1894 49,000 High flood. 1895 37,000 Medium flood. 1896 26,000 „ 1897 11,000 Low flood. 1898 19,000 High flood. 1899 8,000 Very low flood. 1900 14,000 Low flood. 1901 9,000 „ 1902 5,000 Very low flood. The figures tell their own tale. Experience and good organization have enormously decreased the number of men called out, so that it is now but a very slight burden upon the people. The date of calling out has also been altered to August 15 instead of August 1. The possession of the Soudan, and an accurate record of the state of the gauges in those regions, will also assist in making more accurate forecasts of the nature of the flood. In 1900 more men were called out than were required, because the flood came down early and promised to be a high one, but failed to fulfil expectations. The levels in the Soudan were known, but, there being no previous experience to judge them by, no inferences could be drawn from them. But the Soudan readings will be more and more useful as time goes on. But even apart from this, there is great hope of a steady diminution in the numbers called out, and even that, except in years of high flood, the corvée may not be required at all. The record of the year 1901 is of remarkable promise in this respect. North of Cairo no corvée was called out at all, for the first time in the history of Egypt. The flood was low, but not exceptionally low. In any year Upper Egypt is responsible for far the greater number, and this is largely due to the extent of the basin banks which have to be guarded. It is remarkable that the two districts in which the greatest number were called out were in charge of native Egyptian inspectors, who were no doubt influenced by the old tradition that vast numbers of men should be employed. At any rate, as perennial irrigation increases in Upper Egypt, fewer men will be required, concurrently with the disuse of the basin dykes. And it seems likely that in years to come the whole task will be performed by contract labour; though the power of calling out the corvée will always be held in reserve in case of any specially dangerous flood. But in such a case the difficulty will be rather to prevent the work from being hindered by excessive numbers. There will be no doubt of the willingness to serve of practically the whole population. CHAPTER VII RESERVOIR PRELIMINARIES