Notes

n.1On the topic of caste in the Buddhist tradition, see De Jong 1988, Eltschinger 2012, and Silk 2020.

n.2For a detailed analysis of these Buddhist philosophical arguments against caste, see Eltschinger 2012.

n.3It should be noted here that there are several suttas in the Pali canon in which Ānanda emerges from the situation quite differently, perhaps because they depict events that occurred after Ānanda had become an arhat. In the Bhikkhunī Sutta (AN II 145–46), for example, at Kosambī a certain nun, under the false pretense of being ill, asks for Ānanda to visit her. When Ānanda arrives and finds the nun lying in her bed with her robes seductively arranged, he calmly teaches her on giving up craving and desire, at which point she immediately confesses her transgression, which is expressed in much the same way as Prakṛti’s confession in The Exemplary Tale of Śārdūlakarṇa.

n.4Incidentally, Śrāvastī is the setting for several accounts of caste rigidity in the suttas of the Pali canon. See, e.g., the Vasala Sutta in the Suttanipāta (Sn 21–25).

n.5For further references, see Mukhopadhyaya 1954, pp. x–xi.

n.6See DN I 87ff., DN I 235, MN II 200ff., and Sn 594.

n.7This expression could also be rendered as “commonplace designation,” in reference to such Buddhist epistemological thinkers as Dharmakīrti, who explain sāmānya[lakṣaṇa] as an unreal conceptual generality or universal that the mind superimposes on real particulars. It seems, in fact, that sāmānyasaṃjñā is an expanded Sanskritization of the Pali samaññā (“designation”), the term used in the Vāseṭṭha Sutta (Sn 108), for instance, and whose Sanskrit equivalent is actually samājñā.

n.8This attribution to Aśvaghoṣa is doubtful, however, in view of the fact that the Chinese translation ascribes the text to Dharmakīrti, the well-known Buddhist epistemologist who taught at Nālandā during the sixth or seventh century. See De Jong 1988, pp. 426–27 for a brief discussion on the uncertainty of both attributions. For a full edition and translation of the Vajrasūcī, see Mukhopadhyaya 1960.

n.9For a discussion of the historical place of the astrological ideas presented in The Exemplary Tale of Śārdūlakarṇa, see Kotyk 2018, pp. 151–54.

n.10See The Chapter on Going Forth (Toh 1, ch 1) for the full story of Śāriputra’s origins.

n.11For an overview, see Zinkgräf 1940.

n.12This attribution to An Shigao has been questioned by some scholars, who instead place the text in the third or fourth century on account of the idiom of the translation. See Karashima and Vorobyova-Desyatovskaya 2015, p. 257, n. 148.

n.13For an English translation of this version, see Mukhopadhyaya 1954, pp. 213–17.

n.14A full translation in Tibetan, made from Chinese in the eighteenth century, is included in the Narthang, Urga, and Lhasa Kangyurs, but not in the Degé. For this episode see Lhasa Kangyur, (mdo sde, wa), F.386.b ff.

n.15For an English translation, see Giebel 2015. As noted by Giebel (p. 31), the attribution of this translation to Zhi Qian has been questioned by the Japanese scholar Tomojirō Hayashiya on the basis of the translation idiom and the records of the Chinese scriptural inventories. Hayashiya instead argues that this translation was made by an as-yet-unidentified translator during the second half of the fifth century or later, after the time of Kumārajīva (344–413 ᴄᴇ).

n.16Cf. Kotyk 2018, p. 152.

n.17See Bronkhorst 2016 and Joseph 2018, especially the epilogue.

n.18For a useful overview of the shared and added chapters, see Mukhopadhyaya 1967, pp. 61–64. Some of the added chapters are summarized in Sharma 1992.

n.19These manuscript fragments were discovered during the nineteenth century in a place called Kugiar, now known as Kekeya, which lies on the southern rim of the Tarim Basin in present-day Xinjiang province, China. Only recently were they put together in a complete edition, for which see Miyazaka et al. 2015. Because this more Prakritic version of the text differs quite significantly from the Nepalese Sanskrit version and the Tibetan translation, we have made only very occasional reference to it.

n.20The Tengyur contains a related text titled An Explanation of the Nature of the Planets and Astrological Houses Found in [the Life Story of] Ārya Śārdūlakarṇa (Arya SardulakarNa’i nang nas ’byung ba gza’ dang rgyu skra gyi rang bzhin bshad pa), which consists of a short commentary of just over thirty pages on the astrological section of the Śārdūlakarṇāvadāna. This text is not mentioned in the Tōhoku catalog, but it is found in the Technology and Arts (bzo rig pa) section of the Degé and other Tengyurs and has been newly labeled by 84000 as Toh 4321a.

n.21NGMPP A 38-14.

n.22For ease of reference, we have provided the page numbers to Mukhopadhyaya’s Sanskrit edition within the text of the translation with the siglum [M.].

n.23See Burnouf 1844, pp. 183–87. For a recent English translation, see Burnouf 2010, pp. 222–25.

n.24See Zinkgräf 1940, pp. 126–27.

n.25Many renditions of this piece can be found on YouTube. See for example this 2018 performance at Visva-Bharati University in Shantiniketan (accessed March 2, 2024).

n.26The Sanskrit manuscript kept at the Asiatic Society of Bengal (one of the two Nepalese Sanskrit manuscripts included in Mukhopadhyaya’s edition) starts with oṃ namo ratna­trayāya (“ Om. Homage to the Three Jewels!”

n.27The Tibetan translation lacks a rendering of the Sanskrit pāsyāmi (“I would like to drink”), but the Sanskrit reading here is supported by both Chinese translations.

n.28The extant Sanskrit manuscripts again contain the verb pāsyāmi (“I would like to drink”), but here both the Tibetan and the Chinese translations lack an equivalent rendering.

n.29We have followed the Tibetan reading der song ste phyin nas, on the basis of which Hiraoka (2010, p. 54) has rightly proposed emending the Sanskrit edition to tenopasaṃ­krāntaḥ | upasaṃ­kramya.

n.30While in the Degé Kangyur we find the phrase dpyod kyis (“through your thinking”), and in the Narthang Kangyur spyod kyis (“through your practice”), the Stok Kangyur lacks any such phrase, which is in agreement with the extant Sanskrit manuscripts and the Chinese translations.

n.31The Tibetan lacks a rendering of the Sanskrit vocative putri (“my child”).

n.32We have followed the Tibetan dge sbyong gau ta ma ni ’dod chags dang bral ba thams cad zil gyis gnon zhes thos so, except that we have understood the underlying verb śrūyate (“I have heard”) as only applying to the statement that the ascetic Gautama is free from desire. The reading of the extant Sanskrit manuscripts vītarāgasya punaḥ sarva­mantrān abhi­bhavanti is faulty, and in his edition Mukhopadhyaya (1954, p. 2) has tried to emend this by adding mantrāḥ after vītarāgasya. However, we do not find any mention of mantras here in either the Tibetan or the Chinese translations. Hiraoka (2010, p. 54) has pointed out that Cowell’s (1886) emendation to vītarāgas sa does not work given the plural ending of the verb, and he has therefore opted for Mukhopadhyaya’s reading. It seems to us, however, that the plural verb is more likely to be a later, erroneous scribal emendation. The Tibetan rendering suggests vītarāgaś ca punaḥ sarvam abhibhavati as the underlying Sanskrit reading, or perhaps sarvasattvān instead of sarvam, as this is what seems to underlie 一切眾生 in Dharmarakṣa’s Chinese translation.

n.33The Tibetan has rendered the Sanskrit madhye gṛhāṅganasya (“in the middle of the courtyard”) as khang pa’i dbus (“in the middle of the house”).

n.34The Tibetan transliteration of this spell is clearly faulty, so we have tried to reconstruct the underlying Sanskrit form of the spell on the basis of the readings found in the extant Sanskrit manuscripts (in which the spell contains several added words). It can be translated as follows: “O stainless one, immaculate one, saffron one, good-minded one! O lightning, by which you are bound! According to wish, the god rains, strikes lightning, and thunders forth so as to make the great king’s perplexity increase toward gods, humans, and gandharvas. O gods of celestial bodies with flaming tails, O gods of celestial bodies that are tailless, I offer oblations so that Ānanda comes and proceeds! Svāhā!” Hiraoka (2010, p. 54) has followed Mukhopadhyaya’s emendation of the extant Sanskrit buddho to baddhā, but to us baddho seems in better agreement with the Tibetan ban dho and the transliteration 非頭 in Dharmarakṣa’s Chinese translation.

n.35The Tibetan lacks a rendering of the Sanskrit sayyām (“bed”) and has translated prajñapaya (“prepare”) as shes par gyis shig (“you should know”).

n.36All the extant Sanskrit manuscripts read anīti (“freedom from calamity”), which is also the underlying reading of the Tibetan transliteration. In his Sanskrit edition, Mukhopadhyaya (1954, p. 3) has emended this to sunītiḥ (“good conduct”), apparently on the basis of Dharmarakṣa’s Chinese translation. Zhi Qian’s translation, however, reads the same as the Sanskrit and the Tibetan. We see no reason for Mukhopadhyaya’s emendation when anīti, to be understood as an-īti (“non-calamity”), makes good sense.

n.37This entire “utterance of truth” (Skt. satyavākya) has been transliterated in the Tibetan translation, but the rendering found in the different Kangyurs shows several mistakes and elisions. We have therefore followed the extant Sanskrit manuscripts, which appear to have correctly preserved this verse and whose readings everywhere seem to underlie the Tibetan transliteration (including taṃ at the beginning of the third line, which Mukhopadhyaya has emended to tad in his edition). The full utterance can be translated as follows: “May there be stability, freedom from ruin, freedom from calamity, and well-being for all living beings! “A clear lake without blemish, calm and without peril all around, Where calamities, perils, and disturbances become pacified‍— To that, indeed, gods and fully accomplished yogins pay homage. By this utterance of truth may the monk Ānanda be well!”

n.38The Tibetan ’jig rten na sngags mthu dang ldan pa’i sngags gang yin pa indicates that the Sanskrit mantrāḥ (“mantras”) was rendered twice. We have followed the reading of the extant Sanskrit manuscripts mantrāḥ sarvalokasya prabhavanti.

n.39The Tibetan adds bcom ldan ’das la tshe dang ldan pa kun dga’ bo la bka’ stsal ba (“The venerable Ānanda then asked the Blessed One about the incident”), but this does not have an equivalent either in the extant Sanskrit manuscripts or in the Chinese translations.

n.40The Tibetan lacks a rendering of “Śākyamuni.”

n.41We have reconstructed the mantra according to the form that underlies the Tibetan transliteration, which, despite several omissions and misplacements of syllables, largely agrees with the transliteration found in Zhi Qian’s Chinese translation. Several of the readings in the Tibetan transliteration are also confirmed by what seems to be the oldest Nepalese Sanskrit manuscript of the text (NGMPP A 38-14). The mantra can be translated as follows: “ Aṇḍare, paṇḍare! O bracelet, O one on the forearm, O one whose neck has a string, O lady of kin, contain the poison! Cili mili! When inclined to grant according to what is bestowed, O lord of globes, make the boil dissolve!” It may be noted that the mantra is longer than the initial six syllables of aṇḍare paṇḍare, a combination of words that often figures at the beginning of such protective mantras. After paṇḍare, which is somehow lacking in the Tibetan transliteration, the extant Sanskrit manuscripts add karaṇde or kāraṇde, probably due to the influence of equivalents in similar protective mantras in the Mahā­māyūrīvidyārājñī (see The Queen of Incantations: The Great Peahen , 2023) and the Mahā­sāhasra­pramardanī (see Destroyer of the Great Trichiliocosm , 2016). See also Strauch 2014, pp. 75–81. In addition, these later Nepalese manuscripts include an entire string of words between cili mili and sātinimne, which has been adopted by Mukhopadhyaya in his edition (1954, p. 5).

n.42The extant Sanskrit manuscripts add pari­bhāṣaṇārho romaha­rṣaṇena romaha­rṣaṇārhaḥ punar eva mucyate (“If one is to be reprimanded, one will be released with having one’s hair standing on end; if one is to have one’s hair standing on end, one will again be released from that”). Both the Tibetan and the Chinese translations, however, lack an equivalent for this.

n.43The Tibetan lacks a rendering of the Sanskrit sāsurāyām (“with their asuras”), which has an equivalent rendering in both Chinese translations.

n.44The extant Sanskrit manuscripts read śiraḥsnātā (“washed her head”), but the Chinese translations agree with the Tibetan.

