Note On Class Of Ascetics Called Unmajjaka

Author(s): A Wezler
Publisher: A Wezler
Catalog link: https://jainqq.org/explore/269624/1

Page #1

BEI 9 (1991): 217-234
Albrecht WEZLER
A note on the class of ascetics called unmajjaka
1. The well-known dichotomy of the vänaprasthas into pacamānakas' and apacamānakas of BaudhDhS 3.3.16.2 is so obviously based on the distinction between an essential property of the food taken by these two classes of ascetics, respectively, that one hesitates to state this explicitly. Similarly it need hardly be added that the apacamānakas can not by any means be regarded as forerunners of the modern "Rohköstler", i.e. people who (only or partially) live on uncooked fruits and vegetables - as part of a dietetic treatment prescribed by a medical doctor or voluntarily for various reasons — and it is certainly not necessary either to mention that the eating of raw vegetables and fruits exclusively as part of or as the very substance of an ascetic practice can still be observed in present-day India. It is evidently the extraordinary and very special significance attached by the Vedic Indians to the fire, the God Fire, its/his polyvalent manifestations and the various kinds of fireplaces belonging to the house and to the sacrificial ground, which led them to pay all this attention to the manner in which
This is the slightly revised version of a paper which I had the pleasure of reading at the Instituts d'Extrême-Orient of the Collège de France on the 19th of May 1992, at the kind invitation of Association Française pour les Etudes Sanskrites and URA 1058 “LACMI" (University of Paris-3/CNRS (head: Prof. Dr. Nalini Balbir]). In it most of the suggestions given and questions put by my French colleagues on this occasion are reflected in one way or the other.
1. Debrunner, Altindische Grammatik 11,2, Göttingen 1954, p. 538 (8 362d) notes that "-ka- tritt ... schon früh gern auch an Partizipien", but the only example he adduces where this suffix is added to an ātmanepada participle is pravartamānaká (cf. RV 1.191.16), and as for its meaning, he does not say more than "und zwar zunächst nur in deminutiver Bedeutung". With the help of W. Schwarz, Rückläufiges Wörterbuch des Altindischen. Reverse Index of Old Indian, Wiesbaden 1974-1978, p. 18, however, more relevant examples can be found, and some of them (e.g. vardhamanaka) confirm the assumption, suggested by (a)pacamānaka, that -ka- is added in order to distinguish a particular group of people characterized by a common practice, or lack of it, from persons who “cook for themselves", i.e. that it forms what nowadays would be called a designation of an occupation. - Note that the ātmanepada makes real sense only in the case of pacamānaka. - C. Lévi-Strauss', Mythologica I. Le cru et le cuit, Paris 1964, does not deal with this Indian dichotomy.
2. The text reads: atha vānaprasthadvaividhyam // 1 // pacamānakā apacamānakās ceti // 2 //.

Page #2

218
Albrecht WEZLER
vānaprasthas took along "fire" when leaving the world', i.e. their home in the village or town). As many other features of Indian asceticism (in general), renouncing the use of fire, too, gives the impression of leading back to a much earlier stage in the cultural development of mankind, viz. when man had not yet learned to 'tame fire". To try to imagine this stage, and the later one when man had acquired enough knowledge about this elementary force to be himself able to produce fire at his own will, etc., helps in any case to achieve a fuller and deeper understanding of the Aryan and Vedic ideas about agni Agni.
As for the adhyāya of the BaudhDhS at issue here, the dichotomy just mentioned is noteworthy also because the principle on which it is based is followed consistently when subdividing the pacamānakas into five subgroups; but it is now their particular type of food by which the sarvāranyakas, etc., are defined and distinguished from each other: the use of fire to prepare their respective food is, of course, per definitionem common to all of them. The subdivision of the class of the apacamānakas, however, is much less clear in itself, i.e. is not also based on this, or a similarly consistent and plausible, element of the (physical) 'basic needs. Only in two cases, viz. that of the toyāhāras" and that of the vāyubhakşas, what is decisive is once again what these ascetics take. In two other cases, viz. those forming the second and the third sub
3. In reality, however, one of the reasons for regarding a diet consisting of uncooked food as ascetic in substance may have been that the Indians of those early days were still conscious of what it means to be able to prepare food by cooking it in so far as fire, not to speak of other forms of energy, was after all not so easily available as nowadays, even if we do have good reason to assume that the forest covering in the subcontinent then was far more extensive than it is today. Most probably it was a common experience for a rather large group of the population not to eat a warm meal every day.
4. For sūtra 3.3.3 reads: tatra pacamānakāḥ pancavidhāḥ sarvaranyakā vaitusikāḥ kandamūlabhaksāh phalabhaksā, sākabhakşās ceti //.
5. Taught at 3.3.9: parcaivāpacamānakā unmajjakāḥ pravíttāśino mukhenādāyinas toyāhārā vāyubhakşās ceti //.
6. In contradistinction to the use made of this term in recent years by officials of so-called 'developing countries', what I mean is nothing than the quantity of food (of any kind) and of water absolutely necessary to keep a person alive.
7. Cf. sūtra 3.3.13: toyāhārāḥ kevalam
toyāhārāh //.
8. Cf. sutra 3.3.14: vāyubhaksā nirāhārāś ca //. See on these two latter sūtras, but also on 3.3.11 and 12, the important remarks by my friend S.A. Srinivasan, Studies in the Rāma Story ..., Wiesbaden 1984, notes 154156, (Vol. II, p. 46f.). - As for nirāhārāś ca I am, however, not convinced that Srinivasan is right when he states (oote 153) that “This introduces a sixth group of ascetics not covered by the five groups only enumerated at ... sutra 9"; to me it would seem that nirāhārās in fact is meant to explain väyubhaksā - just as kevalar toyāhārāh the definiendum toyāhārāḥ in the preceding sutra -, i.e. that in both cases, though in different ways, it is made clear that the ascetics referred to do not take in anything but water or air (“wind") in accordance with the less problematic paraphrase given by Patañjali, the Mahābhāşyakara (see note 11). The ca may have crept in secondarily, or even have been repeated by the author himself, in view of the function it has in sūtra 3.3.9 (cf. also 3.3.3), viz. to mark the end of the enumeration or explanation, though no ca is admittedly found at the end of 3.3.8. Note, however, that 3.3.15 (iti vaikhānasānām vihitā daśa dikşāh) shows that the classification as a whole is completed by 3.3.14.

