1 18 PANJAB CASTES.
2 tribes, who were orginally five. Of these two, the Rind and Lashari, rose
3 to prominence and divided the nation into two eoiresponding sections. As
4 time went on the nation broke up into a number of independent tribes, each
5 with a separate ttrritoiy and organisation of its own ; and now, though every
6 Biloch i-efers himself to either Rind or Lashari stock, the names are liut a
7 tradition of origin, and in the Panjab at least no Riud or Lashari tribe can
8 be said to exist as such. The groups of tribes found indifferent parts of the
9 Province who claim comtnon descent from some one of the great Rajput races,
10 the Bhatti, Chauhan, Pnnwar, and the like, are instances of the same process.
11 The local tribes are now independent units, and can hardly be included under
12 the original tribal name save as a symbol of origin. Thus the line of demar-
13 cation between tribe and clan is no better defined than is that between caste
14 and tribe. As soon as a section of a caste abandons the customs of the parent
15 stock, whether as regards hereditary occupation or social habits, it tends to
16 become a new caste. As soon as a clan separates itself from the territory and
17 organisation of the parent tribe, it tends to become a new tribe. Whei-e
18 the Indian tribal and easte restrictions upon intermarriage are still oT)served,
19 the best definition would probably be obtained by taking endogamy and
20 exogamy as the differentia3 of the caste and tribe respectively ; a caste being
21 the smallest gi-oup outside which, and a tribe the largest group within which
22 marriage is forbidden. But in a great part of the Panjab this test does not
23 apply.
24 351. Tribal divisions among tlie priestly and mercantile castes.— In the
25 case of the castes or classes who, not being essentially landowners, possess no
26 political or temtorial organisation, the basis of tribal division is very different.
27 Here we have no compact tribes based upon real or fictitious communitv of
28 blood and occupying tribal territories. The Brahman has almost invariably
29 accompanied his clients in their migrations ; and indeed it will sometimes be
30 found that the Brahmans of a tribe or of a group of village communities,
31 being too small in number to lie independent, have kept up the connection with
32 their place of origin long after it has fallen into neglect or even olilivion
33 among the landowning communities with whom they dwell. Thus we find
34 Brahmans of different gotras or clans scattered haphazard over the country
35 without any sort of tribal localization, and the same is true of the mercantile
36 classes also. In both cases the divisions are wholly based upon real or
37 imaginary common descent. The gotras of the Brahmans, the clans of the
38 Khatris and Aroras are innumerable ; but they are not localised, and are there-
39 fore probably more permanent than are the territorial tribes of the landowners.
40 This al sence of tribal organisation is perhaps one of the reasons why, of all
41 classes of the community, the Brahmans and traders observe most strictly the
42 ai-tificial mles which preserve the integrity of caste organisation. How far
43 the Brahminical fiotra is really tribal is a distinct question to which I shall
44 presently return.
45 But in the case of both the priestly and the mercantile classes, we find
46 that their castes are broken up into sections, too large and too devoid of
47 cohesion to be called tribes, and approaching much iriore nearly to separate
48 castes, both in the actual effect of the divisions upon social intercourse and
49 interman-iage, and probably also iu their origin. These divisions are generallv
50 known by geographical designations, such as the Gaur Brahmans of the
51 ancient Gaur and the Sarsiif Brahmans of the Saniswati and the Panjab, the
52 Uttar^dhi Aroras of the north and the Dakhani Aroras of the south, the