1 20 PANJAB CASTES.
2 who lives in the service of the Daiilatkhel will call himself Lobar Daulatkhel
3 There can be no doubt that the connection between the village incuials and
4 the agricultural eoiumunlties whom they serve was in old times hereditary and
5 not voluntary, and that the former were in every sense of the word »(1^^>Y\
6 (ikha. In fact, as I shall presently explain in greater detail, we still hnd
7 the tribal organisation of the ten-itorial owners of a tract perpetuated in
8 ffi-cat integrity by the territorial organisation of the village menials, wliere
9 all but its memory has died out among their masters. It seems to me more
10 than probable that in old davs, when menials were bound more closely to the
11 tribes they served, the names of those tribes were used to distinguish the
12 several groups of menials ; and that for instance Chamars serving Bhattis
13 would be called Chamiir tribe Bhatti, and those serving Khokhars called
14 Cham^r tribe Khokhar. When the bonds grew less rigid and a change ot
15 masters became possible, the old name would be retained though the reason
16 for it had ceased to exist, and thus we should find Bhatti and Khokhar [p,
17 Chamars scattered throughout the Province. In fact the process would be
18 simply another instance of that substitution of the idea of subjection to a
19 common authoritv for that of common blood as the l^asis of tribal _ division,
20 regarding which I have already quoted Sir H. Maine's language in section
21 :349.
22 353. The Brahminical gotras. — I have said that among the priestly and
23 mercantile castes we tind a set of divisions corrcsiionding with the true
24 tribal divisions of the landowning classes, which runs through the great
25 geographical or social sections which I have deseribed above. These divisions
26 are, among the Khatris and Aroras, in all probability real tribes denoting
27 common descent, or at any rate special association of some sort, at an earlier
28 stage in the history of the caste, of the ancestors of all those who now
29 bear the same tribal iiauie. Among the Brahmans and Banyas these divisions
30 are known as gotras, and it is not so certain that their origin, among the
31 Banyas at least, is tribal. The word gotra, more commonly known under
32 the c©rrupted form of got, means a family or lineage, the descendants from
33 a common ancestor, and it also means a flock, those who shelter within a
34 common fold. The Brahmans say that their gotras are named after the
35 great Hindu Rishis, though it does not clearly appear whether the members
36 of each gotra claim descent from the Rishi whose name it bears as from a
37 carnal or as from a spiritual father. It is curious 1 hat the names of many
38 of the founders of these gotras occur among the ancient genealogies of the
39 prehistoric Rajput dynasties, the Rajas in question being not merely name-
40 sakes of, but distinctly stated to be the actual founders of the gotra ; and
41 it would be strange if inquiry were to show that the priestly classes, like
42 the menials just discussed, own their tribal divisions to the great families
43 to whom their ancestors were attached.' At any rate, whatever their origin,
44 the Brahminical gotras have among the Brahmans become absolutely heredi-
45 tary ; and every Brahman, whether Gaur, Siirsut, Diikaut, or otherwise belongs
46 to some one or other of these gotras. Thus, taking these great sections as
47 tribes, the gotra is wider than the tribe ; and while new tribes and clans can
48 be and are constantly being formed, no new go'ra is possible.^
49 » For a curious instauee of classification of BrahinaiiB into tribes by the command of a Rajput
50 ruler, see the qnotation from Mr. Bnrne* given on page 179. [Census Report.]
51 » Is it possible that the jo<ra is a relio of descent through the female liue, like the corre-
52 sponding phienomenon among the Australian and North American Indians ? [Census Report.]