the valley are honey-combed with Mahomedan tombs, and the face of every flat-spaced eminence is scarred with them. A hundred generations of Moslems are buried there. The rocky declivities which hedge in this remarkable site may give some clue to the yet more ancient name of Talara which this place once bore. Talar in Baluchi bears the signification of a rocky band of cliffs or hills. The obvious reason why the port of Tiz was chosen for the point of debarkation for India is that, in addition to the general convenience of the harbour, the monsoon winds do not affect the coast so far west. At seasons when the Indus delta and the port of Debal were rendered unapproachable, Tiz was an easy port to gain. There must have been a considerable local trade, too, between the coast and the highly cultivated, if restricted, valleys of Northern Makran, and it is more than probable that Tiz was the port for the commerce of Seistan in its most palmy days. From Tiz to Kiz (or Kej, which is reckoned as the first big city on the road to India in mediæval geography) was, according to Istakhri and Idrisi, a five-days' journey. Kiz is doubtless synonymous with Kej, but the long straight valley of that name which leads eastwards towards India has no town now which exactly corresponds to the name of the valley. The distance between Tiz and the Kej district is from 160 to 170 miles. No actual ruined site can be pointed out as yet marking the position of Kiz, or (as Idrisi writes it) Kirusi, but it must have been in the close neighbourhood of Kalatak, where, indeed, there is ample room for further close investigation amongst surrounding ruins. About the city, we may note from Idrisi that it was nearly as large as Multan, and was the largest city in Makran. "Palm trees are plentiful, and there is a large trade," says our author, who adds that it is two long days' journey west of the city of Firabuz. From all the varied forms which Arab geographical names can assume owing to omission of diacritical marks in writing, this place, Firabuz, has perhaps suffered most. The most correct reading of it would probably be Kanazbun, and this is the form adopted by Elliott, who conjectures that Kanazbun was situated near the modern Panjgur. From Kej to Panjgur is not less than 110 miles, a very long two-days' journey. Yet Istakhri supports Idrisi (if, indeed, he is not the original author of the statement) that it is two days' journey from Kiz to Kanazbun. This would lead one to place Kanazbun elsewhere than in the Panjgur district, more especially as that district lies well to the north of the direct road to India, were it not for local evidence that the fertile and flourishing Panjgur valley must certainly be included somehow in the mediæval geographical system, and that the conditions of khafila traffic in mediæval times were such as to preclude the possibility of the more direct route being utilized. To explain this