62 



Mr. W. Makower on the Molecular 



on both sides o£ the porous plug a glass tube was attached 

 by rubber over the plug, and a slow current of hydrogen from 

 a Kipp maintained through it. In this case the value of X 

 for the emanation was found to be "277, i. e. about twice as 

 great as when the emanation was mixed with air. 



To test this point still further a series of experiments was 

 undertaken in which the pressure of the air with which the 

 emanation was mixed could be varied, precautions being 

 taken to insure that the pressures on the two sides of the 

 porous plug were always equal. The value of X was found 

 to vary considerably with the pressure, as will be seen from 

 Table I. 



Table I.— Plus I. 



Emanation mixed with Air. 



Pressure in 

 centimetres mercury. 



X. 



760 

 376 

 36-5 



18-6 



158 



15-6 



9-2 



5-4 



5-3 



138 

 •181 

 •177 

 •222 

 •240 

 •307 

 •350 

 •737 

 •577 



These results were confirmed by using another diffusion- 

 vessel of 250 c.c. capacity and a new plug (Plug II.) of 

 porous porcelain attached to the diffusion-vessel by sealing- 

 wax. Similar results were also obtained when the emanation 

 was mixed with hydrogen at different pressures. 



There are tw T o ways of explaining this change in the value 

 of \ with the density of the gas with which the emanation is 

 mixed. We may either suppose that the gas really exerts 

 some influence on the apparent molecular weight of the 

 emanation, say by condensing round its molecules ; or it 

 may be that the pores of the plugs employed are large com- 

 pared with the mean free path of the gas molecules. In this 

 case, the number of collisions of a molecule of emanation with 

 a gas molecule while escaping through the plug would not 

 l)o small compared with the number of collisions with the 

 substance of the plug; and therefore the value of X for the 

 emanation would depend upon the nature of the gas with 



