Notes

n.1Caṇḍa­mahā­roṣaṇa­tantram (Toh 431).

n.2The exception to this statement is chapter 15, “An Account of Royal Lineages,” in which Mahākāla provides an account of the rise and fall of various royal lineages, both human and nonhuman, across the Indian subcontinent and beyond.

n.3The female beings who encircle Mahākāla in his maṇḍala are referred to interchangeably as either goddesses (devī, lha mo) or yoginīs (rnal ’byor ma).

n.4The Tantra of Glorious Mahākāla (Śrīmahākāla­tantra, Toh 667).

n.5Vajra­mahākāla­krodhanātharahasya­siddhi­bhava­tantra (mgon po gsang ba dngos grub byung ba’i rgyud, Toh 416).

n.6The Dhāraṇī of Glorious Mahākāla (Śrī­mahākāla­nāma­dhāraṇī, Toh 668).

n.7The Mahākāla Dhāraṇī: A Cure for All Diseases and Illnesses (nag po chen po’i gsungs rims nad thams cad las thar byed, Toh 669).

n.8dpal nag po chen po’i rgyud drag po’i brtag pa dur khrod chen po zhes bya ba’i ’grel pa (Toh 1753).

n.9Sanderson 2009, p. 45–50.

n.10See 3.­2.

n.11Stablein 1976, pp. 85–88.

n.12Stablein 1976, pp. 90–91. All the Newar vajrācāryas consulted by Stablein in the early 1970s noted that their traditions are based on a twenty-nine-chapter recension of The Glorious Sovereign Tantra of Mahākāla.

n.13Stablein 1976, p. 7. Stablein does not specifically identify the text that he saw being used as part of a “book offering” (pustakapūja) in Tuṇḍikhel as the twenty-nine chapter version Mahākala­tantra­rāja, most likely because he was not allowed to examine it closely.

n.14Both Sanskrit witnesses for The Glorious Sovereign Tantra of Mahākāla in the Bibliothèque nationale de France are signed by Eugène Burnouf and dated to the year 1845. BnFS 85, which does not provide a clear notation of the date it was completed, is nevertheless clearly not a very old manuscript, given that its orthography is perhaps the closest of all to modern Devanāgarī. Unfortunately, neither of the UT manuscripts contain any mention of their specific date of composition.

n.15For Sāṅkṛtyāyana’s own account of his documentation of this and other important Sanskrit manuscripts in Tibet, see Sāṅkṛtyāyana 1935, pp. 21–43 and Sāṅkṛtyāyana 1937, pp. 1–57. For an updated and complete catalogue of these sources see Bandurski 1994.

n.16RST15 reads paramasaugata­parama­bhaṭṭārakamahā­rājā­dhirāja­śrī­manmadana­pāla­deva­pādīya saṃvat 5. The copy of this manuscript made available to me was unfortunately missing the final plate containing the last two or so folios of text, so this colophon reading is taken from Herr Gustav Roth’s transliteration in Bandurski 1994, p. 47.

n.17Bandurski cites Dines Chandra Sircar, “Three Inscriptions from Valgudar,” Epigraphica Indica 28 (1958): 145, note 3.

n.18The first four chapters in RST15 do not correspond to the other Sanskrit witnesses or to the first four chapters of the Tibetan translation. RST15 begins to align with both the Sanskrit and Tibetan witnesses toward the end of chapter 4, and the correspondence between all three continues through the end of chapter 14. RST15 again diverges from the other witnesses through chapter 28. Chapters 29 through the majority of chapter 32 are unfortunately missing in Sāṅkṛtyāyana’s photographs, but it is clear that the material in chapter 33 of RST15 also diverges from the Tibetan. RST15 then aligns with the Tibetan again in chapter 34. The readings in our copy of RST15 for chapters 35 and 36 were unfortunately not very clear. Chapter 35 does not have a title in RST15, but the title of chapter 36 does match the Tibetan. The readings of these chapters in our copy of this witness are also unfortunately not very clear, and, as noted above, the last few folios of the text are missing from our copy as well. As a result, we have not closely consulted the material in chapters 35 through 50 of RST15, and this remains a topic for future research.

n.19Following Tib. ye shes su. The Sanskrit witnesses omit this.

n.20Following Tib. rgyu ba gcod pa la spyod pa. ST reads utsāha­yantre caret. UTM 286 reads utsāha­yantracaret. UTM 288 reads utsaho yantre caret. ND 44-5 reads utsāho yantre recaḥ. RASH 47 and BnFS 85 read utsāho yantra caret.

n.21Following F lus brtan pa’i dgos pa gang lags. D reads lus bstan pa’i dgos pa gang lags. ST reads śarīradṛḍhena kim prayojanam. UTM 286, ND 44-5, and RASH 47 read śarīradṛḍheṇa prayojanaṃ. UTM 288 reads śarīradṛḍheṇa prajojanaṃ. BnFS 84 reads śariraṃ dṛdhena prayojanaṃ. BnFS 85 reads śarīra dheṇa prayoyanaṃ. The reading śarīradṛḍha in the Sanskrit witnesses is supported by the reading lus brtan pa in the Phukdrak Kangyur.

n.22Following ST darśayitukāmāya. Tib. reads lta ba’i don du ’dod chags kyi phyir.

n.23Following H and N rgyu mar sems dpas. D reads rgyu mar sems pas. BnFS 85, RASH 47, ST, UTM 286, and UTM 288 read mayāsattvaº.

n.24Following BnFS 85, RASH 47, ST, UTM 286, and UTM 288 helayā. The Tibetan witnesses read tshogs med par. The reading helayā from the Sanskrit witnesses suggests that the Tibetan reading tshogs med par is a corruption of tshegs med par.

n.25Translation tentative. ST reads praghatita­khurara­vapātho, and D reads rmig pa’i sgra ltar brlag pa.

n.26Following C, J, K, S, and Y nas. D reads na.

n.27While the term kāla is often translated in English as “black” (Tib. nag po), this explanation of Mahākāla’s name reflects the fact that the Sanskrit term kāla can also mean “time,” and thus one way of interpreting the term mahākāla is “a great amount of time.”

n.28D and S read rnal ’byor gyi dbang phyug rnams kyis ji ltar rigs pas. N and Y read rnal ’byor gyi dbang phyug rnams kyis ji ltar rig pas. The Sanskrit witnesses read yogeśvarīṇaṃ yathānyāyaṃ. The translation “queens of the yogas” is informed by the Sanskrit witnesses. The Tibetan translation does not preserve the gender of this compound.

n.29Following F, H, K, N, S, and Y dus kyi gtso bo. D reads dus kyis gtso bo, and Skt. reads mulakālo.

n.30Following K, S, and Y be tA li. D reads bai tA li, and the Sanskrit witnesses read vetālī.

n.31Throughout this translation, it is assumed that the opening statements in each chapter that are marked as first-person speech in the Sanskrit are the words of the same Blessed One who spoke in the first chapter.

n.32Following S sa sbyang ba. D reads las sbyang ba. The Sanskrit witnesses read bhūmiṃ śodhayet. This translation follows the reading in S, which allows us to correct a minor scribal error in D and is supported by the Sanskrit witnesses as well as numerous points at which this same phrasing occurs in this text.

n.33Following S sa sbyang ba. D reads las sbyang ba. The Sanskrit witnesses read bhūmiṃ śodhayet. See n.­32.

n.34The Tibetan reads zhi ba’i sngags, and the Sanskrit reads śivamantraº.

n.35The mantras for the ten- and twelve-armed forms of Mahākāla are reversed in Sanskrit manuscript ND 44-5.

n.36The Sanskrit amuka, sometimes rendered in Tibetan as che ge mo, is a term marking the point at which one should insert the name of the intended beneficiary or target of a rite.

n.37The phrase “while reciting the following mantra” is not included in the Tibetan or Sanskrit text, but it has been added for clarity.

n.38D and S read dngos grub thams cad ’grub bo/ sgyid snyoms dang bcas par. K and Y read dngos grub thams cad ’grub bo/ skyid snyom dang bcas par. ST, RASH 47, and BnFS 84 read sarvaṃ sidhyati helayā. ND 44-5, UTM 286, UTM 288, and BnFS 85 read sarvaṃ siddhyati | helayā. The Tibetan reading in D and S preserves a scribal error, while the reading in K and Y reflect what we find in the Sanskrit witnesses. All the Tibetan witnesses, as well as ND 44-5, UTM 286, UTM 288, and BnFS 85, construe the phrase “easily” or “with ease” with the subsequent statement, but it is far more likely that it is intended to be read as an adverbial form modifying the verb that precedes it, as Stablein has rendered it in his edition.

n.39The name of this mantra (agnimantraḥ) is only mentioned in the Sanskrit witnesses.

n.40The Tibetan reads de ba da t+taM, and the Sanskrit reads devadatta. Like the phrase amuka (Tib. che ge mo), the name devadatta functions in as a placeholder for the intended target of the rite.

n.41Following Skt. māṁsamāraya. D and S read mAM maM mA ra ya.

n.42Following K and Y tsan De shwa rI and Skt. caṇḍeśvarī . D reads tsaM ha De shwa rI. S reads tsa De shwa rI.

n.43Following Skt. lāṃ kaṃ. D reads lA ke. S reads lA kaM.

n.44Following Tib. dum bu can. The Skt. witnesses do not identify the specific goddess to whom this mantra is addressed, and the mantra itself references Camuṇḍā.

n.45ST reads anena niyutaṃ japtvā śmaśāna­bhasmanā puttalikāṃ kṛtvā tasya dvāre gopayet. The Sanskrit sources preserve a line explaining the use of this mantra that translates as, “If one incants an effigy made of ash from a cremation ground with this mantra many times over, it will protect one’s doorway.” The Tibetan witnesses all omit this line.

n.46This mantra is not included in the Sanskrit witnesses.

n.47Following RASH 47 and BnFS 84 sarvapaśūn. D reads sarba pA shu na. This transliteration follows the readings in RASH 47 and BnFS 84, which preserve the correct Sanskrit spelling.

n.48This mantra does not appear here in the Sanskrit witnesses. In the Sanskrit sources, it follows the sword-paralyzing mantra.

n.49Following Skt. naramukhastam­bhana­mantraḥ. Tib. reads mi’i ngag rengs par bya ba la. This translation follows the Sanskrit witnesses, which include the term mantra. A literal translation of this phrase would be “the mantra for paralyzing a person’s mouth.”

n.50Following F bsgrub pa thams cad la cher dga’ ba’i gzungs and ST, RASH 47, BnFS 85, and UTM 286 sarva­sādhana­mahānanda­dhāranī­mahākālasya mantraḥ. D and S read sgrub pa thams cad la cher dga’ ba’i gzugs nag po chen po’i sngags so.

n.51Following K, Y, F, and S yang dag pa’i rigs pas. D reads yang dag pa’i rig pas.