Page #3

The class of ascetics called unmajjaka
219
category, however, it is not the (raw) food itself that distinguishes these ascetics from each other and from the others, but the manner in which it is gathered or taken'. Yet the latter pair as such not only forms a reasonable subdivision by itself, but can also be said to refer at least indirectly to ingestion of food". The fifth subcategory - according to the sequence of enumeration of the BaudhDhS it is actually the first of the 5 types of apacamānakas -, however, is of an entirely different nature, i.e. does not at all go with what clearly serves as the basis of the dichotomy as well as of its corresponding subdivisions. For unmajjakāḥ of Baudh DhS 3.3.9 is, rightly to some extent, not explained semantically in the subsequent sūtra – in contradistinction to the terms denoting other classes of ‘non-cooking' vānaprasthas: there cannot be any doubt that this expression was regarded by the author as sufficiently clear in itself in that it is immediately recognized as a nomen agentis" of ud-Vmajj meaning "one who emerges (out of water)". Clearly this designation has nothing at all to do with food or the manner it is gathered or taken.
Unfortunately sūtra 3.3.10 (tatronmajjakā nāma lohāśmakaraṇavarjam)12 does not give any clue for understanding the nature of this class of ascetics, and Govindasvāmin is equally of no help at all. Bühler14 simply retains the Sanskrit term, and P.V. Kane's in fact follows his example.
9. Srinivasan's critical remarks apart (for which see the reference in note 8), in the case of the mukhenādāyins at least a more appropriate expression than German "essen” would certainly be "fressen". For, to be sure, the manner of eating typical of animals is deliberately imitated by these ascetics; cf. in this connection not only the verse
mrgaih saha parispandah, etc., found at BaudhDhs 3.2.19 and 3.3.22, but also the pāda Mbh. 12.250.19d as well as 7 App. 8, 170 post.
mrgaih saha cacāra sā //.
10. On food as a 'cultural construct in India see now'P. Olivelle's contribution "From Feast to Fast: Food and the Indian Ascetic” to the book Rules and Remedies in Classical Indian Law, ed. by J. Leslie (Panels of the VIIth World Sanskrit Conference .... Vol. IX), Leiden (etc.), 1991, pp. 17-35.
11. And at the same time as a designation like (a)pacamānaka (on which see above note 1)? Note also that it cannot be simply taken for granted that it is a fixed, i.e. well established and generally known, term; after all instead of toyāhāra Patañjali, e.g., uses abbh akşa (Mahābhāşya I 6.23; III 180.15 and 333.6). - The remark of Sprockhoff ("Aranyaka and Vänaprastha in der vedischen Literatur ... Zweiter Hauptteil” in WZKS XXVIII (1984), p. 27), “Daß die unmajjakāh durch Untertauchen im Wasser baden, besagt bereits ihr Name", I fail to understand for more than one reason.
12. Note that tatra ("among them") and nāma ("(those) called ...") are together used in sūtra 3.3.4, too, and nāma alone in 3.3.5f.
13. For he confines himself to explaining only lohāśmakaranavarjam; see below note 95.
14. Sacred Laws of the Aryas as taught in the schools of Āpastamba, Gautama, Vāsishtha and Baudhāyana, Pt. II, (SBE Vol. XVI), Oxford 1882, p. 292f.
15. History of Dharmaśāstra ..., Vol. II, 2nd ed., Poona 1974, p. 922.

Page #4

220
Albrecht WEZLER 2. Of the dictionaries it is, if I am not mistaken, only Apte's 16 which really contributes to a clarification. For he quotes the verse
kanthadadhne jale sthitvā tapaḥ kurvan pravartate/ unmajjakaḥ sa vijñeyas tāpaso lokapujitaḥ 11,
and in a later edition"7 the reference "Rām. 3.6.3" is added. But this is obviously an error because the passage of the Rāmāyaṇa referred to (Baroda Ed. 3.5.3) is entirely different although in a long list of designations of various ascetics 18 it does contain our unmajjakāḥ also. The commmentator Rāma, however, in his Tilaka seems to quote19 the beginning of this verse20, the text of which is given in full also in the Vācaspatya? and in the Sabdastomamahānidhi, but without any indication of the source. Other commentators substantially agree with Rāma, but add also some further interesting details. Thus Mādhavayoginas explains: kanthadadhnajale hemanteșu sthitvā tapassādhakāḥ (= unmajjakāḥ); Govindarāja says24: siromätram uddhrtya sada jalāvagādhāh, while Sivasahāyas formulates the same facts in yet another manner, viz. thus: kanthaparimitajale sthitirūpataponiratä”.
16. The Practical Sanskrit-English Dictionary..., 2nd ed., Bombay 1912.
17. Viz. The Revised and Enlarged Edition prepared by P.K. Gode and E.G. Karve, Poona 1957.
18. Cf. also M.G. Bhagat, Ancient Indian Asceticism, Delhi 1976, p. 262.
19. He does not, however, use an iti.
20. Viz. from kanthadadhne to kurvantah (he uses, of course, the plural).
21. Ed. by Sri Taranatha Tarkavachaspati, repr. Varanasi 1962; most probably this was also Apte's source!
22. Compiled by Tarkavāchaspati Sri Tärānātha Bhattacharya, Varanasi 1967, s.v.
23. Śrimadvālmikirāmāyaṇa with Amstakataka of Mādhavayogi (Vol. IV - Aranyakanda), ed. by N.S. Venkatanathacharya, Mysore 1965, p. 41.
24. Srimad Valmiki Rāmāyana, A Critical Edition with the Commentary of Sri Govindaraja ..., ed. by T. R. Krishnacharya and T. R. Vyasacharya, Bombay 1981, Aranyakānda, p. 20.- Cf. also the verse
hemante jāhnavitoye harakāmy adhifitale / akanthamagna dhyāyanti
sansthită candrasekhare //. quoted by M.R. Kale at the end of his notes on Kumāras. 5.26 (see below p. 226) (Kalidasa's Kumārasambhava ... ed. by M.R. Kale, repr. Delhi (etc.) 1967, p. [88]). His reference "S.P." (= Sivapurāņa) I have not however been able to verify.
25. Rāmāyana of Valmiki with the Commentaries Tilaka of Rāma, Rāmāyaṇasiromani of sivasahāya and Bhusana of Govindaraja, ed. by Shastri Shrinivasa Katti Mudholakara (repr. Parimal Sanskrit Series No. 11). Delhi 1983, Vol. III Aranyakānda, p. 1065.

