of course the water for the engine well must be conducted across the river. The difficulty of such a transmission, at first sight so great, was obviated by the genius of Watt, to whom application was made. 16 62 PRELIMINARY ESSAY. " This celebrated engineer devised a flexible main of iron pipes, so connected as to adapt its form to the bottom of the river. That part laid across the bed of the river, consisted of pipes, nine feet in length, exclusive of the joints, and having a diameter of 15 inches. Some of the joints were formed in the usual manner, others were made similar to those commonly called ball and socket, or universal joints.* By means of these, the whole train of pipes being properly and firmly connected, the main was laid upon massive wooden frames, consisting of logs laid parallel, and joined together by very strong iron hinges. The pipes and frame were put together on the south side of the river, and the end of the pipe intended for the north side, was stopped with a plug, when a trench having been previously prepared to receive them, by the assistance of machinery, the flexible conduit, with its bed of wood, was hauled across the river, the moveable joints of the pipes, and the hinges of the frames allowing the whole range to assume the form required by the bottom of the Clyde. The operation was aided by pontons, and the machinery employed to haul it into its place, was of course, fixed on the north side of the river. When the plugged end of the main emerged from the current, it was imme- diately opened and connected with the pipe leading to the reservoir of the forcing pumps. For the purpose of protecting this main from any injury from vessels passing along the Clyde, the whole of the part under water was covered with gravel and stones. This ingenious and useful contrivance was executed in 1810 — but one main being found insuf- ficient for the demand on the works, in 1818, another of 28 inches diameter, arid subse- quently a third of 36 inches, was laid across the river. The two last conveyed filtered watered only."t There are six engines at Dalmarnock, which propel the water across the Clyde, which is there about 100 yards broad, into the reservoirs in the city of Glasgow, to a height of 157 feet above the level of the river. The houses are generally very high, * Hydraulia, p. 149. t An early instance of this mode of transmitting water across a navigable river is recorded by M. Gautier, a French engineer, in a work published in 1778, upon the construction of roads. M. Gautier had been employed to devise means of supplying the harbor of Rochefort with good water. He discovered a copious source on the side of the Charente, opposite to that on which the town stands. He proposed accordingly, to bring it across by iron pipes laid on the bottom of the river, protected by wooden frames against the risk of accident from the anchors of vessels. The project was rejected as impracticable or inexpedient. " Some years after," says M. Gautier, " when I had charge of the roads on the Rhone, and other works in Languedoc, while at Aries, I heard that a vessel had cast anchor in the Rhone, opposite the city, but when the commander wanted to sail again, he could not raise his anchor. This circumstance attracted much attention ; and the captain, unwilling to lose his anchor, sent down a man to find what was the matter. The diver reported that the anchor was hooked under something round, but he could not tell what it was. By aid of a capstan the anchor was raised, and brought up a leaden conduit pipe from the bottom of the Rhone, which crossed it from the city of Aries towards Trinquetaillade, at a depth of 42 feet, and where the river is some 550 feet broad. I saw some pieces of this conduit of lead, five or six inches in diameter, about one third of an inch thick ! in joints of six feet, soldered lengthwise, and covered by a strip or sheet of lead of the same thickness, covering the first solder about two inches. The conduit was soldered at the joints, six feet apart, by the same material, which made a swell at that distance. On each joint were these words in relief, C. CAINTIUS POI HINUS. P., apparently the name of the maker, or of the architect who laid down .the pipes in the time of the Romans. My project of laying pipes along the bottom of the Charente, would not have been half so difficult, PS it had no