326 GEORGE F. BECKER AND ARTHUR L. DAY 



from 1 to 2 mm. Through the crystallization of the chrome alum, therefore, 

 both the porcelain support and the beaker-glasses were actually separated 

 from the bottom and lifted up" [p. 108]. 



This elevation, they assert, has nothing to do with any force of 

 crystallization, and is to be accounted for by ''forces of capillarity 

 and adsorption," but just how these forces produce this result they 

 do not explain. 



If two parallel plane surfaces are separated by a drop of liquid 

 which wets them both, it is well known that the effect of capillarity 

 is to urge them together with a force 2TV/d 2 , where T is the surface 

 tension, V the volume of the drop, and d the distance between the 

 plates, and this force may be great enough to change the melting- 

 point of blocks of ice or to break sheets of heavy plate glass. Thus 

 a flat disk of porcelain moistened with alum solution and resting 

 on a flat glass plate experiences a downward pressure as if it were 

 heavily loaded. 



If a solution has a greater or smaller surface tension than the 

 solvent, the solute tends respectively to desert the surface of the 

 solution, or to seek it, and in the physics of colloids the concentra- 

 tion of solute at such a surface is known by analogy as adsorption. 

 As E. Dorsey 1 and others have shown, aqueous solutions of salts 

 such as chlorides and carbonates of the alkalies and zinc sulphate 

 have a surface tension greater than that of water, and according to 

 Poynting and Thomson, 2 the surface tension of salt solutions gener- 

 ally exceeds that of water. The adsorption in this case, considered 

 as a possible lifting force, is therefore negative. Adsorption films 

 in their relation to crystal growth have been investigated experi- 

 mentally with great thoroughness by Marc in a series of four papers, 

 "On Crystallization from Aqueous Solutions" (1908-10) ; 3 and 



1 Phil. Mag., XLIV (1897), 369. 



2 Properties of Matter, 1902, p. 181. 



3 The chief conclusions reached by Marc are contained in the following brief 

 citations (Zeitschr. f. phys. Chem.): 



"It was found that the rate of crystallization, so far as it could be determined, 

 of all the substances investigated was proportional to the square of the supersatura- 

 tion" (op. cit., LXVII [1909], 500). 



". . . . that a very rapid change precedes the crystallization proper, which is 

 interpreted to be an adsorption phenomenon. Support is given to this view by the 



