PTOLEMETA TO MERGE. 397 been different places As the distance of Scylax from the port to the city of Barca is given in distance from the sea, we must look for the latter (supposing the one hundred stadia to be correct, which we have no sufficient reason to doubt) in some part of that range of mountains which bounds the country between Bengazi and Ptolemeta to the southward; and it appears extremely probable, that its site should be fixed for in some part of the plain of Merge : it may be, in the remains which have already been mentioned in the southeastern part of the valley. Here, however, a considerable difficulty occurs, of the accounts of the city of Barca, which are given by early Arab writers, are at all to be depended upon as correct ; for although these accounts, in many respects, will be found to be extremely unsatisfactory, they all appear to concur in placing the city in the neighbourhood of a mountainous country, but, at the same time, rather in a plain at the foot of it than in any part of the mountains themselves.

The distance, however, of Barca from the coast unavoidably places it in some part of the range which we have mentioned ; for there is no part of the plain between this range and the sea, (as will clearly appear by the chart,) which is distant anything like one hundred stadia from the coast, with the exception of that to the southward of Bengazi ; and to suppose the town of Barca there would be absurd, for it would then be fifty miles distant from its port and only nine or ten from the more convenient harbour of Berenice. Edrisi has given us several distances to and from Barca; as — from Barca to Angela ten E)£ Se Xi/xHvor rnr "Kifx.svos tov Kocra Baxvv araSia (p', •h Se tioKis Ba§xs<wv a.710 ’bxXa.airts aracoix g. (Perip. p. 109.)