Chap. 9.] The Air Pump. 425 



the valves, and then no more can be taken our of the 

 receiver. Hence it appears, that there is no fuch 

 thing as making a perfe<5t vacuum in the receiver ; for 

 the air that leaves the receiver is driven out by that 

 which remains behind, and there muft therefore fome 

 portion remain behind at laft. 



Such is the conftruclion and nature of the air-pump. 

 Some inftruments at firft contrived only for explain- 

 ing fcience, become at laft, by freqtrent ufe, a part of 

 the fcience icfelf, and demand an equal explanation. 

 Such is the cafe with this; and the reader muft pardon 

 fome prolixity in the defcription. There is a cock k 

 below the plate LL, which being turned lets air into 

 the receiver again. There is a glafs tube Imn open 

 at both ends, and about thirty-four inches long, the 

 upper end communicating with a hole in the pump 

 plate, and the lower end immerled in quickfilver at n in 

 the veffe,! N. To this tube is fitted a wooden ruler 

 m m, divided into inches and parts of an inch from the 

 bottom at #, where it is upon a level with the furface of 

 the quickfilver, and continued up to the top, a little 

 below /, to thirty or thirty-one inches. Now the 

 quickfilver in this tube rifes as the air is exhaufted in 

 the receiver, for it opens into the receiver through the 

 plate L L. And the more the air is exhaufted, the 

 more will the quickfilver rife, fo that by this means 

 the quantity of air pumped out of the receiver may be 

 very exactly meafured *. 



From all the preceding facts, and efpecially from 

 the experiment of Torricelli, it appears, that air is a 

 PONDEROUS fluid ; in other words, that it pofiefTes 

 gravity, and its weight may be eafily afcertained. 



From a large phial (or rather from a flafk, or any 

 glafs veflel of a globular form, for reafons that will 



* Goldfnjith's Philofophy, vol. ii. p, 56. 



afterwards 



