And still the engineers were not satisfied. So treacherous is the river’s bed that no possible safeguards seemed superfluous. It speaks volumes for the courage and skill of those who in 1885 held up 3 metres of water with the old unreformed Barrage, that in 1897, after the successful execution of such great and costly repairs, it was still thought advisable to undertake completely new works to assist in the task of holding up 4, or at the most 5, metres. It is a principle in hydraulics, not easily understood at first by the layman, that the pressure upon a weir or barrage in a river depends entirely upon the difference in level between the water on the upstream and on the downstream side, and not on the mere volume of water in the river behind it. In December, 1897, the Caisse de la Dette voted £530,000 for the construction of two subsidiary downstream weirs, with the object of relieving the pressure on the Barrage by raising the level of the water on the downstream side, thus dividing the head of water to be held up into two—in other words, by making two steps instead of one. Each weir was to consist of a core of rubble masonry set in cement, sunk well below the bed of the river, and protected up and down stream by a long slope of rough stone blocks or pitching. To make the masonry core thoroughly watertight, a mass of clay puddle was to be put on either side of it. The weirs were thus to be a solid dam, blocking the course of the stream up to such a height that the head of water on the Barrage, at that time amounting to 4 metres, would be reduced to 2·5 metres. The flood would pass freely over the top of the weirs. At the same time the sluice-gates of the existing Barrage were to be heightened, so as to permit the upstream level to be raised 1 metre more in June and July, so as to take full advantage of the rising flood and facilitate the early sowing of maize, a great point with the Egyptian cultivator. By the summer of 1900 this programme had been completed. The building of the weir on the Rosetta branch was an especially fine performance, for which great credit was due to Sir Hanbury Brown and Mr. Brooke, who were in charge of it. Five hundred metres in length, it was begun at the end of December, 1899, and actually finished before the flood began to come down in July. The same Portland cement played a great part in its construction. It may now fairly be said that the Barrage is complete at last, and fully equal to every strain that it can be called upon to bear.