Notes

n.1The title of the extant Sanskrit text in the Divyāvadāna is Kanaka­varṇāvadāna, which places it in the avadāna category. However, this designation is likely to have happened when the Divyāvadāna, the collection of texts in which it has been preserved, was compiled. At this time, the different individual texts, whether strictly belonging to the jātaka, avadāna , or pūrvayoga genres or not, were all given the avadāna designation. Historically, the avadāna and pūrvayoga genres, which involve the past lives of other figures apart from the Buddha, appear to be posterior to the jātakas, which exclusively narrate the Buddha Śākyamuni’s past lives. This development might be suggested by the fact that the avadāna genre is among the three added categories in the Sanskrit list of twelve types of teaching (dvādaśāṅga) that are not found in the Pali list of nine types of teachings (navaṅga), which, however, does include jātaka as a category. Both the avadāna and the pūrvayoga genres seem to have been especially popular in the northwest of the Indian subcontinent, as can be inferred from the relatively large number of texts of this type among the recently discovered Kharoṣṭhī manuscripts in the Gāndhārī language (see Lenz 2003; 2010). In fact, the term pūrvayoga might have its provenance there, as has been suggested by von Hinüber (2001, p. 520) on the basis of the occurrence of the term in the Milindapañha, a Pali text that has its origin in Gandhāra. At the beginning of the Milindapañha, before the narration of the past lives of its two protagonists, Nāgasena and King Milinda, we also find a short definition of the term: “pubbayoga means their past actions” (pubbayogo ti tesāṃ pubbakammaṃ; Mil 2.23). Interestingly, within the surviving Gāndhārī corpus of avadānas and pūrvayogas, which has been dated to approximately the first half of the first century ᴄᴇ, a strict distinction is maintained in designating the two kinds of texts: while the pūrvayoga is an account of someone’s actions in a previous lifetime, the avadāna is always an account of actions performed in the present lifetime (Lenz 2003, p. 92). From this it would appear that the avadāna form that includes past-life stories, as found in such collections as the Avadānaśataka and the Divyāvadāna, arose only in later centuries, during which the avadāna genre came to encompass both jātaka and pūrvayoga stories.

n.2See The White Lotus of the Good Dharma (Saddharma­puṇḍarīka­sūtra, Toh 113), chapters 7, 22, and 25; and The King of Samādhis Sūtra (Samādhi­rāja­sūtra, Toh 127) chapters 2, 16, 21, 29, and 37. There are also two pūrvayoga chapters in the Mahāvastu, one at Mvu III 170–172 and one immediately following it at III 172–175.

n.3It 18–19.

n.4Tripathi 1995, p. 18. There is no Tibetan translation of the Dāna Sūtra, at least as a standalone text, in the Tibetan Kangyur. There is also an extant Sanskrit Dāna Sūtra in the Schøyen Collection, BMSC iv (from Bamiyan). Mentioned in Braarvig 2014, p.163.

n.5This phrase has given rise to considerable discussion in modern scholarship, since the compound word that is used in Sanskrit and Pali, khaḍgaviṣāṇa and khaggavisāṇa, can be interpreted in more than one way. The word khaḍga/khagga itself can mean “rhinoceros,” and since viṣāṇa/visāṇa means “horn,” the compound can be taken as “the horn of a rhinoceros,” which is the interpretation followed in the Pali commentarial tradition. In this interpretation, the entire phrase is understood as “move alone like a rhinoceros horn,” in reference to the fact that the Indian rhinoceros has a single horn, distinguishing it from animals that have two horns. On the other hand, khaḍga/khagga also has the meaning “sword,” and together with viṣāṇa/visāṇa, “horn,” the entire compound khaḍgaviṣāṇa/khaggavisāṇa can mean “[the animal which has] a sword-horn,” thus referring to the rhinoceros itself. In this understanding, which is mostly seen in Buddhist Sanskrit literature, the phrase “move alone like a rhinoceros” alludes to the solitary behavior of the Indian rhinoceros, which is known not to roam about in groups. It should be kept in mind, however, that both meanings could be intended at the same time in this particular phrase, as has been pointed out by Salomon (2000, p. 13; 2018, p. xx). For a recent discussion of this issue, see Jones 2014, and for a wider survey on the imagery of the wandering ascetic as a rhinoceros in Brahmanical, Jain, and Buddhist literature, see af Edholm 2021.

