1820.] Specific Gravity of Gases. 165 



found by experiment, is somewhat higher than the theoretical 

 specific gravity. Now this sufficiently shows that the experi- 

 mental result is not absolutely correct. If air were a conipountl 

 of 21 volumes of oxygen and 79 volumes of azotic gas instead of 



20 volumes of the former and 80 of the latter as theory iudicatea, 

 then the specific gravity both of oxygen and azotic gas would be 

 somewhat less than the theoretical number ; for the specific 

 gravity of air is always unity. Therefore (supposing the specific 

 o-ravity of oxygen = x, and that of azotic gas =■ y) 2\ x + 79 v/ 

 = 100. Let us suppose the specific gravity of oxygen gas to 

 be MIO; then the equation becomes 21 (I'll) + 79 j/ = 100 

 or 23-31 + 79 j/ = 100, from which it is easy to deduce y — 

 0-9700 nearly. It is clear from this, that if air be a mixture of 



21 oxygen + 79 azote, the specific gravity obtained by experi- 

 ment, supposing it absolutely correct, v;ould have been below 

 1-1111 and 0-97"22 instead of above it. Nobody who considers 

 these facts can have any hesitation in concluding that there was 

 an error in our experiments, and that the weight of each gas 

 turned out a little higher than the truth ; but the experiments 

 were the most accurate that my apparatus would admit ; and 1 

 am of opinion that they sufficiently confirm the specific gravities 

 of oxygen and azote as pitched upon by Dr. Prout from theore- 

 tical considerations. Henceforth then chemists need be under 

 no hesitation in fixing upon Mill for the specific gravity of 

 oxygen gas, and 0'9722 for the specific gravity of azotic gas. 



2. Hydrogen is the lightest of all the known gases. 1 have 

 satisfied myself by many experiments that carburetted iiydrogen 

 gas, the next in point of lightuess to hydrogen, is very nearly 

 eight times as heavy as this last gas. This great disparity 

 between the v/eight of hydrogen and the other gases renders it 

 much more difficult to determine its specific gravity thiui that of 

 any of the other gases ; because the least mixture of any of the 

 heavier gases occasions a material difference in t'ae result. 

 Thus, for example, a mixture of 99 volumes of pure hydrogen 

 and 1 volume of common air will have the specific gravity 0-0787, 

 which is more than one-eighth part higher than the true result. 

 A mixture of 1 volume of air with 199 volumes of pure hydrogen 

 has the specific gravity of 0-074, which is the result obtained by 

 Biot and Arrago. Hence I am disposed to suspect that the 

 hydrogen gas upon which they experimented was in reality con- 

 taminated witli -T-i^-c^th of its volume of common air. TI;o pre- 

 sence of a little carburetted hydrogen gas would act precisely in 

 the same manner, though not quite so powerfully as air. If tho 

 hydrogen gas examined by Biot and Arrago contained a mixture 

 of one per cent, of carburetted hydrogen, its specific gravity 

 would have been at least as high as they found it, or 0'074. 



The necessity of operating upon extremely pure hydrogen gas, 

 and the tediousness which necessarily attends ail experiments 

 which aim at precision, will account for the length of time wliicli 