n.52J, C, and H read thub pa ka pi la bar+Na shi ba. D reads thub pa ka pi la bar+N+Na shi ba. S and N read thub pa ka pi la bar rna shi ba. ST reads munikapilā iva. ND 44-5, UTM 286, RASH 47, and BnFS 85 read muniṃ kanilāmāritair iva. UTM 288 reads muniṃ kanilāmārair iva. BnFS 84 reads muniṃ kanilāmālitair iva. This translation is tentative as the reference to “the tawny-colored one” (D ka pi la bar+N+Na) could not be identified.

n.53D and S read zhi ba’i ye shes sgrub pa. N reads ye shes sgrub pa. ST reads siddhijñāna. ND 44-5 and UTM 286 read siddhijñānam. UTM 288, RASH 47, BnFS 84, and BnFS 85 read siddhajñānam. The Sanskrit witnesses preserve two alternate readings of this compound that translate as either “the wisdom of siddhi” (if we follow ST, ND 44-5, and UTM 286) or “the wisdom they have attained” (if we follow UTM 288, RASH 47, BnFS 84, and BnFS 85). It is also possible to translate the reading preserved in the Tibetan witnesses as “the wisdom of Śiva” and not “quiescent wisdom.”

n.54Or more literally “a woman with crow’s legs” (Tib. bya rog rkang ma, Skt. kākajaṅghā).

n.55A similar but not identical set of women is listed in, for example, Hevajra Tantra 1.5.2, and their symbolic import is explained at 1.5.16–18.

n.56As Stablein notes in his dissertation, the concluding statement in chapter 4 marks the initial point at which all the Tibetan and Sanskrit witnesses begin to match. This includes the Sanskrit witness RST15, which until this point has contained an entirely different set of four opening chapters.

n.57D reads ka la sha a b+hi nA ya naM. S, Y, K, J, N, C, and H read ka la sha a b+hi na ya na. The Sanskrit sources read kalaśābhina­yanaṃ. This transliteration corrects the reading in the Degé Kangyur following the reading in the majority of Tibetan witnesses as well as the Sanskrit witnesses.

n.58Compare with Hevajra Tantra 2.4.1–5, where Vajragarbha and a group of ḍākinīs pose a similar set of questions to the deity Hevajra based on material from previous, similarly titled chapters. Here, however, the material causing the goddesses’ confusion has not yet been taught in The Glorious Sovereign Tantra of Mahākāla.

n.59The transliteration of this Apabhraṃśa passage follows D.

n.60Translation tentative. We take goms pa’i rigs pa’i sems as approximately equivalent to ºabhyasana­yogacetasā (RST15).

n.61ND 44-5 reads yoginibhiḥ mātṛbhir. RST15 reads yoginībhir mātrabhiḥ. The Tibetan witnesses read rnal ’byor gyi ma rnams. This translation follows the Sanskrit.

n.62Compare the preceding passage with Hevajra Tantra 2.4.9–11.

n.63RST15 reads nṛtyantu madhyavarttinoḥ bhūya. The Tibetan witnesses read bar du gar yang bya’o/ yang yang du. This translation follows the syntax of the Sanskrit in RST15.

n.64Compare the preceding passage with Hevajra Tantra 2.4.12–14.

n.65Following F glu, and Skt. gītaº. D and S read klu. This reading agrees with the parallel passage in Hevajra Tantra 2.4.13.

n.66Following F and S la sogs pa dam tshig rnams. D reads la sogs pa dag tshig rnams. C, J, K, and Y read la sogs pa dag tshigs rnams. Skt. reads gaṇeṣu.

n.67RST15 reads raktam (“red”).

n.68D reads shar phyogs kyi snam bu la. S reads shar gyi snam bu la. ST and RST15 read purve. ND 44-5, UTM 286, UTM 288, RASH 47, BnFS 84, and BnFS 85 read purvapūṭe. ST and RST15 read simply “in the east,” while ND 44-5, UTM 286, UTM 288, RASH 47, BnFS 84, and BnFS 85 translate as “in the space to the east,” once we emend pūṭe to puṭe. Here the Tibetan witnesses suggest a Sanskrit equivalent of purvapaṭe, but considering that the Tibetan sources alternate between snam bu (paṭa) and ’phar ma / ’phar ma (puṭa), we believe paṭa to be in error and have followed the reading puṭa here and below.

n.69Following Tib. lha mo bzhis. Skt. reads catur­yoginībhiḥ. The Sanskrit witnesses retain the title “yoginīs,” while the Tibetan witnesses read “goddesses.”

n.70This translation follows ST and RST15 in reading the verb bhāvayet here. The Tibetan lacks a final verb for the following passage.

n.71Following S phaT kyi sgra chen po. D reads pheM gyi sgra chen po. ST reads mahāpheṭkāra. UTM 286, UTM 288, RASH 47, and BnFS 85 read mahāphaṭkāraṃ. ND 44-5 and RST15 read mahāphaṭkāraº. This translation follows S and the Sanskrit witnesses in reading phaṭ instead of phem.

n.72Following S phaT kyi sgra. D reads pheM gyi sgra. ND 44-5, UTM 286, UTM 288, RASH 47, and BnFS 85 read mahāphaṭkāra. ST reads pheṭkāra. This translation follows S and the Sanskrit witnesses in reading phaṭ instead of phem.

n.73Following RST15 and UTM 288 kālikā . ST and ND 44-5 read kālīkā. D reads ka ling ka.

n.74This translation, which follows the Tibetan, is tentative due to ambiguities in the Tibetan syntax and the wide variation witnessed across the sources.

n.75D and S read dngos po bsgom pa bsgom pa med pa’o. ST reads bhāvam bhāvyaṃ yad abhāvo ’pi bhāvayet. RST15 reads bhāvaṃ bhavyaṃ yad abhāvo ’pi bhāvayet. This translation is informed by ST and RST15.

n.76Tib. reads gang dngos grub thob pa’i rnam pa ji ltar ’gyur. ST reads kathyate yena siddhir yena syāt. RST15 reads kathyate siddhi yena syāt. RASH 47 reads kathyati siddhir yena syāt. ND 44-5 reads kathyati siddhi yena syāt. The Tibetan term rnam pa, for which there is no equivalent in the Sanskrit witnesses, is read here as equivalent to the Sanskrit ākāra. The Tibetan sources also seem to indicate that the Tibetan translators read katham where the Sanskrit witnesses have kathyati/kathyate.

n.77Tib. reads mdun du blta bar bya’o. ST reads purato dṛṣṭva mahābhairavaṃ. ND 44-5, UTM 286, UTM 288, RST15, RASH 47, and BnFS 85 read purato dṛṣṭvā mahābhairavaṃ. BnFS 84 reads purato dṛṣṭā mahābhairava . This translation is informed by the reading in the Sanskrit witnesses, where it is clear that the form one sees in this meditation is Mahābhairava.

n.78While it is not entirely clear in the text itself, the term gnas pa / sthāna is taken to refer to a set of locations on the body that are associated with the sense organs.

n.79Following ST and RASH 47 bhāvayet. UT288, UT286, and BnF85 read prabhāvayet. RST15 omits. D, F, and S read sngar bzhin bya’o. The verb is supplied here from the Sanskrit witnesses, the majority of which read either bhāvayet or prabhāvayet.

n.80Tib. reads dga’ bo dang. Skt. reads ananteṇaiva, recording the name of this nāga king as Ananta.

n.81Following S phaT kyi sgra chen po. D reads pheM gyi sgra chen po. ST and RST15 read mahāpheṭkāraṃ. UTM 286, UTM 288, RASH 47, BnFS 84, and BnFS 85 read mahāphaṭkāraṃ. The reading in the Stok Palace Kangyur is supported by the reading in UTM 286 and UTM 288. ND 44-5 appears to omit this material and skip to the positioning of the yoginīs in the maṇḍala.

n.82Following ST, ND 44-5, UTM 286, RST15, RASH 47, and BnFS 85 purvadiśi. This translation follows the Sanskrit witnesses in reading “to the east.” There is no mention of a specific direction here in the Tibetan witnesses.

n.83Following ND 44-5, UTM 286, RASH 47, and BnFS 85 lañjanī . ST and RST15 read lañchanī. D and S read lan tshwa ni.

n.84Tib. reads go ku da ha na. Skt. reads gokudahana. This term indicates the five kinds of meat using the first letters of the names of the respective animals: go (“cow”), kukura (“dog”), damya (“horse”), hastin (“elephant”), and nara (“human”).

n.85Following ST, RST15, and RASH 47 gopyena pañcakulaṃ viharet. Tib. reads gsang ba’i rigs la gnas na, which translates as “when one dwells among the secret families.”

n.86Following F and ST in reading “Goddess” in the vocative (devi, lha mo). D reads lha mo la.

n.87Compare the preceding passage, beginning with “A ḍombī…,” to Hevajra Tantra 2.3.62–67.

n.88Tib. reads ’bad pa la ngas byin gyis brlab bo. ND 44-5, ST, and RST15 read prayatnena anuṣṭhanīyam. We understand the Tibetan term byin gyis brlab to translate the attested anu√ṣṭhā, rather than the more expected adhi√ṣṭhā.

n.89Following Tib. ras dmar po. Skt. reads rajovaktra. The alternate reading in the Skt. translates as “soiled cloth.”

n.90The transliteration of this Apabhraṃśa passage follows D.

n.91We emend Tib. sgra snyan pa to sgra nyan pa. This line is absent in the Sanskrit witnesses.

n.92Skt. reads saṃsaret. Tib. reads rgyu. The Sanskrit term implies “wandering in saṃsāra.”

n.93ST reads anuṣṭheyam. RST15 reads anuṣṭhet. Tib. reads byin gyis brlabs pa. This translation follows the Sanskrit terminology.

n.94This line translates the reading mantram mantrapaṭale yathoktaṃ karaṇīyam in the Sanskrit witnesses, which is omitted from the Tibetan. The “chapter on mantras” is chapter 2 above.

n.95Following N and S sarba sha truM mukhaM baM d+ha ya. D reads sarba sha trUM Ni mu khaM baM d+ha ya. ST and RST15 read sarvasatrūm mukha­bandhaya. The rendering in N and S is supported in ST and RST15.

n.96The procedures and recipes here and below are at times described quite differently in the Sanskrit sources. The English translations that follow generally preserve the Tibetan version unless the Sanskrit improves the clarity of the Tibetan.

n.97Following Tib. til ti la. Skt. reads bhūmilatātaila. The Sanskrit suggest that this oil is derived from bhūmilatā , which appears to be the name of plant but is sometimes also interpreted as referring to an earthworm.

n.98Following RST15 and ST māṣakaniyamena. D reads ji ltar goms pa yis man cha nges par byas nas. F reads ji ltar goms pas mnan che nges par byas nas. This translation follows RST15 and ST, as the Tibetan text is unclear and potentially corrupt.

n.99Following ST and RST15 balañjarīdravaiḥ. Tib. reads ca la ta’i khu ba. This plant could not be identified.

n.100Following ST and RST15 ambirolīdrava. D reads po ro li’i khu. This plant could not be identified.