n.6The first texts in which these verses are attributed to pratyekabuddhas are the Pali Cullaniddesa, an old commentary on the Suttanipāta, and the Pali Apadāna, a collection of autobiographical poems ascribed to the Buddha Śākyamuni, the pratyekabuddhas of the past, and the foremost arhat disciples of the Buddha. Both texts are old enough to have been included in the Khuddakanikāya of the Suttapiṭaka of the Pali canon; the Apadāna was probably compiled around the second century ʙᴄᴇ (Walters 1997), which is perhaps also the time period of the Cullaniddesa (von Hinüber 1996, p. 58, §118, and p. 61, §121). In both texts, the verses are simply ascribed to pratyekabuddhas collectively, without narrating the specific occasion of each verse. This is also the case for the Mahāvastu, the voluminous Mahāsāṃghika-Lokottaravāda compilation on the life of the Buddha (parts of which have also been dated to the second century ʙᴄᴇ), in which several adapted verses from the Rhinoceros Sūtra are presented as having been uttered by a group of five hundred pratyekabuddhas living in a forest near Vārāṇasī‍—the site of the future Deer Park‍—just before the birth of the Buddha Śākyamuni in this world (Mvu I 357–359). Upon hearing of the coming of the Buddha to this buddhafield and the necessity for them to vacate, each of these pratyekabuddhas is said to have proclaimed such a verse before passing into parinirvāṇa by rising into the sky and incinerating themselves, after which their bodily remnants fell to the ground. This is then said to explain the other name for the site of the Buddha’s first teaching, the Sanskrit ṛṣipatana, which literally means “the falling of sages.” For a useful discussion of the different versions and a translation of the Gāndhārī verses of the Rhinoceros Sūtra, see Salomon 2018.

n.7This Sanskrit is from the extant Sanskrit version of the story of King Kanakavarṇa, which is included in the Divyāvadāna collection, as discussed further below. Cowell & Neil 1886, p. 294; Vaidya 1959, p. 182.

n.8Sn v. 36. This Pali verse occurs in the same form in the Paccekabuddhāpadāna (v. 11) and the Cullaniddesa (Cn 57). For a translation of the Pali Paccekabuddhāpadāna, the pratyekabuddha section in the Apadāna, see http://apadanatranslation.com/text/chapter-2/poem-001.html.

n.9Salomon 2000. The third line of this verse has not been preserved. Note that the second line is the same as in the Pali version, whereas the first word in the first line is in agreement with the Sanskrit verse in the Kanaka­varṇāvadāna. The Gāndhārī verse is also closely related to the Prakritic Sanskrit verses uttered by the pratyekabuddhas in the Mahāvastu (Mvu I 357–359), such as for instance: saṃ­sevamānasya siyāti sneho snehānvayaṃ duḥkham idaṃ prabhoti | mitreṣu ādīnavaṃ saṃmṛśanto eko care khaḍga­viṣāṇakalpo || “Keeping close company, one comes to have attachment; It is from attachment that this suffering comes about. Fully reflecting upon the danger in friends, One should move alone like a rhinoceros.”

n.10In this later Pali commentary on the Suttanipāta, attributed to the fifth-century Pali scholar Buddhaghosa but based on earlier Sinhala commentaries, each verse of the Khaggavisāṇa Sutta is accompanied by its own background story. In the case of verse 36, this involves a paccekabodhisatta, a being striving for individual awakening, who grows up as a prince in Vārāṇasī. Upon hearing of the death of his future bride, the prince reflects upon the conditions of existence and attains individual awakening, after which he shares the verse in question with his ministers. For the full story, see Bodhi 2017, pp. 425–430.

n.11Salomon (2000, p. 6, n. 4) makes mention of a Sanskrit manuscript from Turfan that is said to contain another such text with pratyekabuddha stories around these verses. He points out that this part of the manuscript does not appear to have been published yet.

n.12This is stated in one of the concluding verses of The Rhinoceros Sutta in both the Pali (Sn v. 73) and the Gāndhārī (v. 36), but it is also expressed at the beginning of the Gāndhārī version and the Sanskrit verses in the Mahāvastu, in which the second half of the first verse reads: maitreṇa cittena hitānukaṃpī eko care khaḍga­viṣāṇa­kalpo, “With a mind of loving-kindness, being compassionate for the welfare [of others], one should move alone like a rhinoceros.”

