EXPERIMENTS WITH THE AIR-PUMP 287 



expands and forces the liquid contents through the 

 orifice, 



Water contains always small quantities of air in so- 

 lution ; under the receiver of the air-pump, when the 

 rarefaction is great, this air separates from the liquid 

 in small bubbles, which collect at the sides of the 

 vessel. 



Liquids, like soda-water, which contain some gas 

 (carbonic acid gas) in solution, manifest much stronger 

 effervescence under the receiver of an air-pump, because 

 the gas escapes under a diminished pressure. 



The experiment with the Cartesian diver may be made without an 

 air-pump, for it requires only a slight rarefaction, if the volume of 

 water in the diver is so adjusted as to make it just sink. The 

 diver is simply placed in a bottle with water, and. the air sucked out 

 by the mouth ; or, more conveniently, the bottle is fitted with a cork, 

 a tube is passed through it, bent at right angles, and the mouth 

 applied at the end of the tube. 



Daring the experiments with water and watery liquids, some 

 aqueous vapour always gets into the pump ; if the exhaustion is 

 tfreat, the water will even boil, for a reason which will be explained 

 further on. These experiments should therefore follow each other 

 in succession, and the pump may then be dried. This may be 

 ione without taking the pump to pieces by placing a shallow 

 saucer full of sulphuric acid and the gauge under the receiver, 

 ind exhausting the latter as far as possible ; the stop-cock is then 

 )rought into the position II., and left for an hour, during which 

 lie piston is repeatedly moved to and fro. The water deposited 

 n the walls of the cylinder and in the channels of the stop- cock is 

 hereby made to evaporate, and is absorbed by the acid. 



It has been stated in art. 10, p. 59, that all bodies 

 all with equal velocity in vacuo. This may be demon- 

 trated either by means of a long glass tube, of wide 

 ore, closed at one end with an air-tight cap of brass, 

 nd at the other by a brass collar, to which is attached a 

 top-cock and a nut for fixing the whole upon the plate 



