city, 'at which the corresponding rows of houses on both hills end.' This was held to be more applicable to a valley running from the Jaffa Gate than to that from the Damascus Gate when the slope is so gradual that the rows of houses now run across it without interruption, besides which it probably had formerly a wall on either side of it. Mr. Lewin[17] speaks positively as to the Tyropoeon commencing at the Jaffa Gate, and says it can be traced thence to the Haram by the _rise_ of ground which is still very perceptible on the right hand, as you walk down the street from the gate to the Haram. He makes this valley the boundary of the high town on the north, and puts his first wall on the southern brow of it. It is difficult to see on this hypothesis how the hill of the high town could be 'in length more direct' than the eastern hill, as Josephus says it was; or how the corresponding rows of houses could meet any more readily than near Damascus Gate. However, Mr. Warren, after excavation, tells us that 'a very decided valley runs down from the Jaffa Gate to the Tyropoeon, near Wilson's arch;' and he found under the causeway leading westward from Wilson's arch, vaults and chambers, and a secret passage, at a depth which serves to confirm his view. There is no disputing facts, though it seems to us still questionable whether this valley is any part of the Tyropoeon of Josephus. The valley running south-east from Herod's Gate, in the north-east part of the north wall, proves to be longer and deeper than any theorist had imagined, running into the Kedron at a point between the north-east angle of the Haram and the Golden Gate, and being filled in with more than 100 feet of _débris_. The Pool of Bethesda, which is 360 feet in length, is imbedded in this valley, and stretches across it, having its ends formed by the rocky sides of the valley, and its sides built up of masonry; and since it is found lined with concrete, it must have been a reservoir, and not the fosse of Antonia, which Robinson supposed it to be.[18]