n.13Taishō 155, 109c1–110b18. While we have taken the Chinese 本行 as a rendering of the Sanskrit pūrvacaryā, these characters could also very well be a rendering of pūrvayoga.

n.14Taishō 162.

n.15Bagchi (1927, p. 262) mentions that according to several Chinese scriptural catalogs, this Chinese translation of the Kanaka­varṇa­pūrvayoga was the second to be made, after a first translation by Dharmaruci, a monk from southern India who resided at Luoyang at the beginning of the sixth century. This first translation is said to have been made in 507 ᴄᴇ and later corrected by Bodhiruci. However, at some point it was lost. On Dharmaruci, see Bagchi 1927: 246–47.

n.16Denkarma, folio 301.a; Herrmann-Pfandt 2008, p. 164; Phangthangma 2003, p. 22. In both the Denkarma and Phangthangma entries, the title is extended with bstan pa, which renders the translation of the entire title as “The Teaching on the Past Endeavor of Kanakavarṇa.”

n.17For this Sanskrit version of the Divyāvadāna, see Cowell & Neil 1886, pp. 290–98; and Vaidya 1959, pp. 180–84. Most of the Nepalese Divyāvadāna manuscripts are on Himalayan paper and date from the seventeenth century onwards. Cowell and Neil’s Sanskrit edition‍—from which Vaidya’s edition is derived‍—is based on four of these relatively late manuscripts. In recent years, however, two older manuscripts of the Divyāvadāna have been brought to light, which are written on palm leaf and dated on paleographical grounds to the eleventh century ᴄᴇ (Formigatti 2016, p. 127). Unfortunately, these palm-leaf manuscripts are only partially preserved and do not contain the text of the Kanaka­varṇāvadāna. The Divyāvadāna was an important source for the nineteenth-century French scholar Eugène Burnouf in his pioneering study on Indian Buddhism, Introduction à l’histoire du Buddhisme Indien, in which he also included a translation of the Kanaka­varṇāvadāna (1844, pp. 79–87; for a recent English translation, see Burnouf 2010, pp. 130–37). A German translation of the Kanaka­varṇāvadāna was published about a century ago by Heinrich Zimmer (1925, pp. 85–101). Apart from a translation into Japanese by Satoshi Hiraoka (2007, I, pp. 51936), there is now also an English translation of the Kanaka­varṇāvadāna in Andy Rotman’s second volume of Divyāvadāna stories (2017, pp. 218–38).

n.18Vaidya 1959, pp. 276–77. For an English summary of this version, see Tucci 1949, p. 485.

n.19On the Avadāna­kalpalatā (byang chub sems dpa’i rtogs pa brjod pa dpag bsam gyi ’khri shing or dpag bsam ’khri shing for short, Toh 4155) and its popularity in Tibet, both in textual and visual form, see Tucci 1949, pp. 437–41. For an example of such a woodcut depiction of the story of King Kanakavarṇa, see Rani 2005, p. 49 and woodcut 14.

n.20See Feer 1891, pp. 119–20 for a short summary of this versified version of King Kanakavarṇa’s story. One notable difference in this version seems to be that the bodhisattva abandons his bodhisattva path after seeing a child who had died due to insufficent breastmilk. The Sanskrit text, however, has not been published yet; the Sanskrit edition by Kanga Takahata is based on another recension of the Ratna­mālāvadāna that does not contain the story of King Kanakavarṇa.

n.21Several of these emendations have also been proposed in Hiraoka 2009, p. 66.

n.22For a French translation from the Tibetan, see Feer 1891, pp. 117–19; for the Sanskrit text, see Speyer 1902–6, pp. 173–76.

n.23See The Hundred Deeds (Karmaśataka, Toh 340), 3.­417.

n.24The Tibetan translators have here rendered the Sanskrit śloka as tshigs su bcad pa, “verse,” but since it is here used in conjunction with kīrti, “renown,” and śabda, “fame,” it is more appropriate to interpret śloka in its wider sense as “acclaim” or “praise.”

n.25The extant Sanskrit of the Divyāvadāna version (henceforth “the extant Sanskrit version”) here adds digvidikṣu, “in [all] directions and subdirections.”

n.26The extant Sanskrit version in the Divyāvadāna here adds dṛṣṭa eva dharme, “in this present life.”

n.27The extant Sanskrit version here adds the verbs upa­saṃpadya pravedayate, “having attained which, he makes it known,” which then concludes the sentence. Both the Tibetan and the Chinese translation lack renderings of these verbs, which indicates that they were absent in the underlying Sanskrit versions from which they were made.

