.' 354 ] 



pin on defcends, frequently (in my trials) fo much as the 50th 

 part of the whole capacity of the cyl"", the contents of the 

 circulating-pipe of air muft be fuppofed as encreafed to that 

 quantity ; and the fccond ratio above mentioned, of the cylr 

 to the pipe, muft be taken as that of the cyl"^ to the 50th 

 part of its own capacity ; and according to this, the utmoft rare- 

 faction which could even by this theory be produced, would 

 be about 176500 times ; and this the pump would produce, 

 fuppofing the pifton to throw up all the air above it, and no 

 external air to intrude unlefs fome other principle operated to 

 prevent it Alas ! it falls far fhort of this perfedlion ; and as I 

 have never found any reafon to fufpefl, that the pifton left 

 air behind it adhering to the cyK, nor that any intruded by 

 cocks or joints, in fuch quantity as to occafion fo fad a defedl 

 in its performance l^but the contrary as will appear hereafter) 

 this defedl cannot, as I conceive, arife from the mechanifm of 

 the pump, but from other caufes. 



It is well known how much any moifture, or any fubftance 

 capable of generating a fupply of aeriform fluids, either per- 

 manently elaftic under the ordinary prefTurc of the atmofphere, 

 or fuch only in vacuo, will prevent exhauftion, and occafion 

 a contradiction in the gages, by which its degree is indicated. 

 I have confidered and tried their various effedls, having three 

 gajes applied to the pump together, viz. a common ftrait ba- 

 rom''-tube, in which the mcrcv afcended, the reduced baromr- 



tube. 



