4
Roots and incorporations of Tvaritā, Kurukullā and Bheruṇḍā
Michael Slouber
Introduction
The prevalence of snakebite in South Asia coupled with the deeply rooted traditions of goddess-worship there has given rise to a variety of Śākta traditions of snakebite goddesses, some of which are popular down to the present day. Although the literature of and references to some of these goddesses is widespread, little scholarship has been done on their textual traditions. This lacuna is due in part to the fact that most of the early primary sources on goddess traditions remain unedited and unpublished, and therefore difficult to access. Over the last several years I have begun editing some of the surviving scriptures of the Gāruḍa Tantras, the most fundamental and influential sources for the topic of snakebite medicine. The Gāruḍa Tantras present us with a variety of traditions used by Gāruḍikas, practitioners who treat snakebite and other envenomings. The Gāruḍikas’ repertoire consisted of herbal medicines, mantras to the bird-king Garuḍa, protective diagrams, and other spell systems such as vidyā1 goddesses. I have been keeping track of references to these goddesses and with this paper I would like to acquaint you with some of my initial findings on three of them – Tvaritā, Kurukullā and Bheruṇḍā. Tvaritā was the subject of a vast literature, and though much of it has been lost, a substantial amount of material reaches us in manuscripts from Nepal and elsewhere. The literature on Kurukullā is also large, although the early Gāruḍa Tantra passages are consistently thin on detail. Bheruṇḍā’s references were widespread, but like Kurukullā, they lack depth. I might have also told you about Śākta snakebite goddesses such as Jhaṃkāriṇī,2 Jāṅgulī,3 or Suvarṇarekhā,4 for example, but for conciseness I limit myself to the above-named three, who became known to later tradition primarily through derivative traditions.
Tvaritā
‘The Swift One’ (tvaritā/tūrṇā/śīghrā) is a fitting title for a goddess whose most celebrated function was saving the lives of those bitten by venomous snakes such as a cobras, vipers, or kraits. Her ultimate origin may be lost in antiquity, but the earliest surviving source is one of the most widely cited canonical Gāruḍa Tantras called the Trottala.5 In time she was identified with various goddesses of extensive renown: Kubjikā, Durgā, and Kālī6 in the wider Śākta traditions, as well as Padmāvatī in the Jaina Tantras.7 In the introduction to his massive fourteen-volume edition and translation of the Manthānabhairava’s Kumārikākhaṇḍa, Mark Dyczkowski has an eight-page section on Tvaritā. His discussion is very informative, and he points to a long and important section of the unpublished Kulakaulinīmata that discusses Tvaritā at length.8 I also discovered a Tvaritāvidhānasūtra ascribing itself to the Caturviṃśatisahasrasaṃhitā (i.e. the Manthānabhairavatantra),9 but it is incomplete and damaged and I have been unable to trace it in other manuscripts of this massive work. We have many texts that know Tvaritā as an autonomous snakebite goddess, but it seems probable that most or all of these derive from the aforementioned Trottala, so I will only briefly mention them in the section below on ‘Borrowings’. Since we have several manuscripts of texts which ascribe themselves to the Trottalatantra, let us turn to them now and explore what Tvaritā was like in these early sources.
Tvaritā in the Trottalatantra
I am aware of two surviving works that ascribe themselves to the Trottalatantra: the 700-verse Tvaritāmūlasūtra and the 200-verse Tvaritājñānakalpa. Both use the titles ‘Trottala’ and ‘Trottalottara’ interchangeably, although the latter is listed as a separate text in the canonical lists. The Tvaritāmūlasūtra positions itself as an extraction from the (presumably mythical) version of over 100,000 verses. I have introduced the text and edited and translated most of the first chapter for Somadeva Vasudeva’s forthcoming Tantrik reader Śivasudhāprapāpālikā, so here I will dispense with introductions and describe the most salient features of each of its nine chapters.
The origin story of Tvaritā in the first chapter is a rather grand tale in ninety-four verses. Some stanzas are directly parallel to the similarly named Niśvāsamūlasūtra, and presumably the Tvaritāmūlasūtra redactor has that text in mind. The frame story is that Garuḍa approaches Śiva after many aeons of penance and requests teachings on a variety of topics. As an aside, I should remind you that Garuḍa is an independent deity in the Śaiva Tantras and is almost never associated with Viṣṇu, despite the present-day popularity of his Vaiṣṇava identity. In response, Śiva begins to narrate Tvaritā’s origination story to Garuḍa: a band of yoginīs, themselves originated from Śiva’s Bhairava form, approached Śiva and requested that he preside over their caru ritual, which, the context implies, involved sexual rites. When Gaurī learned that Śiva was engaged in this ritual, she became dreadfully angry and generated herself as Tvaritā by uttering the nine-syllable Tvaritā vidyā. Here she has eighteen arms, which the text reminds us correspond in pairs to the nine syllables of the vidyā, and she is seated on a lion. This mount brings to mind Durgā, but the similarity stops there and whether Durgā, Gaurī, and Tvaritā are three goddesses or one is a matter for another paper – here I treat them as separate personages.10 Several verses are devoted to describing her fierce lion, and then we return to Tvaritā’s own appearance:11
Tvaritā is mounted on [the lion] with her left shank hanging down and the right folded in so the sole of her foot touches her leg. The nails on her feet sparkle like twinkling stars, are lotus red and marked with lotuses. The soles of her feet are beautiful with long and symmetrical toes. She has well-rounded thighs and fleshy calves, and her buttocks and hips are broad. She has a deep naval, a belly with three folds, and is beautified by a streak of abdominal hair, broad hips and breasts like golden pitchers. She has a shell-like neck, lips like the Bimba fruit, eyes like blue lotus petals, a beautiful nose, curved bow-like brows, attractive ears, and she glows with a crown. She shines with snake earrings and a necklace consisting of the king of snakes, brilliant with a thousand hoods and radiant with glittering jewels. The goddess’s arms shine, beautified with snake bracelets, and likewise the girdle on her hips, frightening because of being made of a great serpent. On the feet of the goddess are two terrible snake anklets, and her left foot is placed in the middle of a thousand-petalled lotus. One garment, adorned with flowers, looks like forked lightning in the sky. Another garment is like a rainbow draped across her breasts. It is studded with numerous flowers, and tied tightly with a great serpent. The goddess’s hands carry weapons and the tips of her shell-like nails are sharp. On the right she holds a vajra, staff, sword and discus, a mace, a shining spear, arrow, and javelin, and she displays the gesture of granting boons. On the left she holds a bow, noose, sword, bell, a threatening finger, a conch and a goad, and she also displays a gesture of goodwill and holds a lotus.
This elaborate description of Tvaritā contrasts with the simple two, four, or eight-armed forms which Dyczkowski references in the Kulakaulinīmata and Tantrarāja,12 a point to which I will return in the section below on ‘Incorporations’.
