242 LECTURE XXVII. 



be extremely inconvenient, since it would collect so much the more copiously 

 as the water in the upper part of the pipe would be more free from pres- 

 sure, and neither of the methods which have been mentioned would be of 

 any use in extricating it. It has been usual in such cases to force a quan- 

 tity of water violently through the pipe, in order to carry the air with it ; 

 but perhaps the same effect might be produced much more easily, by making 

 a small airtight valve in the upper part of the pipe, opening outwards, and 

 a stopcock immediately before it : the stopcock being suddenly turned as 

 often as might be necessary, the momentum of the water in the pipe would 

 probably carry it forwards with sufficient force to throw out the air ; or if it 

 were necessary, external pressure might be added, and the air might even in 

 this manner be discharged by the valve much more readily than without it. 

 But it might be still simpler to have a pretty large vessel of water screwed 

 on to the pipe, which would not be filled with air for a considerable time ; 

 and which, when full, might be taken off and replenished with water. (Plate 

 XXI. Fig. 285.) 



The diameter of a pipe required for conveying a given quantity of water 

 to a given distance may be calculated from the experiments of Mr. Buat, 

 which have been already mentioned. Pipes are usually made of wood, of 

 lead, or of cast iron, but most commonly of lead ; and of late tinned copper 

 has been employed with considerable advantage. A pipe of lead will 

 bear the pressure of a column of water ] 00 feet high, if its thickness be 

 one hundredth of its diameter, or even less than this ; but when any 

 alternation of motion is produced, a much stronger pipe is required, and 

 it is usual to make leaden pipes of all kinds far thicker than in this pro- 

 portion. 



The form and construction of stopcocks and valves are very various, ac- 

 cording to their various situations and uses. Stopcocks usually consist of 

 a cylindrical or conical part, perforated in a particular direction, and 

 capable of being turned in a socket formed in the pipe, so as to open or shut 

 the passage of the fluid, and sometimes to form a communication with either 

 of two or more vessels at pleasure. A valve is employed where the fluid 

 is to be allowed to pass in one direction only, and not to return. For 

 water, those valves are the best which interrupt the passage least ; and none 

 appears to fulfil this condition better than the common clack valve of 

 leather, which is generally either single, or divided into two parts ; but it 

 is sometimes composed of four parts, united so as to form a pyramid, nearly 

 resembling the double and triple valves which are formed by nature in the 

 hearts of animals. A board, or a round flat piece of metal, divided un- 

 equally by an axis on which it moves, makes also a very good simple valve. 

 Where a valve is intended to intercept the passage of steam, it must be of 

 metal ; such a valve is generally a flat plate, with its edge ground a little 

 conically, and guided in its motion by a wire or pin. For air, valves are 

 commonly made of oiled silk, supported by a perforated plate or grating. 

 (Plate XXI. Fig. 286, 287.) 



Before we consider the application of the force of fluids in motion to prac- 

 tical purposes, we must attend to the methods of measuring the velocity of 

 their motions. This may be done either by a comparison with linear mea- 



