________________
206
THE INDIAN ANTIQUARY
OCTOBER, 1930
the Slokas of A group and the Slokas of B group, and so should be placed immediately before the slokas of B group of the Markandeya and Vamana. If it be so placed it will clearly present the boundaries of the Kumâra dvipa, showing it to be the country which stretched from Cape Comorin (Kumarika) to the Himalaya (source of the Ganges-Va.) and bounded by the Turks on the north14 (Vam.), by the Andhras on the south16 (Vam.), by the Kiratas on the east16 and by the Yavanas on the west 1T (Mark.; Vám.). But these boundaries roughly correspond to the boundaries of India proper, and as these boundaries constitute a sort of explanatory note to the navama dvipa, it can be safely argued that the ninth dvípa, i.e., the Kumara or Kumarika, was India proper.18
Then again almost all the Purdnas such as the Brahmanda, Väyu. Märkandeya. Matsya, Vamana, Garuda, etc., after describing the boundaries of the navama dvipa [the slokas of which have been quoted above (pp. 204-05)), go on to give a description of the characteristics of the people of that region, followed by a list of the seven hills known as the kula parvatah,19 wbich evidently must refer to the mountain ranges of the Kumara or the ninth dvipa. This is made more clear by Rajasekhara in his Kavyamimamsa, 20 where he also mentions the very same hills and quotes the same sloka as the Purdnas, but prefixes the specific wordsatra ca Kumari dvipe. And as all the seven ranges belong to India proper, Kumâri dvípa might reasonably be regarded as identical with it.31 The Vamana also in another place explicitly refers to this identity. Thus after the enumeration of all the countries or people
14 The Arab geographer Rashidu'd-din refers to this. Thus, while describing the boundaries of India, he says: “On the north lie Kashmir, the country of the Turks, ....." Elliot, History of India, vol. I, p. 45.
18 Obviously tho Andhras cannot form the southern boundary of India proper, unless a limited sense is understood. That it has not been used in that sense is evident from consideration of the other three boundaries that have been set forth. The statement of the Vamana is, therefore, to be taken as an exception; but at the same time we must keep our mind open to this possible view algo, that the floka (V am. 13, 12) right reflect the political conditions of the time when the Andhra rule was widespread as in the second century A.D., almost throughout the whole of southern India (Ind. Ant., vol. XLVII, 1918,
Dekkan of the Sataváhana period '), and sw such formed the southern boundary of India proper. Regarded in that light, the sloka night offer an important clue to the date of the Purana in question,
16 They may be identified with the Kirrhadia of Ptolemy (Cuaningham's Geography of Ancient India, ed. S. X. Mazumdar, p. 219) located near the mount Maiandros.
17 The inscriptions of Asoka mention the Yonas in connection with the Kambojas and Gandharas. The Mahavainsa also refers to the country of the Yonas (Geiger's trans., P. 85). Their capital was Alagande (=Alexandria, op. cit., p. 194 n.) near Kabul.
18 It is, of course, a fact that dloka No. 81 of the Vayu refers to the nine thousand yojana aren of a country, which as we have seen was India proper. So the ninth doipa with which India proper has been identified must be of the same area. But it has boen definitely stated in all the flokus that the ninth dvi pa was of one thousand ' area. So it might be argued that the mention of the nine thousand yojana (va. 81) in connexion with India proper, distinguishes India proper from the Kumâra or navama dufpa, and lends colour to the opposite view, that perhaps those nine dvipas were so many divisions of India proper included within it. But, as opposed to this, it might be said that the sloka No. 81 of the Vayu with its line dyato hyd inevitably refers to the previous Sloka (No. 80), which describes the navama dufpa. and go the boundary of India proper, which is supplied in dloka No. 81 (combined with the next sloka No. 82) applies to the navama dulpa alone, and as such the two are identical. There is no room for distinguishing one sloka from the other. The three slokas of the Vdyu are to be taken in a connected way. Moreover, the Bd. and the Mat, do not follow Vd. in its statement sahasrdrinavaivatu. Thus the Bd., for instance (49.16), reada sahasrdni traya, instead of navama, and the Mat. reads (114, 10) Sahasrdni dadaiva tu.' So we see that there is no coherency in the statement of the Vayu.
19 In overy Purdna the sloka runs in the same form and in the same language. I may quote the Vayu Purdna (45, 87-88) :
Sapta ca aamin suparudno vidrad kuulaparvatah Mahendra-Malayah-Sahya Suktimdn Rkraparpatah
Vindhyadca Paripdtrasca saptaite kulaparvatdh. 30 Kavyamimdmód (Desavibhaga), p. 92.
91 But it must be noted that none of the seven ranges carry ug beyond the Vindhyas into northern India. [Some scholars, however, would identify the Paripátra mountains with the Aravalli range. -JT, EDITOR.)