At the end of 1886 the work was begun, under the charge of Colonel Western and Mr. Reid, sent specially from India for the purpose. The operations were spread over four years. In the first year the left half of the Rosetta Barrage was taken in hand and finished before the flood, next year the right half. In 1889 and 1890 the Damietta Barrage was similarly taken in hand and finished. Each year the part to be repaired was enclosed by earthen dams, and the water pumped out so as to lay the foundations dry. The whole of the existing floor was raised, both on the upstream and downstream side, and it was also considerably lengthened. It was, in fact, enclosed in a new and reliable suit of armour. The dangers and anxieties of the work were incessant. The protecting dams were always liable to be breached. Spring after spring burst out through the treacherous bed of the river, and threatened the destruction of the year’s work; and again and again each of them was successively stopped by a number of ingenious devices. There is no enemy so persistent and so insidious in its attacks as running water. It is always feeling for and finding out the weak spots. It never sleeps or slackens by day or by night. It can only be met successfully by a corresponding activity. While work was possible, it was carried on unceasingly by night as well as by day. Sometimes as many as 1,600 men worked through the night. The upper brickwork was generally sound, but new iron sluice-gates moving in special grooves were fitted throughout. The whole of the repairs cost £465,000. It was money well laid out. Not only was the safety of the Barrage assured, but it was found possible to hold up yet another metre of water. The area of summer crops matured rose once more from 1,200,000 to 1,520,000. It might have been thought that the work was now complete. Both in 1891 and 1892 (a year of specially low summer supply) all the water in the Nile was held up, and diverted into the canals. Not a drop reached the sea during the summer without having done duty. But the engineers were now looking forward to a time when the supply would be greatly increased. The idea of a reservoir had become an affair of practical politics. It was necessary to make assurance doubly sure. Accordingly, in 1896 a new experiment was tried—namely, stock-ramming with clay.