the Diffusion of Gases. 253 



The Density of any gas diffused into air, both being in the 

 same state as to aqueous vapour, is obtained by the formula 



where G is the volume of gas submitted to diffusion, and A the 

 volume of return-air. In operating upon gases lighter than air, 

 the most useful instrument is a bulb of about two inches in dia- 

 meter blown upon half-inch tube, of which about an inch may 

 be left on either side of the bulb. The capacity of the instru- 

 ment, used as a gas-receiver over water, is most simply deter- 

 mined by filling it with water, and weighing the water which it 

 contains, and which can be poured from it into a counterpoised 

 phial. Then, after any experiment, the return-air may be found 

 from the weight of the water which has entered the instrument, 

 determined in the same manner. By proceeding in this way, 

 we avoid wetting the stucco after every experiment. A hood of 

 damp paper may be inverted over the upper tube while the dif- 

 fusion is going on, and the whole counterpoised in a tumbler of 

 water, being suspended from one of the arms of the beam of a 

 balance, the scale on that side being removed. An experiment 

 with the bulb will generally occupy several hours. But with a 

 plain diffusion-tube, a much shorter time will suffice. 



A peculiar advantage of this mode of taking the specific gra- 

 vity of gases, besides its simplicity, is, that we can operate upon 

 a most minute quantity of gas : it is possible to come within 

 100th of the specific gravity, operating upon no more than one 

 cubic inch of gas. 



It is to be regretted that this method is not so fully available 

 in the case of coal-gas, as might be expected. The density of 

 that gaseous mixture appears to depend, in no inconsiderable 

 measure, upon the presence of a small quantity of the heavier hy- 

 dro-carburets, such as naphtha- vapour ; and these are apt to be ab- 

 sorbed and withdrawn in part by the water, during the continu- 

 ance of a diffusion experiment. I have observed coal-gas to con- 



VOL. XII. PART I. K k 



