VOL. LVI.] PHILOSOPHICAL TRANSACTIONS. 309 



main within the bottle. This was done by first introducing the end of the bent 

 tube within an inverted bottle of water, and letting it remain there, till the air 

 driven into this bottle was at least 10 times as much as would fill the empty 

 space in the bent tube, and the bottle containing the marble and acid. By this 

 means one might be well assured, that the quantity of common air remaining 

 within the bent tube and bottle must be very trifling. The end of the bent tube 

 was then introduced within the cylindrical glass, and kept there till a sufficient 

 quantity of fixed air was let up. After letting it stand a few hours, the division 

 answering to the surface of the quicksilver in the cylinder was observed and 

 written down, by which it was known how much fixed air had been let up. A 

 little rain water was then introduced into the cylindrical glass, by pouring some 

 rain water into the vessel of quicksilver, and then lifting up the cylindrical glass 

 so as to raise the bottom of it a little way out of the quicksilver. After having 

 suffered it to stand a day or two, in which time the water seemed to have absorbed 

 as much fixed air as it was able to do, the division answering to the upper surface 

 of the water, and also that answering to the surface of the quicksilver, were ob- 

 served : by which it was known how much air remained not absorbed, and also 

 how much water had been introduced : the division answering to the surface of 

 the water telling how much air remained not absorbed, and the difference of the 

 two divisions telling how much water had been let up. More water was then let 

 iip in the same manner, at different times, till almost the whole of the fixed air 

 was absorbed. As all water contains a little air, the water used in this experi- 

 ment was first well purged of it by boiling, and then introduced into the cylinder 

 while hot. The result of the experiment is given in the following table ; in which 

 the first column shows the bulk of the water let up each time ; the 2d shows the 

 bulk of air absorbed each time ; the 3d, the whole bulk of water let up ; the 

 4th, the whole bulk of air absorbed ; and the 5th column shows the bulk of air 

 remaining not absorbed. To set the result in a clearer light, the whole bulk of 

 air introduced into the cylinder is called 1, and the other quantities are set down 

 in decimals of it. 



Bulk of air let up = 1 . 



Bulk of water let Bulk of air absorbed Whole bulk of Whole bulk of air Whole bulk of air 

 up each time. each time. water let up. absorbed. remaining. 



.322 374 322 374 626 



.481 485 803 SSQ 141 



.082 048 885 907 O93 



.145.. 079 1.030 986 014 



I imagine that the quantities of water let up and of the air absorbed could be 

 estimated to about 3 or 4 1000th parts of the whole bulk of air introduced. The 

 height of the thermometer, during the trial ot this experiment, was at a me- 