Page #5

The class of ascetics called unmajjaka
221
The verse quoted by Apte and the explanations given by the Rāmāyaṇa commentators are important first of all for the following reason: An examination of all the passages listed in the Vaidika-Padānukrama-Koşa s.v. un-majj26 yields the result that this verb, even if it is used alone, i.e. does not stand in explicit opposition to a form of ni-majj or majj, seems always to imply that the act of emerging is preceded by the contrary act of submerging?7. Therefore one cannot help asking oneself the question whether the class of ascetics called unmajjaka is characterized by repeated submergence - and perhaps remaining submerged as long as possible - and emergence from the water - perhaps only for that much of time which is necessary for breathing?8. It is precisely this (not only semantically possible) assumption that is ruled out by the explanations quoted by me just now29. For all the commentators are unanimously of the opinion that the essential characteristic of this class of ascetics is, on the contrary, that they stand in water which reaches up to their throat, and do not move, at least not up and down. And this is tantamount to stating that these ascetics are unmajjakas not in the literal sense of people who carry out the action of emerging from water, but only look like such people in so far as their head and perhaps part of their neck are above the surface of the water in which they in reality remain standing upright, most probably for a considerable period of time. It would seem that it is this practice which constitutes
26. Three of them, viz. "Baudh 1.5.110" and "Baupi 3.4.16; 6,6", I was not able to identify. The others are AV 10.4.4 (= Paipp. 16.15.3), TB 1.1.3.6., JB 3:43, Tripadvibhuti Mahänārāyana Up. 5.13, ĀgniveśyaGS 2.6.2:13; 3.4.4:19 (unmajjya only); BaudhDhS 2.5.7 and Sankhalikhita [for the edition used see note 65) no. 103.
27. As is also explicitly stated in the Udäna-atthakathā (74,26: ummujjanam pana nimajjanam antarena n'atthi). The wrong - meaning "to dive" given among others by Monier-Williams with reference to "ĀśvG? IV,4,10" has been induced by a corresponding misunderstanding.
28. The exhaling could after all be done by the ascetic while still under water.
29. Note that this statement of mine is not falsified by the “500 hundred matted hair ascetics" who according to Vin. I 31.36 (cf. Udäna 6.15) "on the cold winter nights between the eights in a time of snowfall ... were plunging into the river Neranjarā, then emerging and repeatedly plunging in and out" (... te jațilā sitāsu hemantikāsu rattisu antaratthakāsu himapātasamaye najjā Neranjarāyam nimujjanti pi, ummujjanti pi, ummujjanimujjam pi karonti) For, these ascetics not only tend the sacred fires (Vin I 31.6f., 16, 26f.) - and carry implements for fire-worship (Vin I 33.6: aggihuttamissa)-, but also warm themselves, having come up from the river, at 500 fire-vessels (miraculously created for them by the Buddha) so that the interpretation offered by the Udāna (6.16), and the Udāna-atthakathä (74.25ff.), is very convincing indeed, viz. that what these ascetics want to achieve by this practice is pāpasuddhi - and not the accumulation of tapas-substance. Cf. also I.B. Horner, The Book of the Discipline ..., Vol. IV (Mahāvagga), London 1951 (from whose translation I have also quoted in this note), p. 42 n. 1. In any case, I do not think that this passage, isolated as it is, can be assumed to prove quite another nature of the ascetic practice than the one I am arguing for in the present note. The Jain parallels pointed out to me by Prof. Balbir are of a somewhat similar character in that they, too, do not permit to gain an entirely clear idea about the differences in ascetic practice between the unmajjakas, on the one hand, and the sammajjakas and nimajjakas, on the other (cf. E. Leumann, Das Aupapātika Sütra, erstes Upānga der Jaina ..., Leipzig 1883, p. 68f. ($ 74) as well as Jagdishchandra Jain, Life in Ancient India as Depicted in the Jain Canon and Commentaries ..., Delhi 1984 (repr.), p. 300ff.), inspite of the explanations given by Abhayadeva in his commentary on the Viyahapannattisutta (cf. Vihāyapannattisuttam, Pt. II ed. by Pt. Bechardas J. Joshi assisted by Pt. Amritlal Mohanlal Bhojak, (Jaina-Agama-Series No. 4 (Part 2)), Bombay 1978, p. 517).

Page #6

222
Albrecht WEZLER
the specific means of self-mortification
in their case.
3. The assumption that hence suggests itself as immediately plausible is that unmajjaka refers to the ascetic practice technically called udaväsa elsewhere. To render the latter expression by "residing in water” (“Aufenthalt im Wasser”), as it is done in the Larger Petrograd Dictionary, amounts to a literalness with an almost touching naivity in that the essential element of the semantic content is not even mentioned by way of explanation, viz. that a particular practice of self-mortification is referred to.
The compound as such, i.e. the substitution of uda- for udaka- as prior member, is already taught by Pāṇini in 6.3.58. There is no proof that Pänini had in view the ascetic practice at issue, as the expression is obviously not attested in Vedic literature. But the opposite possibility, too, viz. that he did in fact know this practice and udavāsa as the term used to denote it, cannot, of course, be precluded just for this reason3l. After all the evidence of the Mahābhārata has also to be taken into account, as rightly stated by Oberlies 52 although there are more relevant passages than the two mentioned by him”. Most, if not all of them have, I suppose, been mentioned by M. Shee in her doctoral dissertations, including those where the expression udaväsa itself is not attested but a similar one, or where the corresponding practice is described in other words. I do not deem it necessary on my part to examine these passages in detail here, but regard it nevertheless as advisable to take a closer look at some of them with a view of possibly gathering additional information about the udaväsa form of tapas.
Among the various kinds of "superhuman" self-mortificatory practices (atimānuşa3s tapah) taken up by Ambā one after the other, udaväsa in the Yamunā, i.e. a river, is the second one; the attribute nirāhārā, however, she is given in this connection
30. I agree with Dr. A. Roşu that Indian asceticism, or at least some of its practices, have to be studied also from the point of view of psychology and psychopathology (which knows of similar phenomena although generally individual cases only, in Western societies). But I don't deem the present note the right place to address this problem, too, and not so much because I would then certainly need the help of specialists in the field.
31. I could not find any explanatory remark on this compound in works of the Pāṇiniyan tradition; worthy of note is the vigraha given in the Kāśikā (on Pan. 6.3.58), viz, udakasya vāsah. I do not, however, know why the genitivus objectivus (alone) is assumed.
32. Studie zum Candravyākarana. Eine kritische Bearbeitung von Candra IV.4.52-148 und V.2, Stuttgart 1989, p. 212f.
33. Viz. Mbh. 13.7.11 and 50.3.
34. Tapas und tapasvin in den erzählenden Partien des Mahābhārata, Reinbek 1986, p. 249ff.
35. One wonders whether this attribute is meant to express that the various practices as such are beyond what a human ascetic could perform or whether it refers to the periods of time for which they are said to have been followed by Ambā. But ultimately one cannot help gathering the impression that the expression is rather not to be taken literally as it seems primarily meant to emphasize the allegedly - extraordinary vigour of Ambā's ascetic practices.