n.101Following D and S pags pa ’dul byed kyi sman. This translation is tentative, and we have not been able to identify this substance. There is no equivalent term in the Sanskrit witnesses.

n.102Following ST and RST15 pariśoṣya. D and S read legs par spangs te. F reads legs par sbyangs te. This translation follows ST and RST15 because the Tibetan is not clear and may be corrupt. One can perhaps see how the original Tibetan translator read pariśodhya (F legs par sbyangs te) where the extant Sanskrit witnesses read pariśoṣya. The term legs par sbyangs te may then have been emended or incorrectly copied as legs par spangs te.

n.103Here the Skt. verb patāyet is understood in its sense of “to light” or “ignite.” There is no equivalent verb at this point in the Tibetan, so the Sanskrit has been followed for clarity.

n.104Tib. reads gling dkar po la. ST and RST15 read netrakarpaṭeṣu. This translation, which is tentative, follows the reading netrakarpateṣu, which only occurs in the Sanskrit witnesses. The meaning of the Tibetan phrase gling dkar po la is obscure.

n.105Following K, N, S, and Y bab la, taking it as approximately equivalent to the Sanskrit patāyet. D bla ba would be translated as “yellow orpiment” (haritāla).

n.106Following ST and RST15 sarasanāga. D reads ras dan kha. S, K, N, and Y read ras na kha. F reads ras na ga.

n.107Tib. reads sa’i lcug ma’i mar khu. Skt. reads bhūmilatātaila. The term has been left untranslated as the referent of the Sanskrit term bhūmilatā is unknown. It could interpreted as a synonym of bhūlatā (“earthworm”), but a literal reading of bhūmilatā as “ground vine” or “ground creeper” suggests it could signify a plant.

n.108The reference to lcug ma lnga ’tshad tsam / pañca­latāmānena is unclear and is taken here as a portion of the previously cooked mixture.

n.109Tib. reads lcags kyu lnga. ST and RST15 read pañcasālyaº. This translation is tentative. The Tibetan term means “the five hooks.” The Sanskrit pañcaśālya [sic] can indicate a spear or dart, or a broad range of extraneous sharp objects, such as splinters or thorns, that become lodged in the body and cause pain. We understand the Tibetan lcags kyu to here to be equivalent to the Sanskrit śalya. It is not clear what this is a reference to, but it could be another term for the “five ambrosias.”

n.110D reads de ni mtshams med pa lnga’i las byas par ’gyur ro. ST reads: tadā pañcānantarya­karmakāriṇo bhaveyuḥ. This translation and amends de in the Degé to ngas. It also assumes that the reading bhaveyuḥ is likely a corrpuption for the first-person singular form bhaveyaṃ. Such phrasing is a common way to express the unfailing efficacy of a rite. For another example, see note 144 below.

n.111D reads sa lin tsi dang / ma lin ci dang. S reads sa lin tsi dang / ma li ts+tshi dang. ST reads sālañji. RST15 reads śāleñcimāleñciº. ND 44-5 reads sārañjisārañji. UTM 286 reads sālañjiṃ salañjaṃ. UTM 288 reads sālañjiṃ salañjāṃ. RASH 47 and BnFS 84 read sālañjisaleñcīṃ. BnFS 85 reads sālañji salañjā. This translation is tentative and translates only the term śālañji, assuming that the second member of this compound may be the result of a redundancy or duplication.

n.112D and S read ku mu ha’i sha. ST and RST15 read kumuḍamāṁsaṃ. UTM 286 and UTM 288 read kumuhasāsaṃ; RASH 47 and BnFS 84 read kumahasāsaṃ. BnFS 85 reads kumahamaṁsaṃ. This translation is tentative, and the identity of this substance is unclear.

n.113This translation follows the ST and RST15 in reading nānyam (“only / nothing else”). The Tibetan reads gzhan du na (“otherwise”).

n.114Tib. reads sum nam. Skt. reads māṣaka. The Sanskrit term literally means “bean” but is here understood to refer to the size and shape of the portion to be eaten.

n.115Tib. reads rang byung. ST reads tripura. The identity of this substance remains unclear. This term only appears here in the Tibetan witnesses, but when it appears elsewhere it is generally a coded translation for the Sanskrit strīrajas, which may refer either to menstrual blood or sulfur.

n.116D reads klu ’dul ba. ST reads damana. This translation follows the Tibetan witnesses, which suggest a Sanskrit back-translation nāgadamanī.

n.117Tib. reads mnyes. ST and RST15 read maradaṇiyam. The Tibetan verb mnyes suggests a translation of the Sanskrit verbal root √mad (“to rejoice,” “be glad,” or “delight”). However, it is clear from this passage that a form of the Sanskrit verbal root √mṛd (“to crush,” “to muddle”) is intended. As the Tibetan translators frequently use mnyes where the Sanskrit indicates √mṛd, the English translation follows the Sanskrit in those instances without further notation.

n.118Because the Tibetan does not explicitly identify the reference of this line, it is understood to be nidhi (“treasure”) as indicated in the Sanskrit sources.

n.119As above, the referent “treasure” is supplied by the Sanskrit witnesses.

n.120ST ends here.

n.121Tib. reads rnal ’byor pa thams cad kyi. ND 44-5, RASH 47, RST15, and UTM 286 read sarvayoginīnām. The Sanskrit witnesses read “all yoginīs.”

n.122D and S read zla ba gzas zin pa’i dus su. RST15, ND 44-5, UTM 286, UTM 288, and BnFS 85 read śaśīkiraṇ­avelāyāṃ. RASH 47 reads śaśīkilaṇ­avelayāṃ. BnFS 84 reads śaśīkiraṇ­averāyāṃ. This translation follows the Tibetan witnesses, but the Sanskrit witnesses preserve an equally plausible reading that translates as “when the moon is shining.”

n.123D and S read thog gi lhung ba’i shing. RAS15 reads vajra­patita­tālamūlaṃ. ND 44-5, UTM 286, UTM 288, BnFS 85, and RASH 47 read vajrapātaṃ tālamūlaṃ. BnFS 84 reads vajrapātaṃ toramūlaṃ. This translation follows the reading in the Sanskrit witnesses. The Tibetan witnesses preserve a reading that translates as “a tree that has been struck by lightning,” omitting any indication of the type of tree.

n.124Following D and S dza lu’i rtsa ba. The exact identity of this plant is uncertain.

n.125Following D and S phyag drug pa’i mgon po nag po. F reads phyag bcu drug pa’i mgon po nag po. RST15 and UTM 286 reads ṣodaśabhuja­bhairavaṃ. RASH 47 reads ṣodaśa(bhu)jabhairava. The alternate reading in F suggests that this is a sixteen-armed form of Mahākāla. The Sanskrit manuscript witnesses support that this is indeed a sixteen-armed form, but they also refer to the deity as Bhairava.

n.126This translation follows RST15 and ND 44-5 in reading pañcāmṛta. Tib. omits “five.”

n.127Following D sman ma hU ga ga sa bon. RST15 reads madhukukabījaṃ. UTM 286, UTM 288, and RASH 47 read madhūkūrkka(?)bījaṃ. BnFS 84 reads madhukukta­bījaṃ. BnFS 85 reads madhukukta­bījaṃ. The Sanskrit sources suggest it may be possible to emend this reading to *madhukukkuṭī, which Monier-Williams lists as “a kind of citron tree with ill-smelling blossoms.”

n.128Because this plant could not be identified, and because there is wide variation across the Sanskrit and Tibetan sources, this has been rendered as it appears in D.

n.129Following D and S tsa tra sa ku la mnyams pa. RST15 and RASH 47 read pari­pācitacitra­śakulaṃ. This translation is tentative and follows the Tibetan.

n.130D reads oM ka rA la bi ka~M rA la ma ha na da hU~M gr-ih+Na gr-ih+Na kaM tra gi swA hA. S reads oM ka rA la bi ka ra la ma hA naM da hUM gr-iH UM gr-ih+na kaM Ta gi swA hA. The transliteration given above conforms with version given in UTM 286, UTM 288, RASH 47, BnFS 85.

n.131There are multiple Sanskrit terms used for mercury in this text. To reflect this, we have translated rasa with “mercury” and pārada as “quicksilver.” The Tibetan translators sometimes transliterated rasa as ra sa and sometimes with dngul chu. Pārada was most typically translated with dngul chu.

n.132The identity of this ingredient is uncertain, so this term is rendered as it appears in D.

n.133RST15 reads suravimalām. ND 44-5 and RASH 47 read surāvimalām. D and S read mu rA dang dri ma med pa. F reads dri ma med pa. This translation follows the Sanskrit witnesses in reading surāvimāla as single ingredient, not two as indicated in D. We likewise regard the Tibetan mu rA as an orthographic corruption of su rA.

n.134The identity of this ingredient is uncertain, so this term is rendered as it appears in D.

n.135The Sanskrit witnesses begin this recipe with “a portion of mercury” (rasamāsakaṃ). As noted above, many of the recipes in this and other chapters differ between the recensions represented in the Tibetan and Sanskrit witnesses. The English translations here generally preserve the Tibetan version unless the Sanskrit improves the clarity of the Tibetan.

n.136Tib. reads ’dzo ti sa ma’i mar khu. Skt. reads hyosmṛtitaila. This translation is tentative and based on the Sanskrit witnesses.

n.137The identity of this ingredient is uncertain, so this term is rendered as it appears in D.

n.138D and S read sol ba’i lo ma. F reads so ba’i lo ma. RST15 reads simvipatraº. ND 44-5 reads simbimantraº. As the Tibetan term could not be correlated with an ingredient that has leaves (srol ba typically means “charcoal”), we have used the Sanskrit term. The plant could not be confidently identified.

n.139This identity of this ingredient is uncertain, so this term is rendered as it appears in D.

n.140The identity of this ingredient is uncertain, so this term is rendered as it appears in D.

n.141The identity of this ingredient is uncertain, so this term is rendered as it appears in D.

n.142Following K mig na mi mthong bar ’gyur ro. D and S read mid na mi mthong bar ’gyur ro. Skt. omits.

n.143Following D and S mi’i ’bras bu. UTM 286, UTM 288, BnFS 84, and BnFS 85 read naramuṇḍa. ND 44-5 and RASH 47 read naramudra. RST15 reads śvanmeḍhram.

n.144Skt. reads gūḍamārgeṇa. Tib. reads rnam snang gi lam. The Tibetan literally reads “the path of vairocana,” which utilizes the code word vairocana, meaning “feces.” The extant Sanskrit witness do not use this code word, but rather the standard medical term for “rectum” (guḍa).

n.145Following RST15 yadi na sidhyati tadā [sic?] aham eva pañcā­nantrya­karma­kāriṇī syām. Tib. reads de gal de ma grub na des mtshams med pa lnga’i las byas par ’gyur ro. This reading, which reflects the majority of Sanskrit witnesses despite minor variation, means “If this does not work, it would be as if I [the Blessed One] had committed the five actions entailing immediate retribution.” In other words, “it is impossible it will not work.” This is a standard formulation in Buddhist tantric literature, one in which the Blessed One or other deity confirms the efficacy of a given rite or recipe by claiming that its failure would entail the impossible premise that they had themselves committed the five acts entailing immediate retribution. This formulation will be used later in the Tibetan translation of this text as well. Here the Tibetan witnesses read des (“as if one had committed…”). This is taken here as a possible corruption of the Tibetan ngas, which would provide the first-person agent from the Sanskrit and is a common scribal error in Tibetan translations.

