KUMMA which the sandstone used in building the temples of Semna and Kumma was taken, but we had not time to visit them. At no great distance from Kumma, to the south, there are some mudbrick buildings which are thought to be those of a Coptic church and monastery ; these also we could not visit. We returned to the river, and, having said farewell to the shekh, took our places on the raft, which we noticed was somewhat water-logged. When we pushed off we found that the wind which had helped us to cross to Kumma was now against us, and in spite of all the efforts of the men who were swimming behind the raft and pushing it on, our progress across the river was slow. The raftman persisted in working his paddle on the wrong side of the raft, and this helped the current to carry us down stream some distance. One of the swimmers became numbed with cold, and his clinging to the raft kept one end under water, and allowed the little waves to pass freely all over it. Our ankarib was soon awash, but we reached the west bank again in about an hour, and were glad to step on to the rocks and walk back to the temple of Semna. On our way we noticed some Egyptian tombs hewn in the rock, and near the top of a bold bluff we saw an inscription in large hieroglyphics, which commemorated an official called Ameni : this name suggests that it was cut under the Xllth Dynasty, when probably more than one Ameni was associated with the Egyptian occupation of the Sijdan.

The examination which we made of the temple enclosure at Semna convinced the Inspector and myself that only excavations carried out on a very large scale could reveal the general history of the fortress, and we therefore determined to confine our investigations to the buildings in the immediate neighbourhood of the temple of Thothmes III. Directly opposite to the main door of that temple is a large mud-brick building, which up to the time of our visit had always been declared to be the ruin of a Coptic Church. Its walls are about 20 feet high, and inside it was filled with mud, sand, &c., to a depth of about 12 feet. Beyond the southern end of this building we saw several capitals of sandstone pillars lying about, and some very large stones, and we made some small experimental diggings with a view to finding VOL. I. 481 I i