n.28The Tibetan translation has dam pa’i chos ston te (S, N, H: to), which may be translated as a present or an imperfect past tense. The phrase dam pa’i chos suggests that the translators had read saddharmaṃ, “the good Dharma,” in the underlying Sanskrit text. The standard reading of this set passage, however, is sa dharmaṃ deśayati, “he teaches the Dharma,” which is the reading in the extant Sanskrit. The Chinese translation likewise does not contain an equivalent for the Tibetan dam pa’i.

n.29Following S: sbyin pa la ’gyed pa; D, N: sbyin pa la ’ged pa; Y, J, C: sbyin pa la ’god pa; U: sbyin pa la dged pa. The Stok reading seems to be the older rendering for the Sanskrit dāna­saṃvibhāgasya.

n.30The extant Sanskrit version here lacks an equivalent for the Tibetan bdag gis, “themselves.” The Chinese rendering 自, however, indicates that another Sanskrit version here did read ātmanas, “themselves.”

n.31The Tibetan translation reads gzhan dag la, “to others,” whereas the extant Sanskrit reads sacel labheran dakṣiṇīyaṃ pratigrāhakam, “if they were to meet a recipient worthy of offerings,” which is also the reading in the Dāna Sūtra (Tripathi 1995, p. 18). The Chinese translation here reads 他, in agreement with the Tibetan, which suggests that another Sanskrit version indeed read parān, “to others.” Similar to the extant Sanskrit, the parallel passage in the Pali Dānasaṃvibhāga Sutta (It 26) contains the phrase “if there were to be recipients” (sace nesaṃ paṭiggāhakā assu), and in the following verses the recommended recipients are specified as “noble ones” (ariyesu) and “those worthy of offerings” (dakkhiṇeyyesu).

n.32While the extant Sanskrit version here only reads mātsaryaṃ, “selfishness,” the Tibetan ser sna’i sems kyi dri ma, “mental defilement of selfishness,” indicates that its underlying Sanskrit read mātsaryamalaṃ, “the defilement of selfishness.” This is in agreement with the Pali maccheramalaṃ in the Dānasaṃvibhāga Sutta (It 26). In the first part of the Sanskrit Dāna Sūtra we also find the expression mātsaryamala, but then in the second part and in the verses only mātsarya is used (cf. Tripathi 1995, p. 18). It should be noted here that the Tibetan translators seem to have taken the Sanskrit cittaṃ, “mind,” with mātsaryamalaṃ, “defilement of selfishness,” rather than as the object of the verb paryādāya, “having taken possession of,” with mātsaryamalaṃ being the subject of this sentence. The Tibetan instead reads “they would not remain possessed by the arisen mental defilement of selfishness” (ser sna’i sems kyi dri ma skyes pas kun nas dkris te ma ’dug cig).

n.33The extant Sanskrit version lacks an equivalent for the Tibetan bdag nyid, “themselves,” but instead adds āgṛhītena cetasā, “with a stingy mind.” The latter phrase is also present in the parallel passage in the Sanskrit Dāna Sūtra (Tripāṭhi 1995, p. 18). The Chinese translation, however, has 自, which indicates that another Sanskrit version indeed read ātmanas, “themselves,” as is the case above.

n.34The extant Sanskrit version reads prabhūta­sattva­svāpateyaḥ, “abundant beings and possessions,” but neither the Tibetan or the Chinese translation contain an equivalent for sattva, “being.” The Sanskrit edition, therefore, needs to be emended to prabhūta­svāpateyaḥ, as pointed out by Hiraoka (2009, p. 66).

n.35The extant Sanskrit version adds vaiḍūrya, “beryl,” and śaṅkha, “conch.” The Chinese translation has 珂, “white jade shell,” but, like the Tibetan, it lacks a rendering of vaiḍūrya.

n.36Following S, N: rta pho rgod; D: rta rgod ma. The Stok and Narthang reading would indicate that the underlying Sanskrit text read vaḍava/ vaḍaba, whereas the extant Sanskrit manuscripts read eḍaka, “sheep” or “goats.” The Chinese 騲馬群 also appears to be a rendering of vaḍava, with no indication of any female gender vaḍavā, which the Degé reading rta rgod ma, “mare,” would suggest.

n.37The extant Sanskrit version adds aṣṭādaśa kulakoṭī, but this makes little sense here and has no equivalent in either the Tibetan or the Chinese translation.

