two works has been inverted. Nârada's Vyavahâramâtrikâ shows a far more advanced development of the judicial theories than Bhrigu's Samhitâ, and contains matter which conclusively proves that it cannot date from an earlier time than the fourth or fifth century of our era1. As this test fails, Nârada's statement cannot be used for the determination of the order in which the various versions of Manu's laws were composed. It becomes more probable that it has been framed, with a view to enhance the importance of the Nârada-smriti, on the model of such purely fictitious stories respecting the origin of the Dharma, as that given in the Mahâbhârata XII, 59, 22, and 80 seqq., where we are told that Brahman, assisted by the gods, first produced a Dharmasâstra in 100,000 chapters, which was successively abridged by Samkara in 10,000, by Indra in 5,000, by Brihaspati in 3,000, and by Kâvya in 1,000 Adhyâyas. Against the genuineness of Nârada's story we may also adduce the Paurânik statement, according to which Manu's laws were remodelled first by Bhrigu, secondly by Nârada, thirdly by Brihaspati, and fourthly by Angiras 2.

The third argument, drawn from the fact that the medieval Nibandhas and commentaries quote passages from a Brihat (great) and Vriddha (old) Manu, has still less value. Professor Jolly has of late asserted in his able discussion 3 of the quotations from these works that they are later, not earlier, than the existing text of Manu, because some of their rules resemble the advanced teaching of Yâgñavalkya and Nârada, while others contradict our Manu on points where he holds archaic views. Morcover, a passage of Vriddha Manu, to which Professor Max Müller has first called attention, possibly indicates an acquaintance with Greek astrology. I can only agree with Professor Jolly's conclusions, and add that a comparison of the quotations from Brihat and Vriddha Manu with Bhrigu's Samhitâ produces the impression that both works-if indeed the titles refer 1 West and Bühler, Digest, pp. 48-50, third edition; Jolly, Tagore Lectures, p. 56: Preface to the Translation of Nârada, pp. xv-xvii.

2 Mandlik, the Vyavahâramayûkha and Yâgñ., p. xlvii; Jolly, Tagore Lectures, p. 44.

3 Tagore Lectures, pp. 65-66.

4

* India, what can it teach us? p. 366,

to two and not to one-must have been enlarged versions of the latter1. As it thus appears that there existed recensions posterior to our Manu-smriti, the existence of untraceable or partly traceable quotations from Manu's Dharmasâstra in Asvaghosha's Vagrasûki2 and from Manu in Varâhamihira's Brihat-samhitâ 3 possesses no great significance. With respect to Varâhamihira's reference, it must be noted that, according to Albîrûnî's Indica, two astrological Samhitâs, called after Manu, existed in the eleventh century A. D., the smaller of which was an abridgment made by a perfectly well-known human author1. Hence Varâhamihira may have taken his verses on the character of women from the latter. In both quotations the Slokas, not found in our Smriti, have a very modern. look. The case is, however, different with the quotations from Manu, which, as has been shown above, occur in the Mahabharata. We have been compelled to admit that the existing text of our Smriti is younger than the epic. If, therefore, the law-book referred to in the latter is not the ancient Dharma-sûtra, we must also concede the existence of a secondary recension which preceded Bhrigu's Samhitâ. The solution of this question is, owing to the

1 In order to enable the reader to form his own judgment on this point, I add a list of the quotations which I have noted. Those from Brihat Manu occur, 1. Col. Dig. II, 3, 26; 2. Col. Dig. V, 428 = Gî. Dây. XI, 6, 34; 3. Datt. Mîm. II, 8; 4. May. IV, 5, 53 ;—those from Vriddha Manu, 1. Col. Dig. III, 1, 69; 2. Col. Dig. III, 1, 83 = May. XI, 5 = Viv. Kînt. p. 99; 3. Col. Dig. III, 1, 86 = Viv. Kint. p. 89; 4. Col. Dig. III, 1, 90= May. XI, 5= Viv. Kint. p. 100; 5. Col. Dig. III, 1, 93= Viv. Kint. p. 103; 6. Col. Dig. V, 162 = Viv. K ́int. p. 272 = Varad. p. 21 = Gî. Dây. IX, 17 (where attrib. to Brihat M.); 7. Col. Dig. V, 408 = Smri. Kand. XI, 1, 15 = Sar. Vil. 504 = Varad. pp. 33, 40 = Vîram. III, 1, 2 = Gî. Dây. XI, 1, 7 and Viv. K'int. p. 289 (where attrib. to Brihat M.); 8. Mit. II, 5, 6 = Viv. K ́int. p. 289 and Varad. p. 37 (where attrib. to Brihat M.) = Sar. Vil. 591 (where attrib. to M.); 9. Viv. Kint. pp. 126-7; 10. Viv. Kint. p. 180; 11. Varad. p. 50; 12. Varad. p. 28, where in reality Manu IX, 206 seems to be quoted.

