H 



THE AIR-PUMP 275 



r-pump, that is, the larger the barrel, and that the 

 iceiver should be selected for every experiment so as 

 ot to be more capacious than absolutely necessary. 

 It would appear from what has been stated, that the 

 ^haustion of a receiver may be continued to an un- 

 mited degree; but in reality this is by no means the case, 

 id a diminution of density to T-^Q ^th is practically 

 most as much as can be obtained. In order to account 

 r this, it must be remembered that the channels 

 ii the stop-cock, while the latter passes from the 

 osition III. into position II., are filled with air which 

 iis the density of atmospheric air under the usual 

 assure, and that a definite quantity of such air is 

 :ixed every time with the rarefied air in the receiver. 

 .s the exhaustion proceeds, the quantity of air re- 

 :oved at each stroke becomes proportionally less and 

 Iss, and at last the quantity removed will be equal to 

 1 at which is allowed to enter the receiver from the 

 dannels of the stop -cock ; from that instant the degree 

 <' exhaustion evidently cannot be further increased, 

 'he capacity of the channels in the stop-cock is for 

 tis reason often called the c noxious space,' and it is 

 [nerally further increased by the end of the piston 

 ijt fitting exactly in the bottom of the cylinder, so 

 tat a quantity of air of atmospheric density is left 

 llhind in the space between piston and barrel, which 

 aj the next turn of the stop-cock enters the receiver. 

 r ie noxious space should therefore be as small as 

 jlsBible, and a variety of contrivances have been in- 

 "vnted to reduce its influence or to avoid it. By 

 of more complicated machines the exhaustion of 



T 2 