n.38We have here followed the standard understanding of the Sanskrit koṭi as “ten million.” The Tibetan reads sa ya phrag lnga bdun ’bum, “5.7 million,” which indicates that the Tibetan translators had understood the Sanskrit koṭi as “100,000,” a number which is normally denoted by the Sanskrit lakṣa. In the Chinese translation the Sanskrit koṭi has been rendered as 億, “hundred million.”

n.39The extant Sanskrit version lacks an equivalent phrase for the Tibetan chos kyi rgyal po’i, “as a Dharma king.” It also lacks the following two sentences on the king’s generosity and the people’s lifespan at the time. The Chinese translation, however, confirms that the phrase and these sentences were present in another Sanskrit version.

n.40The Tibetan translators have taken the Sanskrit grāmaṇyaḥ, “sirs,” as the object of the sentence, and have translated it (as grong mi “city people”), understanding it as together with sarvabaṇijo, “all merchants.” However, in the Sanskrit it is in the vocative case, so it should be understood as the king’s form of address to his courtiers, as is the case throughout the narrative.

n.41The reading anekopāyena in Cowell and Neil’s Sanskrit edition of the Divyāvadāna (p. 292) is erroneous and should be emended to anenopāyena in light of the Tibetan thabs ’dis, as also pointed out by Hiraoka (2009, p. 66). It seems that Vaidya has silently made this emendation in his Sanskrit edition (p.181).

n.42Following S, N: ltas ngan pa; D: ltas pa. The Stok and Narthang reading appears to bring out a specific connotation of the Sanskrit naimittika as “one who can read bad omens.”

n.43Following S, N: mos pa rnams dang / bar snang gi; D: mo ba rnams sa dang / bar snang gi. The extant Sanskrit version reads bhūmyantarī­kṣamantra­kuśalā, “skilled in consulting the earth and the sky,” and it seems that sa, “earth,” in the Degé reading and the following bar snang gi, “sky,” in all three Kangyurs reflect the first half of this compound, albeit misplaced within the sentence.

n.44The extant Sanskrit version lacks this second exclamation, but it is there in both the Tibetan and the Chinese translation.

n.45The extant Sanskrit version here adds the sentences: “Let me count all the beings of Jambudvīpa. And having counted them, let me make measurements.” (sarva­jāmbudvīpān sattvān gaṇeyam / atha gaṇayitvā māpeyam). These sentences are also present in the Chinese translation.

n.46The extant Sanskrit version lacks an equivalent for the Tibetan legs par, “properly,” in this sentence, but in the following passage we find the verb gaṇ- prefixed with saṃ-. The Sanskrit version that underlies the Tibetan translation might have had the prefix here. In the Chinese translation the prefix appears to have been rendered as 善, “well,” but only for the second instance of the verb gaṇ- in this sentence.

n.47The extant Sanskrit version here lacks the phrase “starting with myself,” but it is there in both the Tibetan and the Chinese translation. In the following passage the extant Sanskrit version does contain the phrase rājanaṃ kanaka­varṇam ādau kṛtvā, “starting with King Kanakavarṇa.”

n.48The extant Sanskrit version here adds paraṃ deveti, “saying, ‘Yes, Your Majesty.’” The Chinese translation lacks a rendering of this phrase, as is the case in the Tibetan.

n.49The extant Sanskrit version lacks this sentence and any mention of the eleventh month in the following sentence, but it is there in both the Tibetan and the Chinese translation.

n.50The Chinese translation adds that the entire storehouse was empty and that the food that was left was only for one person for one day.

n.51The extant Sanskrit version here adds ime sattvāḥ, “these beings [are].” The Chinese translation contains this exclamatory sentence only once.

n.52The extant Sanskrit version reads atraiva kālaṃ kariṣyatīti, which does not make sense here and should be emended to atraiva evaṃ kariṣyatīti in light of the Tibetan de nyid du de ltar byed. This emendation is also suggested by Hiraoka (2009, p. 66).

n.53The extant Sanskrit version here adds adhārmikair, “unrighteous,” which is also reflected in the Chinese translation as 非法.

n.54We have followed the order of the Tibetan, which begins with phar mi ’dzin pa, “those who do not honor their fathers.” The extant Sanskrit begins with amātṛjñair, “those who do not honor their mothers,” and lacks any reference to fathers, but the Chinese 不識父母 confirms that another Sanskrit version did contain the phrase apitṛjñair, “those who do not honor their fathers.” This emendation is also suggested by Hiraoka (2009, p. 66).

