Turning to the inscriptions of Asoka the southernmost limit reached by them is in the north-east corner of the Chitaldroog district of Mysore where Brahmagiri, Siddhapura and

1 That the Kosar were known in four divisions is clearly stated in 11. 508-9 of the Maduraik-Kanji. The author there institutes a comparison between the appearance of the four groups of councillors, at the Pandyan Court (other than the ministers) to the coming of the four sections of the Kosar "of unfailing word."

2 There is a Sellur between the Railway station Koradacheri on to Tanjore-Negapatam line, and Kodaivasal, a place of some importance now and of great repute in the age of the Sangam. The only objection to the identification is that it is not as near the sea as the texts would require. The local Aiyanar (Sasta) temple seems identifiable with the "sacrificial abode" of Parasurama. There is a Sellur, on the West-Coast associated with Parasurama, in the Keralolpotti. This work calls the place Parum-Sellur, great Sellur. This must have been near Cannanore. The coast near Mont D'Ely is called Ramandali = Tam. Raman and tali or Rama's temple, and must be regarded as a later settlement or a colony. It may seem farfetched to connect the Kosar with the Cushites who were banished from their native land of Elam after the Assyrian conquest till we get more evidence of a definite character, or identify them with Satiyapatre of the Asoka edicts, taking the word as the Sanskrit Satyaputra.

Jatingaramesvara hill edicts were discovered. Rock edict II speaks of "his neighbours such as the Chodas, the Pandiyas, the Satiyaputra, the Keralaputra; "ambapanni, the Yona king Anti-yoka as well as among those who are the vassal kings of that Antiyoka" in connection with the establishment of hospitals, etc. The Vth edict refers to the appointment of the overseers of the Law who were concerned with the "welfare and happiness of my loyal subjects, as also among the Yonas, Kamboyas, Gamdharas, Rishtikas, Pitinikas and all other nations who are my neighbours." In respect of these overseers of the Law a distinction is clearly made between Asoka's loyal subjects forming one class; Yonas, Pitinikas and others forming another class, and his neighbours forming the third class. The second of these have therefore to be regarded as not his subjects, nor exactly his neighbours. The geographical position of these would make them his feudatories, the first three being on the north-western frontier, the last two Rishtikas and Pitinikas in the coast region set over against the Dakhan plateau, being respectively Rashtrikas and Pratishthanakas. In Rock edict XIII referring to conquests through the sacred law he claims having effected that conquest over his subjects in his empire and over all his neighbours for a distance of 600 yojanas of the country of Amtiyoka and the four kings his neighbours. Coming down to the south he refers to the Cholas, the Pandyas and Tambapanni or Ceylon. Then he proceeds to the second class in edict V of whom specific mention is made of Visas, Vajris, the Andhras, and Pulidas (Pulin-das) apparently tributary communities in the neighbourhood of the emperor's regular territory, but politically tributary to him according to the notions of the Artha Sastra. Then follows the important statement "even those to whom the messengers of the 'Beloved of the gods' do not go follow the sacred law, as soon as they have heard of the orders of the 'Beloved of the gods' issued in accordance with the sacred law, and his teaching of the sacred law, and they will follow it in future."1 The corresponding portion of this last statement in Vincent Smith's version of rock edict XIII based on the Shabbas-garhi edicts, reads slightly differently, and the reading may be set down here for comparison "and here too, in the king's dominions, among Yonas, and Kambojas, among the Nabhapantis of Nabhaka, among the Bhojas and Pitinikas, among the Andhras and Pulindas - everywhere men follow his sacred majesty's instruction in the law of piety. Even where the envoys of his sacred majesty do not penetrate, there all men hearing his sacred majesty's ordinance based on the law of piety and his instruction in that law practise and will practise the law." This makes a considerable difference in respect of the recital of the tribes that are concerned. While Smith's version is certainly fuller and more correct than that of Buhler quoted above, the tribes Visas and Vajris are clearly mentioned in the Shabbasgarhi and Mansera edicts as Visha, Vajri. In the Kalsi version however, the reading1 is Visa, Baji. Therefore the Visha of the one version is what exactly is Visa in the other, and the Baji of the other version is Vajra of the former. The point I am particularly concerned with here is that the Visas and Vajris are apparently tributary tribes of whom Tamil literature refers to the latter. One passage in the classical work Silappadhikaram referring to the northern invasion of the great Chola Karikala states that three kings made him presents which formed the ornaments of his capital Kaveripattanam. Of these three kings one was a friend, another an enemy recently compelled into treaty terms, and the third one was neither friend nor an ally, that is, a neutral. The king of Vajra is referred to as the neutral king who was neither his ally nor his enemy. His territory must have reached the sea-shore at any rate, and is explained by the asoka's southern limit of empire 33 commentator as territory on the banks of the river Sone. If we make an inference from the general description given, it would mean the territory of Bengal between the Sone and the Ganges reaching down to the sea as it is clearly stated to be bordered by "great waters on all sides." The next king is the king of Magadha who having been an enemy submitted and became his friend. The next one is the king of Avanti who was his ally. The first one presented a canopy of pearls. The king of Magadha gave him what is called a Vidyamantapa,1 apparently an ornamented platform pillared and roofed over. The king of Malva gave him a triumphal arch by way of a present.2 The Vajras therefore as a people of considerable importance and holding the important territory of Bengal on this side of the Ganges, were known to the Tamils of the first century A.D.3 It was apparently a princess of this kingdom, which seems to have been powerful at the time, that Kharavela of Kalinga married.4 In such a case Simhapura would be the capital of the Gangetic Kalinga included

