although they are so numerous on its eastern side, that is to say on the Bosporian territory."

p. 436. "After the second station [towards Kaffa] we passed another antient boundary or vallum like that which has been described before, on which may be discerned the traces of turrets which were placed along this second barrier of the Bosporians."

p. 442.

"In the last stage to Kaffa we passed the third, that is to say the outer, vallum or boundary of the Bosporians. Its remains, as well as those of the towers placed thereon, were very visible."

There is evidently no trace, on the peninsula of Kertch, of any rampart with the fosse on the eastern or, speaking with reference to the Crimea, the exterior side, as is required by the narrative of Herodotus.

Negatively therefore the evidence leads us back to Perekop: and as there is no evidence that the Tartars or Turks during the times of the Lower Empire executed any other works of the kind, we have some grounds for assigning it to an earlier period. And if we could be sure that the existing rampart was there in the time of Herodotus, whom we know to have been once not more than a hundred miles away from the spot, it would be reasonable to suppose that he was likely to mention so conspicuous a structure.