Introduction

i.1The Sūtra of the Moon (1) is a short discourse set in Śrāvastī that presents a Buddhist account of a lunar eclipse. It describes the asura king Rāhu seizing the moon and the Buddha calling for its release. When the eclipse begins, the god of the moon asks the Buddha for refuge, prompting the Buddha to demand that Rāhu let the moon go. When questioned by the asura Vemacitra, Rāhu explains that if he had not released the moon, his head would have split into seven pieces.

i.2The Sūtra of the Moon (1) is a Tibetan translation of a canonical Theravāda text and thus was probably translated from Pali. It belongs to a group of thirteen late-translated sūtras (gsar du ’gyur ba), as noted in Butön Rinchen Drup’s History of Buddhism and in the Lithang, Degé, and Urga Kangyurs. One of these sūtras, The Sūtra of the Sun (Toh 41), presents an almost identical narrative concerning a solar eclipse. It features one additional verse not found in The Sūtra of the Moon (1), and the same is true of the Pali counterparts of these two texts. The equivalent of The Sūtra of the Moon in the Pali canon, the Candimasutta, is included in the Saṃyutta Nikāya (SN 2.9). Except for some words and one short phrase, the Tibetan and Pali are very similar.

i.3There is another text in the Kangyur titled The Sūtra of the Moon (2) (Toh 331), which was translated from Sanskrit in the time of the early diffusion (eighth–ninth centuries). This text has Sanskrit, Chinese, and Old Uyghur parallels. It shares much of the same content as the sūtra translated here, but it differs in its wording, the number of verses and passages of its prose, its setting, and the characters depicted.

i.4The Sūtra of the Moon (1) and The Sūtra of the Sun are popular among Tibetans today, appearing in Tibetan collections of mantras and protective texts. In the Theravāda tradition, they are included in Paritta collections, which have the function of providing protection through recitation.

i.5Peter Skilling dates the translation of the thirteen late-translated sūtras to the first decade of the fourteenth century. The Sūtra of the Moon (1) is found in the Kangyurs of the Tshalpa, the Thempangma, and the mixed lines, as well as in some canonical collections from Western Tibet.

i.6The colophons of all versions of the Tshalpa line only indicate the conclusion of the sūtra, while the other collections mention the translators and the place where they worked: Ānandaśrī and Tharpa Lotsāwa Nyima Gyaltsen Palsangpo at the monastery of Tharpa Ling in Central Tibet.

i.7Ānandaśrī is described in the colophon of the Ārya­maitrī­sūtra, which he also translated together with Künga Gyaltsen Thupten Palsangpo, as a prominent paṇḍita from Sri Lanka, but little is known about how or why he came to be in fourteenth century Tibet or how long he stayed there. Tharpa Lotsāwa, apart from these thirteen sūtras, also translated several other texts that are preserved in the tantra section of the Kangyur, working mainly with Indian and Nepalese paṇḍitas. Notably, he was one of the teachers of Butön Rinchen Drup (1290–1364).

i.8It is interesting to note that this Tibetan translation gives the Sanskrit title Candrasūtra, while its presumed Theravāda source would have had the title Candimasutta. It is possible that the translators and later editors Sanskritized the title in accordance with the translation practices of the time. The title Candrasūtra is not attested in Sanskrit works, but the title Candra­maṇḍala­sūtra is documented in a Sanskrit fragment from Central Asia that lists different sūtras.

i.9The French Sanskritist and Orientalist Léon Feer studied Buddhist accounts of eclipses, including the legend of Rāhu according to Hindu and Buddhist texts. He translated Daniel John Gogerly’s English translation of the Candimasutta into French and produced his own French translations of Toh 42 and Toh 331. His notes are detailed and rich in explanations. He also finalized and published the unfinished work of Paul Grimblot on some paritta texts, presenting the Pali equivalent to Toh 42, excerpts of its commentary in the Aṭṭhakathā, and its translation into French.

i.10Peter Skilling has studied the thirteen late-translated sūtras (Toh 31–43) in detail, examining the location of these texts in the Kangyur, and their colophons, contents, translators, and possible original language. He notes parallels, dates the translations, and reviews scholarship on these works.

i.11Feer translated Toh 42 from Tibetan into French in 1865. However, there are many translations from Pali. There is Feer’s translation into French (1871), Wilhelm Geiger’s translation into German (1930), Rhys Davids’ translation into English (1950), and a recent translation into English by Bhikkhu Sujato (2018).

i.12Our translation is based on the Tibetan text as found in the sūtra section of the Degé Kangyur, but we note variant readings from the witnesses of the Comparative Edition (dpe bsdur ma), the Stok Palace Kangyur, and the Bardan Collection. In addition, we have compared the Pali with the Tibetan and noted cases where the Pali presents different readings. We occasionally refer to Toh 331 and its Sanskrit parallel.

i.13Our translation benefited from Feer’s notes on Toh 42 and Toh 331, as well as from his translations. We also consulted Geiger’s translation of the Pali Candimasutta into German, and Sujato’s translation into English.