The question who the Pallavas were is one which can hardly be described as being out of the stage of discussion yet. The theory that held the field till recently almost unchallenged was that they were a tribe of foreigners supposed to be of Parthian origin who having effected a lodgment in the part of the country near the mouth of the Indus, moved southeastwards gradually till they came to be found in possession of the region dominated by Kanchi. The main reason for this contention is that a class of people called Palhavas figure among the lists of tribes on that frontier in the Ramayana, the Mahabharata and other such sources of information. They are also found to figure among the enemies overthrown by the Satavahanas, namely Gautamiputra Satakarni and his son. The Ceylon chonicle also mentions a tract of country which seems to be located in that region which is named in the Mahavamsa, Pallavabhogga. This collection of refererces to the Pallavas is held to justify the conclusion that they were a body of foreigners who entered India by way of Baluchistan, and moved on till they hinduised themselves so far as to forget their foreign origin and raise no suspicions among the peoples over whom they imposed their authority. There are, however, grave difficulties in the way of accepting this apparently satisfactory account as we have some information in Tamil literature which militates strongly against this view of their origin. We have already pointed out that in the days of the early Cholas, Kanchi was a Chola Viceroyalty, Palatine Viceroyalty though it was. We have already given a number of references to show that the Tamils regarded Pulikat as their northern boundary, and the people or the tribes that inhabited the region immediately north of it have invariably been referred to as speaking a language different from that of the Tamils. Those people are invariably referred to as Vadukar, which is the name by which the Telugus are ordinarily known in the Tamil country to-day. But in that early age the term Vadukar seems to have been invariably applied both to the Telegus and the Kannada people across the Tamil frontier.1

Even the Periyapuranam a work of the early 12th century observes this classification as it speaks of the Karnatakas as Vadukar.1 That designation is still preserved in the name of the Badagas of the Nilgiris. The region on the eastern side of this portion of the Peninsula occupied by this people is the region where we find the earliest memorials of Pallava rule. When the Pallavas emerge into the full light of history we find them in possession of Kanchi. Whether they were Tamils or Telugus they are people we find along the region between the lower course of the Krishna and the river Palar. To begin with, this region, at least the major part of it, was designated Tondamandalam in those days. In regard to their origin and their previous habitat we have already exhibited a certain number of references from the old classical collection Ahananuru referring to what actually constituted Tondanadu; both Kanchi and (Tirupati) were alike included in this territorial division Tondamandalam. We have also quoted an old passage, from Nachchinarkiniyar's commentary, by an author whose name is not quoted, giving the important equation that the people called Tondaiyar, people of Tondamandalam, were treated as the same as the Pallavas. During the period to which this reference must be held to relate the words Tondaiyar and Pallavas, were considered to be synonymous.2

1 For references see ch. 1 above.

1 See Murti Naynarpranam, stanza 11.

2 Nackhhinarkkiniyar's comment on Sutra 54, Poruladhikaram Tolkappiyam.

On this basis alone there is good reason for regarding the Tondaiyar as the name of the people living in the country who were subsequently called by a Sanskrit translation of the same. This inconvenient position is sought to be got round, by votaries of the foreign origin of the Pallavas, by bringing the Chola occupation of Kanchi and of the literature bearing on the period to a comparatively short period of interregnum that is supposed to have existed between one of the early dynasties of the Pallavas and the later great dynasty; in other words by bringing the Sangam age itself to the fifth century A.D. We have already demonstrated clearly that it would be impossible for a variety of reasons to move the period down by about four centuries in that arbitrary fashion. The question rests still upon the specific Gajabahu synchronism supported by so much of valuable historical evidence that it would require a very strong case on the other side to turn it upside down, not to speak of the insuperable difficulty in detail that would have to be confronted in any attempt at constructive criticism. Kanchi figures in this body of early literature as a viceroyalty of the Cholas and the only Tondaman that figures in the whole body of this literature as the ruler of this part of the country is the Tondaman Ilam-Tirayan of Kanchi who ruled over Conjeevaram not so much in his own right but by right of his

Chola ancestry. We shall come to this point a little later.

The Pallavas. Native to South India.

Among the large number of places in which the Pahlavas get mention in Sanskrit literature they are found mentioned with the well-known tribes of the north-western frontier such as the Sakas and the Yavanas. It would be difficult to find any clear reference to these anywhere in South India. There are a certain number of places in which the south Indian kingdoms are mentioned. We do not find anything corresponding to the Palhava state or tribe in the south. The Asoka edicts do not mention any. Even where the reference occurs in classical Sanskrit literature the Cholas, the Pandyas, the Keralas are referred to and where we should, from geographical position, expect the Pallavas a class of people by name Dravida is mentioned. Dravida or Darmida is generally taken to be the equivalent of the Tamil, Tamila in Tamilakam (Sans. Dramidaka), the whole of the Tamil country. It is also used in a somewhat narrower sense as indicating one of the four kingdoms, a kingdom that would correspond to, and that gets to be known to later history as, the Pallava kingdom with Kanchi for its centre. It would be rational therefore to regard the Pallavas native to South India, and as the people who were before then known by the name Tondaiyar more generally. There are some objections to be met before taking this particular position. By a careful study of the available Pallava records that epigraphy has unearthed, we are able to throw the early Pallavas into three groups. They are found first of all as tribal chieftains ruling their various little states, three or four of them could be specifically mentioned, in the region extending from the lower course of the Krishna to almost the Palar, Dasanapura, Palakkada, Menmattura and Kanchi. The records of some of these rulers happen to be in Prakrit and the others in Sanskrit; and they are found scattered across from the west coast to the frontiers of the Godaveri district in the east, the actual belt of country constituting the Vaduka frontier of the Tamils of the classical Tamil literature. Their association with Amaravati, the discovery of certain statues of Roman workmanship as it is thought and the statement that one of these early rulers attained to his royal dignity by marriage, are all brought into requisition to give them a marriage alliance with the Andhras, to give their art a Roman origin and to make these a foreign people who imposed themselves as rulers over the vast region extending almost from the Godaveri to at least Palar in the south. The matter requires therefore careful consideration. 18