The Tvaritā spell itself is an interesting topic, and here we will briefly describe some features of it. The mūlavidyā usually consists of three times three syllables: HŪṂ KHE CA CCHE KṢAḤ STRĪṂ HŪṂ KṢE PHAṬ, and often it is prefixed with oṃ. This is the form of the vidyā that we get spelled out plainly in the Tvaritājñānakalpa.13 It agrees with a versified enumeration in the same text’s twenty-fourth verse.14 In the Tvaritāmūlasūtra’s second chapter we are given the syllables of the mūlavidyā in a simple code based on the standard layout of the Sanskrit alphabet:15
The last of the heated group with a dot and mounted by the sixth vowel (h+ū+ṃ=hūṃ). The first seed syllable of the basic spell has been told, O Lord of Birds. Now, the second of the [soft-]palatal class joined with the eleventh vowel (kh+e=khe). The first [vowel] in conjunction with [that of] the tongue-palate [class] should be alone (c+a=ca). Beneath that same one, one must use the second of that [class] with the eleventh vowel (c+ch+e=cche). Now one must join the second of the heated beneath the first of the [soft-]palatal class together with the sixteenth vowel (k+ṣ+aḥ=kṣaḥ). One must use the first of the tongue-teeth (t) below the third of the heated (s) and the second of the mixed class again combined beneath with the fourth vowel (s+t+r+ī=strī). One must use the second of the heated joined beneath to the first of the [soft-]palatal class combined with the eleventh vowel (k+ṣ+e=kṣe). The last of the heated together with a dot and mounted by the fifth vowel (h+u+ṃ=huṃ). And the second syllable of the labials is another to be joined to that which is first when the tip of the tongue touches the palate; this is to be extracted with a half vowel (pha+ṭ+a/2=phaṭ). This is the Exalted Spell-Goddess Tvaritā who grants all success. She should be prefixed with oṃ and should always have ‘homage’ (namaḥ) at the end. For fire rites it ends in svāhā.16
Thus the vidyā given here is: (OṀ) HŪṂ KHE CA CCHE KṢAḤ STRĪ KṢE HUṂ PHAṬ (TVARITĀYAI NAMAḤ/SVĀHĀ). This differs from the Tvaritājñānakalpa version in several respects: the syllable strī lacks anusvāra, the syllables kṣe and huṃ are reversed, and the syllable huṃ in the eighth position has a short vowel. The significance of these differences is not currently apparent to me, but we may note that huṃ and hūṃ are elsewhere generally interchangeable, with the latter occurring only moderately more frequently than the former. The fact that the author went to the trouble to spell out that the first is with the sixth vowel whereas the second is with the fifth vowel suggests that the difference was significant in his tradition.
Next the Tvaritāmūlasūtra teaches the ancillary mantras: a three-syllable Heart mantra, a Head mantra with an unclear number of syllables, a five-syllable Crest mantra, a five-syllable Armour mantra, a mantra to the three Eyes whose number of syllables is unclear, and a four-plus-one-syllable Weapon mantra, whose first four syllables correspond to weapons placed in the four cardinal directions. The specific syllables of these ancillary mantras, where they can be clearly determined, differ from those plainly enumerated in the Agnipurāṇa parallel, so for the time being I will leave this puzzle unsolved. The text emphasises that without these ‘secret ancillary mantras’, one cannot have success with the Trottalatantra.
Thereupon the Tvaritāmūlasūtra, still in the second chapter and starting with verse 38, teaches the simpler ‘vidyā-ancillaries’:17
The first and second are the Heart. The third and fourth are proclaimed to be the Head. The fifth and sixth are taught as the Crest. The seventh and eighth are the Armour. The star syllable (phaṭ)18 is the Eye qualified with its half-syllable as being the ninth.
So we have the following scheme for the vidyā’s ancillaries:
Heart
hūṃ khe
Head
ca cche
Crest
kṣaḥ stri
Armour
kṣe hum
Eye
phaṭ
Next the Tvaritāmūlasūtra teaches the ten-syllable spells for each of ten female attendants (dūtī): Śakra’s Vajratuṇḍā, Agni’s Jvālinī, Yama’s Śabarī, Nirṛti’s Karālī, Varuṇa’s Plavaṅgī, Vāyu’s Dhūnanī, Kubera’s Kapilā, Rudra’s Raudrī, Viṣṇu’s Cakravegā, and Brahma’s Brahmavetālinī. Their vidyās are characterised by beginning and ending with the respective syllables of Tvaritā’s mūlavidyā, thus each syllable represents one of the female attendants. The way this works out to fit nine syllables to ten attendants is that phaṭ is taken to be two-in-one, so the pha element is Cakravegā and the ṭ element is Brahmavetālinī. The other syllables of the attendants usually include their name in the vocative, sometimes alias names, and either individual syllables or imperatives appropriate to each. Thus Jvālinī, the attendant of Fire, is told ‘blaze!’ and Manovegā, the attendant of Wind, is told ‘go!’ The first eight attendants clearly correspond to the eight compass points starting in the east and they are placed in this configuration around Tvaritā in many of the rituals taught in the Tvaritāmūlasūtra. For example, in the eighth chapter there are instructions for making a ‘Vajra-bolt’ (vajrārgala) diagram, here for the purpose of killing an enemy, using a circuit of the first eight female attendants surrounding Tvaritā on the petals of a lotus.19 TheVajra-bolt is Tvaritā’s signature maṇḍala. It is the same one that Śiva was presiding over with the yoginīs in the first chapter, and that which Tvaritā was persuaded to enter by all of the terrified gods.
My summary has covered most of the first two chapters of the Tvaritāmūlasūtra. I will now very briefly look at the contents of the remaining seven chapters. The third teaches the installation of Tvaritā’s weapon mantras on the hands and body of the mantra practitioner, which affords him invulnerability from gods, demons, or any evil influences. The fourth Chapter is on mudrās – hand gestures used in the worship of Tvaritā and rituals involving her. Twenty-eight mudrās are described, many corresponding to the eighteen weapons/gestures in Tvaritā’s hands.
The fifth chapter teaches initiation. The ritual begins with an elaborate worship of Tvaritā in the Vajra-bolt maṇḍala, with her mounted on a five-faced Śiva acting as her throne. This pose is probably meant to demonstrate her superiority to Sadāśiva, the prototypical five-faced Śiva of the Siddhānta Tantras. The initiation also involves ritual generation of fire in a vulva-shaped pit and offering grains and ghee into it while reciting the basic vidyā along with the ancillary spells. At one point the text says ‘And he becomes initiated by just one oblation, O Bird; in this way he would be authorised. Now listen further concerning liberation.’20 If my understanding is correct, this ritual departs significantly from the Śaiva norm where the most basic initiation grants liberation and further initiation is required for those seeking powers. Against this interpretation is the fact that the opening of the chapter calls the initiation both power-granting and liberation-granting. Many benefits of initiation are listed, such as obtaining a kingdom, success with mantras, destruction of poverty, and obtaining sons. At one point there is a choice to either dismiss the goddess and dismantle the maṇḍala, or for those who are authorised to continue with offerings that include animal and human blood. The goddess is praised as present in a long list of deities, in fact, as all-pervasive. The Chapter closes with a fascinating discussion of who may be initiated and who is unqualified. That chapter will be a fruitful source for future research.
The briefer sixth chapter gives instructions for locating a site on which to practice, ranging from a dreadful cremation ground to cities, towns, or villages where people are predominantly Śaiva. Instructions are given for several basic rites that I will not discuss here. Chapter 7 begins and ends on the topic of different extractions of the syllables of the vidyā for various purposes, but most of the chapter is rather a detailed description of the creation of various deities culminating in Tvaritā’s appearance. The details of this creation story would certainly be of comparative interest to other scholars of Śaivism and Śākta traditions, but space permits me to note only a few features. The basic image is one of chaos in the universe with various exceedingly powerful forces coming into existence and clashing. A battle between Garuḍa and Viṣṇu ends with Viṣṇu being vanquished and leaving the egg of Brahma. The chaos does not come to an end until Trotalā, Tvaritā’s nom de guerre, is established as the protectress (trāyakā) and terrifier (trāsakā) of the world. This serves as a folk etymology of her name. The creation story in the first chapter has little in common with the one here, so we must assume that the text is preserving two separate accounts of her creation.
The long eighth chapter is a collection of various practical applications (prayoga) of the vidyā. We see instructions for making magical diagrams (yantra) on funerary cloths, skulls, or less grim, walls and leaves. Goals include the standard black magic actions like killing an enemy, sowing dissension, driving a rival out of town, or controlling women; white magic actions such as creating peace and well-being; royal work like defeating an enemy army; and more specialised actions like destroying possessing demons and fevers. We also see, of course, several rituals for destroying poison and healing snakebite victims. We will mostly pass over the ninth and final chapter on yoga, as I am not experienced in this subject. Suffice it to say that here it involves meditation, breath control, and visualisation culminating in a vision of Tvaritā.
