The importance of this philosophical work was recognised more than three decades ago when the three distinguished orientalists Sylvain Lévi, Louis de la Vallée Poussin and Theodor Stcherbatsky put their heads together to bring out an edition of this treatise. Unfortunately, many difficulties stood in the way of its publication and the editors had to be content with publishing the first two KoSgasthanas, and that also at long intervals. The idea of taking up its publication came to my mind more than a decade ago. The late Sir Asutosh had also desired to publish an edition and procured a ferro copy of the transcription made by a Japanese scholar. I utilised the Ms. in the A.S.B. as well as the ferro copy of the transcript. These were subject to the faults usual to a tough text of this type, and we struggled hard with these imperfect copies for settling the text when the call to the Round Table Conference took me to England. I utilized my visit to that country by securing a rotograph of the Cambridge Ms. with the kind help of my friend Dr. E. J. Thomas. Our task became - easy with this rotograph as this copy was not only well written but also preserved an almost accurate text, a rare instance in the collection of our ancient and mediaeval Sanskrit Mss. While our edition had nearly reached its end (in Mss.), the welcome edition of Prof. Wogihara of Japan came to our hands. It is printed in Roman characters but the work has been excellently executed and we have been able to point out only a few better readings here and there.

It is a matter of regret that not only the Sanskrit originals of all the Abhidharma texts of the’ Sarvastivadins but also the Sanskrit originals of the Koga and its bhdsya are lost. Thanks to the extraordinary zeal and energy combined with vast knowledge of the late Belgian savant Prof. Louis de la Vallée Poussin, the Koga has been restored from its Tibetan version and from a few fragmentary finds of Mss. in Nepal, but the Bhasya still awaits that restoration or discovery. Prof.Louis de la Vallée Poussin has made our task easy by his invaluable translation of Hiuen Tsang’s commentary on the Koga, and Rev. Sankrtyadyana by publishing the text of the Kosa with valuable notes and tables. Much spade work was previously done in the field by Prof. Stcherbatsky and the late lamented student of his Dr. Rosenberg. The labours of these scholars have concretised our ideas about the teachings of the Sarvastivadins, but still there remains the weak spot that the studies so far made are based primarily on Japanese, Chinese and Tibetan texts and commentaries.

The insurmountable difficulty in the editing of the Vyakhya is the want of the Koéga-bhasya in original, of which the present text is a commentary. Those who are acquainted with the method and style of writing of our old commentators will realise what a tremendous difficulty one has to face to understand a vyakhya or tika without the original. Compliments are therefore due to Prof. Wogihara and Profs. Stcherbatsky and Sylvain Lévi for overcoming this difficulty with the help of the Tibetan translation of the Bhasya. In our edition too we had torely on the Tibetan text of the Bhasya but our task has been made easier latterly by the appearance of the edition of Prof. Wogihara.

The Koéa, as the name suggests, is an epitome of the Abhidharma literature of the Sarvastivadins, who had seven Abhidharma texts, now available only in Chinese translation, and a short synopsis of which was published by Prof. Takakusu in the J. P. T.S. 1904. These seven texts may not be paralleled to the seven Pali texts but there cannot be any doubt that the subject-matter of both the sets of Abhidharma texts isthe same. The if dnaprasthana-sitra has been partially studied by Prof. La Vallée Poussin in the B.E.F.E.O., vol. XXX (1930) and Mélanges chinois et bouddhiques,.

The Koga may be compared to the section on Pafifia of the Visud- dhimagga. Though the plans of the two masterminds are different, both have tried to compress the whole Abhidhamma literature in one work of theirs. Vasubandhu divides his treatise into eight sections devoted to dhdtu, indriya, lokadhGtu, karma, anuSaya, dryapudgala, jfiGna, and dhyana while Buddhaghosa chooses the following head- ings: Khandha-niddeso, dyatana-dhatu-niddeso, indriya-sacca-niddeso, pafifia=bhimi-niddeso, etc. Each of the items has been taken up by Buddhaghosa, but the method of his exposition follows the old line of the Pali School. He is satisfied with bare enumeration of the sub-divisions and detailing their characteristics (lakkhana), present sense (paccupatthana) and basic meaning (padatthana). Vasubandhu’s method is quite different. He gives a running exposition of the several items, and helps his reader by formulating the ideas. Over and above this, he cites, or refers to, the opinions of the schools or teachers who differ from him and refutes them to the best of his ability, by having recourse to logic and citations from original texts. We propose now to give a synopsis of the contents of each of the Koégasthaanas.

