[ 382 ] 



the others, whatever air is left above the piflon undifcharged, 

 or enters by the valves in the piflon or top of the pump, is 

 difFufed in the cyl' under the piflon (fuppofe it could not get 

 into the rec') ; and the piflon being put down, whatever cavi- 

 ties are under it, have the fime quantity of it rcfl:ing there, 

 when the rec' is opened, as in this pump ; and if air infinu- 

 ates by the perforated piflion, it is an inconvenience from which 

 this is fecure ; alfo it condcnfes with a Angle barrel, and with- 

 out any additional parts, which none other can do. It might 

 be conilrucled with two barrels, and the two cocks might be 

 turned together by one handle, in a manner well known and 

 in ufe ; but in this form, it would neither be convenient nor 

 portable ; however in the form here defcribed, the fingle barrel 

 may be made fo large, as to exhaufl: as fafl: as two of ufual 

 fize • and it is eafy to conceive from the theory, that the large- 

 nefs of the barrel is every way an advantage, for the cavities 

 under the piflon need not be proportionably enlarged. 



As it is fcarely poiTible to prevent the intrufion of fome 

 air while the pump is working, and as the quantity fo in- 

 truded, will be proportional to the time of exhaufl;ing, which 

 time would be injudicioufly lengthened, if the paufe at each 

 ftroke were made longer than is neceflkry for the air to pafs 

 from the reC to the cyU until its denfity is the fame in both 

 in any degree of the rarefaclion ; it would be a defirable 



thing 