Figure 4.1 A twelfth-century folio of the Tvaritāmūlasūtra in the Kaiser Library, Kathmandu
The Tvaritājñānakalpa, which I have mentioned several times already, is very parallel to the Tvaritāmūlasūtra. It is only 200 verses in extent,21 and the colophon places it as the thirty-fifth chapter of the 11,000-verse Troṭalottara. We have three Nepalese manuscripts of it and the earliest is paleographically similar to manuscripts from the eleventh or twelfth centuries. It is not obvious from the parallel passages whether the kalpa depends directly on the mūlasūtra, but it is certainly abbreviated and shares many verses. It is notable in giving various applications (prayoga) of the Tvaritā’s basic syllables, sometimes using only a few of them, and sometimes more. These applications feature snakebite cures much more centrally than the mūlasūtra, although other topics are also given.
Borrowings
We see Tvaritā material in many other texts, but we have also determined that little of her grandeur in the Tvaritāmūlasūtra carries over in subsequent literature. I will discuss her other identities shortly, but here I would like to point out some parallel passages in the Agnipurāṇa that are clearly dependent on the Tvaritāmūlasūtra:
The column on the left is extracted from forty-two verses in the Tvaritāmūlasūtra that are redacted as only five and a half verses in the Agnipurāṇa. This continues, with the next line in the Agnipurāṇa (310cd) picking up with Tvaritāmūlasūtra 2.1. I have given the Tvaritāmūlasūtra’s Sanskrit for verses 2.11–20ab in note 15 on Tvaritā’s mūlavidyā, which one can compare to Agnipurāṇa 310.10–18. It is mostly word-for-word except in lines like Tvaritāmūlasūtra 2.11 where the vocative ‘O Lord of Birds’ (khageśvara) did not fit the context of the Purāṇa and the line was simply dropped. In several cases the redactor tried to clean up the non-standard Sanskrit forms like ūṣmāṇasya by changing it to ūṣmaṇaś ca, but in a few cases we see the redactor corrupting the sense even further; for example: ‘mounted by the fifth vowel’ (pañcamasvara–m–ārūḍhaṃ) → ‘mounted by five vowels’ (pañcasvarasamārūḍhaṃ). Needless to say, M. N. Dutt’s ‘translation’ of the Agnipurāṇa passage in Joshi’s edition is complete nonsense and notably altogether skips this and several other verses in the chapter. The parallels may be summarised as follows:
Table 4.1 Parallel passages in the Agnipurāṇa that are dependent on the Tvaritāmūlasūtra
|
Tvaritāmūlasūtra 122 |
Agnipurāṇa 31023 |
|
|
aṣṭādaśabhujā devī dharmasiṃhāsanasthitā // 41 // … tvaritā tatra cārūḍhā vāmajaṅ;ghā pralambitā / dakṣiṇā dviguṇā tasyāḥ pādapṛṣṭhe samarpitā // 45 // … vajradaṇḍāsicakraṃ ca gadā śūlaṃ mahojjvalam / śaraṃ śaktiś ca varadaṃ dakṣiṇena kṛtāyudhā // 55 // dhanuṣpāśadharaṃ ghaṇṭā tarjanī śaṅ;kham aṅ;kuśam / abhayaṃ ca tathā padmaṃ vāmapārśve kṛtāyudhā … yas tu pūjayate bhaktyā gṛhe nityaṃ svaśaktitaḥ / śatravo vilayaṃ yānti iṣṭasattvaparāṅ;mukhāḥ // 66 jayate pararāṣṭrāṇi līlayā pūjitā tu saḥ / vibhūtayaś ca vipulām ārogyaṃ dīrgham āyuṣam // 67 // sidhyanti sarvakāryāṇi divyādivyāny anekadhā / … taleti sapta pātālān kālāgnibhuvanāntikān oṃkārādi svar ārabhya yāvad brahmāṇḍa–vācakam // 82 // trakārāt trāyate sarvāṃs trāsate caiva sarvataḥ / trotalā tena ākhyātā tantrārtho ’yaṃ pratiṣṭhitaḥ // 83 // |
aṣṭādaśabhujāṃ siṃhe vāmajaṅ;ghā pratiṣṭhitā / dakṣiṇā dviguṇā tasyāḥ pādapīṭhe samarpitā // 3 // nāgabhūṣāṃ vajradaṇḍe khaḍgaṃ cakraṃ gadāṃ kramāt / śūlaṃ śaraṃ tathā śaktiṃ varadaṃ dakṣiṇaiḥ karaiḥ // 4 // dhanuḥ pāśaṃ śaraṃ ghaṇṭāṃ tarjanīṃ śaṅ;kham aṅ;kuśam / abhayaṃ ca tathā vajraṃ vāmapārśve dhṛtāyudham // 5 // pūjanāc chatrunāśaḥ syād rāṣṭraṃ jayati līlayā / dīrghāyū rāṣṭrabhūtiḥ syād divyādivyādisiddhibhāk // 6 // taleti saptapātālāḥ kālāgnibhuvanāntakāḥ / oṃkārādisvarārabhya yāvad brahmāṇḍavācakam // 7 // takārād bhrāmayet toyaṃ totalā tvaritā tataḥ / |
|
|
Tvaritāmūlasūtra → Agnipurāṇa |
||
|
1–4 |
310 |
|
|
5–6 |
311 |
|
|
7–8 |
312 |
|
Chapter 313 in the Agnipurāṇa has no material on Tvaritā, but it resumes with chapter 314, ‘Tvaritājñānam’. Such a title makes us suspect that it may be drawing from the Tvaritājñānakalpa, but I found no parallels. Agnipurāṇa 309 also opens with ‘Now I shall tell the tvaritājñānam’, but it too appears unrelated to the Tvaritājñānakalpa that reaches us. I do, however, see that most of Agnipurāṇa 309 is parallel with Nārāyaṇa’s Tantrasārasaṃgraha 22, starting with verse 47 and going to almost the end of the chapter. Īśānaśivagurudevapaddhati 22 also has some parallels with this chapter. It is not clear to me now whether one of these texts is copying from another or if they independently copy from the same source text or texts, but I would not rule out the possibility that the Agnipurāṇa is drawing on post-canonical digests in some cases. It appears to use Nārāyaṇa’s work in chapter 294, unrelated to Tvaritā but of interest to my broader work on the Gāruḍa Tantras.
These identifications of parallels just scratch the surface of what remains to be discovered by careful textual work taking into account the rich treasure of unpublished sources. We have many Tvaritā texts that I have not yet mentioned and have no time or space to explore in detail now, such as the aforementioned Tvaritāvidhānasūtra,24 Pārameśvarīmata 39 which describes itself as drawn from the Trotalottara, and Śāradātilaka 10 (up to around verse 50) with Rāghavabhaṭṭa’s useful citations of many other Tvaritā texts. In her book The Iconography of Hindu Tantric Deities, vol. II, Gudrun Bühnemann points to the tenth-century Prapañcasāra, verses 13.26–31, as the source of her description in Śāradātilaka 10.25
Incorporations of Tvaritā
Much of the Tvaritā material in other traditions cannot at present be attributed to direct borrowing from the Trottala corpus. In this regard, let us return to the topic of her appearance. Dyczkowski lists three important features common to most of her visualised forms: that she is a tribal woman (śabarī), that she is adorned with snakes, and that she is associated with peacocks.26 These criteria agree with her appearance in Agnipurāṇa 309; however, the Tvaritāmūlasūtra visualisation mentions nothing of her being a tribal woman or associated with peacocks, and so these features are absent in the chapters drawn from that work (310–312). Agnipurāṇa 314 has been taken to be a third unique visualisation of Tvaritā in that text27 – as two- or eight-armed – but details of her visualisation are not given and the eleven attendants match those in chapter 309, and so I think it is safe to assume it is coming from the same tradition. What tradition might that be?
In Kulakaulinīmata 3, the main form in which Tvaritā is visualised agrees with Dyczkowski’s attributes, but it also mentions an alternative eighteen-armed form for use in magical rites.28 This would seem to be a reference to our Tvaritāmūlasūtra version. On the provenance of this chapter, Dyczkowski points out that it does not mention Kubjikā at all, but identifies her as Tripurā, and that he suspects this entire chapter was drawn from a tantra of another school.29 I assume this other school would be some early form of the cult of Tripurasundarī.
