Śā.4.51).³⁵⁹ On the subject of the diet of a pregnant woman, he declares what it should consist of from the eighth month onwards (Śā.3.8).³⁶⁰

v1.c26.s3.105

The chapter on the treatment of poisoning (U.40) relates that Dhanvantari largely agrees with Ālambāyana on the substrates of the seven vegas of poison in those bitten by a snake. The only difference between the two is the substrate of the fourth vega: the snāyus according to Ālambāyana, the koṣṭha according to Dhanvantari (U.40.36).³⁶¹ The next two stanzas, on the role of the kalās, appear to come from the same source and represent views of the Dhānvantarīyas according to Indu.³⁶²

v1.c26.s3.106

The name of Kāśirāja appears once only among deities and sages to be honoured during a treatment with clysters (Sū.28.34).³⁶³

v1.c26.s3.107

This material makes clear that the Saṃgraha borrowed from the Suśrutasaṃhitā, but does not yet provide an answer to the problem whether or not the Uttaratantra was among its sources.

v1.c26.s3.108

First, some examples among the borrowings from Suśruta may be useful in illustrating the way this source is handled in the Saṃgraha.³⁶⁴

v1.c26.s3.109

The divisions of time (Sū.4.3–4) are partly the same as those found in the Suśrutasaṃhitā (Sū.6.4–5). The latter’s muhūrta is replaced by the nāḍikā;³⁶⁵ two of these constitute a muhūrta, which makes its duration twice as long as in Suśruta’s scheme. Four muhūrta, minus one-fourth of it, make a yāma, a measure of time absent from Suśruta’s division. Four yāma make one day and night, which corresponds to the thirty muhūrta of Suśruta.

v1.c26.s3.110

The passage on the divisions of time of the Saṃgraha is more probably based on Suśruta than on a common source, because of the similarities in wording and the additions found in the Saṃgraha.

v1.c26.s3.111

Remarkable is the relationship between several chapters of the Uttarasthāna of the Saṃgraha and the corresponding ones of the Nidānasthāna of the Suśrutasaṃhitā. It should be borne in mind that the main features of this relationship are common to Saṃgraha and Hṛdaya.

v1.c26.s3.112

An example is the connection between the mukharoga chapters (Su.Ni.16 and A.s. U.25 = A.h.U.21). Though the relatedness is clear, there are many differences, both in wording and contents, which exclude that the Saṃgraha version is simply based on the Suśrutasaṃhitā. The conclusion that the author of the Saṃgraha must have disposed of additional sources, next to the Suśrutasaṃhitā and works by which that treatise was influenced, is unavoidable.

v1.c26.s3.113

A related issue concerns the rarely broached question which text of the Suśrutasaṃhitā was preferred by the author of the Saṃgraha. Details found in some chapters of the Śārīrasthāna 366 may indicate that he regarded Gayadāsa’s text as authoritative.

v1.c26.s3.114

A relationship between several chapters of the Uttarasthāna of the Saṃgraha and the Uttaratantra of the Suśrutasaṃhitā is obvious, without permitting a definitive judgment on the anteriority of the one or the other of these sections. A major difference is that the chapters of the Saṃgraha are partly in prose, while those of the Suśrutasaṃhitā are in verse. Even more important is the fact that the text of the Saṃgraha is more elaborate, gives more details, and presents new material. This state of affairs makes it necessary to examine whether the corresponding chapters of Suśruta’s Uttaratantra are one of the sources of the Saṃgraha, more concise versifications of the Saṃgraha versions, or based on unknown common sources.

v1.c26.s3.115

Some verses of the Saṃgraha are closely related to stanzas of the Uttaratantra of the Suśrutasaṃhitā, without being identical. 367 Several mantras are common to the Uttarasthāna of the Saṃgraha and the Uttaratantra. 368

v1.c26.s3.116

The Uttarasthāna of the Saṃgraha and the Uttaratantra of the Suśrutasaṃhitā differ considerably on numerous issues. A considerable number of these differences are common to Saṃgraha and Hṛdaya, 369 but the former also presents many peculiar features. 370

v1.c26.s3.117

The provisional conclusion that can be deduced from this evidence remains that the links between the Uttarasthāna of the Saṃgraha and the Uttaratantra of the Suśrutasaṃhitā do not allow us to prove which of the two is earlier.

