138 PHILOSOPHICAL TRANSACTIONS. [aNNO 1/26. 



ground 5 feet below a, except the depending part og, and the ascending 

 part HF. 



When F is stopped, and (a being kept full) the water runs out at d, the 

 quantity of water given is IQ times more than when d is stopped, and the 

 water runs out at p. The air confined in several parts of the long pipe is the 

 chief reason of this difference. 



In order to get rid of the air which, lodging in the pipe, contracts its bore, 

 and thereby lessens the quantity of water, which is to be delivered at the issue, 

 the Doctor made several experiments to find whereabouts the air lodges, the 

 more easily to let it out; one of which was as follows. He took a glass pipe 

 ab, fig. 4, of about one inch in diameter, 12 feet in length from p to p ; only 

 the parts ap and pb, at the other end, were of lead. Then pouring in water at 

 a, till it came up to b, stopping the end g, the air lodged in the eminent parts 

 of the pipe at the places marked cc, dd, and ee : but when the water was 

 suffered to go out at g, the air came forward towards g, and took up the spaces 

 cc, dd, and ee, contracting the bore of the pipe as before, but stood forwarder 

 in the pipe, so that it generally happened that the space of air began on the 

 upper part of the eminence of the pipe. 



The glass pipe may be made of several pieces joined to each other, and to 

 the leaden pipes and funnels, by brass ferrels and elbows, turning in all manner 

 of angles. These are not represented here. When the velocity of the water 

 is very great, the air will go even beyond the eminence of the pipe. 



To let out the air from the conduct pipes, which obstructs the running of 

 the water, Dr. D. recommends the experiments which he made, and the ap- 

 paratus he applied to a wooden conduct pipe, of g inches bore, which runs a 

 mile and a half from the water engine at York-Buildings, to a reservoir near 

 Cavendish- square ; the surface of the water in the cistern at the water-house 

 being sometimes J 5, and sometimes 20 feet, above the issue at the reservoir. 



On a part of the pipe, such as ab, fig. 5, the Doctor fixed a leaden pipe df 

 of 2 inches bore, by means of 3 ferrels, or short communication-pipes, the 

 first at D, just beyond the beginning of the space cc, that used to be filled with 

 air in tlie running of the water, the second in the middle of the leaden pipe, 

 and the third at the end of it; the length of the pipe itself being from 12 to 

 24 feet, according to the steepness of the descent, the shortest pipe being suf- 

 ficient where the descent is very quick. From the middle of the leaden pipe 

 abovementioned, called a rider, from its being laid along on the main or conduct 

 pipe, there goes another leaden pipe as eh, of the same diameter, rising all the 

 way very gently from e to the cock h, and so on to i ; because, if there was the 

 least descent, water would lodge in it. 



