THE NEW THEORY OE SOLUTIONS. 



3«3 



solved in certain solvents, particularly hydrocarbons, gave 

 almost half the "normal" value. Now van't Hoff's theory 

 and subsequent experiments have proved that the molecular 

 lowering of a solvent has only one definite value which 

 can be calculated from its freezing-point and heat of fusion 

 in the manner already shown, and Raoult's so-called " nor- 

 mal " lowerings agree in all cases with the values calculated 

 in this way. The abnormally small molecular lowerings 

 are ascribed therefore to the exceptional behaviour of 

 the dissolved substance, and find a probable explanation 

 in the assumption that in these solutions its molecular 

 weioht is oreater than the theoretical value. From the 

 above results it would appear that the dissolved substance 

 existed as double molecules ; subsequent experiments by 

 Beckmann (1888, 1890), Eykmann (1889), Paterno (1889), 

 Auwers (1893-94), and others, have proved, however, that 

 this view is misleading, because the molecular weight varies 

 as a rule at an abnormally rapid rate as the concentration 

 alters. Here are the results obtained by Beckmann with 

 solutions of alcohol in benzene, the values of the molecular 

 weight being calculated both by the methods of Raoult (M') 

 and Arrhenius (M"). 



In the most dilute solution the theoretical molecular 

 weight is obtained ; in the strongest, however, the value is 

 four times the theoretical number. The result arrived at 

 by Raoult regarding the existence of double molecules is 

 therefore due to the fact that he used one solution only, 

 and did not study the effect of concentration. 



On contrasting the variation in the values for M' and 

 M" in the above table with that of indifferent solutions in 



general, as typified by the solutions of alcohol in acetic 



27 