v1.c26.s3.118

One testimony may, however, be seen as indicating that the Saṃgraha is later than the Uttaratantra of the Suśrutasaṃhitā. This testimony consists of Śivadāsasena’s remark that Vāgbhaṭa follows Jejjaṭa in a particular verse of the Hṛdaya. 371 Since the Saṃgraha contains a prose passage corresponding to the verse found in the Hṛdaya, 372 the remark applies to both treatises. Jejjaṭa is known to have written a commentary on the whole of the Suśrutasaṃhitā, which leads to the conclusion that the Saṃgraha is later than the Uttaratantra, with the obvious proviso that Śivadāsasena’s testimony can be accepted as trustworthy.

v1.c26.s3.119

The fact that the Saṃgraha ends with an Uttarasthāna, resembling the Uttaratantra of the Suśrutasaṃhitā, has been seen as supporting the anteriority of the latter. 373 It must be conceded that it is probable indeed that an Uttaratantra with the structure of an appendix precedes an Uttarasthāna that forms an integral part of a treatise.

v1.c26.s3.120

Finally, a piece of indirect evidence may be interpreted as indicating that Saṃgraha and Hṛdaya are later than the version of the Suśrutasaṃhitā, together with its Uttaratantra, often ascribed to a Nāgārjuna. Narahari refers in his Vāgbhaṭamaṇḍana (chapter 21) to the fact that Vāgbhaṭa adopted the variant bastidvāra (A.s.Śā.4.52; A.h. Śā.2.53) instead of the reading bastamāra of the Suśrutasaṃhitā (Ni.8.14),³⁷⁴ which was the source of the verse; his information that bastidvāra was the reading accepted by the Nāgārjunīyas may mean that Saṃgraha and Hṛdaya accepted some revised text of the Suśrutasaṃhitā, which was accompanied by the Uttaratantra.

v1.c26.s3.121

Bheḍa is one of the authorities quoted in the Saṃgraha who are left to be discussed. P.V. Sharma identified some passages which may have been influenced by Bheḍa, but without his name being mentioned. One of these is the description of one of the ways of examining food by the royal physician in order to prevent poisoning of the king (A.s. Sū.8.6–8). This method, the examination of the food’s rūpa, called annasvarūpaparīkṣā by P.V. Sharma, absent from Caraka and Suśruta, is found in the Bhelasaṃhitā (Sū.18. 4–6).

v1.c26.s3.122

The other medical authorities mentioned by name in the Saṃgraha³⁷⁵ do not give us clues as to its chronological position. Nevertheless, some of these names are of considerable interest, because they are absent from Caraka- and Suśrutasaṃhitā. These noteworthy names are: Ālambāyana, Bṛhaspati,³⁷⁶ Kapilabala, Māṇḍavya, and Uśanas.³⁷⁷ The recipes ascribed to Kauṭilya are not helpful in determining the date of the Saṃgraha.

v1.c26.s3.123

The external and internal evidence, so far discussed, when taken together, leads to the conclusion that the Saṃgraha is later than Dṛḍhabala’s revised and completed version of the Carakasaṃhitā and, very probably, also posterior to the revised and completed version of the Suśrutasaṃhitā. The reliability of Śivadāsasena’s remark that Jejjaṭa was one of Vāgbhaṭa’s sources whether or not the Saṃgraha is posterior to Jejjaṭa, who lived in the seventh or, at the latest, the eighth century. As to the lower limit, the Saṃgraha is probably earlier than the Mādhavanidāna and certainly earlier than Vṛnda’s Siddhayoga.

Cites (14)