________________
logical theory
(3)
kinds of syllogism, * one of five parts and the other of two it was Diñnāga who took the decisive step of rejecting sqat and fanaa as superfluous even in the Parärthāgumāna. There can be no doubt that the Sätrakāra and the Bhāssakāra of the Nyāyadars'ana recognize only the five-membered syllogism, and yet it is curious to observe that in the numberlese argumente which they employ in justifying their system and convincing their opponents they hardly set forth the five-membered syllogism. In the Vaiseșika school the distinction between frigigata and arritia has been recognized by Pras'astapāda whose corresponding terms are 'स्वनिश्चितार्थानुमान' and 'परार्थानुमान, but be has not like Diabāga reduced the latter to three members. If the
I theory which does away with 3990 und fanga from the Parärthānumāna be sound, -of which I am not altogether sure the credit for the reform evidently belongs to Dinnāga. It is said that "the significance of the middle term (called Hetu) for inference and hence for the theory of reasoning, is for the first time discussed by Dinna [ Dinnāga ) and the result of his study is the famous doctrine of the 'three Phases of Hetu't. This is known as the doctrine of the Dead or the three essential conditions of a good betu, which are enumerated in our text asetadate', 'atah arah' and 'grata'. These are the same as the conditions which were laid down in the couplets of Kās'yapa wbich Pras'astapāda bas quoted and to which we referred in the preceding paragragh. It has been said, however, that there was borrowing from Dirnāga on the part of Pras'astapäda, and "this borrowing | Pras'astapāda bas tried to conceal," and such a thing is in entire accord with the natural wish of a school, when it has to appropriate fruitful ideas from another to disguise and adopt them in form if not in substance." I refuse to base my conclusions on the hypothesis of a moral depravity on the part of Pras'astapāda which there is no reason to assume. I rather agree with Dr. Randle in thinking that
* E.I. L. p. 209. + Bugiura'. "Hindu Logic" P, 35.
Stoherbataty.
