What was said of the origin of the Pallavas in the previous sections would have made it clear that they were in all probability a family of feudatories of the Satavahanas of the Dak-han. These feudatories are clearly described as belonging to the family of the Nagas, whatever that may mean to us now. Northern Mysore and the country set over against it up to the western sea which later on became a fief of the Kadambas was in the possession of a Naga family of Maharatis belonging to the Chutakula, apparently a Naga designation. The Satavahana Rashtra proper, set over against the territory of Kanchi further to the east of this division, was the fief of the great commander (Mahasenapati-Skanda Naga). In the days of the greatest expansion of the Andhra empire under Pulu-mavi II and his immediate successors the whole of the southern frontier of the Andhra country, the region of the Vadukas according to Tamil literature, was held by powerful families of these Nagas. When the Andhra empire broke up early in the third century these powerful feudatories made themselves independent in the regions under their government. Tondamanda-lam which in the reign of the great Pulumavi was under the Satavahanas should have fallen to the lot of the Mahasenapati referred to above, or his successors in the same region, the district which was called peculiarly the district of the Satavahanas. The advance of the Satavahanas themselves under Pulumavi must have put an end to the authority of the Cholas in this particular region. When the Governors set up independently of the Satavahanas, a generation or two later, the Mahasenapati Skanda Naga a himself or one of his successors became heir to this region of the Tondamandalam as well. According to the available inscriptions of the Pallavas, the Pallavas could be divided into four separate families or dynasties. The connection of some of these to one another we know, and of others we do not know. We have a certain number of charters in Prakrit of which three are important ones. Then follows a dynasty which issued their charters in Sanskrit; following this came the family of the great Pallavas beginning with Simha Vishnu; this was followed by a dynasty of the usurper Nandi Varman, another great Pallava. We are overlooking for the present the dynasty of the Ganga-Pallavas postulated by the Epigraphists. The earliest of these Pallava charters is the one known as the Mayidavolu 1 (Guntur district) copper-plates.

These plates contain the charter issued by the heir-apparent (Yuva Maharajah) Siva Skanda Varman making a grant in the division Dhannakada, that is Amaravati, in the tenth year of the reign of his father whose name is not given.

The next record is what is known as the Hirahadagalli plates (Bellary District).2 This record is dated in the 8th year of Sivaskanda Varman and confirms the gift made by his father who is described merely as "Bappa-deva" (revered father). Another copper plate charter found in the Guntur district, is dated in the reign of a Vijaya Skanda Varman and is the record of a grant made by Charudevi,3 wife of the Yuvamaharaja Vijaya Buddha Varman and mother of Prince Buddhyankura. There is no doubt that the Yuvamaharaja of the first record is the same as the ruling sovereign of the second, the name and circumstances of the two records giving us full warrant for the identification. The question is a little less certain in respect of the sovereign mentioned in the third record, namely Vijaya Skanda Varman. Is he the same as the Siva Skanda Varman of the previous two records? Among the records of that age Siva, Vijaya and sometimes even Vijaya-Siva, are used as prefixes indicating the regard or respect in which the ruler was held. Apart from this the use of the attribute Vijaya before Skanda Varman does not alter the name, but only gives an additional circumstance of importance. It would not be therefore doing any particular violence to identify the Vijaya Skanda Varman of the third record with the Siva Skanda Varman of the other two. These three charters all of them refer to the region which was peculiarly the district of the Sata-vahanas. If this identification of Vijaya Skanda Varman turns out true the succession could be arranged in the following table:

1 Eph. Ind., Vol. VI, p. 84.

2 Ep. Ind., Vol. I, p. 2.

3 Ep. Ind., Vol. VIII, p. 2, 143.

The Mayidavol grant was issued, from Conjeeva-ram (Kanchi) by the heir-apparent, to the Governor at Amaravati, and the village granted is described as being in the Andhrapatha (Vadu-kavali of the Tamils, the Andhra country). Thus it is made clear to us that Kanchi was already the capital of a region taking in naturally the Tondamandalam and the districts north of it at least as far as Amaravati or the River Krishna. In the second charter the ruler Siva Skanda Varma lays cairn to having performed the Agnishtoma, Vajapeya and Asvamedha sacrifices. Of these the last could be performed only by a conqueror, or one who set up as such. The way that headdresses his grant to the lords of provinces royal princes, generals, rulers of districts, customs officers, prefects of countries, etc., gives us an insight into the distinct Asokan character of the organization of the government and its affiliation even to the Arthasastra. What is more important it exhibits an organization which is northern in character, perhaps quite distinct from that of the Tamils of the farther south. There is another interesting detail in it that the father of this king, whatever his name, had granted many crores of gold, and what is more important to us in connection with the origin of this dynasty one hundred thousand ox-ploughs. This, if it means anything, indicates undoubtedly the effort made by this ruler for the conversion of the great forests into arable land. It would be well to remember in this context that this part of the country was known to the Tamils as Dandaranyam, the same as the Sanskrit Dandakaranya where cattle-rearing was the principal occupation, and cattle-raiding the principal sport. It was apparently this "Bappa-deva" that made an effort, with what success we are not told, to transform the forest into cultivated country. It will thus be clear that this dynasty of the Prakrit charters beginning with "Bappa-deva" were the historical founders of the Pallava dominion in South India. It is taken here that all the ruler whose charters in Prakrit have come down to us are to be regarded as members of a single dynasty while there is the possibility that they were members of two dynasties which may not after all be connected with each other; but there is little doubt, if this alternative should turn out true, that the two dynasties followed each other without much interval.