2 Weber, Indische Streifen, vol. i, pp. 190, 192, 198.

3

3 Kern, Brihat-samhitâ, chapter 74, vv. 7-15, and Preface, p. 43.

4 Albîrûnî, Indica, chapter xiv; see also Kern, loc. cit. p. 42, where the probability of the existence of a Mânavî Samhitâ has been shown. Albîrûnî says that the title of the two works was Manasa (Mânavî ?), and that the shorter one had been composed by one PNKL, a native of Southern India. I owe these notes to the kindness of Professor Sachau, the learned editor and translator of Albîrûni's important work.

shortness of the extracts, very difficult. But, considering all things, I feel inclined to assume that the author or authors of the Mahâbhârata knew only the Dharma-sûtra. The character of the four verses and a half, quoted verbally, as well as of the paraphrase in the Sakuntalopâkhyâna agrees well enough with this assumption, because the Mânava Dharma-sûtra, as we have seen, certainly did contain numerous Slokas. It is further corroborated by the fact that the Mahâbhârata does not differ in its arrangement, or rather in its want of an arrangement of the civil and criminal law, from the Dharma-sûtras. Though the epic contains numerous verses on these topics, it nowhere shows an acquaintance with the eighteen titles of the law which are so characteristic of the secondary Smritis, the handbooks of the special law schools. On the other side it may be urged that the Mahâbhârata says nothing of Dharmasûtras, and that its general view of the origin of the sacred law coincides with that expressed in the later law-books. It holds that the moral and legal doctrines were revealed for the benefit of the human race, first by Brahman to various mythical Rishis, and by them to mankind. This objection may, however, be met by the not unreasonable assumption that at the time when the Mahâbhârata was composed, the real origin of the old Sûtras had been forgotten, while the text had not yet been materially altered. What has been said above regarding the rise of the special law schools, and the facts known regarding the change in the tradition concerning the Sûtras of Gautama and Vasishtha, make the hypothesis of such a transitional period not at all improbable. Should, nevertheless, the possibility of the existence of a metrical redaction of the Mânava-sûtra, preceding that ascribed to Bhrigu, be considered as not altogether excluded, it would at least be necessary to concede that it could not have contained the present arrangement of the Vyavahâra portion under titles.

While there is thus no proof for the opinion that the modern portions of the Manu-smriti have been gradually added one to the other, or that the present text is one of

the last links in a long chain of metrical recensions, there are several points which tend to show that our Manu-samhitâ is one of the first attempts at remodelling a Dharma-sûtra. The most important argument for its comparatively early date is furnished by the incompleteness and awkwardness of its rules on judicial procedure and on civil law1. If we compare these rules with those of the Dharma-sûtras and with those of the other metrical Smritis, they are somewhat more explicit than the former, but very much inferior to the latter. As regards procedure, the Manu-smriti pays more attention to the moral side of the duties, incumbent on the judge and the other persons concerned, than to the technicalities, which are much more clearly and minutely described in the Dharmasâstras of Yâgйavalkya and Nârada. In this respect it comes close to the Dharmasûtras, with which it particularly agrees in the absence of all mention of written plaints and of documentary evidence, as well as in the shortness of its remarks on ordeals. Among the ancient law-books the Vâsishtha Dharmasâstra is the only one which has allusions to written documents, and names them, XVI, 10, 14-15, as one of the means of legal proof. In the other Dharma-sûtras there is no indication that their authors were acquainted with the art of writing. I have already pointed out in the Introduction to my translation of Vasishtha 2 that most probably this omission has to be explained not by the assumption that in the times of Gautama, Baudhâyana, and Âpastamba writing was unknown or little used in India, but by the consideration that the general character of the Dharma-sûtras, which principally pay attention to the moral side of the law, does not require the introduction of matters belonging more properly to the customs of the country or to the Arthasâstra. Whatever may be thought of the prevalence of writing during the earlier times and of the value of my explanation, it may be regarded as perfectly certain that

1 See on this subject and the following discussion, Weber, History of Indian Literature, pp. 279-281; Stenzler, Yâgňavalkya, pp. vii-x; Journal of the German Or. Soc. vol. ix, on the Indian Ordeals; and Jolly, Tagore Lectures, pp. 45-49. 2 Sacred Books of the East, vol. xvi, p. xxvi.

Bhrigu's Manu-samhitâ belongs to a period when the art of writing was known and generally practised. For, first, we find two clear references to written documents, 'what has been caused to be written by force' (lekhita, VIII, 168) and royal edicts (sâsana, IX, 232). Secondly, we have the expression nibandh, 'to record,' in a passage (VIII, 255) where the context leaves no doubt that a written entry is referred to. When it is said there, that in a boundarydispute the king shall record the boundary, according to the unanimous declaration of the witnesses, together with their names, it is impossible to imagine how he can do so without drawing up a written document, which, of a necessity, must have legal force for the future. This use of the verb nibandh makes it further probable that Medhâtithi is not altogether wrong, when he explains (VIII, 76) the compound anibaddha, a person not appointed (to be a witness to a transaction),' by 'a person not entered (as a witness in the document),' and refers the rule to cases of loans and other commercial transactions. Thirdly, there is the term karana (VIII, 54 and 154), which, though less explicit, likewise points to the use of written bonds for loans. The former passage declares that 'a debt which is proved by karana' (karanena vibhâvitam) must be paid, and the commentators explain karana to mean 'written bonds, witnesses, and so forth.' Hence it has been rendered in the translation by good evidence.' Verse 154 prescribes that a debtor 'who, unable to pay a debt (at the fixed time), wishes to make a new contract, may renew (lit. change) the karanam (karaṇam parivartayet 1).' Two commentators, Kullûka and Râghava, take the word here in the sense of a written bond,' while the older ones, Govinda, Nârâyana, and probably also Medhâtithi, explain it by 'bonds and so forth,' and make it include agreements before witnesses. From these explanations and the use of the word karana in other legal works it would appear that karana may also be cited as a witness for the acquaintance of our author with the art of writing. To the conclusion that writing must have been extensively used in businessNandana's reading kâranam is clearly erroneous; see below, p. cxxxiv.