n.55We have followed the extant Sanskrit prati­mukhaṃ smṛtim upasthāpya in rendering this standard description of establishing mindfulness. The Tibetan translation reads dga’ ba dang bde ba dang dran pa mngon du bzhag nas, “established joy, ease, and mindfulness.” This might reflect a misreading of the Sanskrit pratimukhaṃ as prītisukhaṃ. The Chinese translation also makes no mention of establishing joy and ease.

n.56The Tibetan reads gang cung zad skye ba’i chos de thams cad ni / yun mi ring ba nyid du ’gag pa’i chos yin, “Whatever is of the nature to arise, all that, before long, is of the nature to cease,” but this well known statement, most famously expressed by Ājñātakauṇḍinya after the first teaching of the Buddha in The Sūtra of Turning the Wheel of Dharma (Dharma­cakra­pravartana­sūtra, Toh 31), usually never contains the phrase “before long.” Rather, the Sanskrit acirād eva, “before long,” should here be taken with the verb viditvā, “having realized,” as has been done in the Chinese translation.

n.57The Tibetan rkyen ji lta ba bzhin gyis chos thob pa mthong nas seems to be preserving the full and correct reading, whereas the extant Sanskrit only reads yathāprāptān dharmān avalokya, “after beholding dharmas according to how they are attained.” The Chinese 以是因緣 likewise indicates that another Sanskrit version contained the word pratyaya, “condition,” a term which is traditionally associated with the pratyekabuddha (see Norman 1983), and which is reflected in such Tibetan renderings as rten ’brel bsgom, “one who meditates on dependence,” and rkyen gcig rtogs, “one who realizes every single condition.” On the basis of the Tibetan and Chinese translations, Hiraoka (2009, p. 66) has proposed to emend the Sanskrit edition to yathāpratyayaṃ prāptān dharmān avalokya.

n.58Following S, N: tshigs su bcad pa ’di smras pa; D reads tshigs su bcad pa ’di dag smras pa, with “verses” (plural). The Stok and Narthang reading is in agreement with the Sanskrit gāthāṃ bhāṣate. The Degé reading, however, seems to indicate that at some point the following five-lined verse, with its added line (see next note), was considered to consist of two verses. On this verse, and its Sanskit and Pali equivalents, see the Introduction.

n.59The Tibetan chags las sdug bsngal ’di byung ste is here followed by a further line that reads chags la (S, N, H: las) sdug bsngal ’di ’jug pas. This extra line, which would make for an odd five-lined verse, seems to be a slightly adjusted second rendering of the same underlying Sanskrit line which somehow got included in the final Tibetan translation. In the extant Sanskrit version, the line reads snehānvayaṃ saṃ­bhavatīha duḥkham, “Suffering here [in saṃsāra] comes about as a consequence of attachments,” for which the parallel verse in the Mahāvastu (Mvu I 358) reads snehānvayaṃ duḥkham idaṃ prabhoti, which is in close correspondence with the Pali snehanvayaṃ dukkhaṃ idaṃ pahoti in the Khaggavisāṇa Sutta (Sn 6). The Tibetan ’di could possibly be a shortened form of ’di na for metrical reasons, but in light of these attestations, it seems more likely that the Sanskrit version from which the Tibetan translation was made did in fact have idaṃ, “this,” instead of iha, “here,” which is the reading found in the extant Sanskrit version, as given in the Introduction.

n.60The Tibetan reads de phyir mkhas pas chags sun phyung, “Therefore the wise one, wary of attachments,” but the extant Sanskrit version reads ādīnavaṃ snehagataṃ viditvā, so the Tibetan seems to have misread the Sanskrit viditvā, “realizing,” as vidvān, “the wise one” (“Therefore the wise one, wary of attachments, should move alone like a rhinoceros”). Like the extant Sanskrit, the Chinese translation of this verse makes no mention of “the wise one,” but then it lacks the first line of the verse and differs significantly in the subsequent lines: It is from attachments that suffering originates; One should therefore abandon attachments, One should delight in a solitary place, Like the single horn of a rhinoceros.