Page #7

The class of ascetics called unmajjaka
223
(Mbh. 5.187.20) is almost certainly not meant to point out something which by itself forms a constituent trait of udavāsa, but is rather added to emphasize that in addition to the latter practice she observed at the same time also a quite different one, viz. complete fasting. This is confirmed by the fact that Ambā is characterized as nirāhārā, among others, already in the preceding verse (5.187.19) in which quite a different form of tapas is described. The impression one cannot fail to gather is that the Epic poet mentions udavāsa here as one of various ascetic practices not just because it is a traditional element of the topic 'asceticism', but that he really knew what he was talking about and started from the assumption that those listening to him are equally familiar with this practice, too, so that he need not say more about it.
In the narration about Mahādeva's tapas (Mbh. 10.17.8ff.) the characterizations of the god as magna ambhasi (11 and 13) and jale magna (14) are evidently not to be taken as indicating that he is completely immersed in water, they are rather to be interpreted in the light of what has been said above ($ 2) about the unmajjaka, viz. that, at least, Mahādeva's head rises above the surface of the water.
Similarly in the description of the ascetic practices of the Goddess Mrtyu it is not the term udavāsa itself that is met with; but what it denotes is expressed differently, viz. by saying that “she observed complete silence in the water (most probably of a river) for 8000 years” (12.250.20: ... maunam ātişthad uttamam / apsu varşasahasrāņi sapta caikam ca pārthiva). Jājali* is among other things said to have practised tapas as jalavāsa (12.253.4) and jalamadhye (12.253.5; cf. also 6)40, but it is not immediately clear whether the statement that “he went to the region of the sea” (sāgaroddeśam āgamya: 12.253.2) warrants drawing the conclusion that the udavāsa is performed by him in the sea, i.e. where he can still stand, - just as one wonders whether his inspecting (cf. 12.253.4) the worlds with the swiftness of thought should be regarded as a result or even as the result of his self-mortificatory practice. Yet, as it is stated in verse 9 that the demons "took him out of the sea” in order to bring him to Tulādhāra at Vārāṇasi, at
36. And does not therefore see what Brahmā is doing meanwhile? The latter element of the story is rather meant to indicate the degree of Siva's concentration (most probably on his practice of udavāsa tapas and not as part of a yogic practice).
37. The firtha Dhenuka is mentioned in verse 15, but it is not clear that the goddess has returned to it when she starts her udaväsa tapas. The parallel passage (Mbh. 7 App. I No. 8,171) makes her perform it in the river Nandā (said to be punya and fitāmalodakā). As for the relative chronology between the two versions of this story, see critical ed. Vol. 8, p. (264) (after 7.49).
38. I. Proudfoot's rendering (cf. Ahimsā and a Mahābhārata Story, Canberra 1987, p. 99) of the verse I quote in full is, to say the least, not satisfactory: who could understand "one day ... (as he) stood in the water ..." correctly as referring to a very special ascetic practice?
39. That he is characterized as malapankadharo at 12.253.3 has to be explained as either referring to a different tapas practice or as an epitheton omans (of an ascetic).
40. Viz. apsu vaihāyasarn gacched ...; note also that the parallel of 12.250.20 in the 7th parvan (cf. note 37) reads anayat instead of pārthiva at the end of the sloka.

Page #8

224
Albrecht WEZLER
least the first of these questions can be answered clearly4l. The continuation of the narrative is interesting also in that in verses 15 and 16 Jājali is characterized as varşāsv akāśaśāyr42 sa hemante jalasamśrayaḥ43 // vātātapasaho grişme ..., reminding us thus of the passage quoted from Madhavayogin's commentary on Rām. 3.5.314, i.e. ultimately of the problem whether the practice of udavāsa was confined to the cold season or whether statements like that of Mbh. 12.253.15 are not rather to be considered as aiming at a marked intensification of the self-mortificatory practice called udavāsa.
M. Shee is undoubtedly right, too, when she draws attention also to Mbh. 12.278.22:
purā so (= Siva) 'ntarjalagataḥ sthānubhūto mahāvratah varşāņām abhavad rājan prayutăny arbudāni ca //;
for the ascetic practice referred to cannot in fact be but that under discussion in the present note. The passage is instructive in so far as it confirms the assumption (suggested already by other deliberations)`S that immobility (being "stiff as a poker") forms an essential element of udavāsa46. It is mainly for this reason that one has to render the finite verb in the subsequent verse (udatisthat tapas taptvā duscaram sa mahahradāt)*7, not, to be sure, by "to stand up (from a sitting position)" but rather by "to step/come upwards out of the water (i.e. onto the land which is as a matter of course higher) 48,
The last passage drawn upon by my former student is the Cyavanopākhyāna, a story mentioned also by S. A. Srinivasan who however comes to speak of it in the
41. Cf. also 12.253.48 (cf. Shee, 0.c. (note 34), p. 250, note 43): sāgarānupam asritya tapas taptain tvayā mahat /.
42. Cf. the expression ākāśanilaya at Rām. 3.6.3 and Bhagat, o.c. (note 18], 1.c.
43. I disagree with M. Shee who voices (0.c. (note 34), p. 251 note 51) the opinion that the practice referred to in this verse is another one", i.e. to be distinguished from that of 12.253.3. The season itself, i.e. hemanta, is admittedly no clear indication (see also below $ 4), but jalasamśraya can hardly mean anything else than what is elsewhere called udavāsa.
44. See above p. 220.
45. See above $ 2.
46. Hence sthānu at Mbh. 10.17.14 referred to above p. 223 could be deliberately used not only as a name of Siva, but also in its appellative meaning.
47. Note that the udavāsa is here performed in a lake.
48. Note that the same expression (udatişthat) is used also at Mbh. 10.17.20, and that at Vin I 31.38 uttarati is used.
49. O.c. (cf. note 8], Vol. II, p. 410 note 41.

