§1. The so-called “mixed castes” in classical India have been studied in the past. The most comprehensive collection of data appears in P.V. Kane's History of Dharmaśāstra (1930–62: 2.69–104). In this paper I want to return to the original texts on which these data are based, schematize some of the materials, and try to draw certain conclusions. The texts are the Dharmasūtras, the early Dharmaśāstras, and Kauṭilya's Arthaśāstra, as follows:

– GDh, 4.16–21 (Stenzler's edition, Bühler's translation) = 4.14–17 (Ānandāśrama edition);

– BDh 1.16.6–8 + 1.17.3–8 (Hultzsch's edition, Bühler's translation) = 1.16.7–8 + 1.17.2–8 (Kashi edition);

– VaDh 18.1–6, 8–9;

– ViDh 16.2–6;

– MDh 10.6, 8–9, 11–12, 16–17;

– YDh 1.91–95;

– AŚ 3.7.20–30.

It should be noted that one Dharmasūtra, that of Āpastamba, does not discuss the matter of mixed castes.

§2. On the other hand, Gautama gives two sets. A first set is followed by a second one, introduced by: “Some say…” The interesting point is that not only the content but also the way of presentation in both sets are totally different; I shall return to this point later (§29).

§3. Baudhāyana also introduces mixed castes twice, one after the other. The situation is far more confused in this treatise, and considerable text-critical research will be required to solve the problem. The first passage follows immediately after the rules on the number of wives members of the four varṇas are allowed to have: from four to one.