Satya Sena. His son would be Rudrasena as well. Names ending in Sena are not unknown among the rulers of this dynasty. The date of this Satyasena would be sometime anterior to A.D. 388. If Skandasishya's date be 388, the three generations before him would have for them about 40 years if all three of them did rule. It is Skandasishya's father who according to the Velurpalaiyam plates "simultaneously with the daughter of the chief of serpents grasped also the complete insignia of royalty and became famous." Passing on now to the Velurpalaiyam plates themselves we are provided with the following succession of the early Pallavas up to Simhavishnu :

Along with these have to be taken the table provided by the Chura plates.

The two Ongodu plates give us two genealogies :-

I.

II.

These separate genealogies are obviously intimately connected with each other and have to be worked up into one table as many of the names are common and are apparently connected with each other. This is to a certain extent facilitated by the full list of Pallava succession given in the so-called Vayalur Pillar Inscription. Rao Bahadur Mr. Krishnasastrigal proposes to identify Kalabhartr with the Kanagopa of the Kasakudi plates, and also with Maharajah Kumara Vishnu of the Ongodu plates I. Similarly in respect of the second name Chutapallava which would mean a "tender twig of the mango," he would regard it as a surname of Skanda-varman I of the Uruvappalli grant, the Ongodu plates I, also giving the name Skandavarman. The names that follow do not differ. Virkurcha and Vravarman are not so different, nor Skandasishya and Skandavarman. He is led to this identification of the genealogy of the Ongodu plates with those of the Velurpalaiyam ones as he finds the palaeography of the Ongodu plates No. 1 older in point of character, and almost the earliest known record of the Pallava dynasty of the Sanskrit charters. The Maharaja Vijaya Skandavarman, the donor of the grant would be Skandavarman II, Skandasishya of the Velurpalaiyam plates. If this is agreed to, there is no difficulty in accepting this except for the first name Kumaravishnu which has no affinity with Kalabhartr or Kanagopa. One part of the genealogical tree gets then settled. The genealogy in the Ongodu plates No. II amounts to almost the same as the Mangalur plates giving the genealogy from Vlravarman to Simhavar-man II as in the table below. The Chura plates add to this and carry the genealogy to Maharaja Vijaya Vishnugopa Varman the son of Simha Varman II, the donor of the Ongodu grant No. II and Mangalur grant.

The point that the donor's grandfather Vishnugopa is given the title Maharaja in this may be overlooked as a similar discrepancy is noticeable between the Ongodu plates I and I1 in respect of Viravarm in, the first grant omitting the adjunct Maharaja. The fact that the first Ongodu grant was made from the victorious camp of Tambrapa is taken to warrant the conclusion that it was a subordinate family, by M. Jouveau-Dubrenil, and, taking advantage of the name Kumaravi-hnu, he would make the members of the Pallava dynasty whose names are found on this table another line of descendants of Kumara Vishnu I. That would make a difference of three generations between the Epi-graphist's estimate of time and the Professor's, both of them based on Palaeography and nothing else. Three names being in agreement we are rather inclined to accept the epigraphist's dictum on a questin of palaeography. We arrive then at a consolidated table of Pallavas somewhat as under.

Turning now to the Vayaluiv Pillar the names 31 to 36 are in the recognised order of the later dynasty. The name 30 is a Vishnugopa which may he the Vishnugopavarman of the Chura plates in which case we go on to the Simhavarman II in No. 29. No. 28, Simhavarman seems an additional name. 25, 26 and 27 may he the names Simhavarman I, Skandavar-inan III and Nandivarman I of the table. No. 24, Skanda-varman then would he Skanda-varman II on the table. Then comes in a Viravarman, No. 23, who may he the Virakurcha of the table. He is preceded by a Simhavarman No. 22 for whom and for three preceding names we can find no equivalent on the table. Then follows the names Skandavarman preceded by three names Skandavarman, Kumara Vishnu, Buddhavarman which may be the names Skandavarman II, Kumaravishnu I and Buddhavarman; but the same three names repeat from 12 to 14. These are preceded by two other names 10 and 11, Kanagopa and Virakurcha; then from 9 to

1. Vimala.

2. Konkanika.

3. Kalabhartr.

4. Chutapallava.

5. Virakurcha.

6. Chandravarman. 7. Karala.

8. Vishnugopa.

9. Skandamula.

10. Kinagopu.

11. Virakurcha.

12. Skandavarman.

13. Kumaravishnu.

14. Buddhavarman,

15. Skandavarman.

16. Kumaravishnn.

17. Buddhavarman.

18. Skandavarman.

19. Vishnusropa.

20. Vishnudasa.

21. Skandavarman.

22. Simhavarman. 23. Viravarman.

24. Skandavarman.

25. Simhavarman.

26. Skandavarman.

27. Nandivarman I.

28. Simhavarman.

29. Simhavarman.

30. Vishnugopa.

31. Simhavarman.

32. Simhavishnu.

33. Mahendravarman I.

34. Narasimhavarman I.

35. Mahendravarman II.

36. Parmesvaravarman I.

3 there is a considerable agreement with the table here except that No. 6 Chandravarman has to be taken as a mistake for Skandavarman. The name Karala does not appear in any of the grants at all and the connection of the first two names Vimala and Konkanika do not find reference in any of the grants available to us. In respect of this list of 36 names, it must be borne in mind that it is a list made up in the reign of the later Pallava Narasimhavarman II, and in all probability the list was put together from a comparative study of the various tables discussed above from some record of these various grants; what is worse, put together perhaps without any accurate knowledge of the connec-tion of the various members to each other, or their actual position in the succession. This seems the only explanation for the repetitions and variations that one notices in the list in comparison with the genealogies of the grants. It would be safer to guide ourselves by the various tables discussed above rather than by this one omnibus list which otherwise provides us with no details whatsoever.