124 CcnJlruBkn cf an AW Pump without the Lcwer Valve. 



op when the valve pump is ufed. Inftead, therefore, of placing the three dufls at equat 

 diftances round the cock, I have divided the whole into five equal parts ; leaving the dif- 

 tance of one fifth part between the duds leading from the ciftern and the valves to the 

 cock, and two fifths between each of thefe and the one leading from the cock to the re- 

 ceiver. IJy this adjuftment, when the communication is open between the receiver and the 

 valves for condenfation, the other hole through the cock opens the clflerns to the atmo- 

 fphcre ; but when the communication is made between the cifterns and the receiver for ex- 

 liauftion, a folid part of the key comes againft the dud leading to the valves, and (liuts it 

 up, and the air whicli is forced out of the barrel paflcs through the atmofphere into the- 

 valve pump ; for the valve of the fmall pump may be kept open while the great one is 

 worked. 



Upon this conftruiTtiou alfo we are able to make the pump with two barrels like the com- 

 mon pump, which cannot be done conveniently where the lower valve is retained ; becaufc it 

 would be difficult to make the pifton in one barrel come exadlly to the bottom, at the fame 

 time that the pifton in the other touched as elaflly at the top ; it would at leaft require a 

 nicety in the workmanfliip which would be troublefome to execute. 



In this pump the pillons do not move the whole length of the barrels : there is a horizon- 

 tal fe£lion made in them a little more than half way from the bottom where the top plates 

 are inferred. See Fig. 6. PI. VI. By this means the pump is made more convenient and fimplc, 

 as the head of it is brought down upon tlie top of the barrels in the fame manner as La the 

 common air pump. The barrels alfo ftand upon the fame plane with the receiver plate, and 

 this plane is raifed high enough to admit the common gauge of thirty-two or three inches, 

 to ftand under it w^lthout any inconvenience in working the pump, as the winch moves 

 through a lefs portion of an arch at each ftroke, than it would if the piftons moved the 

 whole length of the barrels. 



There is alfo placed, between the barrels in this pump, on the crofs piece over the 

 valves, a gauge to meafure the degree of condenfation, having a free communication with 

 the valves cock, &c. This gauge is fo conftrudled that it will alfo ferve to meafure the 

 rarefaction above the valves when the air is worked off by the valve pump. It confifts of a 

 pedeftal, which forms a ciftern for the mercury, a hollow brafs pillar, and glafs tube her- 

 metically fealed at one end, which moves up and down in the pillar through a collar of lea- 

 thers. The die of the pedeftal is made of glafs, as well to hold the quicklilver as to expofe 

 Its furface to view, that it may be fcen when the open end of the tube is put down into it or 

 raifed out of it. The body of tlie pillar is partly cut away to expofe tlie tube to view in the 

 fame manner. 



If the pump be ufed as a condenfer, the degree of condenfation is fhcwn by a fcalc 

 marked on one edge of the pillar : if it be ufed as an cxhaufter, the degree of the rare- 

 fa£lion of the air above the valves is (hewn by a fcal, marked on the other edge of the pillar. 

 This gauge will alfo ferve to (hew when the valves have done playing, either with the 

 weight of the atmofphere on them or taken off. If we want to know when they ceafe open- 

 ing with the weight of the atmofphere on them, draw the pifton of the valve pump into 

 its barrel, to prevent any air efcaping through that valve : in this fituation work tlie great 

 pump again ; and if any air paffes through the valves into the pipe, the gauge will rife by 

 iondcnfation. This condenfcd air muft tlien be let out by opening the communication at the 

 5 cock 



