^{page 77 note 1} The substance of this study was read as a paper before the Seventeenth Congress of Orientalists at Oxford unfler the title “Some Buddhist writers and the Kauṭilīya Arthaśāstra.” Of the abbreviations B. stands for Buddhacarita, and S. for Saundarananda. In quoting the Arthaśāstra, I give the sentence numbers of Jolly and Schmidt's edition as the most convenient form of reference.

^{page 77 note 2} So long as it was held that the work translated from the Chinese by E. Huber under the title of Sūtrālaṁkāra was by Aśvaghoṣa, it was difficult to escape the conclusion that he was at least some what later than Kaniṣka. Professor L. de la Vallée Poussin seems to hold that the evidence of the MS. fragments published by Professor Luders giving the name of the book as Kalpanāmaṇḍitikā and of the author as Kumāralātadoes not dispose of the previous view (Vijñaptimātrātasiddhi. La siddhi de Hiuan Tsang, Paris, 1928, pp. 223–4)Google Scholar. Enough, however, is extant of the Sanskrit text to show that the style is devoid of the characteristics that distinguish all Asvaghosa's writings, and the references to Nanda's not having obtained arhatship and to six abhijñās (whereas Aśvaghoṣa only knows five, S., xvi, 1) cannot possibly emanate from the author of the Saundarananda.

^{page 78 note 1} I use arthaśāstrafor the teaching of the school generally and Arthaśāstra for Kautilya's work.

^{page 78 note 2} The one disciplined is himself (xii, 2599), the seven protected the seven constituents of a kingdom (xii, 2659–60), the seven abandoned the seven vices of kings (v, 1061–2), the five observed the five measures (xii, 2156), the three obtained dharma, artha, and kāma (xii, 2150), the three understood sthāna, vṛddhiand kṣaya (xii, 2152 and 2665), the two known are probably the frequently mentioned pair, nayaand apanaya or anaya, and the two abandoned kāma and krodha (xii, 2721 and v, 1160).

^{page 79 note 1} For further remarks on Kautilya's point of view, see p. 89 below.

^{page 79 note 2} Cf. its use as an adjective in MBh., xii, 3567, a passage which explicitly declares that conquest by means of adharmais not permissible but isproper when effected by means of dharma.

^{page 80 note 1} Cowell's MSS. omitted 11 verses in canto ix after verse 41, according to the old MS. in Nepal and the Tibetan translation but, as one of these verses is clearly an interpolation, to obtain the correct numbering of the subsequent verses Cowell's numbers should be increased by ten only.

^{page 80 note 2} Cf. also Jātahamālā, p. 67, 11. 23–4, ānatasarvasāmantāṁ … prthivīṁ. In S., ii, 45, Professor Thomas suggests in a private communication the reading aśakyaḥ śaleya”, which is probably the correct reading and strengthens, if anything, the parallel drawn above.

^{page 81 note 1} According to T. Ganapati Sastri rātrisattracarāḥ, which is perhaps preferable.

^{page 82 note 1} Speyer's, “though knowing that the science of politics follows the path of Righteousness (dharma) only so far as it may agree with material interest (artha)”, does not seem quite to hit off the sense.

^{page 82 note 2} R. Fick, Festgabe Jacobi, pp. 145–159, holds that the Pali version is later than the Jātakamālā; but the evidence seems to me insufficient to justify a definite conclusion.

^{page 84 note 1} There is a double meaning in kṛtāśrayeṣsu and kṛtye. Properly speaking men are kṣtāśraya by having a king as their refuge or support and the use of āśraya in this connexion seems even to be extended in B., xiii, 71, to the meaning “leader”. Its opposite use here is meant to emphasize the contrast between the dharma&āstraand the arihaśāstra. The correspondence of kṛtye and yajńe hints that yajńa is really nothing more than kṛtyā, “magic”. The late Professor Gawroński's conjecture of hi for tu in the last pāda spoils the point of the verse.

^{page 85 note 1} The exact meaning of kṣatravidyā in Chāndogya Up., vii, 1, 2, is uncertain; in Dīgha Nikāya, vol. i, p. 9, 1. 7, khattavijjā is classed among the occupations a Brahmin or śramṇa cannot properly follow but this does not necessarily prove that the reference is to the arihaśāstra, for any of the functions of government are improper for those who lead a saintly life.

^{page 87 note 1} The context shows that this is the correct interpretation of verses 797–800, and that Hauer, J. W. (Das Lańkāvatāra-sūtra u. das Sāṁkhya, Stuttgart, 1927)Google Scholar is in error in taking them to give the name and parentage of the author of the appendix.

^{page 87 note 2} Mrvan Manen, J. says of it in the Foreword to the second edition of the Cāṇakyarajanīśāstram (Calcutta Oriental Series, No. 2, 1926), p. xiiiGoogle Scholar, that it has certain verses which are contained in nearly all Cāṇakya collections and are nowhere else attributed to another. I have made a cursory examination of it in the British Museum copy (fol. 194b–200a), this volume being missing in the India Office set. It is divided into seven cantos, containing some 129 or 130 verses; the exact number is uncertain, as sometimes five or six lines are used to translate a single verse, and all the verses must be identified to attain certainty. The first verse gives Masuraksā's name, and mentions the arthaśāstraas one of his sources, an unusual feature in these collections. The remaining ten verses in this canto contain general rules for the conduct of life, of which I have not identified any. In canto ii verses 8–15 consist of the well-known series beginning siṁhād ekāṁ, describing the twenty qualities of animals which should be imitated. Canto vii describes in 19 verses the qualities of a king and his various servants; they seem superior in quality to the similar verses in Haeberlin's Cāṇakyaśatam and in the above mentioned, the Bhojaraja, recension, while the Vrddhacāṇakya (Bombay, 1852)Google Scholar has not the series at all. I have identified a third of the remainder, almost all in the Bhojarāja recension, though some occur in the other two also. A more prolonged search would probably result in the identification of many of the rest. The text of the verses seems generally good; very few deal with the faults of women, most treating of the behaviour to be adopted towards relations, friends, foes, evil men and servants. It seems to have more unity than the Canakya collections generally.