n.146This translation is tentative.

n.147D and S read su ma ka ta. ND 44-5 and UTM 286 read māgadhā. RASH 47 reads sumāgadhā . BnFS 84 reads māgadhā. RST15 reads sumāśaṭha. The precise identity of this substance is not clear.

n.148Following F sen du ma ni ka, RST15 sendumāṇikyam, and ND 44-5 sindhumānikyam. D reads sems can du ma ni ka, and S reads sems can du ma ni bka’. UTM 286, UTM 288, BnFS 84, and BnFS 85 read sedhumāṇikyam. RASH 47 reads sendhumāṇikyaṃ.

n.149Tib. reads sman ’dzag. Skt. reads śravanti. This ingredient could not be identified.

n.150D and S read sa ma ya go la dag dang. RST15 reads samayagolakaṃ. The precise identification of this substance is not clear.

n.151D and S reads pin ta d+ha ka ra’i rtsa ba. RST15 reads piṇḍata­garamūlaṃ. The identity of this substance is not clear.

n.152D and S read rang byung. RST15 reads rajaḥ.

n.153Following RST15, UTM 286, UTM 288, RASH 47, and BnFS 85 devadhānyaṃ. D reads de barta nA ra dang. S reads de ba rta na ra dang.

n.154Following BnFS 85 suramuli­kādravaṃ. RST15 reads suramūri­kādravaṃ. UTM 286, UTM 288, and RASH 47 read sulamūli­kādravaṃ. D and S read mu ru mu ri ga’i khu ba;

n.155D reads sngags pas g+ho d+ha zos nas. S reads sngags pas go da zos nas. RST15, ND 44-5, UTM 286, UTM 288, BnFH 47, and BnFS 85 read godhāmantritā bhakśayet. This translation is tentative.

n.156D and S read sha du ri. F reads sha tsa ri. ND 44-5 reads sabari. UTM 286, RST15, and BnFS 85 read śabarī . RASH 47 reads śābarī. BnFS 84 reads sabali. This translation is tentative, and the specific ingredient is unidentified.

n.157D and S read mnyam pa ri ka na ba su dang ldan pas. RST15 reads tulya || kiraṇe ca susametaṃ. ND 44-5, UTM 286, UTM 288, BnFS 85, and RASH 47 read tulya­kiraṇasu­sametaṃ. BnFS 84 reads tula­kiranasu­sameta. This translation is tentative and follows the reading in the Sanskrit witnesses, emending it to *tulyaṃ kiraṇeṣu sametaṃ.

n.158This translation is tentative and follows the Tibetan translation.

n.159Tib. reads mi skam pa’i me tog. Skt. omits. This translation is tentative.

n.160D and S read de yang me tog re re’am thams cad kyis bar snang la ’gro bar ’gyur ro. This translation is tentative.

n.161D, F, and S read ha na ha na. RST15, UTM 286, RASH 47, BnFS 84, and BnFS 85 read ghana ghana. ND 44-5 and UTM 288 omit. This transliteration follows the reading in the Sanskrit witnesses.

n.162Following RAS15, UTM 286, BnF84, and UTM 286 śopagalikā . ND 44-5 reads śopagarikā. UTM 288 omits. D and S read sman kun su ma. F reads sman ku su ma. The transliteration of this substance is based on the Sanskrit witnesses. Its identity remains unknown.

n.163D and S read yang dag par rab tu sbyin pas ’grub pa. RST15 reads sāmpratam siddhyati. This translation follows the Tibetan, but based on the reading in RST15 it seems the Tibetan translators read sampradam where the Sanskrit witness reads sāmpratam.

n.164This translation is tentative, and the identification of this plant is not clear in the Tibetan or Sanskrit witnesses.

n.165D and S read su gan d+ha ti ka ta chen po. UTM 286 reads sugandh­mahātikta. A specific ingredient by this name could not be identified, so we have translated it descriptively.

n.166Following ND 44-5, RST15, and UTM 286 hiraṇya­parikara. D reads hri rann+ya pa ri ga. S reads hri ran dang nya pari ga. F reads hi ran ya ba ka ri ka.

n.167D and S read su ga tra mu tra mu khi. As there is wide variation in the Sanskrit witnesses, this transliteration follows the reading in D and S. The identity of this substance is uncertain.

n.168Following RST15 sarvadhalī . D and S read sa ba da li. N and H read so ba da li. ND 44-5, UTM 286, UTM 288, RASH 47, BnFS 84, and BnFS 85 omit. Although this transliteration follows the reading in RST15, the identity of this substance is unknown.

n.169Following RST15 masana. D reads na ma sa.

n.170D reads dbu rtsa ba. This translation is tentative. This Tibetan term seems not to correspond to any of the available Sanskrit witnesses. We tentatively understand this to be translation of either śirṣamūla or śiramūla and then emend to śiśiramūla.

n.171D reads in+da ro. RST15 reads indarī. The identity of this substance is uncertain.

n.172This transliteration follows D ba ta ba’i sa bon, since none of the available witnesses provide a satisfactory reading. This ingredient cannot be identified.

n.173Following D sa ta ni. RST15 reads śātaṇī. ND 44-5, UTM 286, UTM 288, RASH 47, BnFS 84, and BnFS 85 omit. This transliteration is tentative. It is possible that the reading in D and RST15 might be corruption of *śatāvarī.

n.174Following RST15 mayā. Tib. reads khyod kyis. This translation follows RST15 because it aligns with the next statement that also references the Blessed One in the first person. It is worth noting that in the Sanskrit witnesses the preceding statement reads, “If it does not work then it will be as if I had committed one of the actions entailing immediate retribution” (RST15 yadi na bhavati tadā aham eva pañcānantaryakar­mākarī bhavate).

n.175Following RASH 47 śaśaṅkadrava. UTM 286 and BnFS 84 read śaśaṅkandava. RST15 reads sumbhanaṃ gulakam. D and S read zla ba’i chu’i ’gu li ka. This translation follows the reading in Sanskrit witness RASH 47 and similar readings of the term śaśaṅkadrava as the equivalent here for zla ba’i chu, which elsewhere has been translated as “lunar water.” The latter term tentatively translates the Sanskrit equivalent śaśadharajala.

n.176D and S read glu sna tshogs. C, F, K, and Y read klu sna tshogs. Skt. reads nānāgitaº. D and S agree with the Sanskrit witnesses in reading nānagītaº. C, F, K, and Y read “various nāgas.”

n.177This translation is tentative. The Tibetan translation appears to preserve a unique version of this passage without direct equivalent in the available Sanskrit witnesses.

n.178D and S read sbyor ba chen po ’dus pa. RST15 reads mahāmelaka­saṁyogaṃ. ND 44-5 reads mahākāla­saṁyogaḥ. UTM 288 reads mahākāla­saṁyogaṃ. BnFS 84 reads mahākāla­saṁyoga. BnFS 85 reads mahārasaṃyogaṃ. This translation follows the reading in the Tibetan witnesses. A number of Sanskrit witnesses read “union with Mahākāla.”

n.179Here the Sanskrit reads bola-kakkola, two code words used respectively for “penis” and “vagina,” as articulated in the Hevajra Tantra. In the Tibetan this coded language has been “translated” using the more standard euphemisms for the two sexual organs.

n.180D and S read gza’ nyi ma la a sa mi ni byas nas dar ling la sdong bu byas te. ND 44-5 reads ādityavāre netrakalpate varttikā kārayet. RST15 reads ādityavāre netrakarpaṭa­varttī kārayet. This translation follows the reading in RST15.

n.181Following F lcags mchog and ND 44-5 and UTM 286 araparamām. D and S read lcags kyu mchog, which translates as “supreme hooks.”

n.182Following RST15 hṛdi kaṇṭhe yoginaḥ tatra vicitraṃ syāt. D and S read rnal ’byor pa’i snying ga dang mgrin pa’i rnal ’byor pa’i ngo mtshar ba. UTM 286 reads hṛḍikaṇṭhe yoginīḥ || tatra vicitra syāt. UTM 288 reads hṛdikaṇṭhe yoginaḥ || tatra vicitra syāt. The reading in RST15 is close to the Tibetan, but the Tibetan translators appear to have interpreted the Sanskrit term vicitra in the sense as something that is “wonderful” (ngo tshar ba) rather than “spots.” The Tibetan also contains repeated instances of the term yoga/yogin that are difficult to interpret and thus understood to be in error.

n.183D and S read sna ring ba. ND 44-5 reads sighraṃ nāsikā. RST15 reads dīrghanāsikā. The term for “nose” (Tib. sna, Skt. nāsikā) is frequently used as a euphemism for the penis.

n.184D and S read dza ya ra. F omits. RST15 reads jāyacaraº. ND 44-5 and UTM 286 read jāyaphalaº. Because of the ambiguity of this term, it has been transliterated as it appears in D. This ingredient could not be identified.

n.185D and S read cod pan can gyi mkhris pa. Skt. reads śikhipittam. This translation is tentative. The Sanskrit term śikhin, which literally means “having a tuft of hair on the head,” might refer to a number of animals that share this feature.

n.186In this chapter the multiple Sanskrit terms used for mercury‍—most often rasa or pārada‍—were not consistently translated into Tibetan using the same equivalents. In order to disambiguate this use of these terms, the English translation uses “mercury” for rasa and “quicksilver” for pārada regardless of the equivalent term used in the Tibetan translation.

n.187Tib. reads sbas pa rnams. Skt. reads gopena. This translation is tentative.

n.188The translation of this passage is tentative due to numerous ambiguities in the Tibetan translation and the wide variation in the Sanskrit witnesses.

n.189Jayantī is the Sanskrit term for sesbania.

n.190RST15, ND 44-5, UTM 288, and BnFS 84 read ekaviṁ­śatidinena. UTM 286, RASH 47, and BnFS 84 read ekaviṁ­śatidivasena. Tib. reads nyi ma rnams. This translation follows the reading in the Sanskrit witnesses, which specifies a twenty-one-day period.

n.191This mantra follows the version reported in D. There is some variation across the Sanskrit and Tibetan witnesses.

n.192The identity of this substance is uncertain, so the term has been transliterated here as it appears in D.

n.193Following RST15 bhaktauṣaṇapatra­draveṇa. D and S read ba ki ta’i lo ma.

n.194The identity of this substance is uncertain, so the term has been transliterated here as it appears in D.

n.195D and S read go rak+ShaM du la. RST15 reads rakṣaṇḍula. UTM 288 reads gorakhaṇḍula. RASH 47 reads gorakhataṇḍula. While this substance cannot be identified precisely, it is given as an equivalent to nāgabāla (snake mallow) in Cakrapāṇidatta’s Bhānumati, a commentary on the sūtrasthāna of the Āyurvedic treatise Suśrutasaṃhitā. About this see Klebanov 2011, p. 193.