Page #9

The class of ascetics called unmajjaka
context of discussing quite a different problem, but takes also into account the testimony of Asvaghosa. This akhyāna should rather be classed with legends, mythical stories or fairy tales; in any case it goes far beyond what is possible or realistic, and is in a way distinctly different from the 'superhumanly' long periods of ascetic practice(s) which are mentioned in some of the passages examined in the foregoing, for we are told that "Cyavana for twelve years remained asleep under water1 at the point where the Ganges and the Jumna meet [and] was fished out by fishers52", his body being completely covered by aquatic plants, mussels and shells. Nevertheless this akhyana is of no little significance in that it throws considerable light on an aspect that is not mentioned elsewhere in the relevant material as far as I can see, viz. the relation between the ascetic practising udavasa and the living beings whose natural biosphere after all is water. Thus it is stated of Cyavana that jalaukasām sa sattvānām babhūva priyadarśanaḥ (13.50.10ab) and that "(the) fishes sniffed at him" rejoicingly" (upäjighranta ca tada matsyās tam hṛṣṭamānasāḥ: 10cd). This may, of course, be due, to some extent at least, to transferring the idyllic image, or topos, of the peaceful coexistence (of not only vānaprasthas and wild animals but also of animals otherwise hostile among themselves) from the aframa to this very special place of "ascetic efforts"; but nonetheless even this ākhyāna clearly testifies to the idea that the relation between this class of ascetics and the element in and with the help of which they perform their austerities has been considered to be a particularly intimate one.
54
225
4. Another relevant passage referred to already in the Larger Petrograd Dictionary s.v. udavāsa is Kumāras. 5.26. In the course of the description of Pārvati's extended and variegated tapas it is said in the first half of this verse:
.50. Viz. Buddhacarita VII.17.
51. Mbh. 13.50.9cd (tatas cordhvasthito dhiman abhavad bharatarṣabha /) could, of course, be taken to state that Cyavana afterwards emerges out of the water. But if this is true, why is he not seen by the fishers?
52. Quoted from S. A. Srinivasan, o.c. [see note 8], l.c., note 41.
53. The reading kosthaiḥ is marked in the critical edition as being not certain; this is true of its meaning,
too.
54. This gives very much the impression of being a wrong, anthropomorphic, interpretation of a basically correct observation of the behaviour of fishes.
55. Clearly this topos is the Indian expression, or one of the Indian expressions, of man's age-old dream of an entirely peaceful nature, i.e. a nature which lacks one of its most essential characteristics, viz. the 'law of the jungle'. In the European tradition this topos is, as is well-known, primarily connected with the description of paradise.
56. As well as in Apte's Dictionary [cf. notes 16 and 17] and by Oberlies, o.c., l.c. [cf. note 32].

Page #10

226
Albrecht WEZLER
ninäya sätyantahimotkiränilah sahasyarätrir udaväsatatparā /7.
viz. that “she passed the nights of the (winter month of] Pausa, during which the wind whirls up snow in large masses, fully intent on staying in water". Neither Vallabhadevas nor Mallinātha have anything to say on udavāsa not already known to uss. Unfortunately, however, they do not offer any explanation of the fact that Kalidāsa speaks of the nights only. Are we to assume that at daytime she does not practise this form of asceticism, or perhaps none at all? Or is not the verse rather to be interpreted, just as the particular season mentionedól, as meant to emphasize the uncompromisingly firm resolution of Parvati and the extraordinary harshness of her, a female's, self-mortification by stating that she does not interrupt it even at night when it gets still colder62?
At the end of his commentary on this verse of Kālidāsa's Mallinātha quotes
‘apsu väsas tu hemante /
kramaso vardhayet tapah // iti Manuḥ 11.
57. The corresponding verse in the sivapurāņa (Rudrasamhita 2.3.22.42) reads:
site jalämtare fasvat tasthau sa bhaktitatparā anahärātapat tatra
nihāresu nišāsu ca 17; see also note 24 above.
58. The first line as attested by Vallabhadeva however reads:
nināya sätyarthahimottaränilā”. Note that both the editions now available, viz. that by M.S. Narayan Murti (Vallabhadeva's Kommentar (Śäradā-Version) zum Kumārasambhava des Kālidāsa, Wiesbaden 1980 (VOHD Supplementband 20,1)) as well as that of Dr. Gautam Patel (Mahākavikālidāsaviracitam Kumārasambhavam (with the Commentary of Vallabhadeva), Ahmedabad 1986), are of such a quality that the reader is compelled to use both side by side (and sometimes to reconstruct the correct text for himself).
59. It should, however, be noted that they dissolve the compound not as is done in the Kašikā (cf. note 31), but instead as udake vāsah.
60. Both, Vallabhadeva on Kumāras. 5.26 (actually 5.25 according to his counting) as also Medhātithi, and other commentators, on Manu 6.23 (with which I deal in the next paragraph) emphasize that fisira is in fact included; Vallabhadeva also explains why other months, or seasons, are not also mentioned. Note that explanations to the same effect are found also in Nibandha works.
61. See also above note 43.
62. But the former interpretation seems to me more plausible also in view of the context which, by the way, clearly stands against taking rātrir here to mean "for 24 hours".

Page #11

The class of ascetics called unmajjaka
227
In the editions of the Kumāras, accessible to me the precise point in the Manusmặti is not indicated. By a circuitious route I was finally led to Manusmrti 6.23cd which, however, reads:
(grisme pancatapās tu syād varsāsv abhrāvakāśikah) ārdravāsās tu hemante kramaso vardhayams tapaḥ //.
And the reading ārdravāsās is attested also by those commentators who give a full explanation of the verse, viz. Medhātithi, Govindarāja and Kullūka", as also in the parallel passages of other Dharmaśāstra texts. The reading (the only one?) known to Mallinātha - in view of the predicate it has, of course, to be interpreted as apsuvāsas. 66
- is not noted by Jolly67 and G. JhaS. Now it cannot by any means be disputed that wearing wet clothes in winter - most probably because of not changing after taking a bath - has in fact been regarded in ancient India as one of the many forms of tapas, and of asceticism practised for its own sake and not only as a means of atonement". But in view of the special clothes prescribed for vānaprasthas common to which is their non-textile character, one wonders if the reading ärdraväsäs of Manu 6.23 is really the original one. In this connection it should also be noted that the number of passages
63. M.R. Käle repeating this statement of Mallinātha's in his own commentary on Bana's Kādambari [Pürvabhāga Complete) ..., repr. Delhi (etc.) 1968, p. 45 (of the text) is, of course, of no help. The same holds good of the Manusmrtipadyānukrama, attached to the NSP-ed. (with Kullüka's commentary), as it contains only the first pädas; Manasukbarāy Mor's edition (with Medhätithi's Bhäşya, in two parts, Calcutta 1967 and 1971) though includes an alphabetically arranged list of all the pädas, but apsu vāsas ... does not appear in it. Thus I finally consulted the Larger Petrograd Dictionary s.v. hemanta.
64. See Manu-Smrti with nine commentaries ..., ed. by J. H. Dave, Vol. III, Bombay 1978, p. 193f. 65. Viz. Vişnu 95.1-4, Yājā. 3.52 and Sankha (ed. P.V. Kane, ABORI VII (1926) and VIII (1926)), no. 159). Note however that the passage quoted by Apararka on Yājn. 3.52 (cf. Kane no. 369) clearly refers to udavāsa, too, for the expression jalaśayana can hardly be interpreted in any other way.
66. Cf. e.g. apsuyoni.
67. Mānava Dharma-Šāstra. The Code of Manu ..., London 1887.
68. Manu-Smriti. Notes, by G. Jha, Pt. 1: Textual, Calcutta 1924.
69. Note that ardrapafavāsas is mentioned at Rām. 3.6.5 as a separate group of ascetics (Bhagat, 0.c., I.c. [note 18]) and cf. also M. Shee, O.c. (note 34], p. 251 note 46.
70. See e.g. GautDhS 3.1.15 (ardravastratā enumerated together with other forms of tapas in a definition of the latter term as used in sūtra 11); "coils of matted hair soaked with water" are also mentioned by Ašvaghoşa, Buddhacarita VII 17. See also above n. 29.
71. Bāna, however, refers in his Kādambari (p. 49 1. 4; for the edition used see note 80) to tāpasas who wash their valkalas in lake Pampā and thereby colour its water. But there is no indication whatsoever that they are conceived of as putting on their wet clothes nor is there any connection between them and the udaväsitāpasas (on which see below p. 15).