n.61Like the Sanskrit khaḍgaviṣāṇa, the Tibetan bse ru allows for two interpretations: (1) either bse is interpreted as “tanned leather,” which, together with ru, “horn,” makes for a compound word that means “rhinoceros;” (2) or bse is taken as “rhinoceros” for short, which would make for a phrase meaning “the horn of a rhinoceros.” We have chosen to translate the term according to the interpretation commonly found in the Buddhist Sanskrit tradition, since this is the tradition that would have been familiar to the Tibetan translators. For more on the ambiguity of the compound word khaḍgaviṣāṇa, see n.­5 to the Introduction.

n.62The extant Sanskrit version here adds viśuddhena, “completely pure,” which is also rendered in the Chinese translation as 清淨.

n.63The extant Sanskrit version reads “Let me receive and eat alms from the house of King Kanakavarṇa (kanaka­varṇasya niveśanāt).” The Chinese translation agrees with the Tibetan.

n.64The Tibetan de’i rdzu ’phrul ’di lta bus indicates that the Tibetan translators read tasya evaṃ ṛddhyā where the extant Sanskrit has tata eva ṛddhyā, “from there, by miraculous ability.” The Chinese rendering 即以神通 seems to be in agreement with the extant Sanskrit reading.

n.65Following S, N: dbul ba dag gi nang na shi dang mtshungs par gyur pa ste; D: dbul ba dag ni nang na shi dang mtshungs par gyur pa ste. The Sanskrit reads: maraṇasamaṃ dāridryam.

n.66The last line of this verse is not preserved in the extant Sanskrit version. It is, however, present in the Chinese translation, albeit more elaborately rendered for metrical reasons.

n.67The extant Sanskrit version lacks an equivalent api for the Tibetan kyang, “even,” but instead it adds the phrase yadi vā na paribhokṣye, “or if I do not eat it.” The Chinese translation agrees with the extant Sanskrit version.

n.68The Tibetan translation reads ’di ’dra ba’i drang srong tshul khrims dang ldan pa dge ba’i chos can de (S, N: te), without the second half of the sentence that would have contained the verb. We have supplied this from the extant Sanskrit: mama niveśane ’dya yathādhautena pātreṇa nirgamiṣyati, which is also reflected in the Chinese translation. By way of explanation, the Chinese translators have paraphrased the Sanskrit yathādhautena pātreṇa as 其不得飯食空鉢, “without having gotten food in his empty alms bowl.”

n.69The Tibetan sbyin pa gtong ba, lit. “giving away a gift,” indicates that the Tibetan translators read dānātisargaḥ in the underlying Sanskrit, whereas the extant Sanskrit version in the Divyāvadāna reads odanātisargaḥ, “giving away boiled rice.” Rotman translates it “this last bit of rice” (Rotman 2017, 233). The Chinese rendering 最後布施 also suggests an underlying Sanskrit reading with dāna-.

n.70The extant Sanskrit version is more elaborate in the description of this scene: “Thereupon King Kanakavarṇa took the bowl of that great sage (tasya maharṣes), put the single measure of food in the bowl, and holding the bowl in both hands, fell to his knees (ubhābhyāṃ pāṇibhyāṃ pātraṃ gṛhītvā jānubhyāṃ nipatya) and placed the bowl in the blessed pratyekabuddha’s right hand.” Like the Tibetan, the Chinese translation does not describe the king as holding the bowl in both hands and falling to his knees, nor does it here refer to the pratyekabuddha as “that great sage.”

n.71This statement refers to the tradition whereby a Buddhist monk will usually give a short Dharma teaching or recitation for any donor from whom he receives alms. It is said that a pratyekabuddha, however, simply accepts the alms and silently leaves, but in doing so can display supernormal powers as a highly realized being. As stated by Vasubandhu in the Abhi­dharma­kośabhāṣya (Pradhan 1975, p. 183.21; Toh 4090, F.159.b), pratyekabuddhas “are not without compassion, because they display their miraculous power for the sake of benefitting beings” (nāpi niṣkaruṇāḥ / sattvānu­grahārtham ṛddher āviṣkaraṇāt). This statement implied that by manifesting supernormal abilities in bodily form, pratyekabuddhas can instill faith in the Dharma in those who witness their miraculous feats, but without giving any verbal teaching.

n.72The extant Sanskrit version reads piṇḍapātram, “alms bowl,” which is likely a scribal error for piṇḍapātaṃ, “alms,” as reflected in the Tibetan bsod snyoms and the Chinese 所施食. Elsewhere in the extant Sanskrit, an alms bowl is simply referred to as pātra, “bowl.”