Passing on to the Pallavas of the Sanskrit charters we come to a number of dynasties which would at first sight appear to be so many separate dynasties. According to the Uruvapalli copperplates the succession is as follows:-

The Darsi fragment refers itself to the time of the great-grandson of Virakorcha Varman, that is Vira Varma, referring apparently to Simha Varman, son of Yuvamaharaja Vishnugopa. The Chendalur Plates issued from the "victorious Kanchipura" gives and Menmattura. There are considerations which would lead us to consolidate these four separate genealogies into one genealogical table. The Uruvapalli copper-plates record the grant of Yuva Maharaja Vishnugopa, but the grant is dated in the reign of a king named Simha Varman. If Vishnugopa issued the grant as Yuvamaharaja and dates it in the reign of a Simha Varman, Simha Varman must have been the Maharaja, either the father or an elder brother of the donor. According to the grant itself Vishnugopa's father is a Skanda Varman. The only other alternative therefore is that Simha Varman was in all probability an elder brother of Vishnugopa. So the genealogy will have to be extended by the addition of Simha Varman and would stand-

The Udaiyendram grant similarly gives:-

The newly discovered Ongodu-Plates give:-

These four separate genealogies were apparently not altogether separate in respect of the fact that several of these grants were issued from Kanchl, and others from places like Dasanapura, Palakkada.

The Chendalur genealogy contains four names beginning with Skanda Varman. Dr. Hultzsch from palaeographical considerations held that these rulers must have come in between 20

Simha Varman II and Simha Vishnu. There are considerations however which would lead to the identification of the Skanda Varman of these plates with the Skanda Varman the father of Yuvamaharaja Vishnugopa. This arrangement would make Kumara Vishnu another brother of Yuvamaharaja Vishnugopa, with a son Buddha Varman and his son Kumara Vishnu II. The genealogy of the Udaiyendram grant gives again at the top a Skanda Varman followed by three other names ending in Nandi Varman. The Velurpalaiyam plates introduce what appears a gap with Kumara Vishnu II and brings in a Nandi Varman before introducing Simha Varman, the father of Simha Vishnu. The only Nandi Varman referable to this period would be the last name mentioned in the Udaiyendram grant. Therefore it is possible to include this genealogy in that of the line of Simha Varman, the elder brother of Vishnugopa; Skanda Varman being the father, Simha Varman his eldest son and the elder brother of Yuva, Maharaja Vishnugopa, his son Skanda Varman and his son Nandi Varman. This last will, according to the Velurpalaiyam plates bring us on to the line of Simha Vishnu. The Ongodu plates discovered in the year 1915 introduce us to yet another line beginning with Kumara Vishnu. The last of these Skanda Varman issued the document not from Kanchl but from Tambrapa.

None of these names figure in the Velurpalaiyam plates in this order; nor does the Vayalur Pillar contain the four names in this order as given in the Ongudu plates. The Kumara Vishnu at the head of the table therefore may be Kumara Vishnu I or Kumara Vishnu II, and the whole dynasty, a local dynasty having had nothing to do with the regular succession of the main line. If it should actually have been so we get the final genealogy as follows:-

Verse 10 of the Velurpalaiyam plates introduces then, without specifying any connection, Simha Varman, father of Simha Vishnu and that introduces us to the line of the well-known Pallava dynasty. Before proceeding to a consideration of that dynasty we have to consider one or two questions that arise in respect of the dynasty of the Sanskrit charters, and Vishnu-gopa of Kanchi, the contemporary of Samudra Gupta. Incidentally also we shall have to consider the question whether the dating of the

Uruvapalli, Mangalur and Pikira grants respectively from Palakkada, Dasanapura and Men-mattura, all of them places in the Guntur district, warrants the assumption of a Pallava interregnum in Kanchi; if there had been such an interregnum, whether that is the time to which we could refer the ancient Cholas, Kari-kala and others.