The Dakinikalpa: Religious and Astrological Medicine

[Full title: The So-Called Dakinikalpa: Religious and Astrological Medicine According to a North-West Indian Collective Manuscript (I) / By Adelheid Herrmann-Pfandt]

Frau Gertraud Eimer zum 21.9.1997 in Dankbarkeit zugeeignet Introduction. In his work on the magico-religious aspects of Indian medicine, G.U. THITE has - in the words of one of his reviewers² - stated that: 'notwithstanding the scientific elements in ancient Indian medicine, its basic assumption[s] are that the diseases are caused not so much by physical derangements or deficiencies in man as by the influence of evil spirits, irreligious sinful conduct, unfavourable stars and cope black magic and that the cures too are to be effected not so much by means of any physiotherapeutic or pharmaceutical remedies as by means of pacificatory and expiatory rituals and religious observances.' Even if this description on the whole perhaps overstates the role of religion and magic in Indian medicine, there are without any doubt many sources which perfectly fit in with the picture drawn here. Some short sources of this kind are gathered in a Sanskrit codex known by the name of Dakinikalpa,³ the only known manuscript of which is kept by the former Raghunath Temple Library, now called Shri Ranbir Sanskrit Research Institute, in Jammu.4 Among the many manuscripts with a title beginning with the term dakini ¹ G. U. THITE, Medicine. Its Magico-Religious Aspects according to the Vedic and Later Literature, Poona 1982. Cf. the review articles by Rahul Peter DAS (in Indo-Iranian Journal 27, 1984, 232-244) and Arion ROSU ('Pratiques magico-religieuses en medecine indienne', Wiener Zeitschrift fur die Kunde Sudasiens 30, 1986, 83-89). 2 A.D.W. in Praci-Jyoti. Digest of Indological Studies 18-19, 1982-83, 207. 4 Concerning this title, which seems not to be the original one, see below. See M[arc] A[urel] STEIN, Catalogue of the Sanskrit Manuscripts in the Raghunatha Temple Library of His Highness The Maharaja of Jammu and Kashmir, Prepared for the Kashmir State Council, Bombay/London/Leipzig 1894, 229 (Section Tantra); M. M. PATKAR, Descriptive Catalogue of Manuscripts in the Shri Ranbir Sanskrit Research Institute, Jammu (Kashmir), Jammu 1970; 1973; 1984 (3 vols.), vol. 3, 1060 (Section Tantra, no. 201).

mentioned in the New Catalogus Catalogorum³ the so-called Dakinikalpa is one of only two non-Buddhist texts,' i.e., one of the very few Hindu sources which might be expected to deal mainly with these minor divinities. On reading the manuscript, however, this expectation is only partly fulfilled, because the title is misleading: of five texts about medicine contained in the manuscript, four complete and one incomplete, only one, the second, deals with dakinis. Nevertheless, the four pages of this second text consist as a whole of material hitherto unknown. In Hinduism as well as in Buddhism, dakinis are known as very malevolent and dangerous demonesses causing damage of all sorts, including disease. However, in both religions sources about dakinis in the role of demonesses connected with disease or fever are very seldom found: neither in Buddhism,' where the mythical "conversion" of the dakinis into Buddhist initiation goddesses and protectors soon raised their status within the religious cosmos, nor in Hinduism, where they remained small godlings. Thus, in the context of our 5 Cf. K. Kunjunni RAJA (ed.), New Catalogus Catalogorum, Vol. 8: Ta - Da. Madras 1974, 6. ❝I am indebted to Dr. Gustav Roth (Universitat Gottingen) for pointing out the Dakinikalpa to me in 1982 when I was doing research on dakinis in Hinduism and Buddhism; cf. Adelheid HERRMANN-PFANDT, Untersuchungen zur Religionsgeschichte und Mythologie der Dakinis im indotibetischen Raum (unpublished M. A. thesis), Bonn 1983. A part of the conclusions concerning the Hindu dakinis has been published as 'The Good Woman's Shadow. Some Aspects of the Dark Nature of Dakinis and Sakinis in Hinduism', in: Axel MICHAELS, Cornelia VOGELSANGER, Annette WILKE (eds.), Wild Goddesses in India and Nepal. Proceedings of an International Symposium in Berne and Zurich, November 1994 (Studia Religiosa Helvetica Jahrbuch 2), Bern 1996, 39-70. Concerning dakinis in Buddhism, see my book Dakinis. Zur Stellung und Symbolik des Weiblichen im tantrischen Buddhismus (Indica et Tibetica 20), Bonn 1992. A short resume of some of the results of this study in English is included in my article 'Dakinis in Indo-Tibetan Tantric Buddhism. Some Results of Recent Research', Studies in Central & East Asian Religions (Journal of The Seminar for Buddhist Studies, Copenhagen) 5/6, 1992-3, 45-63. 7 The other work is the Dakinitantra, an incomplete text in five chapters, the only known manuscript of which is extant in the library of the Bangiya Sahitya Parisad in Calcutta. 8 My warm thanks are due to Dr. Sharma, Director of the Raghunath Temple Library, who during my very short visit to the library in 1983 at once provided me with a xerox copy of the manuscript, and then went with me through the text to make sure that all parts of the copy were readable. 9 In Tantric Buddhism dakinis sometimes help in healing diseases, cf. HERRMANNPFANDT, Dakinis. Zur Stellung..., 431 f.

