1 :a CorifrirfHon of an Air Pump without the Lctoer Valvt. 



like Smeaton's, it forces out all the air above it ; and, as the air cannot return into the bar- 

 rel on account of the valve on the top pl.itc, wlicn the pifton (lefcends, there will be a va- 

 tuum formed betsveen that and the plate, every thing being fuppofed perfcft. But in 

 \iorking the pump, the pillon is not allowed to defccnd entirely into the cittern Co far as to 

 leave the bottom of the barrel open ; becaufc, as the ciftern, for anotlicr purpofe, is made 

 Jaiper thin the bore of the barrel, this might make tlie pillon-rod work unfteadily in the 

 collar of leathers, and caufc it to leak; but it dcfcends below a hole in the fiJc of the 

 barrel near tlje bottom, which opens a free communication between the barrel, ciftern, and 

 receiver. 'Ihruugh this hole the air rufhes from the ciftern into the exhaufted barrel, when 

 the pifton has dropped below it ; and by its next afcent tllis air is forced out as the other 

 was before. If now, the capacity of the receiver, ciftern, pipes, &c. below the bottom of 

 the barrel, taken together, be equal to the capacity of the barrel, half the remaining air will 

 be expelled by every ftruke. 



But as the working a pump of this kind with a folid pifton would be laborious, on ac- 

 count of the refiftance it would meet with in its defcent fiom the air beneath (though 

 this would be leffcned by every ftroke as the air became more rarefied), the inventor, to re- 

 medy this inconvenience, pierced three holes in the pifton at equal diftances from each 

 other 1 and a circular piece of bladder, which is tied over the top of the pifton, to make the 

 joint more perfeft with the top plate, and to defend them from injury when the pilton is 

 brought up againft it, forms a kind of valves over the holes, which open eafily enough to 

 prevent any labour in working the pump, as it allows the air to pafs through the pifton when 

 k defcends. But the air does not'necelTarily depend upon a paftage through the pifton in 

 order to get into the barrel : for when the air becomes fo weak from its rarefa£lion that it 

 tinnot open this valve, it will ftill get into the barrel when the communication is opened by 

 the hole at the bottom. This pifton, therefore, will defcend as eafdy as any other ; and thefe 

 valves do not impede the rarcfa£tion, (ince it is of no confequence, as to this, whether they 

 open or not. By this conftruifkion the valves, which Mr.Smeaton only made to open with 

 more eafe, arc rendered unneceflary in rarefying the air, and that at the bottom of thebarrelj 

 which is the moft difficult to be made and kept in order, is entirely removed ; the valve on 

 the top plate being the only one necelTary in rarefying the air. 



Bat as in a fmglebarrelled pump of this conftruclion, where there is no valve at the 

 bottom to prevent the air, which follows the pifton in its afcent, from returning into the 

 receiver in its defcent, a fluftuation would be produced which might prove detrimental in 

 feme experiments, this pump is made with two barrels, which rarefy the air at every ftroke 

 of the winch. In this conftru£tion, the capacity of the two barrels taken together belowr 

 the piftons is always the fame; for while one is defcending the other afcends, and what 

 is taken from the one is added to the other. 



Having thusfet afide the valves which in fome meafure prevented the air froiYi entering 

 the barrel above the pifton, he next attempted to expel the air more perfcftly out of the 

 barrel than Mr. Smeaton had done, by making a better vacuum between the pifton and the 

 top plate, which would allow more of the air to expand itfelf into the barrel from the re- 

 ceiver. 



Upon Mr. Smeaton's plan, our author alfo contrived to conneft the valves on the top 

 plates with the receiver occafionall/jby means of a pipe and cock, by the turning of which the 



machine 



