^{page 18 note 1} The student of the śāstra is able to take advantage of the modern trend of social anthropology in India away from the exclusive study of tribes and towards a broader survey of the population.

^{page 18 note 2} Section 14 of Act XXX of 1956. The “women's estate” was a topic of law of such magnitude that it occupied sixty pages in the current edition of Mayne on Hindu Law and Usage.

^{page 18 note 3} The locus classicus on the subject is Jaimini, , Mīmārṃsā-sūtra, vi, 1, 10–14Google Scholar. The text with Śahara's commentary is conveniently printed in Joshī, Lakshmaṇ Śāstrī, Dharma-kośa, Vyavahāra-kāṇḍa, Wai, 1938, 1424Google Scholar. Reference to this most useful modern digest will be made below as Dh.k.

^{page 19 note 1} Manu, ix, 194; Dh.k., 1431b.

^{page 19 note 2} Yājñavalkya, ii, 147;Devala, , Dh.k., 1461Google Scholar.

^{page 19 note 3} Yājñavalkya, ii, 136; Viṣṇu, xvii, 4; Dh.k., 1470a; Vyāsa, , Dh.k., 1524 aGoogle Scholar.

^{page 19 note 4} Aparārka's commentary on Yājñ., ii, 135–6.

^{page 19 note 5} Ibid. This is not quite so foolish as it appears. Public examinations of childless widows were held in southern India according to the testimony of Medhātithi, , Manu-bhāṣya, ed. Jhā, Gangānāth, ii, 73Google Scholar; trans, idem, iv, pt. 1, 10.

^{page 20 note 1} Viśvarūpa (on Yājñ, ii, 139, p. 241) does not allow a non-pregnant widow to inherit, but if a daughter is born she may hold the property in trust for daughter. That niyoga alone qualified the non-pregnant widow was the view of King Bhoja, alias Dhāreśvara, whose opinion is immortalized in the Mitākṣarā (Colebrooke's trans., II, i, 8).

^{page 20 note 2} An artifice of Devaṇṇa-bhaṭṭa, in the Smṛti-candrikā (pp. 290–1)Google Scholar. As an Āndhra or Tamilian he knew perfectly well that most marriages in his part of India were celebrated in the Āsura, an unapproved form.

^{page 20 note 3} The view of a certain Udayakara commenting upon Manu, , and referred in the Vivāda-ratnākara (p. 590)Google Scholar and the Vivāda-cintāmaṇi (p. 237; Dh.k., 151b). The reference in the latter text is obscured by the translation of Gangānāth Jhā, Baroda, 1942, at p. 267, whereas the old translation of P. C. Tagore (Madras, 1865, p. 290) faithfully preserves it. Steele reports a custom in the Deccan tending to confirm such a practice, and there is an inscription in Mysore, one amongst many satī-stones in the north-west of the State, relating that despite the repeated deprecations of her relations the lady commemorated burnt herself on her husband's pyre, having first distributed all the property (evidently her husband's estate).

^{page 20 note 4} Sarasvatī-vilāsa, sections 512–5 (Foulkes' edn.).

^{page 20 note 5} This was the view of Śrīkara and others, refuted in the Mitākṣarā, II, i, 31.

^{page 20 note 6} II, i, 39.

^{page 20 note 7} The Hindu Women's Rights to Property Act, 1937, repealed by Act XXX of 1956.

^{page 21 note 1} Trivandrum edition, ii, p. 14; Dh.k., 1430a. MM. Gaṇapati Śāstrī understood the word para-dvisāhasrā as kārṣāpaṇa-sahasra-dvaya-paramāvadhiḥ, that is to say, “having as its upper limit 2,000 kārṣāpaṇas.” As we shall see the insertion of the coin chosen is wrong, although the Smṛti-candrikā and Nṛsiṃha-prasāda provide precedents (Dh.k., 1454b), but the interpretation of the compound is correct. Dr. Shāmashāstry in his trans., Mysore, 1929, p. 172, wrongly says, “above two thousand.” Since the king is advised as to what amount ought to be given to women the limit must be a maximum and not a minimum, as the reference to ornaments shows.

^{page 21 note 2} Kane's edn., 902; Dh.k., 1454b. The Smṛti-candrikā reads dvi-sāhasraṃ, implying that the full 2,000 ought to be given, and not more.

^{page 21 note 3} As Colebrooke notes on Jīmūtavāhana, , Dāyabhāga, trans., IV, i, 10Google Scholar, this text is variously read. See below re Lakṣmīdhara's reading, and the citations given in Dh.k., 1460a. dvi-śāhasra-paro could mean, “more than 2,000,” and doubtless was adopted by many jurists with that object in mind.

^{page 22 note 1} The connection of this text with Vyāsa's is evident, but as yet unexplained. As Colebrooke notes, ubi cit., many manuscripts read tri-, that is to say, “3,000.” The Dh.k., 1429b, prints the text so (xiii, 47, 23), but when the BORI Critical Edition of the Anuśāsana-parva appears we may confidently expect to find the reading dvi- well represented.

