[ 56o ] 



deprefled the fame number of times as before, and the com- 

 munication of the barrel and gage was opened, the mercy was 

 always found to ftand lower than it did at firft before the pifton 

 had been worked ; fhcwing, ss it fhould fcem, that there had 

 been an adual fixation or abforption of air ; but when another 

 elevation of the pifton had been made, the recr-pipe being open, 

 then the mere)' ftood much higher than at firft ; as if the air 

 abforbed, had been regenerated with increafe ; or as if by the 

 working when the barrel had been clofed, the ingredients for 

 new air had been preparing in it, when a little air from the 

 reC fhould be admitted into it, as a generative ferment. 



In all thefe circumftances there was no communication be- 

 tween the external air and that under the pifton or in the 

 recr ; and the appearances took place uniformly and repeatedly } 

 and the experiments were all repeated together on different 

 days, with long intervals, and fometimes in a reverfed order: 

 fo that whatever may have been the caufe of the phasnomena, 

 it could not have been the irruption of any air ; this could 

 never produce thefe regular effeds ; it could only intrude by 

 the pifton ; in which cafe it would not always be the fame 

 in quantity, but it would always make the mercT afcend in the 

 gage ; whereas in the fecond cafe juft mentioned, it did not ; 

 and in the third cafe, it always funk in it : moreover as the 

 top of the barrel was in all thefe cafes open (the valve having 



been 



