The Path of Gaseous Exchange. 



59 



summarises their results witli regard to carbon dioxide. Similar 

 results were obtained with water vapour. 



Table VI 11. 

 Diffusion of carbon dioxide tliroiigJi npeiinws of various sizes. 



In order to explain this result, Brown and Escombe consider 

 first the case of a disc capable of absorbing carbon dioxide and freely 

 exposed to the air. If the latter is perfectly still, convergent streams 

 of carbon dioxide will creep through the air towards the disc to 

 replace that absorbed, and a steady gradient of density will be 

 established, and if surfaces are drawn passing through all the points 

 of the same carbon dioxide density, these surfaces will form ' shells ' 

 surrounding the disc. If the disc is a perfect absorbent of carbon 

 dioxide, these shells will vary in density from zero at the absorbing 

 surface to a maximum density which is that of carbon dioxide in 

 air. This will theoretically be at an infinite distance from the disc 

 but is practically reached at a point 5 or 6 diameters from the disc. 

 Now Stefan has examined mathematically the exact converse of 

 this case, namely, evaporation from a circular surface of liquid. 

 Stefan obtained the following formula for the amounts of evapor- 

 ation from such a surface : — 



M = 4ka 



P— p" 



Where M is the mass of liquid evaporated in a given time, k 

 the coefficient of diffusion of the vapour, a the radius of the disc 

 of liquid, P the pressure of the atmosphere and p' and p" the 

 pressure of the vapour at the surface and at an infinite distance 

 from it respectively. 



The formula given by Larmor for the absorption of carbon 

 dioxide by a perfectly absorbing disc, assuming the formation of 