n.45The extant Sanskrit manuscripts add āyuṣmantam ānandam āmantrayamāṇā (“trying to speak to the venerable Ānanda”), which has no equivalent either in the Tibetan or Chinese translations.

n.46The Tibetan adds de nas gdol ba’i bu mo gzugs bzang mo tshe dang ldan pa kun dga’ bo’i phyi bzhin du rjes su ’brang ngo (“And the outcaste girl Prakṛti kept following the venerable Ānanda closely from behind”), but neither the extant Sanskrit nor the Chinese translations contain this sentence.

n.47The Tibetan lacks a rendering of the Sanskrit tiṣṭhantam anutiṣṭhati (“stopping wherever I stop”), but both Chinese translations contain this phrase.

n.48In the extant Sanskrit manuscripts this dialogue begins differently: “Then the Blessed One asked the outcaste girl Prakṛti, ‘Prakṛti, what is it that you want from the monk Ānanda?’ ‘Venerable Sir, I want Ānanda as my husband,’ Prakṛti replied.” Dharmarakṣa’s Chinese translation reads like the extant Sanskrit, but Zhi Qian’s Chinese translation agrees with the Tibetan, which is what we have adopted here.

n.49The Tibetan seems to have translated the Sanskrit tena hi sam­mukhaṃ mamānu­jñāpaya tvam (“Then you must have them give their permission in my presence”) as bdag gis slar yang go bar bgyi’o (“Then I in turn give my consent”). The Chinese translations are in agreement with the Sanskrit.

n.50The extant Sanskrit manuscripts add prakṛtim apahāya (“leaving behind Prakṛti”), but both the Tibetan and the Chinese translations lack this phrase.

n.51In this sentence most extant Sanskrit manuscripts contain a lengthy addition that is not found in the Tibetan and the Chinese translations, nor in the oldest Nepalese manuscript (NGMPP A 38-14): atha bhagavān yat tasyāḥ prakṛter mātaṅga­dārikāyāḥ pūrvasañ­citāpāya­durgati­gamanībhūtaṃ tat sarvaṃ pāpaṃ sarva­durgati­pariśodhanyā dhāraṇyā niravaśeṣeṇa pari­śodhya mātaṅgajāter vimocayitvā śuddha­prakṛtim (instead of just prakṛtim) (“Then, after completely purifying, without remainder, by means of the sarva­durgati­pariśodhanī dhāraṇī, all the sin that the outcaste girl Prakṛti had accumulated in the past that leads to a lower and unfortunate rebirth, and after liberating her from her outcaste birth, the Blessed One [said] to the purified Prakṛti…”).

n.52The Tibetan has taken the Sanskrit bhagavatā with evam ukte: “When this was said by the Blessed One.” However, the syntax of the Sanskrit does not allow for this interpretation.

n.53The Tibetan has rendered the Sanskrit kāmaiś cādīnavam (“about the dangers associated with sense pleasures”) as ’dod pa’i ro myang bar bya ba ma yin zhing smad pa’i gtam (“a talk about not indulging in sense pleasures and deprecating them”).

n.54The Tibetan lacks a rendering of the Sanskrit dharmyayā kathayā (“with a dharmic talk”).

n.55The Tibetan bcom ldan ’das kyi bka’ gang yin pa shes nas indicates that the translators read bhagavato deśitam ājñātum (“to understand what the Blessed One has taught”), where the extant Sanskrit manuscripts read bhavyā dharma­deśitam ājñātum. We have followed the Sanskrit, since the phrase recurs in the following sentence.

n.56The Tibetan has rendered the Sanskrit sāmutkarṣikī (“most elevated”) as rab tu ldan pa, and it has erroneously translated the Sanskrit pratibalām (“able”), which is here synonymous with bhavyā (“ready”), as bsgoms pa’i stobs (“power of meditation”).

n.57The Tibetan lacks a rendering of the Sanskrit caturāryasatya- (“the four truths of the noble ones”).

n.58The Tibetan has translated the Sanskrit rajanopagatam (“fit for dyeing”) as dri ma med pa (“stainless”), which suggests that the translators read rajanāpagatam.

n.59The Tibetan lacks a rendering of the Sanskrit akopyadharmā (“had become unshakable in the Dharma”).

n.60The Tibetan lacks a rendering of the Sanskrit arthalābha­saṃvṛttā (“had turned toward the attainment of the goal”).

n.61The Tibetan has rendered the Sanskrit vaiśāradya­prāpta (“had attained complete confidence”) as dri ma med pa thob pas (“had attained the immaculate”).

n.62The Tibetan lacks a rendering of the Sanskrit ājāneyamānā dharmeṣu (“becoming thoroughbred in the teachings”).

n.63The Tibetan omits this sentence.

n.64It seems that the Tibetan has taken the Sanskrit ātyatyāṃ saṃvarāya sthitvā (“as you stand firmly in refraining from transgression”) with the previous sentence.

n.65The Tibetan adds zhing tshul bzhin ma yin pa byas pa (“and acted improperly”), but this is probably due to having read svāmivāde na sam­udācarīti instead of svāmivādena sam­udācarīti. Since the Buddha is simply restating what Prakṛti had said before, we have followed the way it is worded in the preceding passage.

n.66The Tibetan adds chos ’dul ba rtogs nas (“having understood the Dharma and Vinaya”) and several more explanatory phrases that seem to have been added during the translation process.

n.67The Tibetan has rendered the Sanskrit prati­kāṅkṣitavyā (“you should expect”) as bsgrub par bya (“you should achieve”).

n.68The extant Sanskrit reading prahitāni viviktāni viharati sma has to be emended to prahitātmanī viviktā viharati sma in the light of other such descriptions beginning with vyapakṛṣṭa (see BHSD s.v.). The Tibetan reads rab tu nyams pa med pas bdag nyid gnas te (“she dwelled by herself without any failings”), which suggests that the translators read aprahatātmanī. The Tibetan then repeats the entire description of retreating into seclusion at the beginning of the next sentence, but the Sanskrit yadarthaṃ…tad construction excludes such an interpretation.

n.69In this set proclamation of an arhat, the Tibetan translation lacks a rendering of the Sanskrit kṣīnā me jātir (“Birth has come to an end for me”) and prajānāmi (“I know”).

n.70The Tibetan has misrendered the Sanskrit mahāśālakuleṣu as shing sA la chen po lta bu rnams (“who are like great śāla trees”) and pravekṣyati (“could enter”) as blta bar bya (“can be seen”). The same misrenderings occur in the following passage.

n.71The extant Sanskrit manuscripts here add “kṣatriyas,” but this is lacking both in the Tibetan and in Dharmarakṣa’s Chinese translation. In an additional note in his subsequent study (1967, p. 77), Mukhopadhyaya states that it should be omitted in his edition.

n.72The Tibetan has rendered the Sanskrit svakasvakāni mātāpaiṭrkāṇi nāma­gotrāṇi (“their maternal and paternal names and lines of descent”) more elaborately as rang rang gi pha mo’i rgyus dang / ming dang / rigs dang / rus rnams (“their paternal and maternal backgrounds, names, lines of descent, and bloodlines”).

n.73The extant Sanskrit manuscripts here add “kṣatriyas,” but this is lacking both in the Tibetan and in Dharmarakṣa’s Chinese translation. In an additional note in his subsequent study, Mukhopadhyaya (1967, p. 77) states that it should be omitted in his edition.

n.74Instead of the sal tree, which is rendered as sA la in the Tibetan, the extant Sanskrit lists several other trees in a long compound: ati­muktakakadalīpāṭa­lakāmalakīvana­gahana­pradeśe (“an area thick with forests of atimuktaka trees, plantain trees, trumpet-flower trees, and gooseberry trees”). Zhi Qian’s Chinese translation only makes mention of the atimuktaka tree, whereas Dharmarakṣa’s translation does not mention any specific kind of tree.

n.75The Tibetan has rendered the Sanskrit anuśrutam (“as traditionally passed down”) as rten pa med par (“without reliance”), which suggests that the translators read aniśritam.

n.76The Tibetan has here rendered the Sanskrit paramayā śubha­varṇa­puṣkalatayā as kha dog mdzes shing mchog tu kha dog rgyas pa (“resplendent complexion and supreme excellence of complexion”), thus translating varṇa (“complexion”) twice. Further below, however, while rendering the same expression when the brahmin maiden Prakṛti is described, the first kha dog is rightly omitted.

n.77The Tibetan has rendered the Sanskrit sasaptotsadaṃ as khyab par sa, which seems to mean “having abundant land.” The Sanskrit expression, however, appears to be the result of an erroneous Sanskritization of the Prakritic sattussada, which actually means “abounding in beings.” See BHSD s.v. utsada (3).

n.78The Tibetan lacks a rendering of the Sanskrit tṛṇa (“grass”) in this set description of an agriculturally prosperous place, and it has rendered the Sanskrit brahmadeyaṃ dattam (“given as a brahmic gift”) as bram ze de la byin nas (“given to that brahmin”). For a brief discussion of such endowments to brahmins, see Mukhopadhyaya 1967, p. 73.

n.79The Tibetan has omitted several components in this description of the brahmin’s learnedness.

n.80The Tibetan lacks the preceding part of Triśaṅku’s reflection, having apparently conflated it with the preceding description of the brahmin and his daughter.

n.81The Tibetan has rendered the Sanskrit kāryaṃ as bya ba cung zhig (“a small matter”), thus adding cung (“small”).

n.82In this context we consider the English “dear” the most suitable rendering for bringing out the nuances of the Sanskrit vocative bhoḥ, which is a polite form of address among cultured persons, especially brahmins, but which takes on a pejorative note when it is used toward someone deemed inferior and lacking education. Thus, being an outcaste, King Triśaṅku is here reproached by the brahmin Puṣkarasārin for using a form of address that only those properly educated are entitled to use. In the Tibetan translation the Sanskrit bhoḥ has been rendered with the particle kye, often translated as “O” or “Hey!” though in Tibetan this form of address tends to be used toward someone higher than oneself.

n.83The Tibetan has rendered the Sanskrit kulaśulkam, which literally means “family fee,” as rin (“price”).

n.84For the following stanzas we have followed the extant Sanskrit. In the Tibetan translation these verses have been rendered into prose, but they show several omissions and misinterpretations.

n.85The Tibetan lacks a rendering of these last two lines. To place a mustard seed on the tip of a hair is here used as a metaphor for trying to do the impossible.

n.86The Tibetan lacks a rendering of this line. This metaphor returns a few stanzas further on, again preceded by the admonition of soliciting the unsolicitable. It also occurs there in Dharmarakṣa’s Chinese translation, while here in this verse another simile is given: “Do not solicit the unsolicitable, like sowing seeds in water!”

n.87The Tibetan has translated the Sanskrit jugupsitaḥ sarvaloke kṛpaṇaḥ puruṣādhamaḥ as ro dang rigs dman pa gnyis ni ’jig rten thams cad la mi gtsang ba yin gyis (“Both a corpse and a low-caste person are impure to the entire world”). To translate kṛpaṇa (“miserable one”) as “corpse” is a stretch, but perhaps this rendering was influenced by the argument in Triśaṅku’s response that a brahmin’s corpse is as repulsive as anyone else’s. Dharmarakṣa’s Chinese translation agrees with the more straightforward interpretation we have followed here.

n.88The Tibetan lacks a rendering of this line.

n.89The Tibetan byad dang ro langs kyi las (“actions for spirits and zombies”) seems to indicate that the underlying Sanskrit was karmāṇi praitāni, as suggested by Mukhopadhyaya (1954, p. 19, n. 2).

n.90The Degé and the Stok reading is rig byed bzhi las, but the Yongle and Kangxi reading rigs byed bzhi las is in agreement with the Sanskrit caturvidhāḥ. The Tibetan translation here lacks an equivalent for the Sanskrit brāhmaṇeṣu (“for brahmins”), so perhaps the variant reading rig byed (“[knowing] the Vedas”) is the result of an attempt to remedy this. The Sanskrit pātaka (“downfall”) has the connotation of “falling from one’s caste,” which in this case would mean losing the brahmin status. For more on these downfalls, one may consult the Manusmṛti (X 235; XI 54), as stated in an additional note to this edition by Mukhopadhyaya (1954, p. 223).

n.91The Tibetan has translated the Sanskrit brahmaghnatā (literally “brahmic killing”) as sdom pa’i tshogs ni ’jig pa ste (“breaking one’s set of vows”), apparently having understood the Sanskrit as referring to the vow of chastity (brahmacarya) in view of the preceding downfall.