A. Herrmann-Pfandt, The So-Called Dakinikalpa 55 present knowledge about dakinis in Hinduism and especially in Hindu medicine, our manuscript is quite a rare source. Since all five texts included in the manuscript not only deal with medicine as such, but especially with medicine supported by astrological knowledge, and since the manuscript as a whole is not very long, it does not seem to make sense to confine a publication to the text on Dakinis. Therefore, the whole manuscript will appear in annotated editions and translations¹º in this Journal, beginning in the present issue with an introduction and the first text. The manuscript. The Jammu manuscript called Dakinikalpa consists of fourteen folios in the size 24.4 x 12.2 cm with two pages each. Each page has eleven lines. There is no title for the whole manuscript. The text begins on fol. la with an invocation of Ganesa and ends abruptly on fol. 14 a 11. It is written in Devanagari with ink on paper, the sentences being separated from each other by small gaps. The manuscript does not seem to be very old." It is not complete because it ends with a unexplained chapter or paragraph number ||1|| in the middle of the last line of fol. 14 a.12 As this is not the end of that folio, 10 During the summer term of 1993 Prof. Dr. Kameshwar Nath Mishra (Institute for Higher Tibetan Studies, Sarnath) went through the whole of the text with me, commented on my transliteration and translation, and discussed with me at length the many corrupt readings as well as problems of the practical realisation of the cult instructions contained in the manuscript. I am very grateful to Prof. Mishra for his help and for the inspiring hours of working together. My thanks are also due to Dr. Johannes Schneider (Institut fur indische Philologie und Kunstgeschichte, Freie Universitat Berlin) who read very carefully an earlier draft of all parts of this article and made several useful remarks to improve it. However, the publication of this article would not have been possible without the help of Prof. Dr. Rahul Peter Das (Institut fur Indologie, Martin-Luther-Universitat Halle-Wittenberg), which was indispensable especially regarding the identification and description of the New Indo-Aryan elements in Text I; indeed, the majority of the remarks on New Indo-Aryan in this article are based on his observations, though I have, in accordance with his wish, refrained from marking them as such in each individual case. Prof. Das also proposed some additional emendations of the text and improved my English. For additional information I am indebted to Prof. Dr. Srinivasa Ayya Srinivasan (Institut fur Kultur und Geschichte Indiens und Tibets, Universitat Hamburg), to Dr. Dr. Jayandra Soni and Dr. Roland Steiner (both Fachgebiet Indologie, Philipps-Universitat Marburg), to Dr. Chlodwig H. Werba (Institut fur Indologie, Universitat Wien), and to Peter Wyzlic, M.A. (Indologisches Seminar, Rheinische Friedrich-WilhelmsUniversitat Bonn). For all mistakes the responsibility is mine. 11 This is also noticed in the Descriptive Catalogue by PATKAR. STEIN, however, has the remark 'pracina'. 12 This number being visible at once when one looks at the last page, it is remarkable that both catalogues qualify the text as 'complete'.

our manuscript cannot be the original complete compilation, but must have been copied from an already incomplete exemplar of the compilation, so that the copyist had to stop on the middle of the folio at the point where the extant part of his original ended. Contents. As already remarked, the manuscript contains four complete texts and a fragment which is the first paragraph of a fifth text. The first three texts deal with different sicknesses, mostly fever, their diagnosis and therapy. Each kind of fever is to be recognised by the astrological date on which its outbreak is thought to take place, and therefore they are arranged according to their dates of commencement. In Text I, seven sicknesses related to the seven weekdays are described, Text II deals with fifteen diseases in relation to the fifteen tithis or days of each half of the moon month (each tithi being dominated by one dakini), and Text III with twentyy-seven sicknesses related to the twenty-seven naksatras or moon mansions. Text IV contains some very brief information, covering a few lines only, on a method of calculating the lifetime. The extant first paragraph of Text V deals with the effects of planetary constellations on human health, especially concerning, again, the origin of fever. We do not know whether the following lost part contained instructions about therapy. In Texts I-III, which form the major part of our manuscript, one paragraph of about one third of a folio on average is given to each date or sickness respectively, and each such paragraph contains as a minimum a description of the symptoms (sometimes including the cause and mostly the duration of the sickness) and of the therapy, which is without exception magico-religious: blood sacrifices, food oblations at special places, mantras to be recited to the demons or deities which are imagined as being the source of the sickness, purificatory baths and so on. Although Texts I-III and perhaps also Text V display approximately the same principle of classification and presentation of the material, the formal and substantial differences between them are remarkable and make it certain that all of them were separate texts which were arranged together only recently by someone who was collecting information about sicknesses in relation to different sorts of astronomical dates. Most striking are the linguistic differences: Text I is written in a kind of hybrid mixture of faulty Sanskrit prose and a New Indo-Aryan language which most probably comes from North-west India. Text II shows nearly always correct slokas of a once perhaps relatively flawless Sanskrit, with a mantra concluding each paragraph of 2-4 slokas. Text III is composed in an abbreviated, notebook-style Sanskrit prose with several corrupt passages. Text IV contains

A. Herrmann-Pfandt, The So-Called Dakinikalpa 57 a mixture of prose and verses in Sanskrit. The fragment of Text V is confined to three more or less correct slokas. Scribal errors and other mistakes are many in all parts of our codex. The titles of the codex and of its different texts. The titles of the separate texts put together in the codex now called Dakinikalpa are the following: I title in heading: Varasaptadosa title in colophon: Jvaropacara (ms.: Jvaro upacara) II title in heading: Tithijvaropacara (ms.: Tithijvaro upacara) title in colophon: Tithidakinikalpa III title in heading:. title in colophon: IV title in heading: V title in colophon: title in heading: Kalacakra title in colophon: (text incomplete) - The codex is included in the catalogues of the Jammu Temple Library manuscript collection under the title Dakinikalpa. The origin of this title is not clear. Since it is incomplete, the manuscript does not include a colophon, and there is also no heading or title page. 13 Moreover, except for Text II, the contents of the codex do not suit this title, which lets us expect instructions about worshipping dakinis (i.e. minor, malevolent goddesses/demonesses known as belonging to the retinue of Siva-Bhairava) or Dakini (one of the names of the Great Goddess). But dakinis are not mentioned at all in the first, fourth and fifth texts and only three times (III, 10;11;14) in the third text. The second text, on the contrary, mentions a dakini in nearly every paragraph and is called Tithidakinikalpa in its colophon. So it seems probable that the title Dakinikalpa has been taken from the second text and given wrongly to the whole codex either by its compiler or while preparing an inventory of the library. Another problem lies in the non-uniform titles of the different texts. Texts I and II have two different titles each, Texts III and IV have no titles at all. The reason for the double titles of the first two texts may lie in the process of compilation of the codex. The compiler seems to have put the different texts together in one codex because of their common feature of establishing a 13 At least, a title page was not shown to me during my visit to the library and is also not included in the copy of the manuscript given to me.