Page #12

228
Albrecht WEZLER
in later Sanskrit texts (i.e. younger than the two Epics) in which the practice called udavāsa or the ascetics called unmajjakas are mentioned, is to all appearances conspicuously small. It is hence possible that this practice became not obsolete in the strict sense of the word, but receded into the background, so to say, and that the textcritical problem posed by the Manu verse can be solved accordingly by assuming that the reading apsuväsas was secondarily replaced by ārdravāsās. But as there is only the testimony of Mallinātha, I do not, of course, want to overstatè my point although I find it difficult to believe that he himself should have invented the reading he quotes just in order to be able to adduce an authoritative passage which supports what Kālidāsa says about Pärvati's 'winter asceticism'.
Manu 6.23 as attested in the MSS. of this work as well as by the commentators, however, gives cause for still other considerations.
1. Are we really to follow Bühler in interpreting kramaso vardhayams to mean "(thus) gradually increasing (the vigour of) his austerities”? Why is this interpretation to be preferred to the alternative one which in my view) suggests itself even more naturally, viz. "gradually/step by step making [hiş tapas substance] grow/accumulating - [ascetic power/might)”?
2. Not to be doubted at all is that this verse testifies to the idea that a vänaprastha should alter his ascetic practice in accordance with the astronomically determined - course of the year, i.e. the sequence of seasons. But how old is this idea of what could be called the 'seasonal conception of tapas in India? Or when and
72. My attention has kindly been drawn by my friend Catharina Kiehnle to Rāmdās who, at any rate according to V.P. Bokil (Rajguru Ramdas, Poona 1979, p. 58f. and 61f.), is said to have decided to complete a cycle of ... thirteen crores of Rama-nama-mantra in twelve years" and to have performed the ascetic practice of muttering the mantra about thirty thousand times every day while "he stood in waist-deep water in the confluence of the two rivers", viz. Nandini and Godavari. This may indeed be a 'survival of the practice of udaväsa. And most probably this holds equally good for the brahmin giving away the merit (punya) which he had previously gained by muttering the three names "in the water of the river Godāvari for a period of 12 years", mentioned in the Vikramacarita (see my article "On the Gaining of Merit and the Incurring of Demerit through the Agency of others: I. Deeds by Proxy" in the Felicitation Volume for Prof. Botto to be published in 1992).
Finally attention should be drawn here to J. Campell Oman, The Mystics, Ascetics and Saints of India ..., London 1905, p. 50 and to R. Schmidt, Fakire und Fakirtum im alten und modernen Indien ..., Berlin 1908, p. 191 - both of which are evidence for the survival of this particular ascetic practice even at the end of the 19th and beginning of the 20th centuries. - Prof. K.L. Sharma (Jawaharlal Nehru University) kindly drew my attention to cases of a so-called jalasamādhi about which Indian newspapers occasionally report. As far as I can see, if what is meant by this expression is deliberate suicide by drowning in water, for whatever reason, be it basically religious or not, then a connection with the tradition of the ascetic practice of the unmajjakas does not seem very likely at all. This is even more true if what is really referred to is accidental death in water (which one does not want to call an accidental death for reasons of reverence, tact, etc.).
73. O.c. (see note 14), p. 293.
74. Cf. also M. Hara's article "Tapo-Dhana" in Acta Asiatica (Tokyo) 19 (1970), pp. 58-76 as also his book Koten-indo no kugyo, Tokyo 1979.
75. See also note 57. It should be noted that in his description of Pārvati performing various kinds of tapas Kalidasa too follows this sequence at one point.

Page #13

The class of ascetics called unmajjaka
229
why was it developed? Did the udavāsa practice have this seasonal character' right from the beginning, or are not at least some of the Mbh. passages examined above (§ 3) clear evidence that udavāsa was practised by some vānaprasthas through the whole year, even continuously for a number of years?
But one should not, of course, for this reason jump to the conclusion that the 'seasonal conception' cannot but be of later origin. For it is immediately plausible that e.g. the 'five-fire-tapas' is, and hence was most probably also in fact considered to be, most effective (as regards "drying up the body", śarīraśoşana) when the sun sends out heat at its strongest, i.e. in summer; and it is equally clear that practising this kind of tapas at this time of the year was accordingly expected to give the best, i.e. quantitatively largest, result in the shape of accumulated prabhāva/tejas (etc.) substance. Hence it cannot be denied that the connection between certain forms of tapas and certain seasons is indeed a natural one, i.e. suggests itself to anyone who wishes to achieve a maximum by mortifying his flesh in the extreme. It is then not less natural for an ascetic to bring his self-mortification in line with the sequence of the seasons and to make an optimal use of what they offer him.
A further point has also be taken into consideration in connection with the 'seasonal conception of tapas. The testimony of the Epics, at least of their narrative parts, in contradistinction to that of the normative and — in the case of BaudhDhS 3.3.1ff. only too evident - systematic character of Dharmaśāstra texts", clearly points in the direction that at least certain vānaprasthas did not, as the BaudhDhs would seem to suggest, practise just one form of tapas only for whatever period they decided to mortify themselves and to accumulate prabhāva/tejas, etc., but rather followed various practices one after the other, sometimes perhaps in accordance with the rotation of seasons and/or even intensifying their tapas in one way or other.
76. S.A. Srinivasan rightly regards BaudhDhS 3.3. even "to have a certain artificiality about it" (o.c. (see note 8], p. 63).
77. That is to say, that in spite of doctrinal' agreements between the Mbh, and Dharmaśāstra texts I am less reluctant than Sprockhoff stems to be (I.c. (see note 11), p. 23ff.) to assign to the Epic, i.e., of course, its narrative parts, an independent testimonial value, in principle capable of supplementing and correcting the information of the normative and often also systematic Dharmasūtras and -smrtis. In passing only I should like to mention here that Jäjali who among other ascetic practices performs that of udavāsa, too, (sce above $ 3) is said at Mbh. 12.253.40 to have taken a bath tarpayitva hutāśanam.
78. In that it distinguishes between different types of ascetics (and not different kinds of ascetic practices).