n.92In an additional note to his edition, Mukhopadhyaya (1954, p. 223) has pointed out that this observance is referred in the Manusmṛti (XI 72, 105, 122).

n.93The Tibetan has translated the Sanskrit aurasā as lus kyi stod (“upper part of the body”), apparently from uras (“bosom”), but here the adjectival aurasa means “own son,” with the connotation of “legitimate son,” as described in brahmanical law books such as the Manusmṛti (IX 166).

n.94The Tibetan has translated this sentence as “O inferior fellow, do you not see that you belong to this fourth caste?” The extant Sanskrit preserves the correct reading, because outcastes are considered to fall outside the varṇa system, below the category of śūdras. The two Chinese translations read in agreement with the Sanskrit.

n.95The Tibetan has translated this as “It is by you that this entire world is created, proclaiming that it is held that ‘We are his foremost sons’ ” (’dir khyed kyis ’jig rten thams cad sgrub par byed pa de las bdag cag skyes pa yin pas thu bo bdag cag yin).

n.96We have followed the Tibetan reading lag pa dang bcas pa (“having hands”), which is also the reading in Dharmarakṣa’s Chinese translation. The extant Sanskrit reads sanakhāḥ (“having nails”).

n.97The following three verses have close parallels in the Pali Vāseṭṭha Sutta (Sn 608–11), as pointed out by Mukhopadhyaya in the additional notes in his subsequent study (1967, p. 77).

n.98For this verse we have followed the reading of the extant Sanskrit manuscripts, even though the word jātiḥ in the second line seems grammatically problematic: yathā hi jātiṣv anyāsu liṅgaṃ jātiḥ pṛthak pṛthak | sāmānyaṃ kāraṇaṃ tatra kiṃ vā jātiṣu manyase. Mukhopadhyaya (1954, p. 24) has tried to fix this in his edition by emending jātiḥ to yoniḥ, and he appears to have done so on the basis of the Tibetan rendering of this verse: de ltar rang gi bsod nams rigs / skyes pa bud med mtshan mas phye / rkang pa lag pa ’thun pa na / de la rigs su gang gis ’dod (“When thus by one’s merit, clan, and male or female characteristics there is distinction, but the feet and hands are common, then why do you think in terms of castes?” It is clear, however, that this Tibetan rendering is marred by at least two misreadings of the Sanskrit, with jātiṣv anyāsu apparently having been read as jātiḥ puṇyāsu and kāraṇaṃ as kara (“hand”), so we cannot be certain that the second line of the verse in fact had the Sanskrit equivalent yoniḥ for bu med mtshan. Unfortunately, neither Chinese translation can provide further clues here, since they both lack this set of verses. The wording of this particular verse seems to have been strained from the outset, since it is an adaptation of a verse that occurs prior to the preceding parallel verses in the Vāseṭṭha Sutta (Sn 607), in which jātīsu in the first line refers to the different nonhuman animal species: yathā etāsu jātīsu liṅgaṃ jātimayaṃ puthu / evaṃ n’atthi manussesu liṅgaṃ jātimayaṃ puthu (“While among these kinds of birth, the particular distinctive marks are there at birth, there are no such particular distinctive marks there at birth among human beings”).

n.99We have followed the reading of the extant Sanskrit manuscripts: dharmādhipatyāḥ pravarā manuṣyāḥ. In the Tibetan this entire line is rendered as mi rnam la mchog tu gyur pa’i chos kyi bye brag ni med do (“Among human beings there is no distinction in the Dharma being sovereign”). In Dharmarakṣa’s Chinese translation, however, there is no presence of negation; perhaps the Tibetan translators read dharmādhi­patyāpravarā manuṣyāḥ. On the basis of the Tibetan, Mukhopadhyaya (1954, p. 25) has emended the Sanskrit to dharmādhi­patyaprabhavā manuṣyāḥ (“Humans are distinguished by the sovereignty of the Dharma”), but in our opinion pravarā seems more appropriate here in view of the brahmin’s previous assertion that he is in the “preeminent caste” (pravare varṇe).

n.100The Tibetan reads bram ze rnams dang gcig yin te (“brahmins are one”), thus seemingly in agreement with the reading of the extant Sanskrit manuscripts brāhmaṇa ekajātya. However, Dharmarakṣa’s Chinese translation reads, “When you say that Brahmā is of one body, then all his creatures are of one kind. When you speak of one substance, then we, too, are the same. Why is that? You said that Brahmā created the world and all its beings.” And Zhi Qian’s Chinese translation also reads “Brahmā” instead of “brahmins” at the beginning of this passage, which makes better sense. This seems to indicate that the original Sanskrit reading here is not brāhmaṇa but brahmā , which is the emended reading given by Mukhopadhyaya in his edition (1954 p. 25, n. 10).

n.101The extant Sanskrit lacks “eyes.”

n.102The extant Sanskrit lacks “dogs.”

n.103The Tibetan has translated this as “Therefore they are the same as brahmins” (de lta bas na ’di dag bram ze rnams dang ’dra zhing gcig pa yin no).

n.104The Tibetan has “those bearing fruit” (’bras bu dang ldan pa).

n.105The Tibetan here reads me tog dang (“and in flowers”), but this phrase is then repeated later in the enumeration where the Sanskrit reads puṣpataś ca. In the following passages the Tibetan has me tog tsam dang where the extant Sanskrit manuscripts read phalgutaś ca (“and in sapwood”), which is an expression commonly used in the description of trees. We have adopted it here accordingly.

n.106The Tibetan translation adds “One discerns a marked difference between flowers that do not bear fruit” (’bras bu med pa’i me tog rnams sna tshogs par rig par bya’o).

n.107The Tibetan lacks mention of the syandana tree.

n.108The Tibetan has translated this twice as a ma la and as skyu ru ra.

n.109The Tibetan has transliterated this as arushaka.

n.110The extant Sanskrit lacks mention of the white water-lily (kumuda).

n.111Dharmarakṣa’s Chinese translation lacks the following section, up until the comparison of the man with four sons (1.­159).

n.112In the Tibetan these lines are wrongly translated as “the mountains are his stomach, the earth his thighs and feet.” At the very end of the text these verses figure again (1.­727), and there the Tibetan rendering is correct.

n.113In an additional note to his edition, Mukhopadhyaya (1954, pp. 225–26) provides several references to related passages in brahmanical literature: the Ṛgveda (X 90, the Puruṣasūkta), the Vājasaneyi­saṃhitā (XXXI), the Atharvaveda (X 7.32–34, XIX 6), the Muṇḍakopaniṣad (II 1.4), and the Bhagavadgītā (XI 19).

n.114We have followed the reading of the Sanskrit manuscripts, parīkṣasva tvaṃ brāhmaṇa svalakṣaṇataḥ, which in the Tibetan is rendered as tshangs pa rang gi mtshan nyid ni de lta yin par rigs sam/ kyod rang gis dpyod cig, the Tibetan translators apparently having read brāhmaṇa as brahmaṇa and having translated parīkṣasva twice.

n.115For this sentence we have followed the Tibetan translation, which has also served as the basis for the emended Sanskrit reading in the edition of Mukhopadhyaya (1954, p. 28). The entire sentence is missing in the extant Sanskrit manuscripts, which instead begin with the argument on rebirth: “Gods go to brahminhood, kṣatriyas are yakṣas, vaiśyas are nāgas, and śūdras are asuras; if being born were in reality like this, if caste were like this, then how would there be any distinction whereby brahmins are brahmins, kṣatriyas are kṣatriyas, vaiśyas are vaiśyas, and śūdras are śūdras?” In Zhi Qian’s Chinese translation, the verses on Brahmā are followed by, “You brahmin, this teaching is a lie. The world is brought about by the karma of sentient beings. How could Brahmā create this thing?” This appears to provide a suitable link to the verses that follow.

n.116The extant Sanskrit manuscripts read svakṛtena karmaṇā (“by the deeds done by himself”), but a more suitable reading, sukṛtena karmaṇā, is suggested by the Tibetan dge ba’i las byas pas (“by having done good deeds”), which we have followed here.

n.117We have followed the extant Sanskrit tadvad tapaś cārṣam iha praśastam. The Tibetan has translated this as “and hence can also become sages who practice austerities for pacification.” It seems the Tibetan translators read praśāntim for praśastam and took the adjectival ārṣam as the subject, plural in number.

n.118Zhi Qian’s Chinese translation lacks a rendering of the following verses.

n.119The Tibetan has dbang po gnyis (“two sense faculties”), whereas the extant Sanskrit reads catuḥśravaṇa (“four ears”), which we have followed here. The Tibetan has also omitted “many legs” in the next line.

n.120In the Tibetan these two lines are rendered as “By ‘father,’ ‘mother,’ ‘brother,’ and ‘relative’ one indicates family,” and the extant Sanskrit manuscripts similarly read mātā pitā bhrātā bandhuḥ for the first line. We follow Mukhopadhyaya’s emendation to mātā bhastrā pituḥ putro yena jātaḥ sa eva saḥ, which seems to be supported by Dharmarakṣa’s Chinese translation. In a footnote in his edition, Mukhopadhyaya (1954, p. 30, n. 13) mentions that this first half of the verse is found in the Harivaṃśa (XXXII 1724) and the Viṣṇupurāṇa (IV 19.2), and in an additional note in his subsequent study (1967, p. 77), he provides further references to the Matsyapurāṇa (XLIX 12) and the Śakuntalā story in the Mahābhārata (I 74,109).

n.121The Tibetan has misread the vocative brāhmaṇa as the subject of the sentence: “Thus, in transmigration, there is no being brahmin.”

n.122The Tibetan appears to have conflated these different questions starting with that on the inner self by rendering them as nang gi zla ba’i dkyil ’khor khyod kyis bklags sam (“Have you studied the internal disk of the moon?”).

n.123The extant Sanskrit lacks an equivalent for the Tibetan byA ka ra Na rnams (“grammars”).

n.124The extant Sanskrit here adds nyāya (“logic”), but this does not have an equivalent in either the Tibetan translation or Dharmarakṣa’s Chinese translation.

n.125We have here followed the Tibetan, which, like Dharmarakṣa’s Chinese translation, lacks any reference to the evening time. The Nepalese Sanskrit manuscripts read te tatra sāyam āsanahetoḥ prāntavāṭikāṃ prātar aśanahetoś ca grāmaṃ piṇḍāya praviśanti sma (“There, they would come to the bordering houses in the evening for a place to stay, and in the morning they would enter the village on alms round for food”). Another Sanskrit version, preserved in an old fragment from Central Asia (Karashima 2009, pp. 338–39), reads te sāyaṃ ca sāyamāśarthaṃ prātaṃ ca prātarāśarthaṃ grāma piṇḍāyām avatarati (“They descended to the village on alms round in the evening for the sake of the evening meal and in the morning for the sake of the morning meal”).

n.126The Tibetan lacks mention of them being brahmins here, but Dharmarakṣa’s Chinese translation reads like the extant Sanskrit versions, which we have followed.

n.127The Tibetan here adds “This is the cause, this is condition, whereby the vaiśya caste appeared in the world.” This sentence, however, is present in neither the Nepalese Sanskrit manuscripts and the Central Asian Sanskrit fragment (Karashima 2009, p. 339) nor Dharmarakṣa’s Chinese translation.

n.128The Tibetan here adds “This is the cause, this is condition, whereby the śūdra caste appeared in the world.” This sentence, however, is present in neither the Nepalese Sanskrit manuscripts and the Central Asian Sanskrit fragment (Karashima 2009, p. 339) nor in Dharmarakṣa’s Chinese translation.

n.129This passage presents a folk etymology for one of the Sanskrit terms for an outcaste, mātaṅga, which here appears to be etymologized as mA taM gAs (“Do not go there / to him”), or possibly as mā taṅga (“Do not stumble”) (cf. BHSD s.v. taṅgati). As stated by Mukhopadhyaya (1954, p. 33, n. 9), the Nepalese Sanskrit manuscripts are quite defective here, which led him to reconstruct the Sanskrit according to the Tibetan text. We should note, however, that the Tibetan translation lacks mention of the “wheel hub” (akṣo) of the wagon, while we find this word both in the Nepalese Sanskrit manuscripts and in Dharmarakṣa’s Chinese translation. It also occurs in the Central Asian Sanskrit fragment (Karashima 2009, p. 339), which reads, abhū (sic) pūrvvaṃ brāhmaṇānyatarāyā vaddhyo (’)nyatarasmiṃ mataṅgaraṇyeyam (sic) vahyamānāyāṃ akṣo bhagno mahāṅga bhagnā mātāṅ[g]a (sic) bhagnā mātaṅga iti sā saṃjñā loke udupādi. Unfortunately, this Sanskrit version shows several defective readings. Karashima (2009, p. 339, n. 13) notes that vaddhyo (“criminal”) is probably a scribal error for vaddhvāṃ, which would be the locative singular of vadhū (“young wife”). In his subsequent study, Mukhopadhyaya (1967, p. 8) has translated this passage as follows: “A man (Śūdra?) abducted another’s (Brāhmaṇa’s?) wife and went to the forest. There his chariot was broken. He could not go back to society. The following injunction was given: mā tvaṃ gamaḥ, ‘Don’t you go (back to society).’ And he was called Mātaṅgama or Mātaṅga (as untouchable).” For further considerations on this passage, see ibid., pp. 65–66.

