^{page 1125 note 1} Dr. Beckh might have added that the merit of having obtained the copy is due to E. Pander, who at that time was Professor at the Peking University and on very friendly terms with the Lamas (see Zeitschrift für Ethnologie, vol. xxi, p. (203), 1887).Google Scholar

^{page 1126 note 1} Dr. Beckh (p. viii) asserts that an alphabetic list of the Sanskrit titles of the Kanjur had not previously existed. But there is one by L. Feer appended to his “Analyse du Kandjour” (Annales du Musée Guimet, vol. ii, pp. 499–553Google Scholar); Beckh's list, however, is far superior and represents the first complete and accurate inventory of the Sanskrit titles.

^{page 1127 note 1} Dr. Beckh credits this date to Waddell (Buddhism of Tibet, p. 159Google Scholar); but Waddell in this passage exactly copies a statement of A. Csoma (Asiatic Researches, vol. xx, p. 42Google Scholar), and even repeats the misnomer “wooden types” (instead of “wooden blocks”). As has been demonstrated by M. Pelliot, Csoma's dates are all unreliable. In fact, the preface of the index volume of the Narthang edition is dated 1742 (c'u p'o k'yii lo, “water male dog year”). If Beckh had consulted Csoma's work in its original issue, he would have been saved from the incorrect statement (p. vi) that Csoma gives no information as to the Kanjur edition utilized by him. Csoma, indeed, states that his study is based on the Narthang edition in the possession of the Asiatic Society of Calcutta and procured by Hodgson. Moreover, his analysis is in harmony with the index of Narthang. Koeppen also (Lamaische Hierarchie, p. 280Google Scholar) had already pointed out this fact.

^{page 1129 note 1} Zeitschrift für Ethnologie, vol. xxi, p. (203), 1889.Google Scholar

^{page 1129 note 2} Mythologie des Buddhismus, p. 178.Google Scholar

^{page 1129 note 3} “Die orientalischen Religionen” (in Kultur der Gegenwart), p. 161.Google Scholar

^{page 1129 note 4} Loc. cit., p. (201). Pander gives the date as 8th year of Yung-lo, but .wrongly identifies it with the year 1411 instead of 1410.

^{page 1130 note 1} The exact date is ; in Tibetan “On the ninth day of the third month of the eighth year of the period Yung-lo (1410).”

^{page 1131 note 1} The colophon is translated in the writer's Dokumente, i, p. 52.Google Scholar

^{page 1132 note 1} Tib. da-lta c'en-po C'iṅ gur-gyi lo-tsts'a-ba c'en-po.

^{page 1132 note 2} In particular an edition of the dhāraṇī styled Vajravidāraṇā.

^{page 1133 note 1} This has been done by him in several cases; and he observes that the context, as in many colophons, remains obscure (p. 68), or that the entire colophon is very difficult and obscure (p. 136). In such cases it would have been advisable to publish the texts of the colophons in extenso in order to enable future students to make the best use of them.

^{page 1133 note 2} Huth, , Geschichte des Buddhismus, vol. ii, p. 165Google Scholar, and Laufer, , Dokumente, i, p. 53Google Scholar. From the text of Jigs-med nam-mk'a it follows that the first Narthang edition was printed in black by means of Chinese ink. Also the later edition of 1742 was printed in black. This point is mentioned here because Schmidt and Boehtlingk (Verzeichnis der tibetischen Handschriften, p. 4Google Scholar) speak of a Narthang edition of the Kanjur in St. Petersburg as printed in red (that is, vermilion); but no such vermilion print has ever been issued from the press of Narthang. Technically it is impossible to print from the same blocks a copy in black and another one in vermilion; the same blocks can be utilized for impressions either in black only, or in vermilion only. From the summary of contents given by Schmidt and Boehtlingk it follows that the Kanjur in question cannot be an edition of Narthang, for in the latter the section Nirvāṇa (Tib. myaṅ- das) occupies a separate department (No. vi); while in their edition this section is joined to the Sūtra class. It is therefore probable that this edition is the one printed at Derge, which is, indeed, in vermilion.

^{page 1136 note 1} Cf. T'oung Pao, 1914, p. 106Google Scholar. The same error of taking sNa-nam for Samarkand in connexion with a purely Tibetan name is committed by P. Cordier (Cat. du fonds tibétain, ii, p. 84, No. 46Google Scholar). These translations are based on the fact that Jäschke in his Dictionary, with reference to rGyal rabs, assigns to sNa-nam the meaning of “Samarkand”; Chandra Das, without adducing any proof, has merely copied Jäschke. The question is whether Jäschke is correct, and on what evidence his opinion is founded. In rGyal rabs we find mention of a queen from sNa-nam, married to King Mes 'Ag-ts'om. I. J. Schmidt (Geschichte der Ost-Mongolen, p. 349Google Scholar), translating from the Bodhi-m r, the Kalmuk version of the Tibetan work, styles her the chief consort from a clan of Samarkand, without advancing evidence for this theory. In his Forschungen (p. 231, St. Petersburg, 1824)Google Scholar, however, the same author states that the Kalmuk original has in this place Samardshen. The date of the Kalmuk work is not known; since Kalmuk writing was framed as late as 1648, the Kalmuk translation, as a matter of principle, cannot be earlier than the latter part of the seventeenth century. The case therefore hinges on the point whether the Kalmuk rendering of recent date is correct in its understanding of the Tibetan word. Neither Kovalevski nor Golstunski (in their Mongol Dictionaries) has recorded the word Samardshen. Whether sNa-nam ever had such a meaning remains to be proved, if indeed it can be proved. For the time being the matter is open to doubt, and it seems more than doubtful that the Tibetans ever had relations with Samarkand. But the supposition that Tibetan authors living and working on Tibetan soil were born in Samarkand, which would presuppose the existence there of a Tibetan colony in the T'ang period, is somewhat adventurous.