connection between a special astronomical date and a sickness. The difference between the texts, on the other hand, lies in the kind of astronomical dates chosen: the days of the week (vara) in Text I, the tithis in Text II, the naksatras in Text III, and in Text V, as far as we can see, different astronomical constellations. So for the compiler it must have been reasonable to choose headings related to this difference: Varasaptadosa for Text I, Tithijvaropacara for Text II. But at the same time he did not delete the original titles Jvaropacara and Tithidakinikalpa in the colophons. As regards the heading Tithijvaropacara of Text II, it is especially probable that this title was invented in the process of compiling, for when the compiler had written the title Jvaropacara in the colophon at the end of the first text, he possibly realised that this title did not establish any difference between his first and second texts, and so he himself established this difference by inserting the heading Tithijvaropacara at the beginning of the following text. If these suppositions are correct, then the reason for the fact that the third and fourth texts have no title at all could be the carelessness of the scribe, which is a common feature also of the manuscript as a whole. It can also be that the third and fourth texts were added to the compilation at a later stage of its development by a compiler, who, unlike the compiler of Texts I, II and V, did not care about text titles. Age and geographical region of origin of the texts. As the first text is greatly influenced by and interspersed with a New Indo-Aryan language seemingly of Northwest Indian origin (even containing in I,1 a word ultimately of Arabic origin, namely tabakhi, and maybe also Persian ta), at least this part of our compilation cannot be older than the historical state of the modern language it represents. On the other hand, a possible inflected passive form in 1,2 could, but unfortunately need not, speak for a relatively early date. Some traces of Hindi or another New Indo-Aryan language of North India are also found in the other texts, but maybe only because of the linguistic affiliation of the scribe of our codex. Specialists on the languages of North India may be able to find out more about that. It is very difficult to say anything about the age of Texts II-IV. Text V arouses our special interest because of the similarity of its title with the Kalacakratantra. As one of the best-known Tantric systems of late Buddhism, the Kalacakratantra cycle was paid attention to also by Hindu authors, especially in Northwest India. The great Kashmirian philosopher Abhinavagupta (c. 1000 A.D.) made use of it in his Tantraloka. 14 Moreover, 14 Abhinavagupta, Tantraloka 4 and 16. Cf. Jean NAUDOU, Les bouddhistes kasmiriens au moyen age, Paris 1968, 125; Navjivan RASTOGI, Introduction to the Tantraloka. A Study

A. Herrmann-Pfandt, The So-Called Dakinikalpa 59 the Kalacakra cycle is especially famous for its astronomical teachings, 15 which, of course, were also used astrologically. From this point of view, it is not so improbable that our Text V, which is definitely astrological, has some roots in the Buddhist Kalacakra cycle. This would mean that its roots go back to the time between the tenth and the twelfth century A.D. when the Kalacakra system flourished in India, before Indian Buddhism was nearly extinguished by the Muslim invaders at the beginning of the thirteenth century A.D. The geographical region of origin of our texts is most probably Northwest India, not only because of the language of Text I and the possible connections of Text V with the Buddhist Kalacakra system, which was well-known to Hindu circles in Kashmir, but also because of the fact that the only known manuscript of the codex is located in Jammu. The texts may originate directly from (a non-Dardic speaking area of?) Kashmir or from another region in the north-western part of India. Edition and Translation of Text I Contents. Text I of the so-called Dakinikalpa deals with seven kinds of fever, each of which has its beginning on one of the seven weekdays. For each weekday/sickness the following information is given: (1) The symptoms and the duration of the fever. (2) The source of the infection. This is in all seven cases described as an involuntary meeting with other persons or ghosts causing damage. (3) The place of the infection: nearly always a sacred or numinous place. (4) The therapy in the form of offerings (mostly animal sacrifices) to different deities. We can see from the table given on p. 61 that most information about the sicknesses and their therapies are magico-religious and correspond to THITE's results referred to above. The only truly medical information included in our text is that about the symptoms and the duration of the fever named in five of the seven cases. Language. As already indicated, Text I is written in a mixture of faulty Sanskrit and a New Indo-Aryan language of North-west Indian origin. The author seems to have been someone knowing some Sanskrit but "thinking", so to in Structure, Delhi etc. 1987, 249 f. 15 See e. g. Wilfried PETRI, 'Die Astronomie im Kalacakralaghutantra', pp. 381-385 in: Helga Uebach and Jampa L. PANGLUNG (ed.), Tibetan Studies. Proceedings of the 4 th Seminar of the International Association for Tibetan Studies, Schloss Hohenkammer - Munich 1985, Munchen 1988.

say, in his own language, which has not only been the source for several words but also the basis for the on the whole quite "un-Sanskritic" syntax. Apart from many spelling mistakes (e.g. I,2 dukham for duhkham; 1,5 utva for hutva; 1,6 athavi for atavi), there are many words without case endings (e.g. I,1 drsta for drsta; 1,2 madhya for madhye) or with wrong case endings (e.g. I,1 pujyah for pujya; 1,3 velayai for velayam). Sandhi rules are often not observed (e.g. I,1 kastah damtasula adhasira for kasto damtasulo 'dhahsirah°; 1,2 samcara abhut for samcaro 'bhut, udarasulah jvara° for udarasulo jvara°; 1,3 °puruse ucchista° for purusa ucchista°; 1,4 ekah mleccha anyah for eko mleccho 'nyo; etc.). Other traces of New Indo-Aryan influence are to be found in omissions of the virama (e.g. 1,2 pasubalira for pasubalir; 1,7 tatrabhuta for tatrabhut) or the separation of conjunct consonants (e.g. I,4 karapatam and karapatham for karpatam; 1,5 udapadyate for utpadyate), both presumably because of the silent a in the pronunciation of various New Indo-Aryan languages. Moreover, there are several words stemming from a language which shows points of contact with the Punjabi group of languages on the one hand, but also with the (Western) Hindi group on the other (e.g. I,1 ave ta 'if it comes, then', janani 'to be known', i 'this', isaka 'of this', bakkara 'he-goat'; 1,2 hoya to 'if it arises, then'), as well as other language groups of North or West India (e.g. the imperative ending u in I,3). One word (I,1 tabakhi) might have a meaning relating it to Kashmiri. However, I shall not investigate the question of language further, since this is not my main field of research. I have tried to identify and translate all forms and words in a manner giving sense to the text, and for the rest, I hope that this publication might inspire linguists to add what is necessary. Verbal constructions. Especially noteworthy are the predicates in the first paragraphs of each section. As these paragraphs deal with the reasons for the respective fevers, they are antecedent in relation to the rest of the text. If we leave out of consideration the introductory part of each paragraph which tells us, as far as true Sanskrit forms are concerned, 16 in the simple present which weekday's fever is going to be described (i.e. I,3 utpadyate, ms.: udarapadyate; 1,4 grhyate; 1,5 utpadyate, ms.: udapadyate; 1,6 utpadyate; 1,7 grhyate), then we can see that in the first paragraphs the following three kinds of predicates are used, all of which designate the antecedence of the incidents and situations described: 16 In I,1 and 1,2 we have the New Indo-Aryan subjunctive forms ave and hoya respectively.