^{page 88 note 1} The only similar name I can find is Suraksa, the name of Vyāsa in the fourteenth age in Vāywpurāṇa (Ānandāśrama S.S.), xxiii, 162. The verse looks corrupt, having no less than three conjunctions where only one is required, and the name may therefore have suffered mutilation by the loss of a syllable.

^{page 88 note 2} Namely in Bhaṭṭotpala's references in his commentary on the Bṛhajjātaka (quoted in the preface to the first edition of Shamasastry's translation), which suggest that according to the authorities he followed Visnugupta and Cāṇakya were considered two separate persons and that he identified them in accordance with the traditions current in his day. That he meant the two persons under discussion here can hardly be doubted; for there surely cannot be another pair of the same names who had also been confused.

^{page 90 note 1} A recent work dealing with some of the passages discussed here is Breloer, B., Kauṭaliya-Studien, i. Das Grundeigentum in Indien, Bonn, 1927Google Scholar. I find myself unable to accept the theories set out in it and disagree entirely with several of the proposed translations and the deductions drawn from them.

^{page 91 note 1} Jolly and Schmidt print sentence 20 as a whole, omitting the avagraha in °bhagikab.I follow Shamasastry in dividing it into two parts for convenience's sake; it would really be better to take the last two words as a separate sentence too.

^{page 91 note 2} For vāpa “sowing”, “area sown”, cf. Ind. Ant., xv, p. 340, 1. 46, where a field is described as vrīhidvipīṭhakavāpa “having an area which requires two pithakas of seed to sow it with”.

^{page 93 note 1} A description of it, which has been much criticized of late years by one school of officers as idealized, will be found in Grierson, G. A., Notes on the District of Gaya, Bengal Secretariat Book Depot, Calcutta, 1893Google Scholar. The latest and most authoritative study from the economic point of view is in Tanner, E. L., Final Report on the Survey and Settlement Operations in the District of Gaya, Patna, 1919Google Scholar; for further details see the similar reports on Fatna and South Monghyr.

^{page 94 note 1} Cf. aṭhābhagiya in the Aśoka edicts.

^{page 96 note 1} Blochmann and Jarrett, Ain, II, p. 44.

^{page 97 note 1} P. 92, n. 1. Those most qualified to judge believe that in practice under the produce rent system, while the landlord's demands work out at between 40 and 45 per cent, of the crop, the actual collections amount to only 25 per cent, or even less on an average.

^{page 98 note 1} For an instance see Grierson's, Notes on the District of Gaya, pp. 75–6Google Scholar.

^{page 98 note 2} Thus in the Final Report on the Survey and Settlement Operations in the District of Eanchi (Bengal Secretariat Book Depôt, Calcutta, 1912), p. 79Google Scholar, § 188, Mr. Reid, speaking of the khuntkhatti tenures of the Mundaa, says, “The rents payable by the owners of the intact khuntkhatti villages really represent the small tribute which the Mundas or their descendants agreed to pay as a subsidy for the support of their feudal chief.”

^{page 98 note 3} Cf. Moreland, W. H., India at the Death of Akbar, p. 97Google Scholar. Naturally no coherent theory of the various rights of property in land could grow up in such soil. The growth of the king's position seems to me also illustrated by the difference in wording between Manu, viii, 39 (on treasure trove) ardhabhāg rakṣaṇād rājā bhūmer adhipatir hi saḥ, i.e. “overlord of the soil” as its guardian, and the couplet quoted by the commentator on K. A., ii, 24, which is evidently much later in date, raja, bhumeh patir drstah Sastrajnair udakasya ca, though even in this patiis not the same as “owner” and leaves room for other rights in land and water.

^{page 99 note 1} Grierson, Bihar Peasant Life,§ 62, records kiyārī as used for “beds formed in a field for irrigation”, but gives a nearly similar use to the above, which was familiar to me in practice, in Notes on the District of Gaya, p. 53.

^{page 100 note 1} It certainly does not, as has been suggested, include sugarcane, which is propagated by cuttings, not by planting roots, and which in any case is already included under the head kedāra.

^{page 101 note 1} For an admirable description of these embankments, see Grierson, Notes on the District of Oaya, p. 54.

^{page 101 note 2} Provided my explanation of taāṭkais accepted; otherwise the author's ignorance of the āharsystem would seem to imply ignorance of the conditions of Magadha, as the āharcan hardly be a recent invention.

^{page 102 note 1} The dictionaries recognize various forms, Forbes baipārī, byopārī, and beopārī, Fallon byopārī and baipārī, and Ram Lai byaupārī; it is also confused with formations from Sanskrit vyavahāra. I give it here in the form familiar to me in practice.