^{page 7 note 1} As I am writing far away from any Chinese library, nor have I yet received the last volumes of the Taishō edition of the Tripiṭaka, I cannot help using for this note the Nanking edition of the works by Kwei-chi and Shen t'ai: therefore I cannot give any better reference.

^{page 8 note 1} When I first wrote this article, although I could not accept the restoration proposed by Nanjiō (Nyāyadvārakatarkaśāstra), I was inclined to think that the title of Nos. 1223–4 in Nanjiō's Cat. was in Sanskrit Nyāyadvāraśāstra. When I received the proofs, I substituted Nyāyamukha for Nyāyadvāra for the following reasons. First of all, those who know Buddhist Chinese are aware of the fact that Ch. , like Tib. sgo, may be the equivalent of Skt. dvāra, as well as of mukha (cf. expressions like vimokṣamukha, etc.). Therefore, according to the Chinese a restoration Nyāyamukha would be quite as possible as Nyāyadvāra. Moreover, the recently published text of the Tattvasaṅgraha confirms the conjecture that the title of the original book by Diṅnāga was, in fact, Nyāyamukha. In the Tattvasaṅgraha we read (p. 372, 1. 23):—

evaṃ, nyāyamukhagraniho vyākhyātavyo diśānayā |

jñānam ity abhisambandhāt pratītis tatra coditā ‖

Kamalaśīla thus comments on this śloka:—tatrāyaṃ Nyāyamukhagranthaḥ: “yaj jñānam artharūpādau viśeṣaṇābhidhāyakābhedopacāreṇāvikalpakaṃ tad akṣam akṣaṃ prati varttata iti pratyakṣam” viśeṣaṇaṃ jātyādi, etc.

In the index of quotations the editor of the text considers this passage to be from the Nyāyapraveśa; but he adds in a footnote: ādarśapustake pāṭhabhedo dṛśyate (p. 90). The ādarśapustaka is the text of N.P. as it is printed in the same collection, where the definition of pratyakṣa is given in the following terms: pratyakṣaṃ kalpanāpodhaṃ yaj jñānam arthe rūpãdau nāmajātyādikalpanārahitaṃ tad akṣam akṣaṃ prati vartata iti pratyakṣam.

But there is no question of pāṭhabheda, because, as may be gathered from Kamalaśīla himself, we have here not a quotation from the N.P., but from the Nyāyamukha. In the commentary on kārikā, 12 (Yuan Chwang's transl., Tokyō ed., p. 3a) Diṅnāga quotes half a śloka as a mūlakārikā (i.e. taken from the Pramāṇasamuccaya):

Then, commenting on the definition pratyakṣaṃ kalpanāpoḍham, he writes

Now it is evident that this definition tallies fairly well with that given by Kamalaśīla in his quotation from the Nyāyamukha. So that there is no doubt that Nanjiō 1223, 1224 represent the Chinese version of the Nyāyamukha, which is the rigs.pai.sgo. attributed to Dinnāga and quoted very often in the Buddhist logical works preserved in the bsTan-ạGyur.

^{page 10 note 1} In the introductory part of his glosses on the Nyāyapraveśa.

^{page 12 note 1} This thesis can be met with in the first chapter of the fragment of the Tarkaśāstra (?) attributed to Vasubandhu. Its text has been restored into Sanskrit by me, and will very soon be published together with other logical works preserved in Chinese.