We might then classify the early Tvaritā literature into two camps: the Trottala corpus and its borrowers on the one hand, and the texts apparently derived from an early or proto-Traipura tradition. Examples of the latter generally share the following features not present in the Trottalacorpus:
It appears likely, then, that the origin of Kulakaulinīmata 3, Agnipurāṇa 309 and 314, Nārāyaṇa’s Tantrasārasaṃgraha 22, and Īśānaśivagurudevapaddhati 22 all may lie in the early Tripurā cult. This furthermore appears to be the form of Tvaritā in which she was adopted into wider pantheons in the form of a nityā, yoginī, mātṛkā, or śakti in the retinue of another goddess. For example, in later Traipura scriptures like the Jñānārṇava and Śrīvidyārṇava, Tvaritā features as an attendant (here nityā) in the retinue of Kāmeśvarī, a synonym of Tripurasundarī, and her vidyā is the twelve-syllable version enhanced with two hrīṃs.
The key question is whether Tvaritā ultimately emerges from the Trottala corpus or that of Tripurasundarī, and the evidence points to the former. Although it may be tempting to suggest that a simpler visualisation of Tvaritā as tribal snakebite goddess was the source of the more complex and encompassing eighteen-armed Tvaritā, it may not be the case. The cult of Tripurasundarī was devoid of the ferocious hordes of prior Kaula pantheons and was set to be incorporated into mainstream religion, and so had a need for powerful yet non-threatening deities. But perhaps there is a third model. Perhaps there was an original Tvaritā cult attached to an early Trottala scripture wherein Tvaritā had a more humble appearance that was adapted to be more Durgā-like as the cult grew in prominence as we see in the Tvaritāmūlasūtra. We cannot be sure, but perhaps these speculations will be useful avenues for future research.
We may also note that this Traipura Tvaritā was closely associated with the Jain goddess Padmāvatī. The third verse of the Bhairavapadmāvatīkalpa – an important Jain tantric work from the eleventh century with significant dependencies on Śaiva tradition – runs as follows:30
Totalā, Tvaritā, Nityā, Tripurā, Kāmasādhanī: these are names of the goddess Padmā, and so is Tripurabhairavī.
In his Hamburg lecture entitled ‘The Appropriation of Śaiva Sources and Models in the Production of Jain Ritual Paddhatis from the 10th to the 15th Century’, Alexis Sanderson pointed to this verse as one among many pieces of evidence that the Bhairavapadmāvatīkalpa drew on Śaiva sources, in this case, he asserted, the Śākta cult of Tripurasundarī. The evidence which I have offered on the widespread presence of Tvaritā in Traipura sources both supports and is supported by Sanderson’s cogent thesis.
The connection between Padmāvatī and Totalā/Tvaritā must have been an easy one, since Padmāvatī was closely associated with snakes through her previous life’s role as snake-protectress of Pārśvanātha, the twenty-third tīrthaṃkara in Jainism, himself associated with curing snakebite. She remains an important goddess in Jainism even today, often pictured with a series of cobras with flared hoods shielding her from above.
Tvaritā is also worshipped today. We may point to the popular Tulja Bhavani temple in the Tuljapur district of Maharashtra, whose website informs us that Tuljā is a Marathi form of Sanskrit Tvaritā.31 The description of the idol and the descriptions on the website, however, make it clear that she is regarded as Durgā, slayer of the buffalo demon, so we can only wonder about the roots of this particular temple.
Looking for Tvaritā on the internet, one predominantly finds references to her as a nityā goddess, which is the identity that I believe was popularised in the early Tripurasundarī literature. At Celextel.com,32 under the category ‘Yantras’ Tithi-Nitya’, one can buy a copper Tvaritā yantra that they promise protects the owner from ‘poverty and poisonous attacks’. At Shiva-Shakti.com,33 we get a description of Tvaritā as a Nityā goddess, evidently drawn from Traipura sources.
Kurukullā
The goddess Kurukullā is best known as a tantric Buddhist goddess and often identified with Tārā. Her Buddhist identity is so popular, that even so eminent a scholar as Jan Meulenbeld remarked that a reference to her by the ninth century Śaiva physician Māhuka was to a tantric Buddhist goddess.34 He backed this statement up with nearly twenty references to her in the secondary literature.35 What he didn’t know, and what the scholars he cites didn’t know, is that Kurukullā actually has a complex Śaiva identity that cannot easily be reduced to borrowing from the Buddhist traditions. Whether Kurukullā originally sprung from Buddhist or Śaiva roots cannot be easily determined, but we can at least here briefly show that it is not a simple question.
Śaiva references
In Śaiva/Śākta literature, it is useful to distinguish between two Kurukullā identities: the first as a goddess who heals snakebite and keeps a home safe from snakes and harmful influences, and the second as a subsidiary figure in various other goddess traditions, usually not associated with snakes or poison. The oldest references, so far as I can determine, are to the first identity.
The earliest Śaiva references come from the ninth century. We have Māhuka’s citation of the power of Kurukullā and Bheruṇḍā in the opening verses of the first chapter of his Haramekhalā, but these are extremely brief – only one line for Kurukullā: ‘Kurukullā drives away snakes [when] inscribed at the threshold of the house’.36 The anonymous commentator fleshes this out somewhat by telling us that ‘threshold’ means a certain part of the door – I would assume it is the lintel in conformity with the practice one sees in modern Nepal for the Nāg Pañcamī festival – and that one is to post a yantra there on birch bark which has been inscribed with the syllables of Kurukullā’s vidyā on the six corners of two interlocking triangles, as in the following figure:
Judging from the widespread references to it, the apotropaic practice of hanging this yantra in one’s house may be the core of Kurukullā’s fame in the early Śaiva tradition. Śaṅkuka’s Saṃhitāsāra also mentions this practice in his section on Kurukullā as a Gāruḍa goddess. This text is roughly contemporaneous to the Haramekhalā (both c. ninth century), and at six verses, this is sadly the longest passage that I have seen on the early Śaiva Kurukullā. In these verses we also learn that the Gāruḍika mantra practitioner would install the syllables on his body and be able to carry out various magical acts just like Garuḍa, incant a string with the vidyā and ritually place it on a patron to ward off snakes, and incant gravel to be thrown in a house to drive out Nāgas. For details on this text and the practices mentioned, see Slouber 2011: 51–56.
In the Kriyākālaguṇottara, a scripture from around the tenth century drawing on older Gāruḍa and Bhūta Tantras, we get a few more specifics:37
oṃ kurukulle svāhā. This vidyā is to be written on a sheet of birch bark on the door of a house, facing out. She drives off a snake. By facing inward, it would enter again.
Figure 4.2 Kurukullā yantra
Ḍalhaṇa gives us several more references to this Kurukullā in his twelfth-century commentary on the Suśrutasaṃhitā. Regarding 1,46.447 on kings avoiding poisoned food, he says that the mantras used to purify food refers to infallible mantras which render the poisoned food harmless such as those of Kurukullā and Bheruṇḍā.38 Commenting on 5,5.9, he again mentions these two as exemplary of antivenom mantras that he thinks the root text is referring to, but notes that he will not give them since they are taught in other works. He mentions Kurukullā a third time in his commentary on 5,5.51. All of this points to her prominence as an antivenom and anti-snake goddess in medieval Hindu India.
The nineteenth Chapter of the Gāruḍa Purāṇa is a sort of hyper-condensed Gāruḍa Tantra in thirty-five verses. Verses 14–17 are on Kurukullā, and they are so similar to the six verses on Kurukullā in the Saṃhitāsāra, that it makes one suspect that to be the Purāṇa’s source, although it is also possible that each draws on a third source text.