n.73The Tibetan ’di ltar rdzu ’phrul gyis indicates that the Tibetan translators read evamṛddhyā where the extant Sanskrit version has tata eva ṛddhyā, “from there, by miraculous ability.” The Chinese translation reads 即以神通, which seems to be in agreement with the extant Sanskrit.

n.74The Chinese translation here adds 并諸大眾, “together with the great crowd.”

n.75The extant Sanskrit version lacks an equivalent for the Tibetan lha’i bka’, “Your Majesty’s instruction,” and it has these two sentences in reverse order. The Chinese translation is in agreement with the extant Sanskrit.

n.76The extant Sanskrit version reads meghāśca pra­varṣayantaḥ pāṃśūn śamayanti, “and raining rainclouds settled the dust.” The Chinese translation instead repeats the sentence 涼風吹閻浮提其地淨, “cold winds began to blow in Jambudvīpa that made the ground clean.”

n.77Following S, N, H: bca’i bar bya; D: bza’i bar bya. The Stok, Narthang, and Lhasa reading is in agreement with the Sanskrit khādanīyaṃ.

n.78The extant Sanskrit version here includes skandha-, “stalks,” which is not rendered in the Tibetan. The Chinese translation does contain a rendering of this as 莖.

n.79The Tibetan here adds til mar, “sesame oil,” and in the Chinese translation we also find oil included in the list, but this item seems out of place. The extant Sanskrit lacks this item. Note that there will be a separate rain of sesame oil further on.

n.80Following S, N, H, Y, Q, C: bu ram; D: bur mar.

n.81The extant Sanskrit version here concludes the list with piṣṭa­khādanīyam, “edible flour,” and it lacks the following sentence, which is found in both the Tibetan and the Chinese translation.

n.82We have followed the Tibetan for this last sentence. The extant Sanskrit version is shorter, reading only phalam anyad bhaviṣyati, “There will be another fruit.” The Chinese translation likewise makes no mention of leaves and flowers here.

n.83The extant Sanskrit version lacks an equivalent for the Tibetan ’bru’i rigs thams cad dang, “and all types of grains.” The Chinese translation only reads 等, “and so forth.”

n.84Following S, N, H: rdzas rnam pa sna tshogs; D: zas rnam pa sna tshogs, “various kinds of food”. The Stok, Narthang, and Lhasa reading is preferable in light of the extant Sanskrit which reads nānāvidhadūṣya, “various kinds of cloths.” The Tibetan translators appear to have read dravya, “substance” or “material,” instead of dūṣya, “cloth.” The Chinese rendering 種種雜, “various miscellaneous things,” likewise seems to suggest that dravya may have been the underlying Sanskrit reading.

n.85The Tibetan translation has taken this sentence as an exhortation: “Monks, you should not be uncertain and in doubt about whether that king Kanakavarṇa at that time, at that moment, was someone else. Monks, you should see that at that time, at that moment, it was I who was named Kanakavarṇa.” As a standard passage that comes at the end of an avadāna , we have followed the extant Sanskrit version here, with which the Chinese translation is in agreement.

n.86The extant Sanskrit version here includes bhikṣavaḥ, “Monks,” at the beginning, but this is absent in both the Tibetan and the Chinese translation.

n.87Following S: bza’ bar mi bya’o ; D: bar mi bya’o.

n.88The Tibetan text is problematic in this passage with repeated use of mi bya’o. We have, therefore, translated the passage in accordance with the passage at the beginning of the text.

n.89Following the Tibetan ’phags pa’i skyes bos legs par bshad pa. The extant Sanskrit version reads ārya­janeṣu bhāṣitaṃ, “What is said among the noble,” but the Tibetan indicates that the Tibetan translators were reading āryajane subhāṣitaṃ. The Chinese 聖眾中善語 is also a rendering of such a reading.

n.90The extant Sanskrit version and the Chinese translation add another verse that is commonly found at the end of avadānas : “Whether a beautiful good deed Or an unbeautiful bad deed, It has its karmic fruition, It will certainly bear fruit.”

n.91There is no mention of bodhisattvas being present in the extant Sanskrit version or the Chinese translation. These versions instead read: “Elated, the monks, nuns, devoted laymen, devoted laywomen, gods, nāgas, yakṣas, gandharvas, asuras, garuḍas, kinnaras, mahoragas, and so on, and the entire assembly, rejoiced at what the Blessed One had said.”