The first KogasthGna opens with a few verses, in which Yasomitra expresses his great regard and esteem for the author Vasubandhu, and Gunamati and his disciple Vasumitra who commented upon the Abhidharma texts. He states that Vasubandhu has given a gist (pratydsa) of the Abhidharma texts. Yagomitra has consulted the comments of Gunamati, Vasumitra and the authors of Abhidharma- vibhdsad, accepting them wherever he thought fit, otherwise he gave his own reasons for differing from them.

In the first stanza Vasubandhu saluted Buddha as the one who had destroyed ignorance completely and rescued beings from the mire of repeated existences (samsdra). Yasomitra in course of his comments on this stanza points out the distinction made by the Vinayavibhasa writers between Buddha and Bhagavan. Incidentally he remarks that ignorance or lack of knowledge of jfieya means non-comprehension of the characteristics of the twelve dyatanas, i.e., the six organs of sense and their objects. He then passes on to Buddha's extra- ordinary knowledge, which is beyond the scope of knowledge of the Arhats and Pratyekabuddhas (pp. 4-6). He comments on the word ‘samsdra’, the flow of which is dependent on trsnd and which is difficult to cross on account of one’s wrong views (drsti) and lack of faith (vicikitsad) in the Triratna. He contributes a long discussion on grammatical rules in connection with the word namah and explains the distinction between visesasamjfid and sdmGnyasamjna in commenting on the word Sastr.

The second Karika is devoted to the exposition of prajfid, the gulyect-matter of the Abhidharma texts. Prajfd, it says, 1s derived gradually by means of (i) Sruti (listening to discourzes), (ii) cinta (reasoning aad cogitation' and (iii) bhdvand (meditation); these three means are mupere (eaiswece) while praifid is pure (andsrava). Yasomitra now takes up the word ‘samucara." and discusses (p. 9) whether citta is an afttradent of pramed or cice cersa. He tries to explain how the Abbsdibarma texts can be described as sénucara, and incidentally refers to the opsnion of some who state that of the seven Abhidharmas the uameprasthana is the chief or the body of the text, which has for its sapvort sx supplements. viz., Prakaranapada, Vijrianakaya, Dharma- skendhea, Prameptisastra, Dhatukaya and Sangitiparyaya. The Abbedharma texts, it adds, have two meanings (i) pdramdrthika (the baghest, real) and samketika (the popular, unreal), and the present text Kode is a gist of the Abhidharma.

In the third Karika is explained the utility of the Abhidharma texts or the Kosa. It is shown that asa man’s life is dependent on food, so a person's upagama (quietude) or escape from existence is dependent on the dharmapravicaya or analysis of dharmas like rapa, vedand, anitya, duhkha, etc., and that dharmapravicaya is the function of the Abhidharma texts.

Now, the question is raised about the authorship of the Abhidharma texts. The Abhidharmikas hold that Buddha is the author of the texts while the Sautrantikas! do not admit it, saying that authorship of the texts is attributed to certain writers.

Jnanaprasthana to Katyayaniputra, Prakaranapadda to Vasumitra, Vijaadnakdya to Devagarman, Dhatukdaya to Pirna, and Sangitiparyaya to Mahakausthila. The Vaibhasikas contend that just as Dharmatrata collected the Udanavarga from the Uddnas uttered by Buddha, so Katyayaniputra and others made their collections from expository sayings of the Teacher. The Sautrantikas are then asked how they admit three pitakas if there were no Abhidharma. In reply they state that the Abhidharma exists as a part of the Sitrapitaka and that certain sutras which discuss the characteristics of dharmas are distinguished as abhidharma, rather Prakirna.

Karika 4 makes a general statement about the various dharmas which constitute the subject-matter of the Abhidharma. The dharmas are broadly dichotomised into impure (Gsrava) and pure (andsrava), material (ripi) and non-material (ariipi), noticeable (sanidargana) and non-noticeable (anidargana), contaminating (samklegika) and purifying (vyavaddnika) and so forth. The four truths and the eightfold path, being vyavadGnika, are regarded as pure dharmas though they are constituted (samskrta). Yagomitra discusses why these should be distinguished as andsrava. He lays stress on the word ‘anuSerate’ by which he means that only those dharmas which serve as the basis or support for appearance of impurities are only sdsrava and the rest are anasrava. In the latter category are included the truths and the path.

The previous karika has in view only the constituted (samskrta) objects while the fifth karika takes up for consideration the asamskrta dharmas. Unconstituted dharmas are in all three in number, viz., Gkasa and two kinds of nirodhas. The Vatsiputriyas admit only one asamskrta viz., Nirvana while the Vaisesikas admit several, viz., the atoms (paramdnu) etc. There are some who regard akaéa as absence of any obstruction and not as something that exists but the Vaibhasikas on the basis of a discourse of Buddha admit the positive existence of akaga (asty Gkagam iti Vaibhasikah), but it is neither fixed (apratisthita) nor supported (andlambana).