[ 255 ] 



tube, in which it defcended, and Mr. Smeaton's pear-gage ac- 

 curately graduated ; aud when I knew there was none of the 

 above-mentioned caufes to create the great and furprifing va- 

 riations, which at different times occurred in the pump's ex- 

 haufting power, even when the gages agreed fo far that when 

 any one of them indicated a greater or leffer rarefadion, the 

 two others did fo likewife ; I was led to look for the caufes 

 of thefe variations, either in the materials of which the pump, 

 &c. confifted, or in the different qualities in the external air, 

 which was in the recr ; and to fufpedl, that new and perma- 

 nently elaflic air is fometimes generated within the pump, in the 

 working, and fometimes abforbed or fixed again ; and that the 

 quantity of air fo generated, is in certain flates of the atmof- 

 phere and of the pump, fo great, even in a fingle ftroke, as 

 would fill with this faditious air, reduced to the denfity of 

 atmofpheric air, a fpacc . even greater than that of the vacuities 

 under the pil^on, and of confequence to limit the exhauftion 

 of the reC even within what the pump may be fuppofed to 

 efFed, without the addition of the valve and circulating-pipe. 

 Now if fuch produdion of air can be proved to exift, the 

 theory here laid down, may be fo far juft, as the flrudlure 

 of the pump is concerned, and defedlive only with refped to 

 other principles. 



In fourteen feveral trials of exhauftion which I made in the 

 months of July, Augufl and September laft (1795) the air being 



Y y 2 generally 



