Defcriptim of a nenv Air-pump. 507 



TCffel was then clofed at the top with melted cement ; and as It communicates by the pipe 

 E with the cortduit-pipe, it is exhaufted with the receiver, and the mercury in the tube will 

 fink- as the rarefaction advances : if a perfedl vacuum could be made in the receiver, and 

 gage-veflel, it would fmk more or lefs beneath die level of the furrounding mercury, as 

 the tube is narrower or wider, even fo as to difappear if the tube were very narrow ; 

 fo that it muft not be lefs than Ti,ths of an inch in diameter j but it would be better 

 if it were -rstlis. If it were obferved, how much the mercury in a tube open at both 

 ends would fink beneath the level of that in a ciftern in the open air, it might be pretty 

 nearly eftimated how great is the rarefaftion in this gage-veflel, and confequently in 

 the receiver*. 



In order that each of tlie dufts here defcribed fliould alternately communicate to tlie bar- 

 Tel, it is neceflary, that the cock fliould be moveable through a quarter of a turn, and confined 

 to that motion. This is efFe£ted, by a pin, «, fixed perpendicularly in the back of the handle, 

 fig. I . which, when the handle is fixed in its place, defcribes a quadrantal arc, imder the ftiell 

 of the cock, where there is jufl room to allow fuch a motion. 



The aftion of this pump may be readily apprehended from the conftruftion of its parts. 

 When the pifton is at the bottom of the barrel, and the key of the cock turned to its limit 

 which opens the communication, between the barrel and the receiver and at the fame time 

 fhuts that with the circubting pipe •, the pifton being then drawn to the top of the barrel dif- 

 charges the air through the valve, while other air from the receiver ruflies into the barrel, and- 

 follows the pifton. As foon as the pifton has arrived at the upper, or more remote end of the 

 barrel, the key of the cock is again turned, and fliuts off the connexion with the receiver, and 

 opens that with the circulating pipe. The pifton being then depreffed, drives the air through 

 the pipe, into the upper, or remote end of the barrel. When the pifton has arrived at its 

 limit, the fljoke is ended, the key of the cock is to be again turned to open the receiver, and 

 a fecond ftroke may be made, with the feme confequences as the firft ; and in this manner the 

 procefs may be continued at pleafure. The limit of exhauftion will be when the air in the 

 receiver fliall have become as rare as the air in the barrel would be when the pifton is up; 

 fuppofing no communication to have taken place between it and the receiver. 



When the pump is required to condenfe, the air is made to move in a reverfe direftionJ 

 The cock of the gage-veflel is flopped, that it may not be burft ; the valve is taken off"; the 

 communication of the bottom of the barrel, to the circulating pipe, opened by tlie key of the 

 cock, and the pifton raifed to the top. The external air pafles into the lower part of the 

 barrel; and by turning the cock, fo as to open the receiver, and fliut the circulating pipe, and 

 then deprefllng the pifton, the air is forced into the receiver, wliich ought of courfe to be of 

 fufiicient ftrength, and properly fccured. By fliutting the receiver, raifing the pifton, opening 

 the receiver again, and returning the pifton to its former fituation, another ftroke is made, and 

 thefe alternations may be carried on to any deCred or praftlcablc extent. 



• The author prefers the ftiort barometer-gage to the fyphon-gage, beeaufe in this !aft he thinkt the 

 motion IS impeded by the tube, and tbe mercury becomes foul on that fide which communicates with 



the receiver N. 



The 