n.130We have followed the reading of the extant Sanskrit manuscripts, but in his subsequent study, Mukhopadhyaya (1967, p. 14) has proposed to emend Araṇemi to Āruṇi, which is the name of Śvetaketu’s father Uddālaka as mentioned, for instance, in the Kauṣītakī Upaniṣad (I 1). The Tibetan only has “Gautama” in this first instance, but in the next sentence it seems araṇemi is rendered as mnyes byed (“paying respect”), whereas in the related passage at the end of the text (1.­731), the name is literally translated as rtsib kyi mu khyud (“rim of spokes”).

n.131We have followed Mukhopadhyaya’s emendation paṅkti , even though the reading of the extant Sanskrit manuscripts pakṣi appears to underlie the Tibetan phyogs. It is not clear which of the two would be more suitable here, given that there are no corresponding statements in brahmanical literature. For an extensive comparison of brahmanical sources, see Mukhopadhyaya 1967, pp. 9–18, 21–26.

n.132We have followed Mukhopadhyaya’s corrections of this passage as given in his subsequent study (1967, pp. 14–15).

n.133That is, it may be inferred that the many distinctions within the brahmin caste came into being at a later point, as is the case with the caste distinctions among the population at large. What might also be suggested here is that the brahmin should infer from this detailed description that King Triśaṅku is fully knowledgeable in all aspects of the brahmanical tradition.

n.134Here and in the following sentence we have followed the reading of the extant Sanskrit manuscripts, daśadhā bhinnā, which is supported by Zhi Qian’s translation and is the reading adopted by Mukhopadhyaya in an additional note (1967, p. 77). In both instances the Tibetan reads, “divided into twenty-two.”

n.135That is, despite the many divisions within the brahmin caste, they still consider themselves as belonging to one group and suitable for intermarriage.

n.136The Tibetan translation runs differently here, and it seems the translators read praśānta instead of praśaṃsā and citta instead of vitta: “For one who guards his ethical conduct and strives after intelligence, there is threefold happiness: a calm mind, steadfastness, and attaining the blissful states in the world beyond.” We have followed the Sanskrit reading, which is supported by the rendering in Dharmarakṣa’s Chinese translation.

n.137We have followed the extant Sanskrit reading darśanaṃ sarva­vedānāṃ, even though the phrase seems somewhat unusual. The Tibetan properly has lta ba for darśanaṃ but lacks a rendering for sarvavedānāṃ, as is also the case in the Chinese translations. Mukhopadhyaya (1954, p. 39, n. 3) suggests the possibility of emending the Sanskrit to damanaṃ sarva­devānāṃ, presumably with sarvadevānāṃ then being taken with svarga­vratapadāni, but this would likewise be quite unusual.

n.138The following passage has not been properly preserved in any of the extant versions. It should present the names of eight wives of the Vedic seer Kaśyapa, all of whom are all sisters, being the daughters of the creator god Prajāpati Dakṣa, and all of whom are the mothers of the different classes of beings, who, in turn, are therefore all relatives. The Tibetan translation only contains the renderings of six of these sisters. The extant Sanskrit manuscripts omit the second name and read pṛthivī bhūtānāṃ mātā sarva­bījānāṃ marutāṃ at the end of the list, despite the fact that Pṛthivī and Marutā (or Marutvatī) are not wives of Kaśyapa according to brahmanical sources. In his subsequent study (1967, p. 19), Mukhopadhyaya has provided useful background information on this passage and pointed out that the original names are better preserved in Dharmarakṣa’s Chinese translation (though several of its renderings are quite garbled). On the basis of Dharmarakṣa’s renderings, Mukhopadhyaya has offered the following Sanskrit reconstruction (for which we have corrected the typos): aditir devānāṃ, danur dānavānām, ariṣṭā gandharvāṇāṃ, iḷā rakṣasāṃ, surabhiḥ saurabheyānāṃ, vinatā suparṇānāṃ, kadrur nāgānāṃ, surasā sarpāṇāṃ, mahāmahaṃ kāśyapaṃ manasā vidanti (or vindanti) ṛṣayaḥ. We have decided to adopt this reconstruction, since it is clear that by the time of the Tibetan translation this passage had become textually corrupt.

n.139The Tibetan omits these two sublineages.

n.140The Tibetan lacks a rendering of the Sanskrit paurāṇaiḥ (“of yore”), and the translators seem to have misread the Sanskrit ekonapañcāśat (“forty-nine”) as “fifty-one.”

n.141We have followed the reading of the Sanskrit manuscripts na caiṣām ūhāpohaḥ prajñāyate, since this expression also occurs in the previous passage on the divisions in the transmission of the Vedas (1.­181) (Mukhopadhyaya 1954, p. 37).

n.142We have followed the Sanskrit prajāpater hi caikatve. The Tibetan reads skye dgu ’di gcig yin te (“These beings are the same”), having omitted -pater.

n.143We have followed the Tibetan dbang po sna tshogs de dag gis/ bya ba tha dad mthong bar bstan, with which Dharmarakṣa’s Chinese translation is more or less in agreement. The extant Sanskrit manuscripts read sa cendriyāṇāṃ nānātvaṃ kriyāvādena dṛśyate, which has been emended by Mukhopadhyaya (1954, p. 41) to na cendriyāṇāṃ nānātvaṃ kriyābhedaś ca dṛśyate (“There is no difference in faculties, and it is just a division of labor that is seen”).

n.144We have followed the Sanskrit for this verse. The Tibetan reads khyod ltar ’di dag rigs mchog ste// bram ze rgyal rigs de bzhin du// rje’u rigs dmangs rigs de dag la// tha snyad ’jug pa yod ma yin (“According to you, these are of the highest caste, but the conventional ideas regarding the brahmin and kṣatriya castes, the vaiśya caste, and the śūdra caste do not apply.” The translators seem to have read samjñā na pravartate where the extant Sanskrit has samaṃ jñānaṃ pravartate. A similar misreading is found in the Chinese translation by Dharmarakṣa, who seems to have read saṃjñānaṃ pravartate, without a negation. The extant Sanskrit reading seems to be the correct one, both metrically and in view of what follows.

n.145The extant Sanskrit manuscripts read kapiñjalyā janitaḥ, but metrically this falls one syllable short. In his edition, Mukhopadhyaya (1954, p. 42) has emended this to kapiñjalādyā janito on the basis of the Tibetan ser skya la sogs skyes nas ni, but the Tibetan rendering is problematic in having taken kapiñjala as synonymous with kapila. The correct name here is Kapiñjalāda, who is mentioned in a verse of Aśvaghoṣa’s Buddhacarita (4.77, Toh 4156) as the son of sage Vasiṣṭha born from an outcaste woman named Akṣamālā. In an additional note to this edition, Mukhopadhyaya (1954, p. 228) refers to a line in the Mahābhārata in the Tanjore recension (XIII 21) that mentions the brahmic sage Kapiñjalāda as having been born from a caṇḍāla (“outcaste”) woman ( kapiñjalāda brahmarṣiś caṇḍālyām udapadyata).

n.146The extant Sanskrit reads “the Sāvitrī” instead of “this mantra,” but the latter rendering is found in both the Tibetan and Zhi Qian’s Chinese translation. At the beginning of this sentence, the Sanskrit adds ṛddhiparihīnaḥ (“completely deprived of his miraculous power”), but this is absent in the Tibetan and the Chinese translations.

n.147This is the most important mantra in the brahmanical tradition, the so-called Gāyatrī mantra, which is found in the Ṛgveda (3.62). It is directed toward Savitṛ, the personification of the sun’s power, and it can be translated as follows: “ Om. Earth, atmosphere, sky. Let us meditate on what is most exquisite of the vivifier, the brilliance of the god, which may direct our minds.” For a full study on this mantra, see Haas 2023.

n.148The extant Sanskrit manuscripts are clearly corrupt here, and the Tibetan translation is also not without problems. The phrase bram ze sngags ’dis seems to be a rendering of the following sentence, iyaṃ brāhmaṇānāṃ sāvitrī (“This is the Sāvitrī of brahmins”), which is otherwise left unrendered in the Tibetan translation. This sentence occurs both in Zhi Qian’s Chinese translation and in Dharmarakṣa’s translation (where this entire passage is found toward the end of the text at 419a13). It seems to us that the second occurrence of bram ze in the Tibetan renders a misinterpreted vocative brāhmaṇa. Mukhopadhyaya’s Sanskrit reconstruction (1954, p. 45) is therefore unreliable here.

n.149This mantra can be translated as follows: “Sharing in the god’s most excellent food, we will bestow it.” We have followed the reading of the extant Sanskrit manuscripts here, since Mukhopadhyaya’s reconstruction (1954, p. 45) is problematic in its reliance on the Tibetan. In the Tibetan we find the transliteration vaM nya ka a ta sA ka nya ar ta tha phra khri na, apparently starting with vayaṃ and then followed by a rendering of the second line of the Sāvitrī of vaiśyas that follows. It seems this mistake was made because the actual line of the mantra was incorporated into the preceding description by rendering it as dka’ thub byas pas lhas zas mchog byin nas gnas shing sngags ’di blzas so (“being given the most excellent food by a god due to his ardent practice, he uttered this mantra”). The extant Sanskrit manuscripts add another line, sarva­kāma­guṇopetaṃ sukha­bhojanaṃ kṣatriyasya pradāsyāmaḥ (“We will bestow the delightful food of a kṣatriya, which is endowed with all pleasurable qualities”), but this is absent in Zhi Qian’s Chinese translation as well as in the Tibetan. In Zhi Qian’s Chinese translation, which lacks any reference to Prajāpati (“lord of creatures”), this entire passage on the Sāvitrī of kṣatriyas that follows is fully transliterated, and on its basis it seems Mukhopadhyaya’s reconstruction hy ātmārāmas taporataḥ should be replaced with simply prītaḥ. In the additional notes in his subsequent study (1967, p. 78), Mukhopadhyaya cites a verse from the Ṛgveda (V 82.1), from which the mantra here seems to be derived: oṃ tat savitur vṛṇīmahe vayaṃ devasya bhojanam | śreṣṭhaṃ sarvadhātamaṃ turam bhagasya dhīmah i.

n.150Tibetan has here translated sāvitrī as dag byed kyi snags (“purificatory mantra”), whereas in the earlier passage on the syllables and parts it was transliterated as sa bya ti and sa byin tra (1.­221).

n.151This mantra can be translated as follows: “ Om. The vaiśya maiden is brightly bedecked indeed. That maiden is then clever in wealth.” The Tibetan transliteration reads oM tsin tra ma hid dhe be sha ka paM sha sha taM sha tra daM sha dam.

n.152This mantra can be translated as follows: “ Om. Lack of ardor, great ardor. May I live for a hundred years. May I see a hundred autumns.” The Tibetan transliteration reads oM a ta pa pra ta pa dze me dha ha ma. The transliteration in Zhi Qian’s Chinese translation is in agreement with the extant Sanskrit, except that it ends with an inexplicable āmra instead of śaradāṃ śatam. In the additional notes to his edition, Mukhopadhyaya (1954, p. 228) cites a closely related line from the Ṛgveda (VII 66.16) and the Vājasaneyi­saṃhitā (XXXVI 24): paśyema śaradaḥ śataṃ jīvema śaradaḥ śatam.

n.153In his edition, Mukhopadhyaya (1954, p. 46) here adds bhūr bhuvaḥ svaḥ, but this is absent in the Tibetan translation, in Zhi Qian’s Chinese translation, and in the Nepalese manuscript (NGMPP A 38-14).