Day of outbreak, symptoms, duration Place of infection Source of Infection Remedy Sunday water in northern direction yogini killing of a he-goat, Kutoothache, affliction of head, intestines, mari worship 57 times skin, hands, feet etc. Monday garden, forest (twilight) bhutas, pisacas etc. stomach-ache, fever, pain in hands, feet etc. 10 days Tuesday pain in bones, belly great danger for 7 days-one month Wednesday Synoptical table of contents of Text I pain in feet, toothache 10 days Thursday animal sacrifice to the vetalas amongst 11 men meal in an assembly impurity after meal food sacrifice in temple (rice, oily beans, oil) cremation ground two bhutas: a mleccha and a Brahmin preparation of a white or black garment under high tree (twilight) or near water (after fire sacrifice) Brahmin woman or seven Brahmins (male or female) sacrifice of a black animal by nine people lasting 20 days Friday Saturday stomach-ache 7 days-4 (or 2) months way to temple in forest somewhere near a village on path within forest, in hole near cremation ground near water at the root of a (Bhav- ani?) tree bhutas bhuta A. Herrmann-Pfandt, The So-Called Dakinikalpa domestic animal sacrifice (at place of transmission of disease?) sacrifice of a black animal and offering of a bean dish 61

1) predicated past participles: I,1 drsta ... janani 'is (to be) known as [having been] seen'; 17 1,2 gatah 'went'; 1,3 bhuktavat 'ate', praptah 'was incurred', prapannah 'got into'; etc., 2) past tense forms (mostly aorist third person) of bhu and vas: 1,2 samcara abhut 'there was a concourse'; 1,6 devageham asit (ms.: aset) 'a temple was (there)' tatra bhuta abhavat 'bhutas were there'; 1,7 bhutah tatrabhut (ms.: tatrabhuta) 'there was a bhuta in that place', 3) both together, i. e. periphastic constructions with abhut as copula: 1,4 dvau bhutau sthitau abhut for 'two ghosts were present', samagatah abhut 'arrived'; 1,5 gatah abhut 'went'. Concerning participles with or without copula in the function of the predicate, J.S. SPEIJER 18 writes: 'Sometimes participles are expressive of the chief predicate. In this case, auxiliaries are often wanted to denote the person or the tense or the nature of the action. The combination of participle and auxiliary effects a kind of periphrastic conjugation, which sometimes has an emphatic character, and sometimes serves to express special shades of tenses or moods, not to be pointed out by mere flexion.' This use of the periphrastic passive, mostly with forms of vas and vbhu, is known from Vedic times onwards. 19 In an article on the predicated past participle especially in Vedic language, 20 Stephanie W. JAMISON reminds us of 'the grammatical orthodoxy concerning the past participle plus copula in Classical Sanskrit: that the copula does not ordinarily appear with the 3 rd person, but that there must be a surface expression of person for non-3 rd persons, either a pronoun... or a present tense copula.... Past tense copulas do not ordinarily appear, since present asmi, asi, etc. plus past participle have the force of a past. ,21 17 In this case the past participle is combined with a New Indo-Aryan verbal noun-cumparticiple in the function of a Sanskrit participium necessitatis. 18 Sanskrit Syntax, Leiden 1886, §376. 19 Cf. e.g. Berthold DELBRUCK, Altindische Syntax, Halle (Saale) 1888, 392 f. 20 'The Tense of the Predicated Past Participle in Vedic and Beyond', Indo-Iranian Journal 33, 1990, 1-19. For this reference I am indebted to Dr. Werba. 21 Op. cit., p. 2. As JAMISON herself concedes (note 4 ad loc.), there are 'a few examples of aorist and perfect copulas with the past participle in later Sanskrit' quoted by SPEIJER, §376 II; one of the examples, however, is from the Mahabharata. Some more examples have been

A. Herrmann-Pfandt, The So-Called Dakinikalpa 63 Regarding both these points, our examples of periphrastic constructions, e.g. 1,5 gatah abhut, deviate from ordinary Classical Sanskrit usage. 22 Firstly, they are in the third person but nevertheless have the copula, and secondly, the copula is in the past tense. This usage reminds one of New Indo-Aryan languages; 23 it could, but need not be, a sign of New Indo-Aryan influence on our text. In any case, the contexts in which these combinations of past tense and past participle occur in our text show that what we have here are clearly no pluperfect tenses, in the same manner as what are morphologically pluperfect tenses in New Indo-Aryan languages such as Hindi or Bengali are often rather the equivalents of English or German imperfects than pluperfects.24 Considerations on Structure. Looking at the distribution of the different kinds of predicates used in the first paragraphs of each weekday section, we can observe that each paragraph (except I,1 and I,3) has both: firstly one or more predicated past passive participles and secondly at least one abhut (or asit, abhavat respectively). The abhut (etc.) nearly always forms the last predicate (except 1,5); the sentences without any finite verb have predicated participles instead which in three of four cases (I,1 and twice in I,5) are formed from or at least with participation of (I,1) Vdrs 'to see'. These observations encourage us to investigate further the specific use of the different kinds of predicates used in our text, especially in the first paragraphs of each section. Regarding the information the text gives us about the origins of the diseases, this is given by the respective first paragraphs of the weekday sections. The structures used are of three sorts, with the second division in its turn being subdivided into two aspects: collected by Dr. Werba, who was so kind as to share them with me: Buddhacarita 4,16 c: tadito 'bhut (pada) 'was kicked (by her foot)'; Dasakumaracarita 7,6: vijnapito 'bhut 'was informed'; ibid. 18,9: niksipto 'bhut 'was thrown down'. 22 These rules are especially true concerning gam in Classical Sanskrit. Dr. Steiner kindly informs me that such periphrastic constructions are quite common in Classical Sanskrit drama, see e.g. Harsadeva's Nagananda (Trivandrum Sanskrit Series 59), ed. T[aruvai] GANAPATI SASTRI, Trivandrum 1917: agato 'si (Nagananda 3,8+ {i.e. prose part after 3,8}: p. 166, line 4); gato 'si (5,7+: p. 244, line 6; 5,22+: p. 275, lines 12-13); gato 'smi (5,32+: p. 292, line 1), gado si (Prakrit, 5,31+: p. 289, line 6), etc. There are no periphrastic constructions in the third person which always shows only the simple participle. Cf. e. g. gatah in the sense of 'he is gone' (Nagananda 1,17+: p. 76, line 5). 23 Cf., for instance, Hindi gaya tha. 24 Cf. Colin P. MASICA, The Indo-Aryan Languages, Cambridge 1991, 293: 'The Past Perfective is often equivalent to an English Simple Past in Hindi and other NIA languages (as well as to the Past Perfect as shown).'

