Bialytic Separation of Gases by Colloid Septa. 403 



thickness, in carbonic acid over mercury, it was seen that the 

 rubber gradually absorbed 0*78 volume of gas in twenty-four 

 hours at 15°, of which 0*7 volume was taken up in the first 

 hour. The mass of rubber was previously measured with care by 

 the displacement of mercury in a specific-gravity bottle, and again 

 when the rubber was charged with carbonic acid ; it gave the 

 same displacement of mercury within a hundredth of a gramme. 

 No measurable change in the bulk of the rubber, therefore, had 

 occurred. It may be added that the absorbent power of vulcan- 

 ized rubber for carbonic acid appears to be less than that of 

 rubber in its natural state, being found only 057 volume in a 

 comparative experiment. 



The penetration of rubber by gases may be illustrated by their 

 passage into a vacuum, as well as into an atmosphere of another 

 gas in Dr. Mitchell's experiments. The diffusiometer, consist- 

 ing of a plain glass tube of about 22 millims. in diameter and 

 nearly a whole metre in length, closed at the upper end by a thin 

 plate of stucco and open below, is taken advantage of in such 

 experiments. A thin film of ^rubber from a small balloon is 

 stretched over the upper end of the tube, where it is supported 

 by the stucco plate, bound with copper wire, and cemented at 

 the edges in contact with the glass with gutta percha softened 

 by heat. If the tube be now filled with mercury and inverted, 

 a Torricellian vacuum is obtained above, into which the air of the 

 atmosphere gradually penetrates, passing through the film of 

 rubber and depressing the mercurial column in the tube. In 

 order to compare the penetration of different gases, a hood of 

 thick vulcanized rubber, provided with a small entrance- and 

 exit-tube for gas (such as is often used in gas experiments), is 

 placed over the upper end of the diffusiometer described, and 

 cemented to it by means of fused gutta percha. The gas to be 

 operated upon can thus be conveyed from the apparatus in which 

 it is generated, or from a gasometer in which the gas is stored, 

 into the hood or upper chamber of the diffusiometer, and the 

 excess of gas supplied be allowed to escape into the atmosphere 

 by the exit-tube of the hood. The stucco plate used as a sup- 

 port to the film of rubber is so highly porous as not to add sen- 

 sibly to the resistance experienced by the gases in passing 

 through the rubber, and, having no absorbent power of its own, 

 may be left entirely out of consideration. 



A comparison was made of the passage through the rubber 

 film, on the same day, of carbonic acid, hydrogen, oxygeu, and 

 nitrogen; barom. 773 millims., therm. 23° to 23°'5 C. The 

 time during which the mercurial column fell in the diffusiometer 

 from 748 to 723 millims. was noted in seconds, and also from 

 723 to G98 millims. The gases were all carefully dried. 



2D 2 