1 E. Ind., Vol. II. Buhler's edition of the edicts of Asoka.

1 Vide Ramavatara Sarma's Piyadasi Inscriptions, Variant readings, p. 20.

1 For a description of this see-Rajasekhara's Kavyamimanisa. Ch. X.

2 Silappadhikaram, Canto 5, II. 95-105.

3 Vajra was one of the two divisions of Ladha or Radha, the two divisions being Vajjabhumi and Subbabhumi. Ayaranga Sutta and other references quoted in "The Ajivikas" by B, M. Barua, Pt. I, pp. 57.8. Calcutta University Publications.

4 J. B. O.R. S., Vol. IV, p. 378. 5 in the Kalinga1 kingdom very often referred to as the capital of Kalinga as a whole.

Asoka's empire then may he regarded in three parts; the whole of Hindustan, the country between the Vindyas and the Himalayas, with an outstretch along the west coast to take in the Pitinikas, Rishtikas, and along the east coast to take in the whole of Kalinga, would have constituted his own kingdom. Then comes the borderland of the great forest of Dandaka. On the frontiers of the forest were situated the territory of the several of the semi-civilised tribes till we come to about 14° of north latitude roughly. These had been reduced to some kind of allegiance which apparently involved the responsibility of paying tribute and being in the kind of tribal subordination recognised in the Artha Sastra. Then follow naturally the territories of his southern neighbours. This disposition is what is actually reflected in Tamil literature which states in clear terms that Pulikat was the northern boundary in the east and that Tulu-Konkan the kingdom of Nannan was the western boundary, a whole belt of country across being occupied by tribes whom the Tamilians called Vadukar. These were border-tribes engaged in cattle-lifting and waylaying people as occasion offered. Referring to the achievements of a Chera ruler who is known to literature Adukotpattu-Seral-Adan, that is, the Chera who captured cattle, the poetess Kakkaipadiniyar Nachchellaiyar states in clear terms that he captured cattle in Danda-ranyam, distributed them in Tondi, among Brahmans, giving along with these one special kind of a cow classified as kapilai (a cow of a dark colour, the darkness spreading over even to the udder) and a village in Kuda-nadu (western hill country), and having defeated the other Malavar, in battle turned back the kings, apparently their kings. The term Bandaranya in this reference is explained by the old commentator as a nadu or division of country in Arya-Nadu thus confirming the statement in the Sarabhanga Jataka that Dandaka was a Bhoja-kingdom with capital Kumbhavati. This means clearly that the forest of Danda or Dandaka was, according to the political divisions of those days, included in the territory of the Aryas as distinct from the Tamils, the semi-civilised tribe or tribes being interposed between the two frontiers across the whole of the Peninsula.

1 Referred along with Kapilapuram in the Silappadhikaram as the two capitals of Kalinga; Canto XXIII, 11.140-41.