1) Place: First of all, the sick person is said to have gone to a certain place or to have done something there. 25 The locative (1,2 sayampratah veladvayamadhya vatika (gh)atavimadhye; 1,3 kutrapi sa (m)bhayam; etc.) or accusative of direction (1,6 at (h)avimadhyam) is usually combined with a past passive participle of an intransitive or transitive verb (I,2, 1,5 and 1,6 gatah; 1,4 samagatah; 1,5 hutva krtam, ms.: utva ...), or, excepotionally, with a past active participle of a transitive verb (1,3 bhuktavat). In two cases the copula abhut is added to the participle (1,4 samagatah abhut; 1,5 gatah abhut). This category of information is included in the first paragraph of each weekday description; however, in I,1 and 1,7 the participle is missing (not, however, the locative). 2 a) Presence: The presence of someone or something, mostly one or more bhutas, but in 1,6 additionally a temple, at the place mentioned in 1) is noted. This information is always expressed with abhut or abhavat or asit (ms.: aset) respectively; in one case, the abhut is construed with a participle to form a periphrastic construction (1,4 sthitau abhut). 2 b) Incident: Alternatively to 2 a), an incident is described in which the sick person has been directly involved: he sees someone (I,1 a yogini, 1,5 one Brahmin woman, seven Brahmin women) or is touched (1,3 by the disease or by another man²). This information again is expressed with past passive participles; in I,1 a New Indo-Aryan verbal noun-cum-participle janani is added. 3) Consequence: Only once, in I,3, is a consequence of these events described (1,3 bhayam prapanna[h]), - again with a past passive participle. All other weekday descriptions do not have this category, at least not at the end of the first paragraph. But one could say that the disease breaking out on the very weekday mentioned is the consequence. This disease is in each weekday section mentioned in the first part of the first paragraph and described in the second paragraph. The table on p. 65 gives the four categories of information with their respective predicates. It shows the relatively regular structure of our text with only some minor irregularities. One can be seen in the missing participle in the 25 The difference between the action in this case and what is described in 2 b) is that what is done there is something that does not originate from the person concerned, but is, so to say, forced upon him by external circumstances. 26 The fact that in all other weekday sections ghosts or persons are the source of the disease could be taken to speak in favour of the interpretation of puruse in 1,3 as denoting a person different from the sick man.

1) Place 1. Sun uttaradisa jalanikata 2 a) Presence 2 b) Incident yogini drsta... janani (no participle) 2. Mon madhye gatah samcara abhut 3. Tue atura ... sambhayam bhuktavat Categories of information in Text I 4. Wed smasanamadhye samagatah abhut 5. Thu samdhyasamaye... gatah abhut paniyasamipam... utva ... krtam 3) Consequence sa puruse ucchistavelayam bhayam prapanna praptah dvau bhutau sthitau abhut, ekah ... anyah .. ... ... 6. Fri athavimadhyam gatah atura 7. Sat atavimadhye (no participle) tatra tena drstih vipra etasya drstih 7 vipra tatrekam ... aset, tatra abhavat bhutah tatrabhuta A. Herrmann-Pfandt, The So-Called Dakinikalpa 65

first category of 1,7. Has it vanished in the course of textual transmission?27 Another irregularity can be seen in the periphrastic predicates in 1,4 sam- agatah abhut and 1,5 gatah abhut as against the bare participle used as predicate in the beginnings of the other sentences (1,2 gatah, 1,3 bhuktavat, etc.), especially if we take into consideration the fact that all other abhuts occur in the category 2 a). Are therefore the first abhut in 1,4 and the abhut in 1,5 to be shifted to category 2 a) and to be seen not as the last word of the first but as the first word of the second sentence of their respective paragraphs?28 This interpretation of the text indeed removes the above-mentioned structural irregularities of the text, collecting, so to say, all bare participles in category 1) and all abhut, abhavat and asit forms and constructions in category 2 a). However, this change is not possible without creating new irregularities at the same time, 29 and therefore I have decided against this proposition here. Our table shows two alternative sorts of situations as producing the disease: on the one hand the mere presence of malevolent ghosts (bhutas) at a place where the sick person has been staying, and on the other the more or less conscious contact with another person or persons by seeing him, her or them 27 Dr. Werba sees in the seemingly superfluous locative atarde (here interpreted as atarde 'in a hole') the possible result of a spoilt participle in accordance with gata aturah (ms.: gatah atura) in 1,6. The regular structure of nearly all the other sentences could speak in favour of this. For actually carrying out a conjecture, however, I do not see a possibility in this sentence which by itself gives sense without any major change. 28 This alternative was pointed out to me by Dr. Werba. For the differences in translation it requires, see the respective notes on 1,4 and 1,5. 29 Three arguments against this are the following: 1. Of all predicates then included in category 2 a) and even of all predicates of the whole Text I, exclusively these three abhuts of 1,4 and 1,5 would be in the beginning and not, as normal in our text, in the middle or end of the sentences. Would it not look curious if all the three predicates with irregular initial position were preceded by a past pasive participle? - 2. Another irregularity as a consequence of interpreting our three abhuts as the first words of their respective sentences would be that in 1,5, the resulting sentence would be: abhut tatra tena drstih (read drsta) vipra, literally: 'A Brahmin woman seen by him was there' (unless we maintain that what we have here is a relative construction with the correlative missing: 'It was there [that/where] a Brahmin woman was seen by him'). This attributive use of the participle would be singular in the text, and would, moreover, not tally with etasya drstih (read drsta) 7 vipra[h] (literally: 'Seven Brahmin women were seen by this one'), where the participle is predicative. - 3. Another irregularity would be that the sentence 'A Brahmin woman seen by him was there' would be the only one in the text belonging to both the categories 2 a) and 2 b) (cf. pp. 64 f.) at the same time.

A. Herrmann-Pfandt, The So-Called Dakinikalpa 67 by physical contact. 30 These two possibilities as well as the initial event of going to the place are expressed in our text with different forms of predicates. Punctual events (categories 1, 2 b, 3) are expressed by predicative participles, durative events by finite forms of √bhu or vas, while periphrastic forms (category 2 a) have been used in either case. The edition of the text. The edition and translation of Text I of the Dakinikalpa has been made according to the following rules: Since the text does not conform to the grammatical rules of classical Sanskrit, the edited text is given unchanged according to the manuscript. All major emendations necessary to render those portions of the text which seem to be in corrupted Sanskrit into correct Sanskrit are listed in the critical apparatus in smaller letters below the main text; the readings of the manuscript are in italics, and suggested correct equivalents follow these. Minor additions and deletions which could be included in the main text without changing it have been added using the following brackets: [ ] <> additions against the manuscript (also, in the English translation, additions necessary for understanding the contents of the text), deletions against the manuscript. All other additions and deletions are to be found in smaller print under the main text. I should like to emphasise here that all changes of the text through brackets, as well as the critical apparatus, are meant as help for understanding the text according to the reading which for me seems to be the most appropriate, and not as actual emendations. It is impossible to determine how many of the seeming mistakes have been meant to be just as they are by the author. In order to facilitate the overview over the structure and the selection of themes dealt with in each paragraph (cf. also the two tables above), each new subject has received a new line in the edited text as well as in the translation. Gaps in the text of the manuscript itself indicate the end of a sentence. They have as a rule been replaced in the edition by a period sign or, at times, by a comma. Missing, misplaced or problematical gaps are marked in the edition and in some cases additionally discussed in the footnotes. Probable cases of elision of a in the pronunciation of New Indo-Aryan words have been neglected; a has therefore been retained throughout. 30 See also note 25. Concerning I,3, there are also translations possible which rather exclude the involvement of a second person.

