________________
292
EPIGRAPHIA INDICA.
(VOL. XIV.
proceed from Hien-tu' to Kashmir, but towards the west, whereby he would have to cross the Indus a second time. His description would accordingly lead us to infer that Ki-pin was situated to the west of Kashmir.
This well agrees with the information which can be gleaned from older Chinese sources. According to the Han annals Ki-pin was bounded towards the 'south-west by Wu-i-shan-li, towards the north-west by the Bactrian empire of the Yüe-chi, while towards the north-east it was eight days' jonrney distant from Nan-tun, and towards the east 2250 li distant from Wucha. In another place Ki-pin is mentioned between An-si, i.e. Parthin and Wu-i-shan. Wui-shan-li is, as shown by Professor Marquardt, a rendering of the Greek Alexandria, i.e. Kandahar, or, according to M. Chavannes, Herat, and according to Cunningham it comprised the whole of South-Western Afghanistan. This description certainly seems to indicate parts of Afghanistan, to the south of the Hindukush. Of Kao-fu, to the south-west of the Ta Yüe-chi, we hear that it sometimes belonged to Ki-pin and sometimes to T'ien-chu. Kao-fu has been identified with Ptolemy's Kaboura, the present Kabul, and, if we consider Kao-fu as the border land between Parthia and Ki-pin, to which latter country it sometimes belonged, we are led to think of parts of the Kabul valley and neighbouring districts, i.e. territories which we have found formed part of Kapiša. Professor Franke therefore arrives at the conclusion that Ki-pin comprised the north-western portion of the present Kaebmir state, the Indus country down to the Kabul river, the country between the lower Kabul river and the Swat river, and further parts of the Panjab. These must be added, because we learn about Ki-pin that it was a flat and hot country, a designation which does not at all suit Kashmir. I think that we cannot get nearer at the truth. The only territories mentioned by Professor Franke which I do not think can be proved to have belonged to Ki-pin are the north-western districts of Kashmir. On the other hand it is possible that it extended a little further towards the west and the south-west.
Ancient Ki-pin thus included districts which were later on known as Kapiga, and I do not see why it should be necessary to assume that the Chinese at different periods used the designation Ki-pin in two different senses. It seems to me that we shall have to return to the old explanation of Ki-pin as a rendering of the saine word which the Greeks made into Kophẽn, whereby we must bear in mind that the Greek ph was an aspirated labial and not a spirant. The reasons brought forward against this identification are little convincing. M. Lévi thinks that the name Kophon had probably already become obsolete when Megasthenes picked it ap. Ptolemy ignores it and calls the principal river of Afghanistan Koas, and finally Strabo has another form of the name, viz. Köphes. I cannot find anything in these arguments which disproves the old identification of the names Kophěn and Ki-pin, which is in itself much more likely than the supposed renderings of Kapisa and Kasmira by Ki-pin. The fact remains that the Greek know a name which they thought sounded like Kophen, that this name sounds much more like Ki-pin than any other name which has been suggested, and that we have every reason for looking for Ki-pin in the same neighbourhood where the ancient Greeks located Kophêne, the country on the Kophēn. The fact that the Cbinese later on used the name Kapiss of districts which others included in Ki-pin seems to be much more naturally explained by assuming that Kapisa formed part of ancient Ki-pin and was sometimes designated by means of the wider name Ki-pin.
In support of my location of ancient Ki-pin I may mention a small detail. I hope to have proved in another place that the so-called Marundas who ruled in the Ganges valley in
· Untersuchungen zur Geschichte von Eran, H. 2, Leipzig, 1905, p. 176. ? T'oung pao, II, vi, p. 514.
Ancient Geography of India, Vol. I, p. 39. • Specht, Jo. As., VIII, ii, p. 325.
• Franke, p. 77. . Jo. 48., IX, vi, pp. 871 8.
1 SBAW., 1916, pp. 790 e.
