1 the Diffusion of Gases. 227 



meable by gases, even in the most humid atmosphere, if not po- 

 sitively wetted. The tube was finally graduated by means of 

 mercury into hundredths of a cubic inch, and the notation, as is 

 usual with gas-receivers, counted from the top. 



When such a diffusion-tube, six inches in length, was filled 

 with hydrogen over mercury, the diffusion, or exchange of air 

 for hydrogen, instantly commenced, through the minute pores 

 of the stucco, and proceeded with so much force and rapidity, 

 that within three minutes the mercury attained a height in the 

 receiver of upwards of two inches above its level in the trough. 

 Within twenty minutes the whole of the hydrogen had escaped. 



In conducting such experiments over water, it was necessary 

 to avoid wetting the plug. With this view, 

 before filling the diffusion-tube with hydro- 

 gen, the air was withdrawn by placing the 

 tube upon the short limb of an empty sy- 

 phon (see figure), which did not reach, but 

 came within half an inch of the plug, and 

 then sinking the instrument in the water- 

 trough, so that the air escaped by the sy- 

 phon, with the exception of a small measure, 

 which was noted. The diffusion-tube was 

 then filled up, either entirely, or to a certain extent, with the 

 gas to be diffused. 



The ascent of the water in the tube, when hydrogen is dif- 

 fused, forms a striking experiment. In a diffusion-tube fourteen 

 inches long, the water rises six or eight inches in as many mi- 

 nutes. The column of water attains in a short time its maximum 

 height, at which, however, it is never long sustained ; for as in 

 DOEBEREINER'S experiment, air is all along entering mechanical- 

 ly through the porous plug in such circumstances, from the pres- 

 sure of the atmosphere ; and after the diffusion is over, the water 

 subsides, in the course of several hours, to the general level In 