n.196D and S read ’dis kyang legs par byas pa shes so. RST15 reads anena ca sasaṃskāram iti. UTM 286 and BnFS 85 read anyena rasasaṁskāraº. RASH 47 reads anena rasa­saṁskāraṃ gagaṇa­saṁskarāram idāniṃ. This translation is tentative and follows the Tibetan.

n.197Following Tib. da ni zhugs shing gi ’du byed ces bya ste. RASH 47 and RST15 read gagana­saṃskāram idānīṃ. UTM 286 reads ºgagaṇa­saṃskāram idāniṃ. This translation is tentative.

n.198This translation follows the Sanskrit witnesses that read “five nights” (RST15 pañcarātra) as the duration of the drying process. In the Tibetan, “five full days” (D nyin zhag lnga) is the duration of the grinding process.

n.199Following F and S kar don dza na. K, N, and Y read ka ra don dza na. C and J read karo+Any+dza na. D reads kardony+dza na. RST15 reads raudreka­dvañjaṇīṃ. UTM 286 reads raudreka­tvañjanīyaṃ. The identity of this substance is uncertain, so the term is given here as it appears in F and S, which is also close to what is given in C, J, K, N, and Y.

n.200D and S read dngul srang gcig. RST15 reads rasapalaṃ. RST15 indicates that the Tibetan dngul, typically “silver,” should be read as dngul chu (“mercury”).

n.201D and S read rwa dza. Skt. omits. The identity of this substance is uncertain, so the term has been transliterated here as it appears in D.

n.202D and S read dngul du ’gyur ro. RST15 reads stambhayati dhruvaṃ This translation follows the Tibetan. The Sanskrit witnesses suggest the reading, “it will certainly be stabilized.” As above, we interpret dngul as dngul chu. There is no equivalent term in the Sanskrit sources.

n.203D and S read rnal ’byor pa gang gis ra ti re re tsam zos na. RST15 reads piṇḍarasena rātikāmānena bhakṣayet yo yogī. RASH 47 reads piṇḍarasena rāttikāmānena bhakṣayet naraḥ. UTM 286 reads piṇḍale senarātti­kāmānana bhakṣayet naraḥ. This translation is tentative and follows the Tibetan.

n.204D and S read zhi ba dang mnyam par ’gyur. RST15 reads śivasamam.

n.205This translation is tentative. In the Buddhist tantras kakkola is often used as a code word for the vagina. See for example Hevajra Tantra 2.3.60.

n.206This colophon is only found in the Tibetan translations. The Sanskrit witnesses do not mark this as the conclusion of chapter 13.

n.207D reads dkar po dang nag po dang sbyar ba ni ’dab ma gcig tu byas nas. RST15 reads śukla­kṛṣṇayoḥ śodhanaṃ ekapatraṃ. UTM 286, UTM 288, RASH 47, and BnFS 84 read śuklakṛṣṇayogaḥ śodhanaṃ ekapatraṃ. This translation is tentative.

n.208D reads sman so na’i khu. S reads sman sa ni’i khu. RST15 reads saṇīdraveṇa. UTM 286, UTM 288, BnFS 84, and RASH 47 read khaṇīdrave. The identity of this substance is uncertain, so the term has been transliterated here as it appears in D.

n.209D and S read lan dgu ru ’di rnams rdzogs par. RST15 reads navavārān etena niṣpannam iti. This translation is tentative.

n.210D and S read sman ka ru li. The identity of this substance is uncertain, so the term has been transliterated here as it appears in D.

n.211This transliteration follows D with minor emendations based on the mantra as attested in the Sanskrit and Tibetan witnesses.

n.212Following Skt. srtīrajas. D and S read dngul.

n.213The identity of this substance is uncertain, so the term has been transliterated here as it appears in D.

n.214The identity of this substance is uncertain, so the term has been transliterated here as it appears in D.

n.215D and S read de’i dus su zla ba dang nyi ma dag zhi bar gyur na. ND 44-5 reads yadi na bhavati tadā candrādityo vinasanaṃ. UTM 286, UTM 288, RASH 47, and BnFS 85 read yadi na bhavati tadā candrādityo vināśanaṃ. BnFS 84 reads yadi na bhavati tadā cadrātityo vināsana. The reading in the Sanskrit witnesses translates as, “If this does not happen, it means the sun and moon have been destroyed,” which might be taken as a statement expressing the assured efficacy of the rite.

n.216At this point RST15 diverges from the other Sanskrit witnesses and the Tibetan. Therefore we have not consulted RST15 in the translation of this chapter or for the remaining chapters without equivalent sections in RST15. Later chapters that do have parallels in RST15 will be noted. Additionally, many of the proper names in this chapter vary significantly across the Sanskrit and Tibetan sources. In cases where the sources do not offer a consistent satisfactory reading and the name cannot otherwise be clearly identified we have rendered it as it appears it in D. While this at times yields implausible terms, we have preferred to preserve the Tibetan reading rather than arbitrarily choose a different term from among the Sanskrit sources, many of which also demonstrate scribal corruptions. Minor emendations to correct orthographic issues have been made when possible to improve the clarity of the transliteration.

n.217D and S read de na klu’i rgyal po bu ga bo ga d+hi ga zhes pa. ND 44-5 and UTM 288 read tatra yuge bogavidha­vaṃganāma­rajā. UTM 286 reads tatra yuge bogavidhagam nāmarājā. BnFS 84 reads tatra yuge vāga­vidhavaga­nāma­rājā bhaviṣyati. BnFS 85 reads tatra yuge vāga­vidhaṃga­nāma­rājā. It appears that the Tibetan translators read the term yuga as the first syllables of the nāga king’s name, thus yielding de na klu’i rgyal po bu ga bo ga d+hi ga. We have followed the Sanskrit witnesses in reading tatra yuge (“in this eon”), while otherwise leaving the name as it is given in D.

n.218D, F, and S read nga skye ba brgyud nas…dngos grub thob bo. None of the Sanskrit witnesses attest to an equivalent of the first-person pronoun “I” (nga) that appears at the beginning of this line of Tibetan. It is possible that the Tibetan nga could be read as da (“now”), but this is also unsatisfactory. We have followed the Sanskrit witnesses and omitted nga.

n.219Following Skt. andro nāma rājā. D reads ming ni Na Da na dra zhes pa. F reads rding drag zhes pa. S reads ming ni na Den da zhes pa. Because the Sanskrit witnesses are consistent and the Tibetan sources vary significantly, we have rendered this king’s name according to the reading in the Sanskrit witnesses.

n.220Following Skt. kīrttanandano. D and S read ki ta nan da’i dus. This translation follows the Sanskrit witnesses.

n.221Following K, S, and Y be la and ND 44-5, UTM 286, UTM 288, and RASH 47 velo. D reads ba bla. F reads ce la. The Sanskrit vela signifies an exceedingly high number.

n.222D and S read ri ba hu ra zhes pa. F reads ri va hUM ga. ND 44-5 and RASH 47 read vaṅgalo nāma parvataḥ. UTM 286 reads vegaro nāma parvaḥ. UTM 288 reads vaṅgaro ṇā parvaḥ. BnFS 84 reads vaṅgaro nāma parvata. BnFS 85 reads vaṅgaro nāma parvaḥ.

n.223D and S read sa mo ri. F reads pa lo ri. Skt. reads gaurī.

n.224D and S read de’i rgyal po ra sa na. F reads de’i rgyal po rangs na. ND 44-5, UTM 286, UTM 288, BnFS 84, and BnFS 85 read janarā sarvasana. RASH 47 reads janarā savasana.

n.225D and S read mtshan nyid bdun. UTM 286, UTM 288, RASH 47, and BnFS 85 read saptarakṣaṇa. BnFS 84 reads saptarakṣana. This translation follows the reading in the Tibetan witnesses, but it is worth noting that the Sanskrit witnesses consistently render this name as saptarakṣaṇa, not saptalakṣaṇa . The letters ra and la are frequently interchanged in the Sanskrit witnesses.

n.226This translation is tentative, as this line appears to be corrupt in all sources consulted. The Sanskrit witnesses consulted read kāmākṣī/ā where the Tibetan has ’dod pa’i gzugs can ma, but it is unclear whether the Tibetan intends to translate this term or it represents a variant reading in the Sanskrit manuscripts.

n.227Following S dA ri ka. D and F read d+ha ri ka. ND 44-5, BnFS 84, and BnFS 85 read dārika . UTM 286, UTM 288, and RASH 47 read dālika.

n.228D reads phyi nas de dpal gyi tshong ’dus zhes par gyur nas ’gro ste de’i lnga cha gcig lci ba dang yang ba dang / gzhon pa’i a ga ru nag po ’bum phrag dang ldan par ’gyur ba dang. ND 44-5 reads tad anu śrīhahantikā bhaviṣyati | pañcamaṃ | guru ca pūṭarū lakṣa­kamalanaṃ. UTM 286 and RASH 47 read tad anu ca śrī­hahantrikā bhaviṣyati | pañcamaṃ guru ca pūtarū lakṣa­kaṃalanaṃ. This translation is tentative.

n.229We read the Tibetan term yongs su byed pa as equivalent to the attested Sanskrit term parikara.

n.230D reads thams cad yongs su byed pa b+hu dzaM ga po da d+hi ka rnams kyi bar. F reads rdza ga pa+di tri ga rnams kyi bar. S reads thams cad yongs su byed pa b+huM dza ga po da tr-i ka. ND 44-5 reads sakalapari­karabhujaṃ­gāyātikuyatra rājo nāma bhaviṣyati. UTM 286 reads sakalapari­kalabhujaṁ­gāyātikā yatra rājā nāma bhaviṣyati. UTM 288 reads sakalapalikala­bhujaṅgā yātikā yatra rajo bhaviṣyati. RASH 47 reads sakarapari­kalabhujaṅgājā­tikā yatra rājāno bhaviṣyati. BnFS 84 reads sakapali­kalarujaṇgāyati ku yatrayā rājo nāma bhaviṣyati. BnFS 85 reads sakarapari­katva bhujaṅgāyātikā yatra rājānama bhaviṣyati. Because of the wide variation among sources, we have transliterated this phrase as it appears in D.

n.231D and S read kha dog ser po so sto ba de ga ba. ND 44-5 reads garuavarṇaḥ ulatadanta­devagayadvatiḥ. UTM 286 reads gauravarṇaḥ ulutadanta­devagavayadvatiḥ. RASH 47 reads gauravarṇaḥ ulutaddanta­devagavayadvatiḥ. This translation is tentative and follows the Sanskrit for ºdevagavayaº.

n.232D and S read dbyangs yig bcu gsum pas rgyan pa ’byung ste. ND 44-5 reads svarasaṃyutaṃ saha. UTM 286 reads sva 12 svarasaṃyutaṃ saha. UTM 288 reads svaraṃ yutaṃ sa. RASH 47 reads 12 svarasaṃyutaṃ. BnFS 84 and BnFS 85 read svarasaṃyuta. This translation is tentative. Some of the Sanskrit witnesses note that this name is spelled with the twelfth vowel, while others do not provide a specific number for this vowel.

