990 COTTAGE, FARM, AND VILLA ARCHITECTURE. which throw the water up or down, into a variety of forms ; such as fans, parasols, sheaves, showers, mushrooms, inverted bells, &c., or (and which is one of the newest forms) the convolvulus of Mr. Rowley, as shown in fig. 1767. The larger the conduit pipes are, the more freely will the jets display their different forms ; and the fewer the holes in the quill or jet (for sometimes this is pierced like the rose of a watering-pot), tin greater certainty there will be of the form continuing the same ; because the risk of any of the holes choking up will be less. The diameter of a conduit pipe ought in no case to be less than an inch ; but, for jets like those in the preceding figures, the diameter ought to be two inches. Where the conduit pipes are of great length, say upwards of 1000 feet, it is found advantageous to begin, at the reservoir or cistern, with pijies of a diametei somewhat greater than those which deliver the water to the quills, because the water, in a pipe of uniform diameter of so great a length, is found to lose much of its strength, and become what is technically called sleepy ; while the different sizes (juicken it, and redouble its force. For example, in a conduit pipe of 1800 feet in length, the first 600 feet may be laid with pipes of 8 inches in diameter the next 600 feet with pipes of 6 inches in diameter, and the last 600 feet with pipes of 4 inches in diameter. In conduits not •jxceeding 900 feet, the same diameter may be continued throughout. When several jets are to play, in several fountains, or in the same, it is not necessary to lay a fresh pipe from each jet to the reser- voir ; a main of suflSeient size, with branch pipes to each jet, being all that is required. Where the conduit pipe enters the reservoir or cistern, it ought to be of increased diameter, and the grating placed over it to keep out leaves and other matters which might choke it up ought to be semiglobular or conical ; so that the area of the number of holes in it may exceed the area of the orifice of the conduit pipe. The object is, to prevent any diminu- tion of pressure froin the body of water in the cistern, and to facilitate the flow of the water. Where the conduit pipe joins the fountain, there, of course, ought to be a cock for turning the water off" and on ; and particular care must be taken that as much water may pass through the oval hole of this cock as passes through the circular hole of the pipe. In conduit pipes, all elbows, bendings, and right angles should be avoided as much as possible, since they diminish the force of the water. In long conduit pipes, air-holes formed by soldering on up- right pieces of pipe, terminating in inverted valves or suckers, should be made at convenient distances, and protected by shatis built of stone or brick, and covered with movable grat- ings, in order to let out the air. Where pipes ascend and descend on very irregular surfaces, the strain on the lowest parts of the pipe is always the greatest ; unless care is taken to relieve this by the judicious disposition of cocks and air-holes. Without this precaution, pipes conducted over irregular surfaces will not last nearly so long as those conducted over a level. We shall here add a Design by Solomon Cans, fig. 1766, which may be described as a conceit, and by no means in what may be considered the best taste. Conceits, however, are sometimes admissible, since they can be enjoyed by those who have not yet arrived at a just feeling for the simple and grand. 1976. T/ie perpendicular hcu/ht to which loater tvill rise in a jet has a limit, depending on the diameter of the jet, and on the specific gravity of the water, and on that of the air which it has to penetrate. A jet of salt water will rise higher than one of fresh water; a column six inches in diameter higher than one of three inches ; and a jet of water of any dimension, higher at Madrid or Munich, than in Paris or Ijondon, on account of the dif- ference of the elevation of those cities above the level of the sea, and the consequent