The Bhairavapadmāvatīkalpa, which I mentioned before regarding Tvaritā, also includes a single verse along the same lines:39
A nāga would not linger in a house where a Gāruḍika (nāgāribandha) has written the Kurukullā vidyā in the middle of a six-cornered diagram.
I think we need not make much of the fact that this is a Jain text – we need not posit a separate Jain Kurukullā – because so much of this text is drawn from Śaiva sources. Nor do we need to detail all of the other texts which refer to such a yantra, but they include Uḍḍāmareśvaratantra 15.1, verse 121 in the unpublished Yogaratnāvalī of Śrīkaṇṭhapaṇḍita, and Śāradātilaka 24.8.
All of these references from texts from the ninth century up to the sixteenth century know Kurukullā as an independent Śaiva goddess whose vidyā may be used against snakes and poison. But this is not her only Śaiva identity. Like Tvaritā, she is also found in many texts as an attendant goddess to another deity. That the situation was complex is evident when we read Śaktisaṃgamatantra 3,14 where the text emphasises that the Kurukullā it teaches is different than the one in the Śrīvidyā tradition. There she is only an ancillary (aṅga) of Kālī, but here she is a mahāvidyā and a nityā of Kālī.40 Skimming the chapter, it is clear that this Kurukullā has no associations with snakes or poison, but is rather used predominantly for love magic. Going back to the older Śrīvidyā text Tantrarājatantra, we get a 101-verse Chapter (22) on Kurukullā that also focuses predominantly on love magic. It does, however, also mention the yantra to drive off snakes and a few other antivenom-type usages.
Buddhist Kurukullā
The Kurukullā of Tantrarājatantra 22 has a suspicious number of Buddhist features. Her encoded ten-syllable vidyā (oṃ tāre tuttāre ture svāhā) in this chapter is identical to that of Kurukullā in a number of early Buddhist Tantras such as the Kurukullākalpa and the Guhyasamāja. I have not seen this vidyā in any other Śaiva sources. So does this make a Buddhist origin likely? Perhaps, but not necessarily. Wiesiek Mical, a doctoral candidate at the University of Hamburg, is writing his dissertation on Kurukullā, primarily from Buddhist sources. His deep research has also led him to the aforementioned Śrīvidyā and Kālīkula sources, and he has explored the origins of Kurukullā at length. Nevertheless, he remains uncertain as to the tradition in which Kurukullā first arose, and he was only partially aware of her Gāruḍa tantric identity. Chronologically, the Buddhist sources appear to be earlier, but issues in dating most of these texts raise significant doubts. We eagerly await Mical’s forthcoming dissertation.
The early Buddhist Saṃmitīya sect had a subdivision known as Kurukullaka as early as the second century AD. This Theravāda sect bore no resemblance to Mahāyāna or Vajrayāna Buddhism and may have even burned Tantric Buddhist scriptures in the medieval period.41 According to Mical, we have sources linking the name of this sect to a mountain as well as other sources linking Kurukullā to a mountain in Gujarat. Thus, although Mical finds the evidence tenuous, we could theoretically have a tradition of the goddess originating from a mountain independently from the earlier orthodox sect of the same name.42
The main source for the Buddhist Kurukullā cult was the Kurukullākalpa. This seems to be the source of several Kurukullā meditations in the Sādhanamālā, and could perhaps even be older than the Hevajratantra. While I have not read this text in detail, I can say that it focuses on using Kurukullā for both healing poison and for love magic. We also see the Kurukullā yantra to be placed on a door to drive out snakes in verses 4.23–24. On the other hand, Mical sees Kaula influence in some of its chapters, although the chronological feasibility keeps him rightly doubtful.43
To conclude, the question of Kurukullā’s roots are far from solved and depend heavily on relative dating of various under-studied texts and traditions on both the Buddhist and Śaiva sides. It is also possible that the influences went both ways, in which case the relevant question is how the traditions influenced each other and not which came first.
Bheruṇḍā
The name of the goddess Bheruṇḍā immediately conjures avian imagery. According to K. N. Dave, the bheruṇḍa bird was either a bearded vulture, adjutant stork, or dodo (1985: 397–399). The first two seem plausible because of their enormous size and strikingly fierce appearance. The bearded vulture can have a wingspan of up to ten feet, and the adjutant stork over eight feet. Hemacandra’s lexical Anekārthasaṃgraha suggests that it may have referred to both birds: ‘The word bheruṇḍa refers to two fierce birds, [while] bheruṇḍā is a specific deity.’44 We may note in passing the two-headed bird named Gaṇḍabheruṇḍa, associated with Viṣṇu’s Narasiṃha incarnation and part of the official seal of the state of Karnataka, although I see no connection between this mythical bird and the goddess Bheruṇḍā aside from the name.
Features in the early references
As with Kurukullā, it is useful to distinguish the stand-alone snakebite-goddess of the Gāruḍa Tantras and dependent literature from her identity as an ancillary goddess in other Śākta sources. I have already mentioned several texts in connection with Tvaritā and Kurukullā that also feature Bheruṇḍā: the Saṃhitāsāra, Haramekhalā, and Ḍalhaṇa’s commentary to several Suśrutasaṃhitā passages all seem to refer to the independent snakebite goddess. We also have references to her in the Rasaratnākara’s toxicology (viṣacikitsā) section, Yogaratnāvalī 122, and Bhairavapadmāvatīkalpa 10.12. Without exception the passages are brief, sometimes only half a verse. They say that the Bheruṇḍā spell should be chanted in the ear of a bite victim to free him of the venom. With the exception of Ḍalhaṇa, who declines to give the spell directly, all of these sources or their commentaries teach a Prakrit spell for Bheruṇḍā. The longest version is that following Saṃhitāsāra 151, which runs for five lines in my edition. The others are briefer; the Rasaratnākara version is only one line, but clearly based on the same Prakrit spell. In Table 4.2 that follows, I give the spell as found in each of the five sources, though I abbreviate that of the Saṃhitāsāra:
Table 4.2 Prakrit spell teachings for Bheruṇḍā
|
Saṃhitāsāra |
saṃ joe mpae bheruṃḍāe vi_bhariakaraṃḍāe taṃta maṃta visa āhosaï jaṃbhaï mohaï thāvara jaṃgama kiṃtima jaja jāhi re jaja jāhi re mahāpasāu bharāḍīe haru visa karu ṇivvisu hūṃ hūṃ / … [continues] |
|
Haramekhalā |
oṃ joe māe bheruṇḍāe vijjābhariakaraṇḍāe mantaṃ suṇuha jaha ghosaï hikkāraï taha visu ṇāsaï thāvarajaṃgamao thaṃbhaï jaṃbhaï mohaï jāhi re jāhi re |
|
Yogaratnāvalī |
oṃ joyasāsaṇe bheruṇḍe vijjahabhariśca karaṇḍe taṃtu maṃtu āghosaï phekarat viśu nāśayi thāvarajaṃgama māre stambhaṇe mohaṇa jāhire gara jāṃ jaḥ |
|
Bhairavapadmāvatī-kalpa |
oṃ ekahi ekamāte bheruṇḍā vijjābhavikajakaraṇḍe taṃtu maṃtu āmosaï huṃkāra viṣa nāsaï thāvara jaṃgama kittima aṃgaja oṃ phaṭ |
|
Rasaratnākara |
oṃ eha mātra bheruṇḍe aïūṃ bījaṃ bhaviakaraṇḍe tantra mantra agdoṣa īn hūṃkāre viṣa nāśaï sthāvarajaṅgameti manhukaï |
Clearly the vidyā is in need of editing, but I do not feel confident enough to fix it. The Haramekhalā version seems more coherent in some respects. I partially translate:
oṃ joe Mother Bheruṇḍā whose basket is filled with spells (?), listen to the mantra! As you cry out, screech, so must you destroy the poison, be it from a plant or animal – terminate it! Destroy it! Make it fail! Go! re!, Go! re!
A Gāruḍapurāṇa 19 reference sounds like our stand-alone Bheruṇḍā, but the spell is not in Prakrit: OṂ HRĪ HRAU HRĪṂ BHIRUṆḌĀYAI SVĀHĀ.
