First, to deal with the green water. Generally when the rise of an inch or two is reported from the nilometer at Rôda, the waters lose the little of clearness and freshness they still possessed. The green colour is the slimy, lustreless hue of brackish water within the tropics, and no filter that has yet been discovered can render such water clear. The colour is really due to algæ. Happily, the continuance of this state of the water seldom exceeds three or four days. The sufferings of those who are compelled to drink it in this state, from vesical disease, even in this short interval, are very severe. The inhabitants of the cities generally provide against it by Nile water stored in reservoirs and tanks. Colonel Ross, R.E., noticed in 1887 and in 1890, when, owing to the slow retreat of the Nile, the irrigation officers had to hold back many basins in the Gîzeh province, and also in 1888 when the water remained long stagnant, that the basin water got green--showed the algæ and smelt marshy--just as the June green water does. Hence it has been argued that, as the Nile water in the bed of the stream--even in very slow-flowing back-waters--does not become green, the greenness must be produced by an almost absolute stagnation of the water. We know of great marshes up above Gondokoro, and hence it is thought that the green water of summer, which comes on suddenly, is this marsh-water being pushed out by the new water from behind, and that is why it heralds the rise. No one has so far minutely observed the gradual intrusion of the green water. The rise of the river proceeds rapidly, and the water gradually becomes more turbid. Ten or twelve days, however, elapse before the development of the last and most extraordinary of all the appearances of the Nile, thus described by Mr. Osborn[66]:--