n.233Following Skt. rājā bhaviṣyati. D and S read de nas slob ma. This translation follows the Sanskrit witnesses, which note that this individual is the next in a line of kings. The reading in the Tibetan witnesses identifies them as a “disciple” or “student” (slob ma).

n.234Following ND 44-5 and BnFS 84 varmāsanarājā, and UTM 286 and RASH 47 varmāsanaṃ rājā. D and S read bram ze’i ming can gyi rgyal po. This translation follows the reading in the majority of Sanskrit witnesses, where we see this king’s name rendered as Varmāsana.

n.235D and S read yul pa tri ka re ka ra sha b+ha la zhes pa. F reads yul pi Ta ke ra k+Sha sha zhes pa. UTM 286 reads pāṭṭīkelake sārabhūnāma. BnFS 85 reads paṭṭīkelake śārabhūnāma. UTM 288 reads paṭṭikelake 2 sārabhūnāmaṃ. ND 44-5 reads pattikerake śārabhūnāma. BnFS 84 reads paṭṭake 2 sārabhūnāmaṃ. RAS47 reads ṣaṣṭīkelake sālabhūnāmaṁ. The spelling for this person’s name is provided from UTM 286, but the alternative spelling Sārabhū is equally plausible. It is possible that the correct spelling for this place name is in fact Paṭṭikelaka, but we have preserved the reading ra for la in the Tibetan sources. Conflation of these two consonants is common in the Sanskrit witnesses.

n.236The transliterations of these names follow D, with minor emendations.

n.237Following ND 44-5, RASH 47, and UTM 286 mālavī . D and S read mA la lI. C and J read ma la ba. F reads ma la wa. H and N read mA la wi. K and Y read ma la wi.

n.238Following D and S sam bu ka. F reads sam bu kyi. Skt. reads samūkī.

n.239Following UTM 286, UTM 288, RASH 47, and BnFS 84 govardhanādayaḥ rājāno bhaviṣyanti. ND 44-5 reads govarddhanādayaḥ rājāno bhaviṣyati. BnFS 85 reads govandhanādayaḥ rājā bhaviṣyati. D and S read b+ha d+ha na zhes pa dang ldan pa’i ming can ’byung ngo.

n.240The transliteration of this name follows RASH 47, UTM 286, UTM 288, and BnFS 85 kavarttaputraḥ. BnFS 84 reads kevatraputra. F reads ka bar+da pu tra. D and S read ke va ta pu tra.

n.241D and S read sin d+hu zhes par gyur nas ’gro’o / de shi na gar zhes par ’byung ngo. F reads sin d+hu zhes par gyur nas ’gro’o / de shi nas ba ra zhes pa ’byung ngo. ND 44-5 reads tatra rājā bandhūdevyā bhavati taṃ hatvā vīrena bhavitavyaṃ. UTM 286 reads tadantaraṃ sindhū bhaviṣyati || taṃ hatvā vīreṇa bhavitavyaṃ. UTM 288, RASH 47, and BnFS 84 read tadanantaraṃ sindhū bhaviṣyati || taṃ hatvā vīreṇa bhavitavyaṃ. The translation is tentative and follows D.

n.242The transliteration of this place name follows RASH 47, UTM 286, UTM 288, and BnFS 85, which read dakṣiṇāpathi sarṣibhañjikā nagarī. BnFS 84 reads dakṣiṇapathi saṣibhañjikā nagalī. D and S read lho phyogs na grong khyer sa b+hiny+dzi ka zhes pa. F reads lho phyogs na grong khyer pa b+hi b+ha nya dza ka zhes pa.

n.243Tib. reads rdo rje’i khri. ND 44-5, RASH 47, and UTM 286 read vajrāsana. The term vajrāsana is often used to refer to Bodh Gayā.

n.244The Tibetan translation identifies Khasarpāṇi as a “land” or “country” (yul), but it is also the name of specific form of Avalokiteśvara.

n.245D and S read bang ga la dang po o Di yA na. F reads bag la o ta ya na. The term dang po (“first”) has been omitted from this translation because it does not appear in F or any of the Sanskrit witnesses.

n.246ND 44-5 reads aparasya purapravesāt sahedevakaivartta­putro bhaviṣyati. UTM 286, UTM 286, and RASH 47 read aparasya purapraveśāt sahadevakaivartta­putro bhaviṣyati. BnFS 84 reads aparasya purapravesyat sahedevakaivatra­putro bhaviṣyati. D and S read de nas des nyam pa’i shu bi la grong ’jug byas pas rgyal po sa ha de wa zhes par ’gyur ba ’byung ngo. This translation follows the reading in the Sanskrit witnesses.

n.247D and S read bum pa gsum la sogs pa rnams dang / gang dang gang du phung po lnga rnams yod pa la lus zhes bya ste. F reads bum pa la sogs pa rnams dang / gang dang gang du phung po lnga rnams yod pa la lus zhes bya ste. ND 44-5 reads tighaṭādayaḥ kaṭapañca­skandhā sarvasthānā deheṣu prakīrtitāḥ. UTM 286 reads trighadādayaḥ kaṭamata pañca­skandhā sarvasthānā deheṣu prakīrtitāḥ. RASH 47 reads trighaṭādayaḥ | kaṭamata pañca­skaṃdhā sarvasthānā deheṣu prakīrtitāḥ. This translation is tentative and primarily follows the Tibetan with some clarity provided by the Sanskrit sources. The meaning of this line is uncertain. The Sanskrit could be tentatively interpreted as, “[There are places] where the three vases and so forth are renowned as the complete sites for each of the five aggregates within bodies.”

n.248Following D, F, and S rgyal srid rnams. Skt. reads ºrājanyakaº. This translation follows the Tibetan witnesses. The Sanskrit rājanyaka means “warriors” or “soldiers.”

n.249Following D and S le’u sum cu so drug par. ND 44-5 reads ṣadvisatipara. UTM 286 and BnFS 84 read ṣadviśatipare. RASH 47 reads śadvīṁśatipale. BnFS 85 reads ṣadviṁśatipara. This translation follows the Tibetan witnesses. The Sanskrit witnesses read “chapter twenty-six” here. Neither chapter 26 nor chapter 36 of this text addresses the topic mentioned here.

n.250Chapter 25 of this text does not address this topic. Instead, it addresses the topic of the interpretation of signs that one will become king and practices for conferring or assuming kingship.

n.251As indicated in the Hevajra Tantra, catuḥsama (bzhi mnyam) is a code word for feces.

n.252Following D and S sman mu ta ka. The identity of this substance is unknown.

n.253Following D and S mu zi dang / lhang tsher dang / su par+Na ma k+Shi rnams lag pa g.yon pas bzung la. RASH 47 reads gandha­kābhraka­suvarṇa­kalaṃ mīnam vāmahaste gṛhitvā. ND 44-5, UTM 286, UTM 288, BnFS 84, and BnFS 85 omit. This substance could not be identified based on the Tibetan and Sanskrit sources, so it has been transliterated as it appears in D and S.

n.254Following UTM 286, UTM 288, RASH 47, BnFS 84, and BnFM 85 bho putra. D and S read skyes bu ces bya ba. F reads dge’u ’di zhes pa. This translation follows the Sanskrit witnesses in reading the Tibetan skyes bu as kye bu (bho putra).

n.255D reads phyogs gsum du. S reads phogs gsum. F omits. ND 44-5, RASH 47, and UTM 286 read trivācilam. The phrase phyogs gsum (D) is used multiple times in this chapter and appears to indicate the number of times a specific phrase is repeated. The Sanskrit witnesses report a number of equivalents, many of them seemingly corrupt. In some cases, F and S read phogs gsum, as does D in one instance below. Since the term is used in the same way in each instance, we have translated it as “three times” despite the spelling variations in Tibetan.

n.256D reads shing bsrung ba ldan gyi drung ngam me tog can gyi drung dag gong dang mtshung so. S reads shing srung ba ldan gyi drung ngam me tog can gyi drung dag gong dang mtshung so. ND 44-5 and BnF84 read rakṣanaṃ puṣpavanaṃ samaṃ. UTM 286 reads ra–ṇaṃ puṣpavanaṃ samaṃ. UTM 288 reads rakṣaṇaṃ puṣpavana­saṃmasaṃ. BnFS 85 reads rakSaNaM puSpavanaM sama. This translation is tentative and follows the reading in the Tibetan witnesses.

n.257Following Tib. bcud len gyi dngos grub. Skt. reads rasasiddhi. This translation follows the Tibetan witnesses.

n.258D and F read phogs gsum du. S reads phyogs gsum du. ND 44-5 and UTM 286 read vāratrayam. RASH 47 reads vālatrayam.

n.259Following F rdo rje’i khyim du and Skt. vajragṛhe. D and S read de’i khyim du.

n.260D and S read spyi bo la thal mo rdog pa’i phyag rgya. F reads thal mo rdebs pa’i phyag byas. ND 44-5 reads kaṭasphoti­mudrāṃ. UTM 286 reads kaṭasphoṭā­mudrāṃ. UTM 288 reads kaṭasphoṭī­mudrāṃ. RASH 47 reads kaṭasphoṭī­mudrā. BnFS 84 reads kataspoṭimudrā. BnFS 85 omits. This translation follows the reading in D and S, where this mudrā appears to refers to slapping the deity image on the head. The reading in F simply mentions a “hand-clapping mudrā.”

n.261Following F and S oM k+Sha hUM Ta, as well as ND 44-5, UTM 286, UTM 288, RASH 47, and BnFS 84 oṃ kṣaḥ hūṃ phaṭ. D reads oM yak+Sha hUM TaH.

n.262D, F, and S read ca co. Skt. reads kilikilā.

n.263Following F phaT chen po’i sgra grogs. D reads sa dang pheM chen po’i sgra grogs. S and Y read sa dang phed chen po’i sgra grogs. Skt. omits. This translation follows the reading in F because the readings in the other Tibetan witnesses are syntactically unsatisfactory.

n.264Following UTM 286, and RASH 47 oṃ hrīḥ kṣaḥ amukī āgacchantu yaṃ, as well as BnFS 85 oṃ hrīḥ kaḥ amūkī āgacchantu yaṃ. ND 44-5 reads bho kṣeḥ amuki āgacchantu yam(?). D reads oM hrIHk+ShaHa mu ki A gats+tshana ti yaM. S reads oM hrIH k+Sha a mu kI a gats+tshaM ti yaM. The transliteration of this mantra is emended following the Sanskrit witnesses.