Other Śākta identities
We find many other texts using Bheruṇḍā as an attendant goddess. The Tantrarājatantra has Bheruṇḍā as a nityā alongside Tvaritā and Kurukullā. In verses 3.35–37, the text gives her nine-seed-syllable spell in code working out to: OṂ KROṂ BHROṂ KROṂ JHROṂ CHROṂ JROṂ SVĀHĀ. It is completely different from the Prakrit vidyā of her independent identity, but here too she is said to be able to destroy the three types of poison, but rather than chanting in the ear of the victim, it need only be recalled by the initiate. The Tattvacintāmaṇi has a similar series of syllables for its Bheruṇḍā nityā. In the Brahmāṇḍapurāṇa, the Viśvasāratantra, and the Matsyendrasaṃhitā, she is listed as one of fifteen or sixteen nityā goddesses, but no details on her form or mantra are given. In the Matasāra, Bheruṇḍā is one of eight goddesses of speech (vāgeśvarī).
Conclusion
Tvaritā, Bheruṇḍā and Kurukullā each have multifaceted identities in the Śaiva and Śākta literature. We have separate literature, visual forms, spells, and identities in the early Gāruḍa Tantra material and in the early Tripurasundarī literature. Which came first is difficult to prove, but the fact that the Gāruḍa Tantras give no hint that the goddesses are borrowed from another system is suggestive. On the other hand, the Tripurasundarī literature, and later Śākta systems that drew on it, frequently mention the ability of these goddesses to heal poison and drive away snakes. I close with one final passage from the Śrīvidyārṇava:45
May the vidyā who is called Suvarṇarekhā, the one said to be an eradicator of snakes, give ease to me. May the mighty-looking vidyā called Kurukullā, arisen from the mouth of the Lord of Birds, always be present on the tip of my tongue. May the one called Jhaṃkāriṇī always be present in my body. [May the one] named Remover of Poison be a cleaver to the form of the Kali age. May Bheruṇḍā always be present in my throat. May Totalā be present in my head. And likewise may Suvarṇarekhā also always be present at my base. Let Jāṅgulī make my speech perfect for the destruction of poison.
Notes
1 Vidyās are both spells, the female equivalent of mantras, and goddesses. The sonic spell was understood as the embodiment of the goddess.
2 See Slouber 2011 for an edition and translation of the ninth-century Saṃhitāsāra passage on Jhaṃkāriṇī.
3 Predominantly known as a Buddhist goddess from the sādhana literature, she was also known as a Śaiva goddess. Asha Archives manuscript number 3152 contains a Śaiva Jāṅgulīvidyā on folios 5 recto to 8 verso. We also see a Śaiva passage on Jāṅgulī in the Tantrasadbhāva passage starting with 23.296.
4 See the aforementioned Saṃhitāsāra (Slouber 2011) for a passage on Suvarṇarekhā.
5 The spelling of this title varies. We also see it as Trotala, Trotula, Totula, Totala, and Troṭala.
6 Mark Dyczkowski 2009: 83–85 (vol. 2).
7 See Bhairavapadmāvatīkalpa 1.3 in Jhavery 1944.
8 The colophons of the Kulakaulinīmata ascribe it to the Caturviṃśatisahasra (i.e. the Manthānabhairavatantra). Thanks to Mark Dyczkowski for sharing his draft edition of the Tvaritā chapter of this work.
9 Microfilmed in NGMPP A59/13. Folios 16–34 carry the Tvaritāvidhānasūtra passages.
10 The crucial verse on the topic of Gaurī/Tvaritā’s identity is: Tvaritāmūlasūtra 1.37: tataḥ kruddhā tu sā gaurī taḍitkoṭisamaprabhā / tvaritaṃ tvaritā vidyā tribījā triguṇīkṛtā // ‘And then Gaurī was enraged, with a fiery splendour equal to millions of lightning bolts. Instantly (tvaritaṃ) she [uttered] the Tvaritā vidyā, consisting of three times three seed syllables.’ The verse has no verb, so inserting ‘uttered’ is an interpretation. We could alternatively supply ‘became’, but that is also a theologically loaded assumption.
11 Tvaritāmūlasūtra 1.45–56, ff.5v–6v (Here and henceforth I cite folio numbers for manuscript H170/3). I offer the following provisionally edited Sanskrit text, but the grammar is often highly non-standard, sometimes to preserve the meter: tvaritā tatra cārūḍhā vāmajaṅghā *pralambitā(conj., pralambitāṃ Cod.) / dakṣiṇā dviguṇā tasyāḥ pādapṛṣṭhe samarpitā // sphurantārakavad devyā bhrājante pādayor nakhāḥ / ambhojāruṇa*varṇābhāḥ (conj., -varṇābhā Cod.) padmalāñchana*lakṣaṇāḥ (conj., -lakṣaṇā Cod.) // rājate caraṇādhastāt susamāṅguli*–m–āyatā(conj., -māpatā Cod.) / suvṛtaṃ jaṅghapīnoru *vistīrṇa (conj., vistīrṇe Cod.) jaghanoru sā // gambhīra*nābhis (conj., -nābhi Cod.) trivalī romarājīsuśobhitā / vistīrṇā kaṭideśe tu hemakumbhapayodharā // kambugrīvā tu bimboṣṭhī nīlotpaladalekṣaṇā / sunāsā cāpabhrūbhaṅgā sukarṇā *mukuṭojjvalā (conj., makuṭojvalā Cod.) // *visphuranmaṇibhir (conj., visphuretmaṇebhir Cod.) *dīptaḥ (conj., dīptaiḥ Cod.) sahasraphaṇidīptimān / nāgarājakṛtohārakarṇakuṇḍalabhāsinī // nāgabandhakṛtā śobhā *bhujau (conj., bhujo Cod.) devyā virājate / mekhalā kaṭideśe tu mahāhikṛtabhīṣaṇā // pādayor *nūpurau (conj., nupurau Cod.) ghorau devyāḥ kṛtabhujaṅgamau / sahasradale cāmbhoje vāmapādakṛtodarā // viyattaḍillatā*bhāsaṃ (conj., -bhāsāṃ Cod.) vāsaḥ kusuma*śobhitam (conj., -śobhitāṃ Cod.) / indrāyudhanibhaṃ cānyaṃ *vāsaḥ (conj., rāsaṃCod.) kṛtapayodharau // anekapuṣparacitaṃ dṛḍha*bandha (conj., -vadha- Cod.) mahoragam / kṛtāyudhakarā devī tīkṣṇāgranakha*śuktayaḥ (conj., -muktayaḥ Cod.) // vajradaṇḍāsicakraṃ ca gadā śūlaṃ *mahojjvalam (conj., mahojvalaṃ Cod.) / śaraṃ śaktiś ca varadaṃ dakṣiṇena kṛtāyudhā // *dhanuṣ (conj., dhanu- Cod.) pāśa*dharaṃ (conj., -haraṃ Cod.) ghaṇṭā tarjanī śaṅkham aṅkuśam / abhayaṃ ca tathā padmaṃ vāmapārśve kṛtāyudhā //
12 Dyczkowski 2009: 88–89.
13 Following verse 46, ff.4v–5r in manuscript A59/15.
14 Tvaritājñānakalpa A59/15, f.2v: (oṃ) hūṃkāradvayasaṃyuktaṃ khe ca cche padabhūṣitam / vargātītaṃ visargaś ca strīṃ hūṃ kṣe phaṭ ca vai smṛtāḥ // 24 // Taken literally, we might assume that hūṃkāradvayasaṃyuktaṃ means that the vidyā begins hūṃ hūṃ, but I think it is rather just indicating that the entire vidyā will have two hūṃ syllables. The oṃ at the beginning does not fit the meter and may have been added later.
