to mark a conspicuous religious movement. Yet some authorities refer them to the third century and others to the eleventh, nor has any standard been formulated for distinguishing earlier and later varieties of Tamil.

I have already mentioned the view that the worship of śiva and the Linga is Dravidian in origin and borrowed by the Aryans. There is no proof that this worship had its first home in the south and spread northwards, for the Vedic and epic literature provides a sufficient pedigree for śiva. But this deity always collected round himself attributes and epithets which are not those of the Vedic gods but correspond with what we know of non-Aryan Indian mythology. It is possible that these un-Aryan cults attained in Dravidian lands fuller and more independent development than in the countries colonized by the Aryans, so that the portrait of śiva, especially as drawn by Tamil writers, does retain the features of some old Dravidian deity, a deity who dances, who sports among men and bewilders them by his puzzling disguises and transformations.¹ But it is not proved that śiva was the chief god of the early Tamils. An ancient poem, the Purra-Poruḷ Veṇbâ-Mâlai,² which contains hardly any allusions to him mentions as the principal objects of worship the goddess Koṭṭavai (Victorious) and her son Muruvan. Popular legends³ clearly indicate a former struggle between the old religion and Hinduism ending as usual in the recognition by the Brahmans of the ancient gods in a slightly modified form.

We have no records whatever of the introduction of Brahmanism into southern India but it may reasonably be supposed to have made its appearance there several centuries before our era, though in what form or with what strength we cannot say. Tradition credits Agastya and Paraśu-Râma with having established colonies of Brahmans in the south at undated but remote epochs. But whatever colonization occurred was not on a large scale, An inscription found in Mysore⁴ states that Mukkaṇṇa Kadamba (who probably lived in the third century A.D.) imported a number of Brahman families from the north, because he could find none in the south. Though this language may be exaggerated, it is evidence that Brahmans cannot have been numerous at that time and it is probable that Buddhism and Jainism were better represented. Three of Asoka's inscriptions have been found in Mysore and in his last edict describing

------------------------------------------------------------------------------------------------------------------

¹ The old folklore of Bengal gives a picture of śiva, the peasant's god, which is neither Vedic nor Dravidian See Dinesh Chandra Sen, Bengali Lang, and Lit, pp. 68 ff. and 239 ff.

²J.R.A.S. 1899, p. 242.

³ See some curious examples in Whitehead's Village Cods of South India.

⁴ Rice, Mysore and Corg from the Inscriptions, pp. 27 and 204.