n.265D reads oM stri k+ShIHa mu ka~M aHs+yaHbi tA bi bA ho nam hy+o~M rda~M pa ya swA hAH. S reads oM stri k+ShIH sa mu kaM as+yaH pi tA bi ba ho na ma h+yaM daM pa ya swA hA. F reads oM sti kSi a ma ka a ki a bya pi ta vi ba na na makhyaM dA ba la sva hA. ND 44-5 reads oṃ strī amuki amuki asyā pitā vivāhena mahya(ṃ) dadāya svāhā. UTM 286 and UTM 288 read oṃ strī kṣīḥ amukī amukī asyā pitā vivāhena mahya(ṃ) dadāya svāhā. RASH 47 reads oṃ srī kṣīḥ amukī amukī aśyā pitā vivāhena mayu dadāyā svāhā. BnFS 84 reads oṃ srī kṣīḥ amuki amuki asyā pitā vivāhena madu dadāya svāhā. The transliteration of this mantra largely follows the Tibetan witnesses with the Sanskrit consulted to clarify ambiguities in the Tibetan rendering.

n.266D and S read a mu ha. F reads mu ha. ND 44-5 and BnF84 read paramasivaṃ mūha. UTM 286 and UTM 288 read paramaśivaṃ mūhaṃ. RASH 47 reads paramaṃ śivaṃ mūhaṃ. BnFS 85 reads paramaśivaṃ muhaṃ. This substance could not be identified and has been transliterated here as it appears in D and S.

n.267The Sanskrit witness BnFS 85 concludes at chapter 18, but the material in its chapter 18 appears to be a combination of the opening material in chapter 19 and a fragment of material from chapter 30.

n.268D reads grong khyer gang na lha mos ’dod pa gsum dang ldan pa’i sangs rgyas kyi mtshan nyid byed pa yod pa de rengs par byed pa’i gnas te/ de ’di rjes su ’gro’o. S reads grong khyer gang na lha mo ’dod pa gsum dang ldan pa’i sangs rgyas kyi mtshan nyid byed pa yod pa de rengs par byed pa’i gnas te/ de ’di rjes su ’gro’o. UTM 288 reads trikama­devīnagare yatra buddhasya lakṣaṇa recayanti hi | stambhanapade yadānusaret. RASH 47 reads trikāma­devīnagare yatra buddhasya lakṣaṇa recayati hi | stambhanapade yadānusaret. BnFS 84 reads tikāmadevinagare yatra buddhasya rakṣana racanti hi stabhanapade yadānusaret. This translation is tentative and generally follows D. However, because the Tibetan syntax is problematic, the translation is also informed by the Sanskrit witnesses, particularly RASH 47.

n.269D reads ka Nardu ka ya ra ka’i sim bU ri. S reads kan d+hu ka ya ra ka’i siM bU ri. F reads sman ka Ta hUM ka ra ya ka’i sim b+hi ra. ND 44-5 reads gaṇḍakapāla­kasibira. UTM 286, UTM 288, RASH 47, and BnFS 84 read gaṇḍakapāla­kasimbira. BnFS 85 omits. This translation is tentative and is based emending the reading from the Sanskrit witnesses to gandhakapāla­kaśimbī.

n.270Following D and S phu la Di. ND 44-5 reads phūlati. UTM 286, UTM 288, and RASH 47 reads phūllaṭī. BnFS 84 reads phurati. This substance is unidentified.

n.271Following D be li d+ha. F reads vi li Ta. S reads be la d+hi. RASH 47 reads cyalī. UTM 286 reads vyāli. This substance is unidentified.

n.272Following D and S ka la Da dang ka kSa twi. F reads ka la ya Ta kSi ta. ND 44-5 reads kanāṃrakaṇiṭī. UTM 286 and RASH 47 reads gaṃnārakaṇiṭi. This substance is unidentified.

n.273Following D si dzi dA. F reads sa tsa ra. S reads si dzi rA. ND 44-5 and UTM 286 read sajīra. RASH 47 reads sūgrā. This substance is unidentified.

n.274This substance is unidentified.

n.275Following UTM 286, RASH 47 maṅgala­mahākṛṣṇāṣṭamyāṃ. D and S read bkra shis pa’i yan lag tu nag po chen po’i brgyad la. In the Sanskrit witnesses, it is clear the phrase nag po chen po refers to the particular day of the month and is not a translation of the name Mahākāla.

n.276Following D and S oM buM aHb+hai ra wa swA hA. Skt. reads oṃ sūṃ āḥ bhairava svāhā. The translation of this line is tentative due to ambiguities of the syntax in the Sanskrit and Tibetan sources.

n.277D and S read oM maHhU~M kha kha khA hi khA hi/ mA ra mA ra sarba tra wa ma hA b+hai ra wa tra yats+tshaM tu swA hA. ND 44-5, UTM 286, and RASH 47 read oṁ maḥ hūṁ kha kha khāhi khāhi māraṇa sarvaśatravaṃ mahābhairavaṃ prayaccha tu svāhā. The transliteration of this mantra is informed by the Sanskrit witnesses.

n.278D reads chos gcig gi gong ma la sam bcug gis. F reads tshes gcig gi gong ma la sa phug gis. K and Y read tshes gcig gi gong ma la sam bcug. S reads tshes gcig gi gong ma las sam bcu gcig gis. Skt. omits. This translation is tentative and adopts the reading tshes gcig from F and S.

n.279The object in this section is not clear, but it is presumed to be an effigy of the target of the rite.

n.280Following RASH 47 saptakaṁṭaka. ND 44-5, UTM 286, and UTM 288 read saptakaṭaka. D reads shi sa d+ya kan+Tha rnams. F reads shing sar+ya kan tha ka. S reads shi sa d+ya kaN Tha rnams. This translation is tentative.

n.281The term “target” has been added to the English translation here for the sake of clarity.

n.282Following D shing ba ra ya’i ’bras bu. F reads shing pa la ya’i ’bras bu. S reads shing ba ya ra’i ’bras bu. ND 44-5, RASH 47, and UTM 286 read badarī.

n.283Following D kaN+Da pha la. The identity of this plant is tentative. Alternately, it may be identified with Pueraria tuberosa, commonly known as kudzu.

n.284D and S read phag gi sa. ND 44-5 reads śukladantaṃ. UTM 286 and UTM 288 read śukaradantaṃ. RASH 47 reads śūkaradantaṃ. BnFS 84 reads śukraradantaṃ. This translation is tentative and emends the Tibetan reading to phag gi so (sūkaradanta).

n.285Following D, S, and F sa la yi ge raM gsum.

n.286Following RASH 47 and UTM 286 dantacatuḥ­daṃkāraṃ. ND 44-5 reads dantaracatuḥ­daṃkāraṃ. D and S read sa la yi ge de bzhi. F reads sa la yi ge bzhi. This translation follows the Skt. in reading so (danta) instead of sa, and daṃ instead of de.

n.287Tib. reads dri shim po. Skt. reads sugandha. This translation is tentative.

n.288The translation of this passage, based on the Tibetan, is tentative. The Sanskrit witnesses only approximate the reading given in the Tibetan sources.

n.289Following D and S sgrub pa po gang gis bsgrub pa ni ’dir dgun zla tha chungs kyi nag po’i tshes brgyad la ’dzam bu’i gling du skyes pa’i dus la rnal ’byor pas bsgrubs na dngos grub brgyas de yis ’grub. F reads sgrub pa po gang gi sgrub pa ’di ni/ dgun zla tha chungs kyi nag po’i tshes brgyad la ’dzam bu’i gling du/ nga skyes pa’i dus la rnal ’byor pas bsgrubs na / dngos grub brgyas de ’grub po. ND 44-5 reads yena sidhyanti śādhakāḥ | atra māghamāsi­kṛṣṇāṣṭamyāṃ jambudvīpe asya utpattiḥ | tatra yo yogiṇī aṣṭamahā­siddhi sidhyanti. UTM 286 reads yena sidhyanti sādhakāḥ | atra māghamāsi­kṛṣṇāṣṭamyāṃ jambudvīpe asya utpattiḥ | tatra yo yoginī anuṣṭhet | aṣṭamahā­siddhiḥ sidhyanti. RASH 47 reads yena sidhyanti sādhakāḥ | atra māghamāsi­kṛṣṇāṣṭamyāṃ jambudvīpe asya utpattiḥ | tatra yo yo(?)gī anuṣṭhet | aṣṭamahā­siddhiḥ sidhyanti. This translation is tentative.

n.290D reads oM k+ShaM U U U s+pho Ta s+pho Ta ya/ mA ra ya mA ra ya gar+dz+dza gar+dz+dza ru ta ru ta haHhU~M phaT/ aSh+Ta nA gA nAM kha kha khA hi khA hi/ UHUH. S reads oM k+SaM U U U/ s+pho Ta ya s+pho Ta ya mA ra ya mA ra ya/ gar+dz+dza gar+dz+dza/ ru ta ru ta/ ha hUM phaT/ a ShA nA gA nAM kha kha khA hi khA hi UH UH. RASH 47 reads oṁ kṣaṁ ha ha ha ha sphoṭaya sphoṭaya māraya maraya garja garja turū turū haḥ hūṃ phaṭ aṣṭanāgānāṃ kha kha khāhi khāhi haḥ hoḥ. ND 44-5 reads oṁ kṣa ha ha ha ha sphoṭaya sphoṭaya māraya maraya garjja garjja turū turū haḥ phaṭ | aṣṭanāgānāṃ kha khaḥ khāhi khāhi haḥ hoḥ. UTM 286 reads oṁ kṣaṃ ha ha ha ha sphoṭaya sphoṭaya māraya maraya garjja garjja tura tura haḥ hūṃ phaṭ aṣṭanāgānāṃ kha khaḥ khāhi khāhi haḥ hoḥ. This transliteration follows D with some minor revision based on the Tib and Skt. witnesses.

n.291Following C, K, S, and Y lag par seng ge bsam pa. D and F read lag pa seng ge bsam pa. ND 44-5, UTM 288, and RASH 47 read hastasiṁha dhyāyāt. UTM 286 reads hastasiṃhaṃ dhyāyāt. This translation is tentative.

n.292D reads khar rje ngar gyi khrag blugs te. S reads khar rje ngar gyi khrar blugs te. ND 44-5 reads jaṅghārakta­mukhaṃ prakṣipya. UTM 286 reads jaghārakta­mukheprakṣipe. UTM 288, RASH 47, and BnFS 84 read jaṅghārakta­mukhaṃ prakṣipe. This translation is tentative. Neither the Tibetan nor Sanskrit witnesses make it clear whose mouth or calves are referred to here.

n.293D and S read brun. F reads byi brun. RASH 47 and UTM 286 read indūlamṛttikayā. This translation follows D and S, but it seems apparent from F and the Skt. witnesses that some versions read “mouse dung.”

n.294Following ND 44-5 sphoṭaya. UTM 286, UTM 288, RASH 47, and BnFS 84 omit. D reads s+phA Ta ya. S reads s+pha Ta ya. F omits. This transliteration follows the reading in ND 44-5.

n.295Following ND 44-5, UTM 286, UTM 288, RASH 47, and BnFS 84 pravarṣaya pravarṣaya. D and S read pra sha pra bar+Sha pra bar+Sha. F reads pra ba sha. This transliteration follows the reading in the Sanskrit witnesses.

