explanation ; — when imruus ima u\\^x:iivjas 

 gas are in the two phials, the residuary gases 

 after the experiment are nearly as pure as 

 before ; because those portions of them that 

 me^t in the tube, form nitrous acid vapour, 

 which is absorbed by the moisture in the 

 phials, and therefore docs not contaminate 

 either gas. 



1 . Two 1 ounce phials were connected with 

 the small tube, the under containing nitrous 

 gas, the upper atmospheric' air ; after three 

 hours, the upper phial was taken off when 

 a quantity of air was perceived to enter, as 

 was expected ; the air in the upper phial was 

 scarcely distinguishable from what it was at 

 first ; that in the under phial was still so much 

 nitrous as to require its own bulk of common 

 air to saturate it. 



2. The above experiment was repeated, 

 and the upper phial drawn off when the 

 whole was under water, in order to prevent 

 communication with the atmosphere : about 

 i' of an ounce of water entered the phials, to 

 compensate the diminution. Remaining air in the 

 upper phial was a very little worse than com- 

 mon air, it being of the standard 1,47 when 

 the former was 1,44. The gas in the under 

 phial was still nitrous and nearly of the same 



l1 