I. Varasaptadosa/Jvaropacara³¹ Text I: Harm of the Seven Weekdays/Treatment of Fever <la> OM sriganesaya namah. ii atha varasaptadosa likhyate. ii 'dosa likhyate: °doso likhyate/dosa likhyante/dosam [pustakam] likhyate. ¡OM! Homage to Lord Ganesa. Now the Harm of the Seven Weekdays is being written. I,1. Adityavara/Itavara - Sunday iatare tapa ave ta yogini drsta uttaradisa jalanikata janani. "dasadinata i kastah damtasula adhasiratabakhidukhe lutadikasta hastapadadikastah. isaka upaya bakvara marana. kumari 57 pujyah. [1] i drsta: drsta, uttaradisa: uttaradisi (or uttaradisayam), nikata: "nikatam or °nikate. ii kastah: kasto, damtasula adhasiratabakhidukhe: damtasulo 'dhahsirastabakhiduhkhe, kasta: kasto. iii bakvara: bakkara, pujyah: pujya. 33 'If the fever comes on Sunday, 32 then a yogini is (to be) known as [having been] seen [by the sick person] in the northern direction near water. "For³4 ten days this affliction 35 [lasts]: toothache, suffering due to a drooping head 36 and intestinal sickness 37, affliction through cutaneous disease 38 etc., 31 Colophon title of this text, see above, introduction. 32 atare is obviously related to Hindi itavara (= Skt. adityavara), but used with the Sanskrit locative suffix, in accordance with most of the other names of days in this text. 33 janani: Note the adjectival use of the so-called verbal noun or infinitive. This usage is also known in Hindi (though today rare except with par); cf. Kamataprasada Guru, Himdi vyakarana (Sastravijnana Gramthamala 1), 8 th reprint, Kasi V.S. 2022, 271 (§372). 34 ta here probably is a New Indo-Aryan equivalent of tavat (see 1,4 and 1,5), or else derives ultimately from Persian (though note its postpositional use then), whereas in ave ta we have an equivalent of Hindi to. 35 Since several diseases follow, it bears consideration whether we should not translate: 'these afflictions'. This would mean that the New Indo-Aryan plural kasta (with mute a at the end, and similar in form to the singular) was "Sanskritised" incorrectly into a singular. 36 This probably means that the person is unable to keep his head upright. 37 Arabo-Persian tabaq in New Indo-Aryan usually means 'dish, bowl' or the like, but also 'surface; heap' and so on. George Abraham GRIERSON, A Dictionary of the Kashmiri Language, compiled partly from materials left by the late Pandita Isvara Kaula (Bibliotheca Indica N.S. 1405), Calcutta 1915-1932 (four parts), 962 b, also lists 'a certain severe disease

A. Herrmann-Pfandt, The So-Called Dakinikalpa affliction of hands, feet, etc. 69 The remedy for this is to kill a he-goat 40.41 Kumari is to be worshipped 57 [times] 43. 1,2. Somavara - Monday 'somavare jvara hoya to sayampratah(.) veladvayamadhya vatika (gh)atavimadhye gatah. tatranekabhutapisacadinam samcara abhut. "udarasulah jvarahastapadadidu[h]kham. ekadasapurusesu vetalaprityartham pasubalir (a) deve. [2] i jvara: jvaro, madhya: 'madhye, pisacadinam: pisacadinam, samcara abhut: samcaro 'bhut. ii udarasulah: udarasulo. involving intestinal tumour'. Though here this specialised meaning cannot be assumed, especially since the affliction lasts only for some days, tabaqi (the derivative suffix i could be Persian as well as Indo-Aryan) too could refer to something similar, maybe some more general affliction of the intestines, as has been assumed here. But Persian dictionaries also give the meaning 'vulva' for tabaq (cf. e.g. F. Steingass, A Comprehensive Persian-English Dictionary. Including the Arabic Words and Phrases to be met with in Persian Literature. Being Johnson and Richardson's Persian, Arabic, and English Dictionary Revised, Enlarged, and Entirely Reconstructed, reprinted Beirut 1975, 809 a); so one may debate whether here an affliction of the vulva or (as words denoting the vulva and the anus may interchange in many languages) the anus might be meant. Anyway, the fact that tabaq(i) as a loan-word in New Indo-Aryan does not as a rule seem to refer to a disease, but is or was found in such a meaning in Kashmiri (and so maybe also in neighbouring languages), does seem to point to a north-western origin of this text too. 38 The primary meaning 'spider' of luta might also be considered. As to the disease luta, Monier MONIER-WILLIAMS, A Sanskrit-English Dictionary, Oxford 1899, 905 c writes: 'a kind of cutaneous disease (said to be produced by the moisture from a spider)', with only the Rajatarangini as a source, which could be taken as further proof of our text's north-western origin. But Prof. Das informs me that the disease luta is known to medical texts too. n 39 On n cf. the retroflexed forms from western and northern New Indo-Aryan languages e.g. in R.L. TURNER, A Comparative Dictionary of the Indo-Aryan Languages, Oxford etc. 1966, no. 10066 (also in the Addenda and Corrigenda, edited by J.C. Wright, London 1985). 40 bakkara is an obvious emendation for bakvara, as the ligatures kv and kk are often confused or even identical in Devanagari manuscripts. Note that the doubling of the consonant k in bakk° is more reminiscent of Punjabi or Lahnda than Hindi. 41 This sentence is purely New Indo-Aryan with no Sanskrit elements (except for the tatsama word upaya). 42 Kumari is a name of the goddess Durga. It cannot be excluded, however, that here the worship of a young maiden is meant: 'a kumari. 43 This translation of the number 57 is only a guess.