n.154This mantra can be translated as follows: “ Om. Sense pleasures are supreme to beings in the world. When they are not abandoned, they are obstacles to beings. Therefore you should all abandon sense pleasures. You will then surely attain the incomparable world of Brahmā.” In an additional note to his edition, Mukhopadhyaya (1954, p. 228) points out that the extant Sanskrit reading teṣām prahāṇāya abhūtāntarāyaḥ makes little sense and is moreover metrically defective. In the Tibetan, however, we find the second line transliterated as de sha ma a pra ha na ca bhu bA (= dA in the Urga Kangyur) na ta ra ya, on the basis of which the problematic reading of the extant Sanskrit manuscripts can be emended to teṣām aprahāṇe bhūtāntarāyāḥ, leaving out the ca in the Tibetan transliteration for the sake of meter.

n.155The Tibetan here reads rgyu skar gyi dbang rnams (“the influences of lunar asterisms”), which indicates that the translators read nakṣatravaśaṃ instead of nakṣatra­vaṃśam. The latter is undoubtedly the correct reading here. For an extensive discussion of nakṣatras in brahmanical sources, see Mukhopadhyaya 1967, pp. 28–48, 51–55.

n.156We have followed Mukhopadhyaya’s emendation pañcadaśa (“fifteen”), which is the number found in Dharmarakṣa’s Chinese translation and which seems to be the correct one, given that this is a small asterism of only one star. The extant Sanskrit manuscripts read pañcatvāriṃśan (“forty-five”), while the Tibetan translation has sum cu rtsa lnga (“thirty-five”). Zhi Qian’s Chinese translation does not properly list the number of muhūrtas in this description of the lunar asterisms.

n.157Following the additional note in the edition of Mukhopadhyaya (1954, p. 229), we have adopted the reading of the extant Sanskrit manuscript śūlabhṛd (“Spear Holder”), a name for Śiva. The Tibetan and Zhi Qian’s translation read “the sun,” whereas Dharmarakṣa seems to have read śabda in the Sanskrit.

n.158The Tibetan has rgyas pa (Vāsiṣṭha).

n.159The Tibetan here has “fifteen,” but the extant Sanskrit manuscripts and Dharmarakṣa’s Chinese translation read “forty-five.”

n.160The Tibetan and Zhi Qian’s translation have Bṛhaspati, but they also list this as the deity of Puṣya, the following asterism.

n.161The Tibetan reads dbang byed.

n.162The Tibetan here reads rlung lha (“wind god”).

n.163The Tibetan and Zhi Qian’s Chinese translation here seem to give “Nāga” (Tib. sbrul) as the lineage.

n.164We have followed an additional note in the edition of Mukhopadhyaya (1954, p. 229) that, according to the Taittirīya­saṃhitā and the Taittirīya­brāhmaṇa, the deity should be Bhaga here. The Tibetan skal ldan can be a rendering of this name. The extant Sanskrit manuscripts read pitṛ-.

n.165The Tibetan reads khri (“seat”).

n.166The Tibetan reads rdo sbal kyi sha (“tortoise meat”).

n.167The Tibetan reads bya sna tshogs dang sred (“various birds and millet”).

n.168The Tibet reads pad ma’i snying po.

n.169Instead of all these foods the Tibetan only reads shing tog sna tshogs (“various kinds of fruits”).

n.170The Tibetan reads nyi ma (“the sun”).

n.171The Tibetan reads mon sran sde’u dang ’bras dang mar (“mung lentils, rice, and ghee”).

n.172The Tibetan here reads yid can gyi rgyud, whereas in the previous asterism Kātyāyanīya was rendered as nag pa ya na.

n.173The Tibetan reads rwa’i rten mgo (“a head with[?] a horn”).

n.174The Tibetan reads glang po che’i mgo dang sna (“the head and trunk of an elephant”).

n.175The Tibetan reads phur bu (“Bṛhaspati”).

n.176That is, it is thick in the middle and thin at both ends. The Tibetan reads them skas (“ladder” or “staircase”).

n.177Following the Sanskrit. The Tibetan reads ’bras dang nas dang gro (“rice, barley, and wheat”).

n.178In the Tibetan it has nine stars, while in Dharmarakṣa’s translation it has three. Zhi Qian’s translation is in agreement with the Sanskrit.

n.179The Tibetan reads dmangs rigs sdom pa can (“the one bound to the śūdra caste”). Dharmarakṣa’s Chinese translation agrees with the extant Sanskrit.

n.180The Tibetan reads ba lang gi mjug ma bsgreng ba (“a raised cow tail”).

n.181In an additional note to his edition, Mukhopadhyaya (1954, p. 229) states that, in accordance with the Taittirīya­saṃhitā, the deity should be Viśvedeva (sic). The Tibetan has rgyas pa.

n.182Zhi Qian’s Chinese translation has rendered this as “It has no food,” as pointed out by Mukhopadhyaya (1954, p. 50).

n.183The Tibetan reads rnga (“a drum”). The Chinese translations are in agreement with the extant Sanskrit.

n.184The Tibetan reads kha zas ni rgya sran dang sran chung ngo (“It has horse gram and red lentils as its food”).

n.185The Tibetan here again renders the Sanskrit yavāgu (“gruel”) as nas dang gro (“barley and wheat”).

n.186In the Degé Kangyur this is rendered as rta ba (“a horse”), whereas in the Narthang we read ta ba (“an iron pan”). Zhi Qian’s Chinese translation is in agreement with the extant Sanskrit.

n.187The Tibetan reads lus gi sha (“sheep meat”), but both the Sanskrit and the Chinese translations only have “meat.”

n.188In an additional note to his edition, Mukhopadhyaya (1954, p. 229) states that Ahirbudhnya should be Ajapād or Ajaikapād. The Tibetan, however, reads klu dang gza’ lag, which is a rendering of Ahirbudhnya.

n.189The Tibetan reads bal sran (“wool lentils[?]”). Dharmarakṣa’s Chinese translation reads “cow meat.”

n.190In an additional note to his edition, Mukhopadhyaya (1954, p. 229) states that Aryaman should be Ahirbudhnya here. The Tibetan, however, reads ’phags pa, which is a rendering of Aryaman.

n.191The Tibetan reads rta ’gro ba (“a horse’s movement”). Zhi Qian’s Chinese translation is in agreement with the extant Sanskrit.

n.192The Tibetan reads dri za (“ Gandharva ”), which is in agreement with the reading of the Asiatic Society of Bengal manuscript (one of the two Nepalese Sanskrit manuscripts included in Mukhopadhyaya’s edition) and Dharmarakṣa’s Chinese translation. In an additional note to his edition, Mukhopadhyaya (1954, p. 229) states that the Paris manuscript (the other Nepalese Sanskrit manuscript included in his edition) reading aśvi- (“Aśvi”) is the correct one.

n.193Tibetan reads sa ga (“Viśākhā”) for Śatabhiṣā.

n.194The Tibetan reads “the four of the eastern quarter, as well as Viśākhā, Anurādhā, and Svātī,” thus omitting Punarvasu.

n.195The extant Sanskrit manuscripts have “Aśleṣā” here, but in an additional note to his edition, Mukhopadhyaya (1954, p. 229) points out that Aśleṣā has already been mentioned and that it should therefore be Revatī here. In both the Tibetan translation and Dharmarakṣa’s Chinese translation we find Revatī in this list.

n.196The Chinese translations leave out Rāhu and Ketu in this list. In an additional note to his edition, Mukhopadhyaya (1954, pp. 229–30) explains that these two were counted as planets only at a later stage in Indian astronomy.

n.197The Tibetan here reads rgyu skar ’jig rten du ’dzom pa bshad cing (“as the asterisms are shown to come to the world”), apparently due to having left out vipari­vartamāne in the translation, which is part of a locative absolute construction with loke.

n.198The Tibetan reads gdun zla tha chungs kyis (“with the winter month of Māgha”). The extant Sanskrit manuscripts read doṣaḥ, which makes no sense here. In the additional notes to his 1967 study (p. 78), Mukhopadhyaya cites a relevant verse from the Vedāṅgajyotiṣa (verse 5): māghaśukla­prapannasya pauṣakṛṣṇa­samāpinaḥ yugasya pañca­varṣasya kāla­jñānaṃ pracakṣyate. According to this verse the year beginss at the start of the bright fortnight of Māgha, at the conclusion of the dark fortnight of Pauṣa.

n.199The Tibetan has taken this passage as solely consisting of questions, when in the Sanskrit it is interspersed with King Triśaṅku’s answers.

n.200The Tibetan omits this question.

n.201The Tibetan here reads nyi shu rtsa drug (“twenty-six”), but further below it is rightly stated to be brgya nyi shu. In an additional note to his edition, Mukhopadhyaya (1954, p. 230) points out that in the Abhidharmakośa it is the other way round, with one tatkṣaṇa being defined as consisting of one hundred twenty kṣaṇas.

n.202The Tibetan here only reads drug (“six”).

n.203The Tibetan here only reads gsum (“three”).

n.204The Tibetan here reads drug cu rtsa dgu (“sixty-nine”).

n.205The Tibetan here reads gcig (“one”). According to the added statement in Zhi Qian’s translation, this is in fact the measurement at noon.

n.206The Tibetan here reads brgyad (“eight”).

n.207The extant Sanskrit manuscripts read tārāvala, but we have followed Mukhopadhyaya’s emendation to tārāvacara on the basis of the Tibetan skar ma rgyu ba.

n.208The extant Sanskrit manuscripts here add (“then it is the muhūrta called Sānuka,” but this is not mentioned in either the Tibetan or Chinese translations.

n.209The Tibetan here reads sring mo (“sister”), which is possibly a scribal error for srin po ( rākṣasa ).

n.210The Tibetan omits kṣaṇa .

n.211The Tibetan here again reads drug (“six”).

n.212The Tibetan here rightly reads sum cu. One should note that according to the following passage on the derivation of time, a muhūrta can be calculated to consist of 7,680 lavas. In an additional note to his edition, Mukhopadhyaya (1954, p. 230) has therefore argued that this present paragraph must be a later interpolation, even though the definition given here is supported by the Abhidharmakośa . This entire passage is absent in the Chinese translations.

n.213The Tibetan omits to render “Śarapatha.”

n.214For a discussion of the following passages on the divisions of time, see Mukhopadhyaya 1967, pp. 48–50.

n.215The Tibetan has rendered the Sanskrit dve palaśate as srang brgya (“one hundred srang”), which is considered to be equal to about forty grams.

n.216In an additional note to his edition, Mukhopadhyaya (1954, p. 230) cites a similar set of definitions found in the Suśrutasaṃhitā (VI), and in another note at the end of his 1967 study (p. 78) he provides further citations from the Harivaṃśa (VII 3–4), Viṣṇupurāṇa (II 8.55), and Manusmṛti (I 64).

n.217The Degé reading drug cu is in agreement with the reading ṣaṣṭi (“sixty”) of the Asiatic Society of Bengal manuscript (one of the two Nepalese Sanskrit manuscripts included in Mukhopadhyaya’s edition), but the reading catuḥṣaṣṭi (“sixty-four”) of the Paris manuscript (the other Nepalese Sanskrit manuscript included in Mukhopadhyaya’s edition) is supported by the Narthang Kangyur reading, as pointed out by Mukhopadhyaya (1954, p. 230). In addition, we may add that the latter number is also attested in the Choné, Lhasa, Yongle, Lithang, and Kangxi Kangyurs.

n.218The following definitions can be compared with those given in Abhidharmakośa III 85–86, as pointed out by Mukhopadhyaya (1954, p. 230).

n.219In an additional note to his edition, Mukhopadhyaya (1954, p. 230) rightly states that the Sanskrit edition needs to be emended to sapta gorajāṃsyekā likṣā | sapta likṣā ekā yūkā | sapta yūkā eko yavaḥ.

n.220As stated in an additional note to the edition of Mukhopadhyaya (1954, p. 230), the reading aṅguliparva of the extant Sanskrit manuscripts should be adopted here. The Sanskrit manuscripts further add the sentence triparvāṇy ekāṅguliḥ (“Three digits make one finger”).

n.221The extant Sanskrit manuscripts here add yojanasya pramāṇaṃ piṇḍitam (“The measure of a yojana has thus been added up”).

n.222Following the Sanskrit koṭiśata­sahasrāṇi catuviṃśatiś caikona­triṃśatkoṭi­sahasrāṇi dvādaśa ca śatasaha­srāṇi. The Tibetan reads bye ba phrag ’bum phrag nye shu rtsa bzhi dang / bye ba brgya stong phrag sum cu rtsa dgu, which seems to make 24,390,000,000,000. Dharmarakṣa’s Chinese translation has 16,950,012,000. Following the measurements given just before, however, it should be 15,495,785,088,000.

n.223Based on the following calculations regarding the measurements for palas and liquids, it should be 21,176,820.

