1 CASTE IN THE PANJAB. 19
2 Agarwal Banyas of Agroha and the Oswal Banyas of Osia. But the present
3 distinction between these sections is as a rule based upon difference of social and
4 religious x;ustoms. It is not unnatural that, in the course of ages, the strictness
5 wfth which the artificial restrictions which regulate social and caste matters
6 are observed should vary in different parts of the country j and it is no less
7 natural that, where the two standards come into contact, those whose standard
8 is the stricter should look down upon those whose practice is more lax. The
9 Gam- Brahman sees with hoiTor his Sarsiit brother eat bread from the hands of
10 other than Brahmans, and do a thousand things which to him would be pollution.
11 The result is that the Gaur refuses to eat or intermarry with the Sai-sut, and
12 that for all practical pui-poses the sections are not one but two castes ; far
13 more so indeed than, for instance, the Jat and the Giijai'. Nor does it seem
14 to me impossible that these sections may in sMiie cases represent real diversity
15 j o£ race or" origin ; that the Gaurs may have been the Brahmans of Gam-
16 and the Sarsuts the Brahmans of the Panjab, both called Brahmans because
17 they were priests, but having nothing else in common. Again, among some
18 of the Panjab trading castes great sections have been fixed within recent times,
19 which are based not upon geographical distribution, but upon voluntary diver-
20 gence of social custom. Such are the great Dhaighar, Charzati and other
21 sections of the Khatris described under that caste heading. Thi-oughout all
22 these great sections, whether geographical or social, the same tribal divisions are
23 commonly found unchanged. The tribes or clans of the Gaur and Sarsut
24 Brahmans, of the Uttaradhi aud Dakhani Ai'oras, of the Agai'wal and Oswal
25 Banya are in great part identical. Now where these divisions are really tribal,
26 and based upon common descent, this must mean that the tribal divisions
27 preceded the divergence of custom which resulted in the formation of what
28 I have here called sections, and that the original stock was one and the same.
29 But where, as is often the case, they are mere Brahmnical gotras, I do not
30 thirdi that this necessarily follows.*
31 352. Tribal dmsions among artisan and menial castes.— Among the
32 artisan aud menial castes we find precisely the same great sections, based either
33 upon differences of custom which in turn depend upon geographical distribu-
34 tion or, I believe in very many cases indeed, upon difference of origin, one
35 section of an industrial caste being descended from Jats who have sunk in the
36 social scale, another perhaps from Ahirs, while a third is the original stock to
37 which the industry has been hereditary beyond the memory of the tribe. The
38 Chamar of the middle Satluj will not intermany with the Jatia Chamar of
39 the Dehll Tenitory because the latter works in the skins of impure animals ;
40 I the Snthiir carpenter from Sindh looks down upon and abstains fi'om marriage
41 with the Khiiti of the Malwa ; and so forth throughout the list. Among the
42 menial castes moreover, as among the priestly and mercantile, we have a double
43 classification ; and by the side of the great sections we find what correspond
44 with tribal divisions. But among the menial castes, or at least among those
45 who occuiDy the position of hereditary village servants, I believe that these
46 divisions often have their origin rather in allegiance to the tribal master than
47 in any theory of common descent. It has often been noticed that the menial
48 castes denote then- tribal sub-divisions by names famous in political history,
49 such as Bhatti, Khokhar or Chauhan ; and our present papers furnish abundant
50 instances. Now on the frontier a Lobar who is attached to the village of
51 the JNluhammadzai tribe will call himself Lobar Muhammadzai, while one
52 ' See further sectiou 353 on the next page.