Dialytic Separation of Gases by Colloid Septa. 509 



Heated platinum tube containing hydrogen ; air outside. 



Time. 



Rise 



of mercury in gauge 

 barometer. 



minute. 









millim. 



10 minutes. 





115 



millims. 



20 „ 





245 



>> 



30 „ 





400 



)> 



40 „ 





535 



}> 



50 „ 





645 



)) 



60 „ 





710 



)> 



the actual height of the atmospheric barometer being 750 mil- 

 lims. at the same time. The tension of the residual gas was 

 therefore no more than 40 millims. of mercury. The ratio be- 

 tween the volume of gas at the beginning and end of the hour is 

 here as 18*75 to 1 ; whereas in a diffusion experiment of hydro- 

 gen into air, the ratio would be as 3*8 to 1. Further, the resi- 

 dual gas in the platinum tube still retained a small portion of 

 hydrogen. Withdrawn by means of the Sprengel pump and ex- 

 amined, the residual gas in the platinum tube amounted to 3*56 

 cub. centims., and consisted of 



Nitrogen . . . 3*22 cub. centims. 

 Hydrogen . . . 0*34 cub. centim. 

 3-56 cub. centims. 



The available capacity of the platinum tube was 113'1 cub. 

 centims. ; and when the tube was heated, the gas driven out by 

 dilatation measured in the cold 39"5 cub. centims., leaving in the 

 hot platinum tube 73"6 cub. centims. of gas estimated at 20° C. 

 and barom. 760 millims. It was found necessary in these expe- 

 riments to stuff that portion of the platinum tube that was placed 

 across the furnace and strongly heated, with asbestos, to give 

 support to the tube when softened by the heat of ignition, and 

 to prevent the tube from collapsing. 



It is difficult to say where the small volume of nitrogen found 

 in the platinum tube, amounting to 3*22 cub. centims., actually 

 came from. It appears too great in amount to have formed an 

 impurify in the original hydrogen gas, or to have gained access 

 to the vacuum through defective joinings in the apparatus. Its 

 presence suggests the inquiry, admitting that nitrogen cannot 

 pass alone through platinum into a vacuum, whether the same 

 gas may not be enabled to pass, in some small proportion, while 

 hydrogen is simultaneously travelling through the platinum in 

 the opposite direction. The liquid or the gaseous hydrogen oc- 

 cupying the platinum septum would thus form a vehicle or chan- 



