[ 38o ] 



ngain, as never to admit air : alfo the bafe of the ftopple fnufl 

 be broad, that it may be eafily difengaged from its cavity 

 which enlarges the furface or joint that is to be made im- 

 pervious to air .■ this ftopp'.e is in reality a kind of valve, refting 

 on the eritent of its whole conical furface as a bafc, and it 

 was before obferved how difficult it is to make a valve air- 

 tight, vi'hen the fmallcft folid particle of duft, &:c. will prevent 

 the conta(5l of the furfaces though ever fo extenfive, and that 

 the danger of this encreafes with the enlargement of the fur- 

 face ; and that were it not from this natural imperfedlion of 

 valves, Mr. Smeaton's air-pump ought to be inferior to none ; 

 its mode of action being precilely the fame as that of Mr. 

 Cuthbertfon, except that in the latter, metallic valves or flop- 

 pies are fubflituted for the flexible valves in the former : 

 whereas the excellence of a ftop-cock is, that the furfaces of 

 the key and fliell are nevtv feparated : no dirt can get between 

 them, nor oil be blown away ; but fliff ointment may be ap- 

 plied inflead of oil, to make them air-faft : to the ftop-cock 

 therefore, and the unperforatated pifton, I attribute the excel- 

 lence of the pump here defcribed above others confiituted on 

 the like principle ; for in it, though air fhould get in through 

 the upper valve, &c. it cannot get into the rec' ; whereas in 

 the others, if air infinuates through the pifton, it mufl: do fo ; 

 and if to prevent this, in Mr. Cuthbertfon's machine, it were 

 provided to have a pool of oil above the pifton, I imagine 



it 



