Most of us were weary, and many preferred lying on the floor of the corridors or vestibules at the end of the cars, to sitting straight up in the cramped compartments. We made several halts, and it was near midnight when we arrived. Our guards, a few gendarmes, seemed to have no idea where we were going, or what was to be done with us. Eventually we were told to leave our kit, which was to be brought along later, and were guided down towards the river. After walking a mile, we found ourselves in a small Arab village on the river bank, and were conducted into a courtyard some 40 yards square, where we were told we were to stay. There was a rough shelter round three sides, formed by brushwood supported on a rough wooden framework; this promised a certain amount of shade, and we were all glad to be in the open air rather than in another barrack building. There were no signs of any transport fetching our kit, so the most enterprising managed to procure two trollies, and trundled them up to the station along a narrow-gauge line. The Turks used this line for taking stores, ammunition, etc., to the railway, from the rafts on which they were floated down from Mosul. By dawn, nearly all the kit had been collected, and we had settled down as best we could. There was a certain amount of food obtainable from Arab vendors, and as we had our Indian servants, and a few things left from stores received at Shamran, we were fairly comfortable. As usual, no one seemed to know how long we were to be there, before our journey by road across the desert began. Fortunately, we were not guarded very strictly, and were allowed to go outside the courtyard, and down to the river to bathe; the current here was very strong, and only the most powerful swimmers could make any headway against it, and that only for a few yards. The town of Samarra was on the other bank, and some little height above the land on our side. It stands back from the river, and contains a fine mosque, with a golden dome. The inhabitants cross the river in gufahs--the large round coracles which are used all down the Tigris. Owing to the current a start always has to be made very much higher up-stream than the point where it is desired to land on the other side.