'If the fever arises on Monday, then [he] went in the evening or the morning between the two times of day (i.e. in the evening or morning twilight) into the garden or the forest. There was a concourse of many bhutas, pisacas etc. there [at that time]. "[This affliction consists of] stomach-ache, fever, suffering in hands, feet etc. iii An animal sacrifice is offered amongst 45 eleven men to please the vetalas. I,3. Bhaumavara - Tuesday ibhaumavara jvara udarapadyate, atura[h] kutrapi sa(m/bhayam bhuktavat[.] sa puruse ucchistavelayam praptah. tatraiva bhayam prapanna [h]46. "asthisulam udarasulam mahabhayam dinani sapta (h), athava masam 1. tadartham tamdulam annam tailamasa (m/m annam taila[m] upau. 7 e<lb>katra paviyitva (.) kutrapi devasthanam samvibhajya deyam. 3 i bhaumavara: bhaumavare, udarapadyate: utpadyate,47 bhuktavat: bhuktavan. puruse: purusa. ii asthisulam: asthisulam, saptah athava: saptathava. iii tamdulam: tandulam, paviyitva: pacayitva, devasthanam: devasthane. 48 [If] the fever comes up on Tuesday, [then] the sick man ate somewhere at an assembly. It (the fever) was incurred in the man at the time of (i.e. when he was in the state of) impurity after eating [and before washing].49 Just then he 44 Though deve looks like an active optative form, the preceding nominative and the Sanskrit deyam in exactly the same place in the next section (1,3) make it seem more probable that what we have here is actually an old inflected passive; similar forms are indeed found in older New Indo-Aryan, though in the majority of the modern languages the inflected passive, at least in a garb similar to the above, has died out except for some rudimentary remnants. This could be useful for dating our text. However, according to Prof. Mishra passive forms such as the above still exist in Hindi dialects. Should this indeed be the case, then the mere existence of this form could by itself not be used as a criterion for dating the text. 45 The translation strives to be as ambiguous as the original. But cf. manusyanavakena in 1,5; this could speak for translating with 'through' or 'by' here. 46 Also possible: prapanna[m]. 47 See e.g. I,6,i. The interpretation pavayitva 'having cleaned, purified', which requires less emendation, does not make sense here. 49 For this translation I am indebted to Prof. Srinivasan. Though this seems to be the most elegant solution, one might also consider translating: 'He (i.e. the patient) was reached (i.e. attained [by the disease] or touched [by another man]) when the man (the patient or the other person) was at the time of impurity after eating.' In that case syntactically the construction would have to be regarded as an equivalent of *purusa ucchiste (sati), or as a parallel to *purusa ucchistamatre. On the locative absolute construction (purusa ucchista-

A. Herrmann-Pfandt, The So-Called Dakinikalpa got into danger.50 71 "[There are] pain in the bones, pain in the belly, [and] great danger³¹ for seven days or else for one month. 52 53 For this one must prepare rice food, oily bean food, and oil. Having cooked [all this] together 7 [times],54 it is to be offered somewhere at a temple after having divided [it].55 1,4. Budhavara - Wednesday budhavare jvaranaya grhyate[,] smasanamadhye samagatah(.) abhut, yatra dvau bhutau sthitau(.) abhut, ekahh(.) mleccha anyah (.) vipra[h.] "padasulam(.) damtasulam dasa dinani tavat. velayam) cf. also J.S. SPEIJER (see note 18), §367: 'It is not necessary, that the predicate of the absolute locative be a participle. It may be also a noun (adjective or substantive).' One could also consider emending the text to read sa purusenocchistavelayam (i.e. purusena ucchistavelayam) praptah, which would then mean: 'He was reached (i.e. touched) by a man at the time of impurity after eating', it in this case also being unclear who is in the state of impurity. Or else one could take puruse to stand for the instrumental (which would do away with the need for emending the text), in this connection pointing to ekadasapurusesu in 1,2; the translation would be the same as in the preceding case. See also note 26. 50 Or, if we read prapanna[m] (= prapanna[m]): 'Just then danger was attained.' 51 Alternative translation: 'great fear'. The translation 'great danger' follows bhayam in the preceding sentence. 52 Or else simply: 'Prepare'. The use of the imperative ending u is well-known from late Middle Indo-Aryan and the transition stage to New Indo-Aryan; cf. e.g. Ganesh Vasudev TAGARE, Historical Grammar of Apabhramsa, reprinted Delhi/Varanasi/Patna/Madras 1987, 297 ff., Subhadra Kumar SEN, Proto-New Indo-Aryan, Calcutta 1973, 90. In New Indo-Aryan, too, it is relatively wide-spread both in time and area; see on this e.g. S.H. KELLOGG, A Grammar of the Hindi Language ..., first Indian edition, New Delhi 1972, 299, 315, 346, John BEAMES, A Comparative Grammar of the Modern Aryan Languages of India ..., second Indian reprint, New Delhi 1970, III, 108. It thus cannot be used as a criterion for dating our Concerning the New Indo-Aryan Vupa 'to produce, create' (= Skt. causative utpadayati), cf. Monika THIEL-HORSTMANN, Crossing the Ocean of Existence: Braj Bhasa Religious Poetry from Rajasthan. A Reader, Wiesbaden 1983, 156, s.v. upa- (from Dadu Dayal's sakhis). This last reference was pointed out to me by Dr. Steiner. text. - 53 tailamasa is probably the same as masataila 'an oily preparation from beans'. 54 This translation of the numeral seems to give the most sense here. 55 One could also consider translating: 'Having cooked [all this] together, it is to be offered somewhere at a temple after having divided [it] 7 [times] (i.e. into seven parts).' Though this seems to make better sense than the translation further above, the position of the numeral tends to speak for the former translation.