n.296Following S oM muH haH and RASH 47 oṃ muḥ haḥ. D reads oM huHha. ND 44-5 reads oṃ muha. UTM 286 reads muha muha. UTM 288 and BnFS 84 read oṃ muhaḥ.

n.297F reads spen pa bkrad pa’i le’u, ND 44-5 reads śaniścara­nibandhana­paṭala, and UTM 286 and RASH 47 read sanaiścara­nibandhana­paṭala. D and S read bskrad pa’i le’u. This translation follows F and the Sanskrit witnesses in including Śaniścara in the line spoken by the Blessed One but otherwise follows D and S.

n.298Following Skt. anākāla. D and S read rnyed dka’ ba’i dus su. F reads rnyed par dka’ ba’i dus su. This translation conveys the sense of the Sanskrit term anākāla, which indicates conditions that are “untimely” or “unseasonal” and thus difficult and fraught.

n.299Following RASH 47 vajrāgraye. ND 44-5 and UTM 286 read vajrāgraya. D and S read badz+ra a g+ha ye. F reads badz+ra ar ga ye.

n.300D and S read ka da la sun da la. F reads bkang la bsnun da la. ND 44-5 reads kaṭagaṇḍa. UTM 286 reads kaḍargaṇḍaṣu. RASH 47 reads kaḍarśuṇḍa. UTM 288 reads kadurgaśaṇḍa. BnFs 84 reads kadagaśuṇḍa. This ingredient is unidentified and has been transliterated here as it appears in D and S.

n.301This translation is tentative and follows the reading in RASH 47: pataṅgopadravo bhaved yadā. BnFs 84 reads paṭañjāpadravo bhaved yadā. ND 44-5 reads sarvopadravo bhavedyadā. UTM 286 and UTM 288 read ṣaṭaṅgopadravo bhaved yadā. D reads pe chag pa’i dgra byung na. S reads pe chag pa’i sgra byung na. F reads spe chag pa’i sgra byung na.

n.302D and S read ts+tshe daM yak+Shi. F reads tsh+tsha da ya kShi. ND 44-5 and RASH 47 read cchādaya cchādaya kṣīṃ. UTM 288 reads cchādaya cchādaya kṣī. UTM 286 reads cchādamaya cchādamaya kṣīṃ. BnFS 84 reads cchādaya cchādaya kṣi phaṭ. The transliteration of this term is informed by the Sanskrit witnesses. Otherwise, the mantra follows D.

n.303Tib. reads bar du gcod pa. Skt. reads sarvopadrava. Here we accept the Tibetan term as equivalent to the attested Sanskrit.

n.304D and S read gal te nyan pa dang klog pa mi byed na de’i tshe rnal ’byor pa ’chi ba dang / zhar ba dang / sgur bar bya ba rnams cung zad tsam dang mi ’grub par ’gyur. F reads gal te nyan pa dang klog pa dang / mi byed na de’i tshe rnal ’byor pas ’chi ba dang / zhar ba dang / sgur bar bya ba rnams cung zad tsam yang mi ’grub par ’gyur. ND 44-5 reads yadyavam mantraṃ dṛṣṭvā paṭhitvā śrutvāpi na kṛyante | tadā rogī mṛtyūś ca kāṇūkuṃja na sidhyati. UTM 286 reads yadyanam mantraṃ dṛṣṭvā paṭhitvā śrutvāpi na kriyante | tadā rogī mṛtyūś ca kāṇakubja na sidhyeti. RASH 47 reads yadyena mantraṃ dṛṣṭvā paṭhitvā śrutvāpi na kriyate | tadā rogī mṛtyūś ca kāṇakubja na sidhyati. This translation is tentative and generally follows the Tibetan but is clarified by the Sanskrit.

n.305Following ND 44-5, UTM 286, and BnFS 84 oṃ hrīḥ sarva­satvānu­kampayā hrīḥ hūṃ phaṭ svāhā. RASH 47 reads oṃ hrīḥ sarva­sattvānu­kampayā hīḥ hūṃ phaṭ svāhā. D and S read oM hrIHsarba sa twa na nu paM pa ya hrIHhU~M phaT swA hA.

n.306Following F, H, S, K, N, and Y shan pa. D reads bshen pa. UTM 286 and RASH 47 read cāṇḍālā. UTM 288 reads caṇḍālā. The reading in UTM 286, UTM 288, and RASH 47 indicates “a caṇḍāḷa,” referring to a person from a low caste outside the traditional four-caste system.

n.307Following UTM 286 and RASH 47 nāḍiṃ ghṛdhrena bhakṣayet. Tib. reads bya rgod kyi rgyu ma za ba.

n.308D reads dpyid zla tha chungs san+ta’i. F reads dbyid zla tha chungs las Ta’i. S reads dpyid zla tha chungs san ta’i. ND 44-5 reads vaiśāṣānte. UTM 286 reads vaiśaṣāntam. RASH 47 reads vaiśāṣāvantam. This translation is tentative.

n.309This translation of this passage is tentative and follows the Tibetan witnesses. Both the Tibetan and Sanskrit witnesses present an array of ambiguities that are not easily resolved.

n.310Following Tib. ston zla ra ba. The Sanskrit witnesses consulted all read mārgaśīrṣa, which would be dgun zla ra ba mgo in Tibetan.

n.311Following UTM 286, UTM 288, and RASH 47 suparva. D and S read su sarba.

n.312Following K, N, Y, and S maN+Dala gru bzhi pa bcus par. D reads maN+Dala gru bzhi pa gcus par. F reads dkyil ’khor gru bzhi pa bcas pa. ND 44-5 reads maṇḍalayitvā | caturasram. RASH 47 reads maṇdalaṃ kālayitvā caturasram. UTM 286 reads maṇḍalaṃ kārayitvā caturasram. This translation follows K, N, Y, and S, with bcus pa (’chu ba) understood in the sense of “ladling out” or “pouring out.”

n.313The translation of this passage is tentative.

n.314The transliteration of this Apabhraṃśa passage follows D.

n.315Tib. reads zhing las skyes. Skt. reads kṣetrajam. This is a class of yoginī or ḍākinī that takes birth in a human form. There are multiple types of such beings listed in both Buddhist and non-Buddhist texts, but the Buddhist tradition often employs a threefold typology: those born from sacred spaces, those born from mantra (mantrajā, sngags las skyes), and those born from the natural state (sahajā, lhan cig skyes). All the Sanskrit sources consulted report the masculine/neuter form kṣetrajam instead of the expected feminine kṣetrajām, but the content of the chapter describes only women.

n.316This point marks the conclusion of Sanskrit witnesses ND 44-5, UTM 286, UTM 288, RASH 47, BnFS 84, and BnFS 85. Manuscript BnFS 85 jumps from a fragment of Chapter 19 to a fragment of Chapter 30 on its final folio and mislabels all this material as Chapter 18.

n.317Following F, N, S, and Y dA ya kA. D reads dA yak+Sha.

n.318D and S read gsum ’dres pa’i rnal ’byor dang ldan pa. It is unclear what “the three” refers to.

n.319The transliteration of this Apabhraṃśa passage follows D.

n.320This translation is tentative.

n.321The transliteration of this Apabhraṃśa passage follows D.

n.322Following S oM kA lA ya yak+ShAya. D reads oM kA lA ya ka yak+ShA ya.

n.323Following F a mra’i ’bras bu. D reads aM pa’i ’bras bu. S reads aM ba’i ’bras bu.

n.324Following N and S ka tha ya. D and F read ga kha thA ya.

n.325Tentative for dbugs phyed.

n.326Following S shu tsi ra spu ra ban+d+ha na. D reads shu tsi ra pu wi ra spu ra bar d+ha na.

n.327We have not been able to identify viable Sanskrit equivalents for the Tibetan transliterations of many of the names of the substances that follow. All transliterations follow D unless otherwise noted.

n.328The term badara does not appear in the Tibetan witnesses, but this translation assumes that this rite uses the same substance as the previous rite.

n.329This and the following two chapters are attested in RST15, but they only partially align with what is reported in the Tibetan translation.

n.330Divination practices that rely on the medium of a young girl ( kumārī ), or sometimes a young boy, are well attested in Buddhist and non-Buddhist tantric literature. These rites often make use of a mirror, bowl of water, painted toenail, or other reflective surface, upon which the child sees visions related to a petitioner’s questions. On this practice see Smith 2006, chapters 11 and 12, Vasudeva 2015, and Orofino 1994.

n.331The transliteration of this Apabhraṃśa passage follows D.

n.332This chapter is reported in RST15 but does not have a title. The Tibetan title is mnyams su sbyor ba’i le’u.

n.333The translation of this chapter is tentative and follows D.

n.334D, F, and S read kha dog ’byung ba. RST15 reads varṇavāṃsa. The translation of the title of this chapter is tentative. Only the final procedure in this chapter seems concerned with the complexion.

n.335Following RST15 sāṣṭa. D and S read sA ShA. F reads swA s+thA. This substance is unidentified.

n.336Following N, S, and Y ha pu ri. D reads ha pu ru ru. F reads ha su ri. RST15 reads uparī.

n.337Following F a ka sha mu li and RST15 ākāśamūlī. D and S read a sha mU li.

n.338Following RST15 piṇḍa­tagaramūla. D and S read paN+Da ta ka ra mU la. F reads ka ra mu la.

n.339Following RST15 guḍamārga. D and S read gsang ba’i lam. F reads gsangs ba’i gnas.

n.340Following F lus lus la byugs na, which describes the application of this substance to the body. D and S omit. RST15 is illegible in places and possibly corrupt, but it seems to report a similar statement.

n.341The term “water” is added to the English translation for the sake of clarity. Neither the Tibetan nor Sanskrit sources specify what is to be incanted.

n.342This is the final line of the photo reproductions of RST15, to which we had access for this translation.

n.343D reads che ge mo las, indicating that the target’s name should be declined or expressed in the ablative case.

n.344Following F, H, K, N, Y, and S blun po. D reads blon po.

n.345D and S read zla ba’i sbyor ba bya ba. F omits. The practice referred to here is unknown.

n.346The translation of this sentence is tentative, and the object of the verbs uncertain.

n.347The Tibetan witnesses do not indicate where this response begins, so this phrase has been added to the English translation for the sake of clarity.

n.348Following F myur bar mi rtag pa bstan na. D and S read myur ba ni rtag pa bstan na.

n.349The colophon in F reads lha sa’i gtsug lag khang chen po/ lha sa ra mo cher mkhas pa rnams kyis zhus pa’i don du/ paN+Dita chen pos sa man ta shrI dang / zhu chen gyi lo ts+tshA wa dge slong chos rab kyis bsgyur cing zhus te gtan la phab pa’o/ phal che bas ’di pha rgyud du bshad pa (“This was translated, edited, and finalized by the great paṇḍita Samantaśrī and the great editor and translator Gelong Chörap at the request of the learned ones at Lhasa Ramoché, the great temple of Lhasa. Most say this is a father tantra”).