PROGRESS IN PHYSICAL CHEMISTRY. 295 



the vessel eventually become the same? In all cases this 

 was nearly but not quite the case, the ratio of the inside to 

 the outside pressure varying between 0*9 and 0*97. In 

 the same direction Moore (90) has put forward the view 

 that osmotic pressure may be due to differences between 

 the capillary pressures of solvent and solution acting through 

 pores of molecular dimensions in the semi-permeable wall. 

 Ostwald points out, however (91), that as solvent and 

 solution are miscible with one another, there is no surface 

 of separation between them, and consequently that capillary 

 action in the ordinary sense is impossible. 



For some time past a discussion (92) has been going 

 on regarding the validity of Arrhenius' deduction from 

 theory that the osmotic pressure of a dilute aqueous solution 

 is proportional to the lowering of the freezing-point. 

 Dieterici by using Fischer's numbers for the vapour- 

 pressure of ice and water found this was not strictly true, 

 but Arrhenius by means of Juhlin's more recent observa- 

 tions showed that the proportionality was valid within the 

 limits of experimental error, and Dieterici has admitted that 

 for depressions of the freezing-point up to io° this is the 

 case. 



In his last paper, Dieterici gives a relationship between 

 osmotic pressure and depression of the freezing-point which 

 applies to concentrated solutions. Ewan (93) has also dealt 

 with strong solutions. He finds that the depression of the 

 freezing-point of a non-electrolyte will be greater than van't 

 Hoffs theory of dilute solutions demands if the heat of 

 dilution be positive and vice versa. The observed depres- 

 sions of the freezing-point of cane sugar and those calculated 

 by this formula agree but approximately. A similar and 

 detailed inquiry dealing with the properties of electrolytic 

 solutions of finite concentration has been published by van 

 Laar (94). 



As regards molecular weight estimations by means of 

 osmotic pressure, it may be noted that Lob (95) has 

 described a method employing red blood corpuscles. A 

 small quantity of blood is added to the solution and the 

 flask containing them is rapidly rotated for three minutes, 



