Dialytic Separation of Gases by Colloid Septa. 407 



sion in the second hour, 2 in the third hour, 3 in the fourth 

 hour, and 51 divisions in the first twenty-four hours taken 

 together. Then the rise in the following* successive days was 

 42, 59, 37, 29, 13, 5, 1, 0'5, 0-5 (in two days), and 0-0, the 

 original volume of 249 volumes of hydrogen being finally re- 

 placed by 53 volumes of atmospheric air; barom. 747 millims., 

 therm. 21°*1. The ultimate replacing volumes are here as 1 to 

 4" 7. In gas-diffusion they are as 1 to 3*8. 



A balloon filled with air subsided in forty-eight hours from 

 150 to 147 millims. in diameter, from the mechanical effect alone 

 of the elasticity of the membrane in compressing the enclosed 

 gas. These little balloons vary from 0*75 to 1 grm. in weight. 

 Supposing the form to be truly spherical, a balloon of 150 mil- 

 lims. in diameter would have a surface of 0*0706 square metre 

 (5*905 inches in diameter and 0*0844 square yard of surface). 

 Supposing the balloon to be 1 grm. in weight, the thickness of 

 the membrane will be T of a millim., with a specific gravity 

 = 1, or n T of a millim., with a specific gravity = 0*93, the 

 admitted density of pure rubber. This last is a thickness of 

 19 o . 6 °f an i ncn j or it would require nearly 2000 such films, 

 laid upon each other, to form the thickness of a single inch. 

 Yet such a film of rubber appears to have no porosity, and to 

 resemble a film of liquid in its relation to gases — differing en- 

 tirely in this respect from a thin sheet of paper, graphite, earth- 

 enware, or even gutta percha, as will appear hereafter. These 

 last enumerated bodies appear all to be pervaded by open chan- 

 nels or pores, sufficiently wide to allow gases to be projected 

 through by their own proper molecular movement of diffusion. 

 But liquids and colloids have an unbroken texture, and afford no 

 opportunity for gaseous, diffusion. They form even in the thin- 

 nest film an impervious barrier to gas* 



The penetration of rubber is much affected by temperature, 

 and apparently in two different ways at the same time. An in- 

 crease of temperature no doubt renders all gases less easily lique- 

 fied by pressure, and consequently less considerably absorbed by 

 any liquid or colloid. But such an influence of heat appears to 

 be counteracted in rubber by the tendency of that colloid to be- 

 come more soft when heated, and to acquire more of liquid 

 and less of solid properties. Certainly the rubber film be- 

 comes more and more permeable to gases as the temperature is 

 elevated, within a moderate range. This was distinctly observed 

 in operating with silk cloth varnished on one side with rubber, 

 such as is sold as a waterproof material. Without anticipating a 

 detail of the experiments, it may be stated in general terms that 

 the same specimen of rubber was penetrated by air from the at- 



