PROFESSOR THOMAS GRAHAM S SCIENTIFIC WORK. 



207 



and inverting it into a cup of mercury, a Torricellian vacuum was left 

 at tbe toji, into wbich the external air, or any external gas experimented 

 on, gradually found its way by passage through the caoutchouc film, so 

 as to cause a depression of the mercurial column. By experiments made 

 in this manner, it was found that difierent gases penetrated the rubber, 

 and entered the vacuous space with the following relative velocities, 

 differing widely from the velocities of diffusion and transpiration of the 

 same gases given in the other two columns of the table : 



Oxygen 20.8 X 2.55 = 53.04 ~ 



Nitrogen 79.2x1.0 = 79.20 



f 40 

 CO 



Bearing in mind the partial separation of gases from one another at- 

 tainable by reason of their unequal diffusive velocities, the possibility 

 of effecting a similar separation of gases by reason of their unequal 

 velocities of transmission through India rubber was easily to be fore- 

 seen. For example, atmospheric air consisting of 20.8 volumes of oxj^- 

 gen and 79.2 volumes of nitrogen, and the transmission velocities of 

 these two gases being respectively 2.55 and 1.0, it follows that the air 

 transmitted through India rubber into a vacuum should consist of 40 

 per cent, oxygen and GO percent, nitrogen, thus: 



132.24 J 1 100 



In subjecting this conclusion to the test of experiment, Mr. Graham 

 availed himself of Dr. SpreugeFs then newly invented mercurial puuip 

 or exhauster, an instrument which also stood him in good stead in his 

 subsequent work, and to which he freely acknowledged his obliga- 

 tions. By a slight alteration in the pump, as originallj^ constructed, 

 Mr. Graham made it serve not only for its original purpose of creating 

 and maintaining an almost perfect vacuum, but also for delivering pari 

 pasKH any gas penetrating into the vacuum through its caoutchouc or 

 Oither walls. 



The caoutchouc films employed in these experiments were of various 

 kinds ; but the most readily practicable and, on the w-hole, successful 

 results, were obtained with India-rubber varnished silk made up into a 

 flat bag, exposing on each side about 0.25 meter of square surfiice. 

 The interior of such a bag being in communication with the Sprengel 

 pump, the. constituents of the external air were graduallj^ sucked 

 through the walls of the bag and delivered by the turned-up fall-tube of 