n.224We have followed Mukhopadhyaya’s emendation on the basis of the number given in Dharmarakṣa’s Chinese translation. The Tibetan here gives 84,407,280, whereas the extant Sanskrit manuscripts have 84,007,280.

n.225We have again followed Mukhopadhyaya’s emendation on the basis of Dharmarakṣa’s Chinese translation. The extant Sanskrit manuscripts have omitted the last 974,720, and the Tibetan here gives 2,033,974,720.

n.226The Tibetan here reads zho gnyis (“two karṣas”).

n.227We have adopted the number preserved in the extant Sanskrit manuscripts (as emended by Mukhopadhyaya in his edition by adding śata). The Tibetan here gives 1,508,261,500, thus showing two obvious omissions. Both Chinese translations here read 1,282,261,530.

n.228In view of the Sanskrit yuddhārthī, the Lhasa reading ’thab mo du gnyer ba is to be adopted over the Degé reading thab mo du gnyer ba.

n.229The Narthang reading zas skom ’byung bar ’gyur ro is in better agreement with the Sanskrit than the Degé reading khro ba blun zhing longs spyod dang ldan par ’gyur ro.

n.230The Tibetan has omitted the Sanskrit utsāhavān (“enthusiastic”).

n.231The Tibetan has rendered the Sanskrit mahātmā as mthu che bar ’gyur ro (“will have great power”).

n.232The Tibetan reads cho ga dang rig pa dang ldan (“and possess prescriptions and knowledge”), thus lacking a rendering of ābharaṇa (“adornments”) in the compound ābharaṇa­vidhijña.

n.233The Tibetan lacks a rendering of alpāyuṣko (“having a short life”).

n.234The Tibetan here reads rig pa chung bar ’gyur ro (“will have little knowledge”).

n.235We have followed the Sanskrit mūliko. The Tibetan here has phreng rgyud mkhan (“proficient with the rosary”), which suggests that the translators had read mālako.

n.236The Tibetan lacks a rendering of the Sanskrit gandharva , which in this context must mean “musician.”

n.237In view of the Sanskrit sañcaya, the Lhasa reading gsos is to be adopted over the reading sogs found in Degé and several other Kangyurs.

n.238The Tibetan reads dpal dang ldan (“will be prosperous”), so it seems the translators read śrī instead of strī. The following two lines are rendered as “having garlands and enjoyments, it will be full of bulls,” but the reading of the extant Sanskrit is confirmed by Dharmarakṣa’s Chinese translation.

n.239The Tibetan omits this line.

n.240The Tibetan has omitted an equivalent for the Sanskrit krūrapuruṣā (“cruel people”).

n.241The Tibetan has rendered this entire verse as mi gzi brjid dang ldan la der phyis ni ’bru’i bcud thams cad nub par ’gyur ro (“The people will be radiant, and afterward all the sap of the crops will disappear there”).

n.242The Tibetan has rendered the last two lines as zhing skyes bu dang / bud med rmongs pas gang bar ’gyur ro (“And it will be full of foolish men and women”).

n.243The Tibetan has rendered this as bud med dang / skyes bu thams cad dpal dang ldan (“Women and men will all be prosperous”).

n.244The Tibetan reads mes tshig par ’gyur ro (“it will burn with fire”), a misinterpretation of the Sanskrit jvalantam. Dharmarakṣa’s Chinese translation has the same misreading, whereas Zhi Qian’s translation is in agreement with the Sanskrit.

n.245The Tibetan omits this last line.

n.246This verse is absent in the Tibetan translation as well as in Zhi Qian’s Chinese translation.

n.247The Tibetan has rendered the Sanskrit murkhāmāna­kāmavi­varjitā as rmongs pas bde ba spong bar ’gyur ro (“fools will avoid pleasure”), but in view of the context the Sanskrit must be understood as murkhā amāna­kāmavi­varjitā (“fools who are not free from pride and sensual desire”).

n.248The Tibetan has rendered the Sanskrit rāṣṭrāṇi (“kingdoms”) as rgyal po’i pho brang (“royal palaces”).

n.249For the following, Mukhopadhyaya (1954, p. 231) refers to a similar discussion in chapter 15 of the Bṛhatsaṃhitā.

n.250The extant Sanskrit reads, “Ārdrā is the asterism of kṣatriyas and brahmins. Punarvasu is the asterism of the people of Suparṇa. Puṣya is the asterism of all who live purely and of royal servants,” but this is probably an erroneous expansion, as noted by Mukhopadhyaya (1954, p. 231), who refers to a correct rendering of this passage on p. 79, line 13 of his edition, which supports the Tibetan reading adopted here. Mukhopadhyaya also notes that there is a parallel to this entire passage in chapter 15 of the Bṛhatsaṃhitā.

n.251In accordance with the Sanskrit nāga , the reading klu of the Yongle, Lithang, Kangxi, and Choné Kangyurs is adopted over the Degé reading klung.

n.252The Tibetan has translated the Sanskrit gauḍika as shar phyogs kyi mi rnams (“people of the east”), presumably in reference to the eastern location of the Gauḍa region, situated in present-day Bengal.

n.253This sentence is missing in the extant Sanskrit manuscripts.

n.254The Tibetan reads pho nya rnams (“messengers”), which suggests that the translators read dūtakānāṃ, but the extant Sanskrit manuscripts have dyūtakānām (“gamblers”).

n.255The Tibetans reads kau shi ka pa (“Kauśikas”), but the reading of the extant Sanskrit manuscripts is maulika . The Chinese translators seem to have taken the Sanskrit term as referring to a medical practitioner who works with roots and herbs.

n.256The Tibetan reads zhing pa bzang po rnams, whereas the reading of extant Sanskrit manuscripts here is bhadra­padakarmaṇāṃ bhadra­kāyakānāṃ ca. In another passage on the connection between the asterisms and certain peoples (F.260.a, 1.­364), the Tibetan reads zhing bzang po rnams, but there the extant Sanskrit has the reading bharukacchānāṃ, of which the Tibetan must be an etymologized rendering, having interpreted the place name Bharukaccha as consisting of a Prakritic form of the Sanskrit bhadra (“good”) and kaccha, presumably “(watery) land.” The Chinese translations also suggest bharukacchānāṃ as the underlying reading here. The extant Sanskrit reading here must be the result of scribal attempts at Sanskritizing the original reading bharukacchānām. One should note here that the Newari bhāro, derived from the Sanskrit bhadra, used to be a term for the landed nobility in the Kathmandu Valley.

n.257For the following, Mukhopadhyaya (1954, p. 231) refers to a similar discussion in chapter 23 of the Bṛhatsaṃhitā.

n.258The Tibetan renders this as “One should know that when Venus leads, it will rain incessantly,” which suggests that the translators misinterpreted the Sanskrit saṃjanayati as “one should know.”

n.259The Tibetan has Vaiśākha (dpyid zla tha chungs).

n.260The Tibetan has omitted the Sanskrit jāyāpatikānām (“wives and husbands”).

n.261The Tibetan has “eighty.”

n.262The extant Sanskrit reads sukhino bhavanti (“will go well”), but in view of the context and the Tibetan rendering ’jigs par ’gyur (“will go to ruin”), this needs to be emended to sukhino na bhavanti.

n.263The Tibetan has “fifty-nine.”

n.264The Tibetan has “fifty-nine.”

n.265The Tibetan has “fifty.”

n.266The Tibetan has rendered the Sanskrit rājānaś cātra chidrayuktā bhavanti as rgyal po dang skye bo rnams sngags par ’gyur te (“Kings and people will be terrified”).

n.267The Sanskrit mitrāṇi cātra dṛḍhāni bhavanti needs to be emended to pattrāṇi cātra dṛḍhāni bhavanti in view of the neuter case ending and the Tibetan rendering lo ma rnams.

n.268The Tibetan has omitted the Sanskrit kṛṣi- (“agricultural”).

n.269The Tibetan has “sixty-four.”

n.270The Tibetan omits Proṣṭhapada .

n.271This entire passage on the start of the rainy season under Abhijit is missing in the Tibetan translation and in the Chinese translations.

n.272The Tibetan has “sixty-one.”

n.273The Tibetan reads char cung (“little rainfall”).

n.274The Tibetan has “twenty-nine.”

n.275This sentence is missing in the Tibetan.

n.276The Tibetan adds nad chen pos ’chi ba mang zhing (“There will be many deaths due to epidemic”), for which Dharmarakṣa’s Chinese translation has “Elephants and domestic animals die.”

n.277The Tibetan has “sixty.”

n.278The extant Sanskrit here adds udvignāś ca dānapatayo bhavanti (“and those who are charitable will be distressed”), which seems rather misplaced in this context.

n.279This sentence is omitted in the Tibetan and in Zhi Qian’s translation, but it is contained in Dharmarakṣa’s translation.

n.280The Sanskrit does not mention any “second” rains, but the Tibetan reads gnyis su char phun sum tshogs.

n.281The Tibetan has omitted the Sanskrit bhaya (“fear”).

n.282The Tibetan has rendered the Sanskrit putra­pautrāṇāṃ ca kalaho bhavati as bu dang bu mo rnams ’chi bar ’gyur (“Sons and daughters will die”).

n.283The Tibetan reads gser ’tshong ba rnams (“gold traders”) instead of “maulikas and Atharvavedins,” and it omits the next sentence.

n.284The Tibetan omits this sentence.

n.285For the following, Mukhopadhyaya (1954, p. 231) refers to a similar discussion in chapter 98 of the Bṛhatsaṃhitā.

n.286We have followed the Tibetan translation and Mukhopadhyaya’s Sanskrit edition, but the extant Sanskrit manuscripts have a negation na in this line (“One should not wear new clothes”).

n.287The Sanskrit navaṃ kārayet and the Tibetan rendering gsar bya ba (“one should make new”) can also be translated as “one should repair.”

n.288The Tibetan reads, “One should not wear adornments.”

n.289The extant Sanskrit manuscripts add naṣṭaṃ viddhaṃ kṛtaṃ cāpi na tad astīti nirdiśet (“One may foretell that there will be no loss or damage done”), but this is absent in both the Tibetan and the Chinese translations.

n.290We have followed Mukhopadhyaya’s emendation to sarvaṃ varāṅgakam. The extant Sanskrit manuscripts read tathā sarvā varāṅganā (“thus all women”), which also seems to underlie the Tibetan translation.

n.291The Tibetan and Zhi Qian’s Chinese translation do not have a negation in this sentence.

n.292For this line the Tibetan reads, “It rains for four days.”

n.293The Tibetan has rendered this as pho nya btang ba (“One should send off messengers”).

n.294The Tibetan has rendered the Sanskrit dīrghaśrotra as rna ba gsang (“hidden ears”).

n.295After this stanza, the extant Sanskrit manuscripts add another three lines: “One should anoint the king. One should plant what is wet (praklinnāni). One should buy all kinds of jewels, a servant, and ornaments. One should make immovable property. One should not create employment (?) (prayuktaṃ na māpayet).” These lines are not present in the Tibetan and Zhi Qian’s Chinese translation.

n.296The Tibetan and the Chinese translations read “gold,” whereas the extant Sanskrit manuscripts have ratna.

n.297We have followed the Tibetan in including this entire stanza, seemingly consisting of six lines, which is lacking in the extant Sanskrit manuscripts.

n.298The Tibetan reads smad gsum (“three lower ones”).

n.299The Tibetan has bcing ba (“bondage”), which suggests that the translators read baddhaṃ instead of viddham.

n.300The Tibetan reads stod gsum (“the three upper ones”).

n.301Both the Sanskrit and the Tibetan have a negation in this sentence, but this does not seem to make sense considering the harsh and negative nature of these asterisms.

n.302The Tibetan again reads smad gsum (“three lower ones”).

n.303The extant Sanskrit manuscripts read abhijin muhūrta­samastayogaḥ (-bhāgaḥ in the Paris manuscript, one of the two Nepalese Sanskrit manuscripts included in Mukhopadhyaya’s edition) pañca­bhāgakṣetrikā. The Tibetan lacks a rendering for pañcabhāga and seems to have read kṣetrika as kṣatriya .

n.304The Tibetan omits this sentence.

n.305The Tibetan omits this sentence.

n.306The Tibetan has “Maghā” instead of “Puṣya” and refers to seventeen muhūrtas and thirteen muhūrtas, respectively.

n.307The Tibetan reads “six fingers.”

n.308The Tibetan specifies this asterism as Pūrvāṣāḍhā.

n.309The Tibetan specifies this asterism as Pūrvabhādrapadā.

