214 J. A. POLLOCK. 



concentration, up to the same point, will be constant, but 

 will then, until the surface is reached, continuously 

 decrease. 



The term concentration gradient will here be used only 

 in connection with variations of concentration within the 

 surface stratum in the direction of the outward drawn 

 normal. For descriptive purposes it is more convenient, in 

 many cases, to consider the average concentration in the 

 surface layer ; this will be referred to as the surface con- 

 centration in order to distinguish it from the bulk concen- 

 tration, which is that of the general mass of the liquid. 



It may be said, then, that solutions which show a decreas- 

 ing surface tension with increasing concentration will be 

 more concentrated in the surface stratum than in the 

 interior part of the liquid, whereas the concentration in 

 the surface layer will be less than in the body of the liquid 

 in those solutions which exhibit, with increasing concen- 

 tration, an increased surface tension. 



From the experimental side Lord Rayleigh 1 has made an 

 important addition to the discussion ; in an ingenious way 

 he has shown that newly formed surfaces, in the case of 

 dilute aqueous solutions of oleate of soda and of saponine, 

 have, within one hundredth of a second of their formation, 

 surface tensions not very different from that of the solvent, 

 although under ordinary conditions the values are notably 

 less than that of water. With a 10 per cent, aqueous 

 solution of alcohol, a considerable diminution of the surface 

 tension, if not the whole change, was found to have occurred 

 in the same interval. 



The important fact is here established that time is 

 required for the full development of the concentration 

 gradient ; the value of this time factor is shown to depend 



1 Lord Rayleigh, Proc. Roy. Soc, Vol. xlvii, p. 281, 1890; Scientific 

 Papers, Vol. in, p. 341. 