Page #14

230
Albrecht WEZLER
As noted already by Böhtlingk", Bāṇabhatta mentions in his Kādambari, viz. in the description of lake Pampā, udavāsatāpasas - who use the flowers of the trees on its banks for paying homage to deities - but it is by no means sure that he himself still knew such vānaprasthas; this may well be just the skilful use of a knowledge acquired from 'classical texts"82.
The reference to still another relevant expression I owe to my friend Oberlies83, viz. to antarudakavāsin as it is used at Arthaśāstra 13.2.16. The whole chapter deals with "Drawing out (the Enemy) by means of Stratagems" (yogavämana), and among these are various allegedly miraculous events that are calculated to persuade a king to stay for seven nights at a particular place outside his fort and thus to offer an opportunity to assassinate him84. The sentence reads:
jațilavyanjanam antarudakaväsinam vā sarvasvetam tatasurungābhūmigrhāpasaranam varuņam nāgarājam vā sattrinah kramabhinitam rājñaḥ kathayeyuḥ //;
it has been translated by R.P. Kangle® thus: "Or, as an agent appearing as an ascetic with matted locks, all white, is staying in water, with the means of getting away to an underground tunnel or chamber under the bank, secret agents should tell the king, after gradually making him believe, that he is Varuņa or the King of Nāgas." This certainly already contains elements of interpretation, but in my view it is acceptable by and large,
79. Smaller Petrograd Dictionary (s.v. udavāsin); cf. also Oberlies, 0.c., 1.c. (see note 32); as for his interpretation of the word, I should like to just note in passing 1) that it could also be a kyt-formation (cf. Wackernagel-Debrunner, Altindische Grammatik, II,2, Göttingen 1954, p. 346 ($ 217 c)) and 2) that the taddhita suffix -in does not in each and every case have the meaning analysed and described by P. Thieme (cf. Kleine Schriften, ed. by G. Buddruss, Wiesbaden 1971, p. 670ff.).
80. The reference given in the smaller P.D. viz. 24.23, I for one could not identify even with the help of the concordances in A.A.M. Scharpe's Bāna's Kadambari, Leuven 1937 (see p. 407 and p. 484) as the NSP-ed. used by him is that published in 1932, and the edition of 1948 (which I have in my own collection) has a different pagination. The expression is found there on p. 49 1. 6 (= p. 23.1 in Peterson's ed., etc., see Scharpé, p. 484).
81. The relevant part of the long sentence reads: ...udavāsitäpasänām devatārcanopayuktakusumabhir ... van arājibhir uparuddhafiram ... pampābhidhānam padmasarah /.
82. In any case the explanation of the commentator Bhānucandra, viz. udavāsinām tatra sthitijuşām (tāpasānām) is not apt to dispel such doubts. What M.R. Kale (see note 63] says in his Sanskrit commentary (udake vastum Silam yesām te udavāsinas te tāpasāś ca, teşām / 'pesamvāsavāhanadhişu ca' (Pān. 6.3.58) ity udādeśah /atra Manuh - apsu vāsas tu hemante kramaso vardhayet tapah' iti /) and in his 'Notes' (p. [42]:"udavāsi - (practising penance by) standing in water; (this is an approved kind of penance during winter). Cf. nināya ... / Kum. V.26"), is clearly different, viz. the outcome of modern scholarship.
83. O.c., 1.c. (see note 32).
84. Cf. H. Scharfe, Untersuchungen zur Staatsrechtslehre des Kausalya, Wiesbaden 1968, p. 206. - On surungā see idem, p. 316f.
85. The Kaufiliya Arthaśāstra, Pt. I, Bombay 1960, p. 259 and Pt. II, Bombay 1963, p. 555.

Page #15

The class of ascetics called unmajjaka
231
tura Nasa le inter bly tre
however surprising the English expression may be. On sarvasvetam T. Ganapati Shastri86 aptly remarks (in his Sanskrit commentary based on the Malayalam Bhāşya) among other things: sarvasvetatvam ca devatāvarnaviśvāsārtham. It is at least probable that an unmajjaka ascetic, turned agent, is used here to create the impression that he is the Lord of the Waters" or a Nāga king by seemingly disappearing and reappearing in water. But the alternatively possible interpretation cannot also be entirely ruled out, namely that simply an agent who is especially trained in swimming and diving is selected for this job, and, to be sure, the expression jațilavyañjana speaks clearly in favour of this alternative87
Besides udavāsa the Vaidika-Padānukrama-Koşa (Vedāngas) has also an entry udakavāsa, and refers for this un-Pāņiniyan compound to VasDhS 29.6 nagadhipatir udakavāsāt, translated by Bühler "by staying (constantly) in water he becomes a lord of elephants”. Indeed, the context strongly supports the assumption that it is the particular ascetic practice under discussion in the present note which is referred to here. Bühler's interpretation of nāgādhipatir, however, is most probably, nay clearly wrong. What one expects, and by no means only because of the passage from the AŚ examined just now, is a close relation between the unmajjakas and the Nāgas inhabiting the waters under the earth, or rather representing them . The second reference is to SumantuDhs”: gobrāhmaṇasarire rudhiram utpădya udakavāsam kuryāt /, the only passage where the practice of udavāsa is prescribed as a präyaścitta”; but on the many and close links between tapas qua asceticism and qua atonement I need not dwell here.
86. I use the reprint, viz. The Arthasāstra of Kautalya with the Commentary "Śrimüla" ... With an elaborate Introduction by N.P. Unni, Delhi-Varanasi 1984, Pt. III. p. 186.
87. Cf. also Gaņapati Shastri's explanation (I.c., see note 86) antarıdakavāsinam caranavidyayodakāntarvāsasilam /.
udaka
88. O.c. (see note 14), p. 136.
89. Cf. e.g. Mbh. 1.206.13 (kingdom of Nāgas located antarjale). - In this connection it is also interesting to note that according to Mbh 10.17 (see above p. 223) Mabādeva by his udavāsa practice creates the plants and other food necessary for life.
90. Ed. by T.R. Chintamani, JOR Madras (1934), pp. 75-88.
91. A second instant I have however chanced upon in the meantime, is the verse
brahmacāri yo 'śniyān madhumāmsam katharcana/ so triny ahāny upavased
ekāham codake vaset // quoted from Manu (yatha manuh) by Vrsabhadeva in commenting on the passage yäni loke tapastvenābhikhyātäni brahmacaryadhaḥsayanodavāsacāndrāyaṇādini of the so-called Svopajñavrtti on Vākyapadiya I. 11 (ed. K.A. Subrahmania Iyer, Poona 1966, p. 40.5). Vrsabhadeva seems to have fallen prey to a failure of his memory, for the first half of the verse he quotes corresponds to Manu 11.158ab (followed there by sa ktvā prāktarn kcchram vrataśesam samāpayet / and the second half is actually Manu 11.157cd (preceded there by mäsikānnam tv yo 'śniyād asamāvartako dvijah).