svetakar (a)patam va krsnakarapatham (.) dharmakaryam[.] etatkrte satasubham bhavati. 4 i jvaranaya: jvaranayo or jvaranayo, samagatah abhut: samagato 'bhut, sthitau abhut: sthitav abhutam, ekah mleccha anyah vipra: eko mleccho 'nyo viprah. iii krsnakarapatham: krsnakarpatam, bhavati: bhavati. The course of fever on Wednesday is [now] taken up." [He] arrived in the midst of a cremation ground where two ghosts were present, one [of them] a mleccha,57 the other one a Brahmin. 58 "[There are] pain in the feet and toothache for ten days. [To prepare] a white garment or¹ a black garment is a religious duty. When this is done, it becomes hundred[-fold] auspicious. 1,5. Guruvara - Thursday guruvasare jvara udapadyate, samdhyasamaye unnatavrksatalam gatah(.) abhut[.] tatra tena drstih(.) vipra[.] athava paniyasamipam agnim utva snanam krtam. etasya drstih 7 vipra[h]. "tatraiva madhya[h]ne velayai manusyanavakena krsnapasu (h.)bali[h]62 karyah dinavimsakam tavat. tato api samdehah. (7)[5] 56 One could also consider keeping the reading jvaranaya and translating: '[If he] is seized for feverishness (i.e. so as to be feverish), [then]...'. jvarana would obviously be a nomen actionis, similar to jvalana; on the dative cf. J.S. SPEIJER (see note 18), §91: 'In short, in Sanskrit datives of nomina actionis (bhavavacanani) do often duty of infinitives.' What speaks against this is not only that we would have to assume here a word otherwise not found, in a rather uncommon construction to boot, but also jvaranayah in 1,7 and, generally, the fact that each weekday section of Text I is introduced by an announcement of this kind; cf. also the remarks on antecedence on p. 60. Probably someone not within the fold of the same religious system ('Hinduism'). Or could an outcast be meant? 58 One could also consider dividing the last three sentences thus: smasanamadhye samagatah. abhut yatra dvau bhutau sthitau. abhut ekah mleccha anyah vipra: 'He arrived in the midst of a cremation ground. It was [there] where two bhutas were present. It was [the case that] one [of them was] a mleccha, the other one a Brahmin.' See, however, p. 66. 59 tavat here is clearly used like similar New Indo-Aryan postpositions. 60 See the New Indo-Aryan derivatives of karpata listed by R.L. TURNER (see note 39), no. 2871 (also in the Addenda and Corrigenda). 61 The position of va 'or' is similar to that in New Indo-Aryan. 62 The position of the visarga seems to have been shifted from balih to pasuh as the result of a scribal error.

A. Herrmann-Pfandt, The So-Called Dakinikalpa 73 i udapadyate: utpadyate, gatah abhut: gato 'bhut, drstih vipra: drsta vipra/viprah, agnim utva: agnim hutva, drstih 7 vipra: drstah 7 viprah. ii velayai: velayam, krsna°: krsna, karyah: karyo, tato api: tato 'pi. [If] the fever comes up on Thursday, [then he] went at the time of twilight under a high tree. There a Brahmin woman 63 was seen 64 by him.65 Or else, after having sacrificed into a fire near the water, a bath was taken. 7 Brahmin women were seen by this one.67 66 "In this very regard at noon time a group of nine people should perform a sacrifice of a black animal for twenty days. Even after that there is doubt [whether there will be success]. 1,6. ukravara - Friday isukravare jvara utpadyate, athavimadhyam gramadau kutrapi gata (h.) atura[h][.] tatrekam devageham aset() marge[.] tatra bhuta abhavat. "tatraivakasmat(.) sala (m)pasubali[h] ka<2 a>ryah. 6 i athavi°: atavi, tatrekam: tatraikam, aset: asit, bhuta abhavat: bhuta abhavan. ii. °kasmat sala°: kasmac chalaº. [If] the fever comes up on Friday, [then] the sick person went to a forest 63 It is unclear whether one may assume a sandhi in the manuscript between vipra and athava or not. In the former case we would have several Brahmins (or Brahmin women, if drstih is feminine; see below), in the latter one Brahmin woman. I have chosen to give the translation 'a Brahmin woman' on the assumption that drsti is indeed feminine, and that in the case of a plural number of women a number would have been specified as in the following case. However, the translations 'Brahmins' and 'Brahmin women' instead of 'a Brahmin woman' remain possible. 64 drsti could have been erroneously taken to be the feminine of drsta. 65 One could also consider dividing the last two sentences thus: ... samdhyasamaye unnatavrksatalam gatah. abhut tatra tena drstih vipra: '[then he] went at the time of twilight under a high tree. It was there [where] a Brahmin woman was seen by him.' See, however, p. 66. 66 Here again I presuppose that drsti is indeed feminine; otherwise, one could also translate: '7 Brahmins'. 67 etasya is used here as a subjective genitive; see SPEIJER (cf. note 18), § 114. 68 Cf. tavat in I,4. Another case of "shifting visarga"; compare krsnapasuh bali in 1,5 above.

70 somewhere near a village etc. A temple was there on the way." Bhutas were there. "Exactly there a sacrifice of a domestic animal 72 is to be offered at once. 1,7. Sanivara - Saturday 73 'sanivare jvaranayah grhyate[.] atavimadhye marge smasanasamipe, atarde bhutah.) tatrabhut (a) paniyasamipe bhavanivrksamule[.] udarasulam dina 7 masadyai 2. krsnapasubalir deyah, ma(m)sannam deyam[.] subham bhavati. [7] i jvaranayah: jvaranayo, smasanasamipe atarde: smasanasamipa atarde, bhutah: bhutas. ii dina: dinani, masadyai: masadvayam. iii krsna°: krsna, bhavati: bhavati. The course of the fever [beginning] on Saturday is [now] taken up. On a path in the midst of a forest, in a hole 74 near a cremation ground, there was a bhuta in that place, near the water at the root of a Bhavani's tree. "Stomach-ache [arises] for 7 days [or] for two [times] two months.76 iiiA sacrifice of a black animal is to be offered. Food of beans" is to be offered. 70 An attempt to translate the locative, which seems to refer to the environs of the village. 71 One could also consider reading aset as asret and translating: 'There he entered a temple on the way.' But this requires more emending. 72 This could refer to the sacrifice of a dog, or maybe a cat (cf. salamrga and salavrka), in any case to an animal living in or coming into the house, i.e. not domesticated in the same manner as cattle, goats, or the like. 73 This meaning of akasmat has to be assumed here. It is not found in Sanskrit dictionaries, and also not in most Hindi dictionaries. However, the meaning tatksana 'at once' for akasmat is given in: SYAMASUNDARADASA (chief ed.), Himdi sabdasagara. Prathama bhaga, 2 nd printing of the revised and corrected new edition, Varanasi 1986, 65 b. Could we therefore have New Indo-Aryan influence here? 74 Concerning atarde (read here as atarde), see also note 27. = 75 I have not been able to find such a tree name mentioned anywhere. Could we actually have a mistake for bhava(n)ti bhava(n)ti here? Then we would have to translate: '... there was a bhuta (maybe in the sense of: there were bhutas?) in that place. It/They tend to be near water at the root of a tree.' 76 Should °dyai 2 be just the result of an erroneous duplication, however, then the translation would be: 'for two months'. 77 One could also consider emending to mamsannam 'food (scil. a dish) of meat', which would then probably refer to cooked meat. However, the s makes it more likely that masannam is correct.

A. Herrmann-Pfandt, The So-Called Dakinikalpa 75 Good fortune arises. ¡iti jvaro upacara. ijvaro upacara: jvaropacarah. Thus the Treatment of Fever.