^{page 22 note 2} Kṛtya-kalpataru, Vyavahāra-kāṇḍa, Baroda, 1953, 684Google Scholar.

^{page 22 note 3} At p. 450.

^{page 22 note 4} At p. 281.

^{page 22 note 5} At III, 548–9.

^{page 22 note 6} At p. 510.

^{page 22 note 7} At p. 82; Dh.k., 1521a.

^{page 22 note 8} At p. 377; section 523.

^{page 22 note 9} At p. 376.

^{page 22 note 10} At p. 154 of Kane's edn.

^{page 22 note 11} Dāyabhāga, text (Calcutta, 1930), 117, trans., IV, i, 10Google Scholar.

^{page 23 note 1} On Yājñ, ii, 143; Dh.k., 1460.

^{page 23 note 2} At pp. 439–40.

^{page 23 note 3} A digest made in the time of Rāja Śarabhojī of Tanjore (circa a.d. 1820).

^{page 23 note 4} At p. 237 of the P. W. Saras. Bhav. Series edn. of 1934.

^{page 23 note 5} At p. 544.

^{page 23 note 6} At p. 731 of the Chowkāmbā Sanskrit Series edn. of 1914.

^{page 23 note 7} A Digest of Hindu Law … translated by H. T. Colebrooke …, London, 1801, iii, 583Google Scholar; Madras, 1865, ii, 600–1 = V, ix, text 482.

^{page 23 note 8} For example Mādhava, Madanasiṃha, and Nīlakaṇṭha.

^{page 23 note 9} At p. 293. For the text see Dh.k., 1402a, where the variant reading, “34,” shown. Devaṇṇa-bhaṭṭa says that paṇa means kārṣāpaṇa; that may have satisfied his contemporaries, but to us it is merely misleading.

^{page 24 note 1} Barnett, L. D., Antiquities of India, 1913, 208Google Scholar, gives several equations showing the weight of an āḍhaka, but certainty has yet to be achieved, due to the variety of standards. Maity, S. K., The economic life of Northern India in the Gupta Period, Calcutta, 1957, at p. 39Google Scholar deals carefully with the āḍhaka and finds it equivalent to about 16 seers (of rice). Manu (vii, 126 recommends or prescribes just twice this allowance of grain to a working man.

^{page 24 note 2} Section 527; Dh.k., 1428a. One must record, in all fairness, that the Sarasvatīvilāsa cites Viṣṇu and indeed other smṛti-writers quite frequently in cases where no other writer cites them and for opinions which do not always harmonize readily with the views of the same authorities in their recognized and frequently-cited texts. This has led to a suspicion that the compiler manufactured texts and attributed them to known authors of smṛtis. Puzzling as the problem is, it is unlikely that this is the solution, since (i) such frauds were relatively easily detected and detection meant the literary death of the work, which was already dangerously long for its hopes of survival; (ii) the names of smṛti-writers whose works did not then survive in entirety would have been chosen; and (iii) the forger would not have fabricated texts containing rules the meaning of which was ambiguous and required explanation and, as happens in several instances, the comparison of the views of other commentators thereon. It seems that better, if less simple, explanations must be found. See “Kuttā: A class of land-tenures in South India,” BSOAS., xxi, 1958, 61Google Scholar, and seqq. at 69.

^{page 25 note 1} Barnett, op. cit., 207: so Nārada-smṛti, xxi, 57, as published by Jolly. However, as Dh.k. shows, 532a, the majority of the sources read ṣoḍaśaiva, “16.” Viṣṇu's own description of a kārṣāpaṇa is not particularly helpful, but he understood it to be a copper coin: iv, 13: Dh.k., 526. The reading “20”, if it is genuine, is evidently very much older than the proportion 16:1 which appears to be almost universal in later text-books. But the equation 20 māṣas = 1 paṇa which is found in Nārada and other smṛtis (Dh.k., 532–3) would seem to throw additional doubt upon this otherwise helpful reading; particularly since, to make confusion worse confounded, Kātyāyana says that a kārṣāpaṇa contains 20 māṣas (Dh.k., 533—note variant readings). However, if we are to assume, as Kātyāyana's words very strongly suggest, that he understands māṣa here not as a subdivision of the paṇa, but a variant of it (he says the kākiṇi is a fourth part of a māṣa and of a paṇa), his text is even more helpful to us than Nārada's, and he appears to be supported by a text of Uśanas cited by Haradatta on Gautama (see Kane's edn. of Kātyāyana, 493, p. 213, and note thereon). Maity, op. cit., 171, accepts the equation of 20 māṣa/paṇa = 1 kārṣāpaṇa, and on p. 172 comments, “It . . . seems that the Smṛiti writers are not thinking in terms of coinage, but rather of goldsmith's weights.” He is relating the texts to Gupta coins.