[ 



1 



Accordingly this pump in this ftate (i. c. the valve being taken 

 offj will fo exhauft a rec'' as to make the mere'' in an open 

 tube connedted with it rife as high as and fometimes higher 

 than that in a barom^" of equal bore, placed parallel to it in the 

 fame ciflern, and filled with mercY as carefully as it could be 

 ■without boiling it in the tube ; while the merc^ finks in the 

 mean time in the gage to the level of the external mercX and 

 even beneath it ; and the pear-gage, in the dried ftate of the 

 air and reC (in which condition only of the atmofphere fo 

 great rarefadlion could be effedledi will then indicate a i-are- 

 fadlion of between 2000 and 3000 Such limited performance 

 was nearly what this theory promifed ; and I alfo found the 

 refidual air under the pifton, to elevate the mere)' in the gage 

 fo much at every ftroke, when, on turning the key, it was 

 let into a fmall gage-veffel (no rec" being \ifed) that it was 

 plainly neceffary to leffen the refiduum of air under the pifton, 

 by applying the circulating-pipe, to tranflate the air from the 

 fpace under to that above the pifton ; by which the ufe of a 

 valve, either in this or in the bottom of the barrel, was avoided ; 

 and at the fame time, the top of the pump could be covered 

 with a plate and valve, in order to take off the great wei2,ht of 

 the air incumbent on the pifton, which makes the exhauftion by 

 a fingle barrel (and fo wide a one) too laborious. 



The effedl produced by this addition is, that whatever before 

 was the refidual air under the pifton, is now diminiflied to fuch 



a part 