Page #16

232
Albrecht WEZLER
5. The main result of the foregoing examination of — mostly - Epic text material is, no doubt, that the expression unmajjaka of the BaudhDhS (to which I should like now to return) in fact denotes that class of ascetics who are in the other sources said to practice udavāsa. Apart from this the results are rather to be styled as disappointingly meagre. that udavāsa can be performed in running and also in stagnant water, that the ascetic does not move, only his head, and throat, reaching out of the water, that the season considered to be particularly 'suitable' is winter, etc.
Many problems, however, remain unresolved. E.g. what do the unmajjakas live on? Do they leave their place in the water temporarily, e.g. in order to collect food? Indeed, just like the root vas itself and its derivatives so does ovāsa as secondary member of the compound udavāsa not necessarily imply semantically that any change of place whatsoever is entirely excluded”, nor does it only intend "staying overnight”24. That is to say, this compound may refer, and most probably does in fact refer, to nothing but a person's staying for most part of time in water; that he stands upright etc., is not directly expressed, but clearly associated with this term.
This assumption is supported by the fact that at BaudhDhS 3.3.10 "(the use of iron and stone implements” is prohibited for the unmajjakas (lohāśmakaranavarjam). For this statement makes sense (practically) only when it is - in accordance with Govindasvāmin - taken to mean that only wooden implements are allowed them. And the fact that nothing else is said in this sūtra or elsewhere in the BaudhDhs about the unmajjakas is indeed strong evidence that Bühler is right when he takes this prohibition to refer to the collection and preparation (of course not by cooking) of food by these vänaprasthas. Yet he does not give any explanation for this peculiarity
Now, I do not think that the prohibition of the use of implements made out of these two kinds of material is due to the idea that an implement meant for people who stand more or less all the time up to their neck in water should have one very important trait, viz. that of floating, i.e. not instantly sinking down to the ground. That is to say, I do not think that the wooden implement, ladle or whatever it may be, is used by the unmajjakas in order to collect and eat food while standing in water. For, if these ascetics live on certain aquatic plants, one expects that this kind of foodstuff is also made the essential criterion of classification. Besides Buddhacarita VII 14 shows that annam salile
92. Because - as already noted above the author(s) do not give a detailed description, but rather seem to start from the assumption that their listeners are anyway familiar with this and the other forms of tapas.
93. As can be observed also in the case of similar compounds such as vanavāsa, antevāsa/-väsin, parkavāsa, "crab" (according to lexicographers).
94. As it is given for durgaväsa with reference to Mbh. III 12344 (= 3.174.6 a) in Monier-Williams's Sanskrit-English Dictionary (following the Smaller Petrograd D.) in that this expression is rendered by "staying overnight in an unhospitable place".
95. For his commentary on 3.3.10 reads thus: lohakaranam darvyādi / aśmakaranam apy evamākịtikam eva kincit/kāsthāny eva karanam ādāna ity arthah //.
96. And this holds good also for what is said by Sprockhoff, 1.c. (see note 11), p. 27.

Page #17

The class of ascetics called unmajjaka
233
prarūḍham forms part of ascetic diet in general. Therefore I deem it a much more plausible solution that the unmajjakas are allowed to use only implements that are made of a material which is directly and intimately connected with, and dependent on water, viz. material obtained from vegetal life, plants in general and in particular the largest and most important among them, namely trees. It is in this connection that I regard the 'myth' about Cyavana as particularly instructive: The relation between the vanaprasthas practising udavasa and 'their' chosen element", water, is so intimate, their becoming a part of acquatic life so intensive, that any material alien or even opposed to water is ruled out as a matter of course, so to say.
In concluding the present note let me draw attention to yet another problem connected with this particular form of self-mortification which too seems to have been overlooked in previous research, if I am not mistaken. It is true that in India since of old fire and water have not been viewed exclusively, perhaps not even mainly, as elements. opposed to each other, but fire even as originating from water. Nevertheless, even if Vedic mythical ideas should still have been a living tradition at the time when the practice of the unmajjakas came into existence, the very idea of tapas as accumulation of heat, fiery energy, within oneself is not really, at least not simply, compatible with a practice of self-mortification that consists in udaväsa. But that this latter practice formed part of what was called tapas stands beyond doubt. One cannot hence help drawing the final conclusion that is suggested by other forms of asceticism, too, like e.g. living on water or air only, etc. that the conception of tapas qua asceticism did not, or perhaps did not any longer at a certain period of time, by itself exclusively refer to the practice of heating oneself up in the literal sense of the word, and this is what is also actually attested in the texts at our disposal. But it is quite possible that one has to go even a step further and to draw the more far reaching - conclusion that the subtle substance accumulated by practising tapas was not only, or always, considered as having a fiery nature, and that hence text passages which speak of tapasvins sending forth flames of fire, etc., are to be relativized as regards their testimonial value: they may simply be due to the fact that the old, and original, idea of asceticism as tapas, i.e. "heating oneself up", was kept alive as just one of the various ideas, or strands, which taken together have ultimately formed the highly complex phenomenon which continued to be called tapas in spite of the fact that it had assumed a markedly different character in the course of its historical development.
97. On the elements water, fire and wind as welcome means for ascetics to end their life by see e.g. Mbh. 15.45.27.