n.310The Tibetan refers to sixteen muhūrtas and fourteen muhūrtas, respectively.

n.311The Tibetan reads “Pūrvāṣāḍhā on the new moon” and refers to fifteen muhūrtas for both day and night.

n.312The Tibetan reads “three fingers.”

n.313The Tibetan refers to thirteen muhūrtas and seventeen muhūrtas, respectively.

n.314The Tibetan reads “four fingers.”

n.315The Tibetan refers to thirteen muhūrtas and seventeen muhūrtas, respectively.

n.316The Tibetan has “five fingers.”

n.317The Tibetan refers to twelve muhūrtas and eighteen muhūrtas, respectively.

n.318The Tibetan reads “six fingers.”

n.319The Tibetan refers to thirteen muhūrtas and seventeen muhūrtas, respectively.

n.320The Tibetan reads “five fingers.”

n.321The Tibetan refers to fourteen muhūrtas and sixteen muhūrtas, respectively.

n.322The Tibetan reads “four fingers.”

n.323The Tibetan specifies this asterism as Pūrvāṣāḍhā.

n.324The Tibetan reads “Pūrvabhādrapadā” instead of “Uttarabhādrapadā,” and it refers to fifteen muhūrtas for both day and night.

n.325The Tibetan reads “three fingers.”

n.326The Tibetan reads “Pūrvāṣāḍhā” instead of “Phalgunī.”

n.327The Tibetan refers to sixteen muhūrtas and fourteen muhūrtas, respectively.

n.328The Tibetan reads “two fingers.”

n.329The Tibetan reads “Pūrvabhādrapadā” instead of “Uttarabhādrapadā.”

n.330The Tibetan refers to seventeen muhūrtas and thirteen muhūrtas, respectively.

n.331The Tibetan reads “two fingers.”

n.332The Tibetan reads “Jyeṣṭhā” instead of “Viśākhā.”

n.333The Tibetan reads “Puṣya” and “Pūrvāṣāḍhā,” respectively.

n.334The Tibetan refers to eighteen muhūrtas and twelve muhūrtas, respectively.

n.335The Tibetan reads “half a finger.”

n.336The Tibetan reads “Pūrvāṣāḍhā” instead of “Jyeṣṭhā.”

n.337The Tibetan reads lo lnga lnga zhing zla ba re re lhag par ’gyur (“Every five years an extra month is added”).

n.338We have followed Mukhopadhyaya’s emendation on the basis of the Tibetan translation. The extant Sanskrit manuscripts only read ṣaṇmāsāt samudre udaka­parimāṇo bhavati (“For six months there is a measure of water in the ocean”).

n.339The Tibetan has omitted the Sanskrit samudrodaka (“the water in the ocean”).

n.340The Tibetan omits “Śukra.”

n.341The extant Sanskrit manuscripts add Ketu, Dhūmaketu, and Rohitaketu.

n.342The Tibetan has rendered this as ’di rnams kyi gtso bo ni phur pu’o (“Chief among them is Bṛhaspati”).

n.343The Tibetan here adds “Śravaṇā.”

n.344We have adopted the Tibetan drug pa (“sixth”) against the extant Sanskrit saptamam (“seventh”), since according to Mukhopadhyaya (1954, p. 105, n. 8) the Tibetan reading is supported by the Bṛhatsaṃhitā. The extant Sanskrit manuscripts read, “Abhijit, Śravaṇā, Dhaniṣṭhā, Śatabhiṣā, and the two Bhādrapadās‍—this is the sixth general circle. Revatī and Aśvinī‍—this is the seventh general circle.”

n.345In an additional note to his edition, Mukhopadhyaya (1954, pp. 231–32) points out that some of the names of the muhūrtas in the following section do not agree with the names listed earlier. Given that this section is moreover absent in the Chinese translations, he argues that this section must be a later interpolation to the text. For extensive citations of passages from other texts containing the names of the muhūrtas, see Mukhopadhyaya’s further notes at pp. 232–33.

n.346The Tibetan omits this enumeration.

n.347The Tibetan has “seventeen.”

n.348The Tibetan omits this sentence.

n.349The Tibetan has “thirteen.”

n.350The Tibetan reads “Āṣāḍha” (dbyar gyi zla ba gnyis pa).

n.351The Tibetan has “sixteen.”

n.352The Tibetan has “fourteen.”

n.353The extant Sanskrit adds vāyavo, with which this sentence translates as, “At midnight it is the perilous hour called Vāyava.” The Tibetan has sgyu ma chen po (“Mahāmāya”).

n.354The Tibetan has “fifteen.”

n.355The Tibetan has “fifteen.”

n.356The Tibetan has “Aśvayuja.”

n.357The Tibetan has “Aśvayuja” (ston zla ’bring po).

n.358The Tibetan has translated this as dga’ bral (“Without Delight”).

n.359The extant Sanskrit adds saṃyama , with which the name could be rendered as “Continuum Restraint.”

n.360The Tibetan has rendered this as nam langs pa na (“At daybreak”).

n.361The Tibetan reads “Mārgaśīrṣa” (dgun zla ra ba).

n.362The Tibetan has translated this as dga’ bral (“Without Delight”).

n.363The Tibetan has rendered this as nam langs pa na (“At daybreak”).

n.364The Tibetan reads “Pauṣa” instead of “Māgha.”

n.365The Tibetan has rendered this as dga’ byed (“Gladdening”).

n.366The Tibetan has translated this as dga’ bral (“Without Delight”).

n.367The Tibetan has rendered this as rdul med (“Free from Dust”), which suggests that the translators read virajas.

n.368The Tibetan has rendered this as nam langs pa na (“At daybreak”).

n.369The Tibetan reads differently here: “As it is in the first month of the hot season, so it is in the first month of the rainy season. As it is in the fourth month of winter, so it is in Kārttika. As it is in Pauṣa, so it is in the first month of winter. As it is in Pauṣa, so it is in the second month of the hot season.”

n.370The extant Sanskrit adds muhūrtānāṃ, which would translate as “the course and movement of the muhūrtas of the asterisms.”

n.371The Tibetan lacks “fourteenth.”

n.372The exact meaning of this passage evades us, so our translation of it remains uncertain. The Tibetan rendering is as unclear to us as the extant Sanskrit.

n.373For the following, Mukhopadhyaya (1954, p. 233) refers to a similar discussion in chapter 32 of the Bṛhatsaṃhitā.

n.374The Tibetan has rendered this as shin tu ’chi ba (“Extreme Death”).

n.375The Tibetan has rendered this as tha dad pa’i rim pa (“Distinct Succession”).

n.376The Tibetan has rendered this as “the middle month of the hot season” (dpyid zla ’bring po) or Caitra.

n.377The Tibetan has rendered this as “the first month of the hot season” (sos ka’i zla ba dang po) or Phālguna.

n.378The Tibetan has rendered this as “the second month of the hot season” (sos ka’i zla ba gnyis pa) or Caitra.

n.379The Tibetan has rendered this as “in the month of Jyaiṣṭha” (dbyar zla ra ba la).

n.380The Tibetan has rendered this as “in the month of Āṣāḍha” (dbyar zla ’bring po la).

n.381We have followed Mukhopadhyaya’s alternative reading of the extant Sanskrit manuscripts here as ṛtur (1954, p. 109, n. 7). This entire line is missing in the Tibetan translation.

n.382This line is also missing in the Tibetan translation.

n.383The Tibetan has rendered this as “the fourth month of winter” (dgun gyi zla ba bzhi pa).

n.384The Tibetan has “twenty-five days.”

n.385The Tibetan has “Phālguna” (dpyid zla ra ba).

n.386The Tibetan has rendered these two lines as de dag ni sa g.yos pa’i dus ji tsam na smin pa yin par bstan to (“It is taught that at the time of the earthquake, they are the result”).

n.387This line is missing in the Tibetan translation.

n.388This line is missing in the Tibetan translation.

n.389The Tibetan has rendered the Sanskrit anāvāsā as mi bde ba mi ’byung ngo, perhaps to be understood as “difficult and impossible.”

n.390The Tibetan has rendered the Sanskrit daṃṣṭripālikāḥ as yul ’khor chas pa (“those who have set off to kingdoms”).

n.391These two lines are placed at the beginning of the stanza in the Sanskrit and the Tibetan, but they interrupt the sentence there. We have therefore followed Zhi Qian’s Chinese translation in placing them at the end of the stanza.

n.392The extant Sanskrit manuscripts here add tiryakvātaś caiva vāti kṛtaṃ naśyati śāśvataṃ | pathikāś copatapyanti māṣayācyopa­jīvikāḥ (“There will be horizontal wind, deeds will be forever lost, and travelers will be in distress, as will those who subsist on asking for beans[?]”). These two lines are missing in both the Tibetan and the Chinese translations.

n.393These three professions are not attested in the available Sanskrit lexica. Probably the Sanskrit manuscripts are corrupt here. The text may have already been corrupt at the time of the Tibetan translation, which omits these two lines.

n.394We have followed the reading of the Tibetan and Chinese translations. The Sanskrit here has two different stanzas: “ ‘When the earth shakes under Citrā, Artisans, craftsmen, Maidens, and all jewels, Along with the sowers of crops“ ‘And the peoples of Vaṅga, Daśārṇa, Kuru, and Idimāhiṣaka, Will all undergo tribulations Once shaken by that earthquake.”

n.395The Tibetan has shan pa’i bdag po (“the chiefs of murderers”), whereas the extant Sanskrit manuscripts read abhyarthi­kamūṣika, which we have followed here. While Mukhopadhyaya has opted for abhyarthita­mūṣika in his edition, he has suggested an emendation to apy arthi­kamūṣika in an additional note (1954, p. 234).

n.396The Tibetan omits this line.

n.397The extant Sanskrit manuscripts lack the last four lines.

n.398The Tibetan reads byi ba (“rats”), whereas the extant Sanskrit reads vṛṣabhā (“bulls”).

n.399The last two recurring lines are lacking in the extant Sanskrit and the Tibetan, but they occur in Zhi Qian’s Chinese translation.

n.400We have followed the Degé reading ’jigs pa chen po ’byung, which is in agreement with the reading of extant Sanskrit manuscripts eteṣu mahābhayam. In his edition, Mukhopadhyaya has emended the Sanskrit text to mahā­vidyākarāś ca ye (“who are great sources of knowledge”) on the basis of the Narthang reading rig pa chen pa ’byung zhing.

n.401The Tibetan omits the last two lines.

n.402The Tibetan omits this line.

n.403The Tibetan has rendered the Sanskrit mantrapāragāḥ as glu mkhan rnams (“experts in song”).

n.404The Tibetan omits “Yugandharas.”

n.405The Kiśaṭṭas are not attested in the Sanskrit lexica. Mukhopadhyaya has emended the reading of the Sanskrit manuscripts kiśaṭṭa to kuśaṇḍāḥ, even though the former seems to underlie the Tibetan rendering‍—gi Sa ta in the Narthang Kangyur and gi Si ta in the Degé Kangyur.

n.406We have followed the reading of the extant Sanskrit manuscripts tad eti yad anīpsitaṃ, which in the Tibetan seems to have been rendered as de dag gi bsam pa ’grub bo, thus lacking a negation. In an additional note to his edition, Mukhopadhyaya (1954, p. 234) has stated that the manuscript reading is in fact preferrable over his emendation to ye narā rājapūjitāḥ in his edition. We have also left out the following sentence in his edition, since it is not found in the Tibetan.

n.407The reading cakravākān (“cakravāka birds”) of the extant Sanskrit manuscripts seems out of context here. In view of the Tibetan ’khor lo gcig pa, which suggests some sort of vehicle, the Sanskrit probably needs to be emended to cakravāhān.

n.408The Tibetan has omitted the Sanskrit aurabhrikān (“shepherds”).

n.409The Tibetan adds “horses” (rta rnams).

n.410The Tibetan reads mon gru (Śatabhiṣā or Dhaniṣṭhā).

n.411The Tibetan omits the Śibis and Vatsas.

n.412We have followed the reading of the extant Sanskrit manuscripts na hi tāṃś citragarbhāś ca ye ca (emended to na) cāñjanakā janāḥ. The Tibetan rendering seems problematic here.

n.413In the additional notes in his 1967 study (pp. 78–79), Mukhopadhyaya cites several passages from the Atharvaveda­pariśiṣṭa, the Jyotirnibandha, and the Agnipurāṇa that are related to the following verses.

n.414The Tibetan here reads me yis (“by fire”).

