________________
270
I'HE INDIAN ANTIQUARY
[ OCTOBER, 1923
greater familiarity between cross-cousins, and even that is restricted by the respect for age, which is such that * a man will address his servant as ayya (elder brother), if he is older."
Thanks to R20 Saheb S. Krishnaswami Aiyangar, I am able to produce more definite evidence from South India. I will quote his letter :-" Whether they actually marry or no, these cross-cousins usually enjoy that license, particularly as between men, to indulge in free talk, which between others would be regarded as insulting. As between these cousins there is infinitely more freedom of talk. This habit has even invaded the castes to whom marriage between cross-cousins is a prohibition, such as, for instance, the Brahmans. The habit is almost generalamong all classes other than that of the Brâhmans."
Another way of approaching the problem is by looking for divisions that fight one another. The only case I know of is the hostility between the right-hand and left-hand factions of Southern India, as described by Dubois in his Hindu Manners and Customs (Oxford ed., p. 25). The left-hand includes the Vaisyas, a high caste, and also the lowest of all. The right-hand consists of most of the higher Südras and of the Parias. Their disputes centre, it, should be noted, round religious ceremonies. It may be objected that these two groups do not intermarry and that there is no evidence that they ever did ; on the other hand there is no evidence that they did not. The rigidity of caste is admittedly not early. Even at the present day cases of intermarriage are not uncommon, and I need not dwell upon them beyond quoting Mr. H. Codrington's inforination as regards Ceylon:-" The castes used to intermarry, 1.c., a higher caste man took a wife from the caste next below. This is still done in parts of Ceylon by the Hali (Salagama. Tamil, Saliyar) and Vahunpurayo."
But whether caste ever intermarried or not, the Tamil and Sinhalese kinship system is there to prove that there must at one time have been in the South intermarrying groups like the Sakya and Kôli, for the Tamil system is based on the dual organization and is sufficient evidence of its former existence. If in Tamil land this system divided the clans into two intermarrying groups, we should get back to a state of society such as exists in Fiji. There each state is divided into two groups of clans : the nobles and their councillors or heralds are always in one, to the vanguard in the other; it can be shown that marriage into the other half was, until recently, the proper thing ; but the nobles have tended to form alliances with the nobles of foreign states and thus to become endogamous within their rank or caste; the car. penters are strictly endogamous because no one will marry into them, they are so despised.
The Todas, who have the cross-cousin system are divided into Tartharol and Teivaliol. These two divisions do not now intermarry, but the following custom is significant. When a girl reaches a certain age "a man belonging to the Tartharol, if the girl is a Teivali, and to Teivaliol, if she is a Tarthar, comes in the day-time to the village of the girl, and, lying down beside her puts his mantle over her so that it covers both and remains there for a few minutes. Fifteen days later she is cleflowered by a man of either division."'11 This looks very much like a survival of the time when a woman's proper husband came from the opposite division. She still, in the majority of cases, finds her official para mour in the opposite divi. sion.12 The Tôlàs therefore constitute the first link in the chain with which we want to connect the Tamil social organisation with the Fijian. Students of Indian society may well find soine more links among the backward tribes of India, for those who are out of the swim of civilization move more slowly and are often to be found now exactly where their neighbours stood thousands of years ago.
The use of the terms right and left as applied to social divisions, lends probability to my suggestion. Ainong the Elema of New Guinea the clubs are divided into right and left.
19 It is the posible that they have the same origin as the Koattriyas and Brāhmans of India I! Riscrs' Th Tnas, . 303.
12 Ibid., p. 526.