15 The text divides the Sanskrit syllables into eight groups, as is typical, but the names of several vargas appear to be peculiar to this text and the Agnipurāṇa passage that draws on it. The verses preceding the vidyā clearly list the groups in order: svaravarga, tāluvarga, jihvatāluka, tālujihvāgra, jihvadanta, oṣṭhapuṭa, miśravarga, and ūṣmāṇa. In manuscript H170/3 it is ff.5v–6v (AP=Agnipurāṇa 310 parallel, which I only report when I accept its reading over ours): ṣaṣṭhasvarasamārūḍhām *ūṣmāṇāntaṃ (AP, ūṣmāṇānta- Cod.) sabindukam / mūlavidyādikaṃ bījaṃ kathitaṃ tu khageśvara // tāluvarga*dvitīyaṃ (AP., -dvitīyas Cod.) tu svaraikādaśayojitam / jihvātālu*samāyoge (AP, samāyogaḥ Cod.) prathamaṃ kevalaṃ bhavet // tad eva taddvitīyaṃ tu adhastād viniyojayet / ekādaśasvarair yuktaṃ prathamaṃ tāluvargataḥ // ūṣmāṇasya dvitīyaṃ tu adhastāt tasya yojayet / ṣoḍaśasvarasaṃyuktam ūṣmāṇasya tṛtīyakam // jihvādantasamāyoge prathamaṃ yojayed adhaḥ / miśravargadvitīyaṃ tu adhastāt punar eva ca // caturthasvarasambhinnaṃ tāluvargādimaṃ punaḥ / ūṣmāṇasya dvitīyaṃ tu adhastād viniyojayet // svaraikādaśasambhinnam ūṣmāṇāntaṃ sabindukaṃ / pañcamasvara–m–ārūḍham oṣṭhasampuṭayogataḥ // dvitīyam akṣaraṃ cānyaṃ jihvāgre tāluyogataḥ / prathamaṃ *yac ca saṃyojyaṃ (conj., pañcame yojyaṃ Cod.) svarārdhenoddhṛtā imā // tvariteyaṃ mahāvidyā sarvasiddhipradāyikā / oṃkārādisamāyuktā–m–ante namo japet sadā // svāhāntamagnikāryeṣu…
16 I assume we should also supply tvaritāyai before namaḥ/svāhā.
17 Folio 13v: ādidvihṛdayaṃ proktaṃ tricatuḥ śiramiṣyate // pañcaṣaṣṭhaśikhā proktā kavacaṃ saptamāṣṭakam / *tārakā (conj., tārakāṃ Cod.) tu bhavennetraṃ navārdhākṣara*lakṣaṇam (AP, rakṣaṇam Cod.) //
18 The context makes it clear that tārakā refers to phaṭ, but we may also point to the similar sounding code word for phaṭ found in Dakshinamurti’s Uddhārakośa: turaga.
19 See ff. 62v–63v.
20 ff. 26r–v: ekayā caiva āhutyā dīkṣito *bhavate (conj., bhavete Cod.) khaga / adhikāro bhaved *evaṃ (conj., devaṃ Cod.) mokṣaṃ śṛṇu ataḥparam //
21 The text gives this figure itself. It is actually around 150 plus prose, but the traditional way of measuring the length of a verse-text is by counting the average syllables per line, which would include the prose too.
22 For the following words I deviate from H170/3: pralambitā, B126/9, pralambitāṃ H170/3; dhanuṣpāśadharaṃ B126/9, dhanupāśaharaṃ H170/3; iṣṭasattvaparāṅmukhāḥ B126/9, duṣṭasattvā parāṅmukhāḥ H170/3; trakārāt conj., oṃkārāt B126/9, aiṃkārāt H170/3; sarvāṃs trāsateB126/9, sarvān trāśate H170/3.
23 Here are the words that I take from the Agniopurāṇa etext (‘APe’) rather than Joshi’s printed edition: samarpitā APe, samīpsitā Joshi; oṃkārādisvarārabhya APe, oṃkārād īśvarād ārabhya Joshi. It is evident from citations of these chapters by other scholars that the transmission of the Agnipurāṇa includes a great deal of variation.
24 Although the Tvaritāvidhānasūtra is put in the mouth of Śrīvakrā (Kubjikā), she and her interlocutor Śrīkaṇṭha reference the ‘Trottalāmata’ as a source and the material appears more closely aligned to the Tvaritāmūlasūtra and Tvaritājñānakalpa than other Kaubjika material I have seen.
25 Bühnemann 2000: 207 (vol. 2).
26 Dyczkowski 2009: 89 (vol. 2).
27 Dyczkowski 2009: 88, citing Mallmann 1963: 160.
28 Ibid. 88.
29 Dyczkowski, personal communication.
30 Jhavery 1944: 373 (p. 1 of the edition’s pagination): totalā tvaritā nityā tripurā kāmasādhanī / devyā nāmāni padmāyās tathā tripurabhairavī //
31 http://ncbinfotech.appspot.com/tulja/tulja.php?pageid=TV20.
32 www.celextel.com/tvarita-yantra-320-mm-p-916.html.
33 www.shivashakti.com/nitya.htm.
34 Meulenbeld 1999; vol. IIA: 134.
35 Ibid. vol. IIB: 151.
36 Edition of Bhattarai 1972: 5–6.
37 Kriyākālaguṇottara 35, mantra seventeen: oṃ kurukulle svāhā / bhūrjapatre iyaṃ vidyā lekhyā gṛhadvāre parāṅ-mukham // sarpam uccāṭayati / saṃmukhena punaḥ praviśati /
38 siddhair avyabhicāribhiḥ kurukullābheruṇḍāprabhṛtibhir hataviṣam annam iti sambandhaḥ /
39 Verse 10.41 in Jhavery 1944: ṣaṭkoṇabhavanamadhye kurukullāṃ yo likhed gṛhe vidyām / tatra na tiṣṭhati nāgo likhite nāgāribandhena // With Bandhuṣeṇa’s commentary: ‘ṣaṭkoṇabhavanamadhye’ ṣaṭkoṇacakramadhye / ‘kurukullāṃ' kurukullānāmadevyā mantraḥ / ‘yo likhed’ yaḥ ko ’pi mantravādī likhet / kva? ‘gṛhe’ gṛhadehalyām, svavāsottarāṅge / kām? ‘vidyām’ kurukullādevyā vidyām / ‘tatra’ tasmin gṛhe / ‘na tiṣṭhati’ na sthāti / kaḥ? ‘nāgaḥ' sarpaḥ / kasmin kṛte sati? ‘likhite’ sati / kena? ‘nāgāribandhena’ garuḍabandhena // mantraḥ––oṃ kurukulle hūṃ phaṭ //
40 Śaktisaṃgamatantra, Sundarīkhaṇḍa, Chapter 14, verses 8–9. Thanks to Wiesiek Mical for pointing out this chapter to me and sharing his draft translation of it.
41 According to Tārānātha’s History of Buddhism in India, translated by Chattopadhyaya 1970: 279.
42 Personal communication.
43 Personal communication.
44 3.173cd: bheruṇḍau bhīṣaṇakhagau bheruṇḍā devatābhidi //
45 The numbering in the electronic edition is odd, but it cites it as p. 342 of the 1947 Shrinagar edition: suvarṇarekhiṇī proktā vidyā yā procyate kila // nirmūlinī bhujaṅgānāṃ sā karotu sukhaṃ mama / kurukulleti vikhyātā pakṣirājamukhodbhavā // yā vidyā sā mahārūpā jihvāgre sthātu me sadā / *jhaṃkāriṇīti (conj., oṃkāriṇīti Cod.) vikhyātā dehe sthātu sadā mama // *viṣāpahāriṇī (conj., vidyāpahāriṇī Cod.) nāma kalirūpavidāriṇī / bheruṇḍā sthātu me kaṇṭhe *totalā (conj., toralā Cod.) sthātu mastake // tathā *suvarṇarekhāpi (conj., śavalarekhāpi Cod.) mūle sthātu sadā mama / *jāṅgulī (conj., jāṅgalī Cod.) viṣanāśāya vācāṃ siddhiṃ karotu me //
References
Primary sources
Agnipurāṇa, see Joshi 2001.
