The account which Josephus gives of the site on which Jerusalem was built is explicit and easily understood. It was placed on two hills (B. J. v. 4) opposite each other, with a valley between. The hill of the Upper City was the highest and largest; the second, that of Acra, was lower; a third hill, lower still, was to the east, separated by another valley, which was filled up by the Hasmoneans. The first valley--the Tyropœon, which divided the Upper and Lower City--ran down to Siloam. Other deep valleys with precipices existed beyond the city on all sides, except on the north where three successive lines of fortification protected the town. Turning to the plan of the rock beneath modern Jerusalem, which is given in illustration, we see just such a site before us. On the south is a large and high hill, the top 2540 feet above the sea, with a deep valley to the south and west, and a second valley, almost equal in size, to the north and east. Down the last-mentioned valley David Street now runs, but the accumulation of rubbish is in parts fifty feet deep. By the observations taken in making excavations in the old Hospital of the Knights of St. John, and in a vault farther east, as well as at the foundations of the Bishops’ Palace and of the hotel near David’s Tower, we ascertain the following details: that the valley, breaking down suddenly eastward, has its head at a narrow saddle at a level about 2500 feet above the sea, and that this saddle separates the head of the eastern valley from that of Wâdy Rabâbeh, which runs to the west of the Jaffa Gate: the eastern valley proves to have a depth of more than 100 feet below the summit of the southern hill. Other observations, farther east, show that the precipice, visible just opposite the great bridge from the south-west corner of the Haram, runs north and turns westward, where either a vertical scarp, or a very steep slope, forms the north-east angle of the southern hill above the corner where the great valley sweeps round southwards descending towards Siloam. The plan further shows that the ground rises again north of the valley, and forms a small knoll in the neighbourhood of the Church of the Holy Sepulchre, with a second valley head to the east. This knoll is actually fifty feet lower than the top of the southern hill, and, from the lie of the ground, it appears to be still lower than it really is. The second valley on the east of the knoll separates off a third hill now occupied by the Mohammedan quarter of Jerusalem, and this is divided from the Temple Hill, of which it is really a part, by the rock-cut trench forty feet deep, hewn on the north side of the scarp, which I have endeavoured to show was the Castle of Antonia. The third hill is lower again by fifty feet than the knoll last mentioned.