^{page 1137 note 1} By the third Dalai Lama bSod-nams rgya-mts'o (1543–88); see Huth, , Geschichte des Buddhismus, vol. ii, p. 224Google Scholar. The foundation of the monastery, accordingly, falls within the Wan-li period (1573–1620), during which the Berlin Kanjur was presumably copied. It is therefore impossible to assume that a translator named in this edition could have come from Li-t'ang.

^{page 1137 note 2} The Life of the Buddha, p. 230.Google Scholar

^{page 1137 note 3} See T'oung Pao, 1908, pp. 20, 22.Google Scholar

^{page 1138 note 1} Huth, , loc. cit., p. 118Google Scholar. In the index volume of the Kanjur of Derge (fol. 97b), where he has the attribute sNar-t'aṅ-pa, “the man from Narthang,” he is expressly listed among the collaborators of the Kanjur.

^{page 1138 note 2} Ibid., p. 122. Tib. zlos-gar is not “art of dancing”, as translated by Huth, but “dramatic art” (nāṭaka).

^{page 1138 note 3} Tib. brda sprod-pa c'en-po, which does not mean “der grosse Erklärer von Symbolen”, as Dr. Beckh (p. 128) translates. A title which has greatly embarrassed the author occurs in the same colophon, in the form sgrai gtsug lag lam rmoṅs-pa, tentatively translated by him “one who obscures the road of linguistic science”, and accompanied by a note to the effect that this might possibly be a proper name, though somewhat strange. It is not, however, a proper name but rather a title. The word rmoṅs-pa was indicated as an epithet of Tāranātha by A. Schiefner (Tāranātha's Geschichte des Buddhismus in Indien, p. vii, n. 2Google Scholar), and rmoṅs-pai gñen-po, “the adviser of the ignorant,” is a title bestowed on members of the clergy (Dokumente, i, p. 61Google Scholar). Thus the above title apparently means “one who is a guide along the dark points in the science of language”. For the rest, rmoṅba, rmoṅs-pa, is not a transitive verb, and never means “to obscure” (which is rmoṅs-par byed-pa), but “to be obscured, obscurity”, etc. The monastery Sar-sgreṅ on p. 9 is to be corrected into Ra (or Rva)-sgreṅ, as shown by a Peking print of the work in question containing the same colophon. This follows also from the historical context of the passage, owing to the mention of Brom-ston, who was the founder of the monastery Ra-sgreṅ.

^{page 1139 note 1} After the above was written, I had meanwhile an opportunity of exactly collating the Index of the Kanjur of Derge with that of Berlin, and may now positively state that the two editions are independent, and that the Berlin version cannot be traced to that of Derge. There are treatises in the latter wanting in the former and vice versa; above all, the arrangement of the works in the section Tantra is widely different in Derge from the Berlin copy and other editions of the Kanjur. I hope to come back to these questions in detail in a future bibliographical study of the Kanjur. The collation with other editions bears out the fact that many colophons of Berlin are sadly deficient, and especially that numerous proper names are disfigured. A few examples may suffice. On p. 76a (below) we read of a monastery Yu-tuṅ-lhan in Nepal; the real name is Yu-ruṅ, while lhan is an error for lhun, which does not belong to the name, but to the following -gyis grub-pai gtsug lag k'aṅ (“miraculous monastery”). On the same page, and again on p. 77, we meet the wrong name of a translator in the form La-bciṅs-yon-tan-bar; it should read Yon-tan-bar from C'iṅs (written also C'iṅs). Byai gdod-pa-can (on p. 95b) should be gdoṅ (“the Bird-faced one”). The name of the translator K'u-ba-lha btsas (p. 126a) is correctly K'ug-pa lhas-btsas; instead of Klogs-skya (ibid.) read Glog-skya; instead of Ḍo-ma-bi (p. 87a) read Ḍombi. In many cases the Berlin colophons are incomplete, or there are none at all where they can be supplied from other editions. It is therefore unsafe to found a study of the translators on the work of Beckh. The colophon on p. 106 (No. 29) has been entirely misunderstood by the author: he distils from his corrupted text a monastery Dbe-rñid in Kashmir, and makes it the place where the translation took place. Neither, however, is the case; dbe rñid is an error for dpe rñiṅ (“old book”), and the passage means, “The Paṇḍita Parahitaprabha [thus written in the Index of Derge] and the Locāva gZu-dGa-rdor have translated the work, and edited it on the basis of an ancient book hailing from the monastery Amṛitasambhava (Tib. bdud-rtsi byuṅ-gnas)in the country of Kaçmīra.” A wrong translation occurs on p. 67 in the colophon of mDsaṅs blun, which does not mean “seems to be a translation from Chinese”, but “it has been translated from Chinese”. The verb snaṅ-ba never assumes the significance “to seem”.