Anekārthasaṃgraha, see Zachariae 1893.
Bhairavapadmāvatīkalpa, see Jhavery 1944.
Brahmāṇḍa Purāṇa, GRETIL Etext Archive. URL: http://fiindolo.sub.uni-goettingen.de/gret_utf.htm#BrahmandaP, accessed 20/08/2011.
Caturviṃśatisahasra, see Manthānabhairava.
Gāruḍa Purāṇa, see Sresthi 1906.
Guhyasamāja, see Matsunaga 1978.
Haramekhalā, see Bhattarai 1972.
Hevajratantra, see Snellgrove 1959.
Īśānaśivagurudevapaddhati, see Sastri 1920.
Jāṅgulīvidyā, unedited Asha Archives manuscript #3152.
Jñānārṇavatantra, see Malaviya 2001.
Kriyākālaguṇottara, unedited. See Slouber 2007.
Kulakaulinīmata, unedited. See Dyczkowski 2009.
Kurukullākalpa, see Pandeya 2001.
Manthānabhairava, see Dyczkowski 2009.
Matasāra, unedited, see NGMPP reel B 28/16.
Matsyendrasaṃhitā, see Sen Sharma 1994.
Nārāyaṇa’s Tantrasārasaṃgraha, see Aiyangar 1950 and Slouber 2009.
Niśvāsamūlasūtra, see Goodall (forthcoming).
Pārameśvarīmata, Muktabodha Digital Library. URL: http://muktalib5.org/digital_library.htm, accessed 20/08/2011.
Prapañcasāra, see Bühnemann 2000.
Rasaratnākara, see Srikrsnadas 1909.
Sādhanamālā, see Bhattacharya 1968.
Śaktisaṃgamatantra, Muktabodha Digital Library. URL: http://muktalib5.org/digital_library.htm, accessed 20/08/2011.
Saṃhitāsāra, see Slouber 2011.
Śrīvidyārṇava, see Sharma 1966.
Suśrutasaṃhitā, see Bhattacaryya 1891.
Tantrarājatantra, see Sastri and Woodroffe 1918.
Tantrasadbhāva, unpublished edition of Dyczkowski based on NGMPP reels A 188/22, A 44/1, and A 44/2.
Tattvacintāmaṇi, see Bhatt 2005.
Trottala, see Slouber (forthcoming).
Trottalottara, see Slouber (forthcoming).
Tvaritājñānakalpa, unedited, see Slouber (forthcoming) and NGMPP reels A 59/15, B 26/14, and B 126/15.
Tvaritāmūlasūtra, unedited, see Slouber (forthcoming) and NGMPP reels B 126/9, H 170/3, and C 6/7.
Tvaritāvidhānasūtra, unedited, see NGMPP reel A 59/13.
Uḍḍāmareśvaratantra, see Zadoo 1947.
Uddhārakośa, see Taki 1978.
Viśvasāratantra, Muktabodha Digital Library. URL: http://muktalib5.org/digital_library.htm, accessed 20/08/2011.
Yogaratnāvalī, unedited, see NGMPP reel A 210/10 and elsewhere. Manuscripts widespread.
Secondary literature
Aiyangar, M.D. 1950. Tantrasarasangraha. Madras: Government Oriental Manuscripts Library.
Bhatt, V.P. 2005. Tattvacintamani. Delhi: Eastern Book Linkers.
Bhattacaryya, J.V. 1891. Nibandhasamgrahkhya susrutasamhitaya vyakhya. Calcutta: Sarasvati Press.
Bhattacharya, B. 1968. Sadhanamala. Baroda: Oriental Institute.
Bhattarai, K. 1972. Haramekhalā saṭīkā. Kathmandu: National Archives.
Bühnemann, Gudrun. 2000. The Iconography of Hindu Tantric Deities, Volume II. Groningen: Egbert Forsten.
Chattopadhyaya, A., Chattopadhyaya, D. and Chimpa, L. 1970. Taranatha’s History of Buddhism in India. Shimla: Indian Institute of Advanced Study.
Dave, K.N. 1985. Birds in Sanskrit Literature. Delhi: Motilal Banarsidass.
Dyczkowski, M. 2009. Manthānabhairavatantram. New Delhi: Indira Gandhi National Centre for the Arts.
Goodall, D. Forthcoming. The Niśvāsatattvasaṃhitā: The Earliest Surviving Śaiva Tantra. École française d’Extrême-Orient and Nepal Research Centre.
Jhavery, M.B. 1944. “Bhairavapadmāvatīkalpa”. In Comparative and Critical Study of Mantrasastra with Special Treatment of Jain Mantravada, being the Introduction to Sri Bhairava Padmavati Kalpa, pp. 367–450. Ahmedabad: Sarabhai Manilal Nawab.
Joshi, K.L. 2001. Agnimahāpurāṇam. Delhi: Parimal Publications.
Malaviya, S. and Malaviya, R. 2001. Isvaraproktam Jñanarnavatantram. Varanasi: Krsnadasa Akadami.
Mallmann, M.T. 1963. Les Enseignements Iconographiques de l’Agni Purana. Paris: Presses Universitaires de France.
Matsunaga, Y. 1978. The Guhyasamaja Tantra: A New Critical Edition. Osaka: Toho Shuppan.
Meulenbeld, G.J. 1999. A History of Indian Medical Literature. Groningen: E. Forsten.
Pandeya, J.S. 2001. Kurukullakalpah = phags-ma sgrol-ma kurukullei rtog pa. Varanasi: Durlabha Bauddha Grantha Sodha Anubhaga, Kendriya Ucca Tibbati Siksa Samsthana.
Sastri, L. and Woodroffe, J.G. 1918. Tantrarājatantra. London: Luzac & Co.
Śāstri, M.M.T.G., ed. 1920–25. Īśānaśivagurudevapaddhati. 4 vols. Trivandrum Sanskrit series 69, 72, 77, 83. Trivandrum: Superintendant, Government Press. Reprint, Delhi and Varanasi: Bharatiya Vidya Prakashan, 1988.
Sen Sharma, D. 1994. Matsyendra Samhita. Calcutta: Asiatic Society.
Sharma, B. 1966. Sri Srividyarnava Tantram. Prayaga: Kalyana Mandira.
Slouber, M. Forthcoming. “Introduction, Edition, and Translation of the First Chapter of the Tvaritāmūlasūtra”. In Vasudeva, Somadeva, Śivasudhāprāpālikā.
Slouber, M. 2011. ‘Śaṅkuka’s Saṃhitāsāra: Edition and Translation of Verses 1–5, 75–85, and 129–154 with an Anonymous Commentary’. Master’s thesis, University of Hamburg.
Slouber, M. 2009. E-text of Nārāyaṇa’s Tantrasārasaṃgraha. Hamburg: Centrefor Tantric Studies. URL: http://tantric-studies.org/e-texts/shaiva/NaTaSaSa.txt, accessed 20/08/2011.
Slouber, M. 2007. ‘The Manuscripts of the Kriyākālaguṇottara’. Newsletter of the NGMCP, 5, 9–16. URL: www.uni-hamburg.de/ngmcp/nl5light_e_e.pdf, accessed 20/08/2011.
Snellgrove, D.L. 1959. The Hevajratantra: A Critical Study. Oxford: Oxford University Press.
Sresthi, K. 1906. Garuda Mahapurana. Mumbai: Sri Venkatesvara Steam Press. URL: http://reader.digitale-sammlungen.de/resolve/display/bsb10211126.html, accessed 20/08/2011.
Srikrsnadas, K. 1909. Rasaratnākara. Mumbai: Sri Venkatesvar Steam Press.
Taki, S.1978. Sri-Daksinamurti-viracita Uddharakosah. New Delhi: Munsiram Manoharalal.
Zachariae, T. 1893. Der Anekarthasamgraha. Vienna: A. Holder.
Zadoo, J. 1947. The Uddamaresvara Tantram. Srinagar: Normal Press.