556 



substances. Bachmann '), who already deteimiiied an isotherm of 

 carbon and water before me, found in cocoanut carbon a maximum 

 water ab.sorption of / = 0,25. Berl and Anukess ') also found in their 

 carbon a considerably smaller value than I in mine. 



The doubie-S-shaped curve of the isotherm obtained is practically 

 the same form as that whicli van Bkmmelrn has observed in gels 

 of silicic acid and of iron hydroxide. The flat portion there corresponds 

 to the part of the curve in wliich the gel, which is first transparent, 

 becomes opaque. 



3. Comparison of Free Energy and Heat Effect. 

 A simple comparison of the curves fig. 1 and fig. 2 sliows that 



ilW 



di 



and log h must have an analogous course as function of i'. Both 



curves have an almost horizontal, almost rectilinear (slowly descending) 

 portion between « = 0.10 and /^0.60 to 0.65; both curves have 

 before and after this the shape as for liquids which mix with water 

 with strong heal effect. By graphical determination of the differential 



quotient — ;— this can be estimated for some values of i, for which 



di 



log h is known. Thus I find: 



These are only rough estimations; but they show nevertheless 

 with sufücient probability that in the large middle portion of the 

 curve (from / = 0.05 to j^O.80) the variation of the free energy 



1) Zeitschr. f. anorgan. Chemie 100, p 32 (1917). 

 ') Zeitschr. f. angevvandle Chemie 1921. Bd. 1. 



I 



