prison for them; failing money, friends and relatives, the prisoners starved to death. I have already said that the best and greater part of the food sent to the prison gates was appropriated by the gaolers, that is to say, after Idris es Saier had seen to the wants of his "starving children" and numerous household first. Idris, even during the worst period of the famine, did not lose flesh; he was always the same tall, stout, flat-nosed black, both when I first saw him on May 10, 1887, and when I last saw him in September, 1898. Nor was Idris quite so bad as he had been painted; he would often — when the Nebbi Khiddr tale had had the desired effect in repentance, or when he was in a good humour after a bout of marrissa drinking — go out of his way to do his prisoners small kindnesses, such as the removal of extra chains, and giving permission to sleep in the open; but the Nebbi Khiddr institution left him so much at the mercy of the Khaleefa's immediate attendants, that his periods of good humour were, in consequence, of very short duration. Some day, if I return to the Soudan, or Idris pays a visit to civilization, I may learn from him whom I have to thank for a few of the unnecessary hardships inflicted upon me.
It might be asked why we, knowing that the guards would purloin the greater part of the food sent in, did not arrange for a larger quantity to be sent. There are two reasons, and the first is the least of the two: the guards knew very well what was the minimum amount of food to keep us alive, and just that quantity of food
