182 Prof. Graham on the Law of the Diffusion of Gases. 



As soon as it was filled, it was placed in a glass-jar, of 

 about the same height, with a little water left in the bottom, 

 and in proportion as the water rose in the tube of A, from 

 the subsequent contraction, the jar was filled up by repeated 

 additions of water, so as to keep the surface of the water, 

 within and without the tube, as nearly as possible at the same 

 level. With the view of having the external air in a constant 

 state in regard to humidity, means were taken to saturate it. 

 A small cone of damp paper was inverted, like an extin- 

 guisher, over the upper part of the instrument; the jar con- 

 taining the instrument was placed on the shelf of the pneu- 

 matic trough, and a bell-jar with an opening at the top, which 

 could be shut at pleasure, inverted over the whole. The re- 

 turn-air must therefore have been in the same state, in regard 

 to humidity, as the hydrogen itself. Aqueous vapour would 

 diffuse neither outwards nor inwards, as it existed in the same 

 proportion on both sides of the plug; but dry hydrogen only 

 would be exchanged for dry air, in the proportion of their 

 equivalent diffusion-volumes. 



In the first thirty-four minutes, the gaseous contents of the 

 bulb were diminished by 95 measures, and ultimately, in 

 twenty-six and a half hours, they were reduced to 227 mea- 

 sures, which were common air. The contraction in this and 

 other cases, in which the water rose into the bulb, was deter- 

 mined by weighing, at the end of the experiment, the water 

 which had entered ; a mode which admits of even greater 

 nicety than measuring the bulk of residuary gas in a graduated 

 vessel. 



With the view of obtaining elements for a correction for 

 any change in the bulk of the gas, which might take place 

 during the continuance of the experiment, from changes in 

 temperature, pressure, or from solution of the gas 

 in water, a receiver was made of the same tube, f X 

 with a bulb of nearly the same capacity as the dif- ( ) 

 fusion-instrument, but close at the top. This re- 

 ceiver was also nearly filled at the commencement 

 of the experiment with hydrogen gas, and the 

 quantity of gas noted, the tube being graduated. 

 The hydrogen in this standard receiver contracted y^nd part 

 during the experiment. We have therefore to increase the 

 (juantity of air found ultimately in the diffusion-receiver by 

 g^-nd part. In this way the residuary air is increased to 229*8 

 measures, 12 of which, or, more correctly, 1 1*85 ( = 12 — ^(1 2)), 

 were present from the beginning. 



The temperature was also 68° at the end of the experiment, 

 the same as at the beginning. The ultimate contents of the 