n.415We have followed Mukhopadhyaya’s emendation of kāṣṭhāni to rāṣṭrāni, which he has provided in an additional notes in his 1967 study (p. 79), in reference to a related verse in Atharvaveda­pariśiṣṭa.

n.416We have followed the Tibetan gnod pa ’byung bar ’gyur zhing, since the extant Sanskrit jayati (“are victorious”) does not seem make sense in this context of calamity.

n.417The Tibetan has “at night.”

n.418The Tibetan has “at daybreak.”

n.419For this phrase, the Tibetan has “farmers, children, and the elderly.”

n.420We have followed the Sanskrit for these last four lines, even though they seem to be out of context here. The Tibetan translation is difficult to make sense of: dman pa dang / dge ba dang dma’ bar ’gyur ba rnams shes shing gang gis snga ma’i rang bzhin du gyur pa shes pa’i bram ze’i nyi mar gyur par blta’o.

n.421The Tibetan lacks this sentence.

n.422We have followed the Tibetan sran sngon gyi khur ba, which we have understood as “lentil broth,” though the Sanskrit here only reads maṇḍa (“cream” or “froth”). Zhi Qian’s Chinese translation, however, similarly has “pulses.”

n.423We have followed the reading ādityo devatā of the extant Sanskrit manuscripts, which is supported by the rendering de’i lha ni nyi ma yin pas in the Tibetan. In an additional note to his edition, however, Mukhopadhyaya (1954, p. 235) proposes an emendation to aditir devatā on the basis of the passage on p. 47, line 6, of his edition.

n.424The Tibetan has rendered the Sanskrit kṛsara as ’bru dang til btags pa’i ri lu (“a lump of ground grain and sesame”).

n.425We have followed the reading pañca- of the extant Sanskrit manuscripts. The Tibetan has “seven days.”

n.426This is not in agreement with the earlier passage, at the beginning of the astrological section, in which Bhaga is said to be the presiding deity of Pūrvaphalgunī, and Aryaman that of Uttaraphalgunī. The Tibetan here reads klu (“serpent” or “nāga”).

n.427This is not in agreement with the earlier passage, at the beginning of the astrological section, in which Aryaman is said to be the presiding deity of Uttaraphalgunī, and Bhaga that of Pūrvaphalgunī. The Tibetan here reads gshin rje (“lord of death”), i.e., Yama.

n.428The Tibetan reads dri za (“ Gandharva ”).

n.429The Tibetan reads nas kyi ljang bu’i (“of barley sprouts”).

n.430The Degé Kangyur reads “twenty-eight,” the Narthang Kangyur “twenty-two.”

n.431The Tibetan has “twenty-one.”

n.432The Tibetan reads nyi ma (“the sun”).

n.433The Tibetan reads dmangs rigs (“śūdra” or “low caste”).

n.434The Tibetan has omitted the Sanskrit madya (“liquor”).

n.435In an additional note to his edition, Mukhopadhyaya (1954, p. 235) states that this should be “Viśvedeva.”

n.436In an additional note to his edition, Mukhopadhyaya (1954, p. 235) states that this should be “Brahma” or “Brahmā” on the basis of the passage on p. 50, line 1, of his edition.

n.437These two verses on Abhijit are omitted in the Tibetan, as well as in the Chinese translations.

n.438The Tibetan reads byug pa dang phreng ba (“fragrances and garlands”), but Zhi Qian’s Chinese translation also has “ghee.”

n.439In an additional note to his edition, Mukhopadhyaya (1954, p. 235) states that this should be “Ajapād” or “Ajaikapād.” The Tibetan omits this entire verse.

n.440The Tibetan has khrums stod (“Pūrvabhādra[padā]”).

n.441In an additional note to his edition, Mukhopadhyaya (1954, p. 235) states that this should be “Ahirbudhnya,” which also seems to be what underlies the Tibetan klu (“serpent”). Metrically, however, this does not fit.

n.442The Tibetan has “twenty-eight days.”

n.443In an additional note to his edition, Mukhopadhyaya (1954, p. 235) states that the Sanskrit should read “the Aśvinis,” and he refers to an earlier passage in which the Paris manuscript (one of the two Nepalese Sanskrit manuscripts included in his edition) reads aśvidaivatām (1954, p. 51, n. 12). However, both manuscripts here have “ Gandharva .” The Tibetan reads nyi ma (“the sun”).

n.444The Tibetan here reads mar dang sran sngon (“ghee and mung beans”).

n.445In an additional note in his 1967 study, Mukhopadhyaya (p. 79) mentions that a similar discussion can be found in chapter 121 of the Agnipurāṇa.

n.446The Tibetan has “twenty days.”

n.447The Tibetan has “three days.”

n.448The Tibetan has “twelve days.”

n.449The Tibetan has “twenty-nine days.”

n.450The Degé Kangyur has “ten days,” the Narthang Kangyur “fourteen days.”

n.451The Tibetan has “twenty-six days.”

n.452The Tibetan omits this clause.

n.453In the Tibetan translation it is stated that under all the asterisms from Uttarāṣāḍhā until Revatī one will be released after fourteen days.

n.454For the following, Mukhopadhyaya (1954, p. 235) refers to similar material in the Jyotiṣa­kalpa­madruma.

n.455The Tibetan has omitted the Sanskrit snigdha (“smooth”), and it has translated the Sanskrit padma­samānavarnaḥ as sa dkar po’i kha dog (“the color of white earth”).

n.456For this line, the Tibetan reads de ni mchu pad ma ltar dmar ’gyur ro (“she will have lips that are reddish like a pink lotus”).

n.457We have followed the reading of the extant Sanskrit manuscripts nābheś ca madhye pratibimbakaṃ syāt. The Tibetan here reads zas dang skom mang du rnyed par ’gyur ro (“she will obtain much food and drink”) and then adds another sentence that seems to be a mix of the previous line and the last line of the second to last verse of this section.

n.458The Tibetan has rendered the last two lines as de ni khyim thab las ’phyo bar ’gyur zhig dka’ ba’i sdug bsngal thob bar ’gyur ro (“She will be unhappy in married life and she will experience unbearable suffering”).

n.459The Tibetan has rendered the Sanskrit śrutidhāriṇīṃ as khyim thab kyi brtul zhugs can (“faithful to her husband”).

n.460In an additional note in his 1967 study, Mukhopadhyaya (p. 79) emends the reading cibuke to cibau, which we have followed.

n.461The extant Sanskrit manuscripts contain several additional verses in this section that are not found in either the Tibetan or the Chinese translations. See the edition of Mukhopadhyaya (1954), p. 130, n. 4; and p. 131, n. 7.

n.462The Tibetan has omitted the Sanskrit nakṣatra- (“under an asterism”). In an additional note, Mukhopadhyaya (1954, pp. 235–36) points out that the following section deals with the same subject matter as the section at p. 61 of his edition (n.­18) but that the statements made are in fact quite different. In view of the absence of this section in both Chinese translations, it seems likely that the following was added at a later point during the transmission of the text.

n.463The Tibetan reads rkun ma la dga’ (“likes to steal”), whereas the Sanskrit has priyasāhasaḥ.

n.464The Tibetan reads pha rol ’jigs par byed (“frightening to others”), whereas the Sanskrit has paramajalpakaḥ.

n.465The Tibetan has rendered this line as khang po bzang po rnyed par ’gyur ro (“and will obtain a fine house”).

n.466The Tibetan has rendered this line as rtag tu bla ma la smod par ’gyur ro (“and will continually slander the guru”).

n.467The Tibetan has rendered the Sanskrit ’steyakarmā as ’joms par ’gyur ro (“will make conquest”), which suggests that the translators did not recognize the elided a-.

n.468The Tibetan reads chu smad (“Uttarāṣāḍhā”).

n.469The Tibetan has rendered the Sanskrit parivādī as byi rgyug pa (“one who commits adultery”).

n.470In an additional note in his 1967 study, Mukhopadhyaya (p. 79) proposes to emend the Sanskrit from nānusūyakaḥ to nānasūyakaḥ, which is the correct form, but this would translate as “not unenvious,” which seems out of place in this positive characterization of a Revatī-born person. It seems to us that the authors of the text here understood anusūyakaḥ as “envious.” The Tibetan reads dga’ bar ’gyur ro, which suggests that the translators read something with nand-.

n.471The extant Sanskrit manuscripts add many more sections on astrology and prognostication here. See the edition of Mukhopadhyaya (1954), pp. 136–203.

n.472Zhi Qian’s Chinese translation here adds “and no dowry is needed.”

n.473Zhi Qian’s Chinese translation lacks the following passages in which some of the earlier arguments are revisited, up until the passage that describes the brahmin pouring water to consecrate the matrimony (1.­739).

n.474The extant Sanskrit manuscripts here add brahmaṇaḥ (“[the particular characteristic] of Brahmā”), but this is absent in the Tibetan and seems to make less sense with what follows.

n.475We have followed the Tibetan bde ba ’dod pas bslus nas. In his edition, Mukhopadhyaya (1954, p. 206) has emended the clearly corrupt reading aśvodana tena of the extant Sanskrit manuscripts to aśvastanās tena, but this does not seem convincing to us.

n.476The Tibetan seems to have rendered the Sanskrit kilāsī twice: as ’brum bu (“pox ridden”) and as sha bkra (“vitiligo”).

n.477The Tibetan and Dharmarakṣa’s Chinese translation omit this sentence.

n.478We have followed the reading of the extant Sanskrit manuscripts (after hṛṣṭacittaḥ): udakenāsau kanyakām anupradāsīd iyam astu kanyakā prakṛtiḥ śārdūla­karṇasya bhāryā, with which the Tibetan translation agrees. For some reason, Mukhopadhyaya (1954, p. 208) understood this narrative passage as forming a stanza and sought to emend the text accordingly.

n.479The Tibetan lacks this sentence, but the extant Sanskrit ugracitta āsīn mātaṅgarājaḥ is supported by Dharmarakṣa’s Chinese translation. Mukhopadhyaya (1954, p. 208) again presented the following sentence as forming a stanza, but in our opinion this narrative passage must all be in prose.

n.480Mukhopadhyaya (1954, p. 208) adopted the reading āśrame of the extant Sanskrit manuscripts, but this does not seem to make much sense here. We have followed the Tibetan rang gi grong khyer du phyir log ste, which is supported by Dharmarakṣa’s Chinese translation.

n.481The extant Sanskrit manuscripts here add kṣemaṃ subhikṣaṃ ca sadotsavādyam (“[his kingdom,] which was peaceful, well provisioned, and always engaged in festivals and the like”), but this short description is not found either in the Tibetan or in the Chinese translations.

n.482The Tibetan rendering of the Sanskrit pūrvakeṇa nivāsena as skye ba sngon gyi bag chags kyis suggests that the Sanskrit nivāsa (“lifetime”) was understood as having the same meaning as vāsanā (“karmic imprint”). We have stayed with the proper meaning of the Sanskrit nivāsa, which can have the added connotation of “living together,” especially when it is prefixed with saṃ-.

n.483The Tibetan text here reads chu nang nas skyes ci bzhin no (“as what is born from within water”), but perhaps nas is to be emended to chu, which would make for the reading chu skyes (“water lily”). This would be in agreement with Dharmarakṣa’s Chinese translation as well as the related verses found in other past life stories, such as in the Mahāvastu (II 98): pūrve vā saṃnivāsena pratyutpanne hitena vā | sarvatāṃ jāyate premaṃ utpalaṃ vā yathodake. The alternative reading of the extant Sanskrit manuscripts candrasya kumude yathā (“like [the love] of the moon for the water lily”) seems to be a later scribal emendation.

n.484In the Tibetan, this sentence has been combined with the following one through a rendering of the Sanskrit śabdāpayāmi (“I call for”) as tshig tu ma yin par rtogs par bya zhing, which is difficult to make sense of here.

n.485The Tibetan lacks a rendering of the Sanskrit drutam (swiftly”), but we find it in Dharmarakṣa’s Chinese translation.

n.486The extant Sanskrit manuscripts add viśuddham (“pure”).

n.487The Tibetan lacks this sentence.

n.488With this sentence we have followed the Tibetan translation. The extant Sanskrit manuscripts only read āttamanasas te bhikṣavo bhagavato bhāṣitam abhyanandan (“Elated, the monks rejoiced at what the Blessed One had spoken”). Dharmarakṣa’s Chinese translation reads, “Elated, the brahmin householders and the monks rejoiced at what the Blessed One had spoken,” whereas Zhi Qian’s translation reads, “Elated, King Prasenajit, the monks, nuns, devoted laymen, and devoted laywomen rejoiced at what the Blessed One had spoken.”