Josephus states that "the beginning of the third wall was at the tower Hippicus, where it reached as far as the north quarter of the city and the tower Psephinus, and then was so far extended till it came over against the monuments of Helena, which Helena was queen of Adiabene, the daughter of Izates; it then extended farther to a great length, and passed by the royal caverns (+spêlaiôn+), and bent again at the tower of the corner, at the monument which is called the Monument of the Fuller, and joined to the old wall at the valley called the Valley of the Kidron[132]." In laying down the course of this wall I differ from all those (in particular Barclay, Schultz and Robinson) who, up to the present time, have written on the topography of ancient Jerusalem. I am led to do this by the careful investigations, which, during a long time, I carried on in the district north of the city. It is my positive opinion that the ancient walls did not extend to the north beyond the present enclosure; that is, that they began at the Jaffa gate, passed by the Damascus gate, and ended at the north-east corner of the _Haram-es-Sherîf_. Let me now state the facts which have led me to this conclusion. In 1860 the Greek convent repaired the building outside the Jaffa[133] gate, now used as a custom-house. Wishing to lay some foundations against the city wall, I came, on digging down, upon those of Agrippa's, which rest upon the rock; now we know that this wall near to Hippicus was defended by the steep slope of the side of the valley, and that where this ceased, towards the north-west corner, a ditch was cut in the rock. This may still be seen, and is a proof that I am right in supposing the present to be the wall that went from Hippicus to Psephinus. At the north-west corner a massive ruin still exists inside the city, rising about twenty feet above the ground, and built of small stones joined with strong mortar; in the south-west corner however are found large stones, rusticated after the Herodian pattern. On digging about the shapeless pile, I discovered that courses of similar stones continued down to the rock. I also found two sides of masonry, and many large rusticated stones buried in the rubbish, and traces of a great cistern. Hence I consider this to be the site of the tower Psephinus, an octagon in form, and seventy cubits high[134]. Beyond these ruins, outside the present wall, is a ditch cut in the rock, unquestionably a work of the Herodian age, for no later conquerors would have had the time or desire to execute such a great and costly work. It is now concealed by rubbish, but it runs eastward parallel to the present wall, which therefore can scarcely have extended beyond it, in the course laid down by Barclay, Schultz, and many others.