120 Dr. M. Wildermann on an Experimental 



and by Ejkman ; and the equation 



0-02 T 2 



t — 



* w 



where T = absolute temperature, w = latent heat of fusion of 

 the solvent, has been experimentally verified and confirmed 

 for several solvents. 



The method of determining the freezing-point of very 

 dilute solutions*, which was devised by my late friend 

 P. B. Lewis, from 0°to — o, 2, and my recent investigations 

 of similar solutions from — 0°'2 to — o, 4 and of the freezing- 

 point of water f, have given us a means of submitting Van't 

 Hoff's equation, 0-09T 2 



t = — , 



W 



to a more accurate verification. 



The value of t is 1*87 if, with Bunsen and others, we take 

 the latent heat of water to be 80 cal., and is somewhat greater 

 if w is put =79*6 cal. 



First come the experimental data. In the following Tables 

 a is the number of molecules in solution per litre (in pure 

 water this is 0) ; 7 is the freezing-point of the water or of the 

 solution, as read on the yooo° thermometer ; 8 the number of 

 readings of the temperature made after the freezing-point has 

 been reached ; e the limits within which the readings of the 

 temperature varied ; (f) is the time over which the readings 

 extended ; b is the amount of over-cooling, or difference 

 between the temperature of the solution before and after 

 the formation of ice, these temperatures being read on the 

 t ^q° thermometer ; c, the temperature of the ice-bath ; 

 c\ the temperature of the room ; k is the reading of the 

 barometer. 



* See 'Transactions of the Chemical Society,' 1894. A fuller account 

 has been given in the Zeitschrift fur phys. Chemie, xv. p. 358. See also 

 the very valuable work of Jones, Zeitschr. phys. Chemie, xi. p. 110. Later 

 investigations of Jones have shown that his method was not brought to 

 the high development we all supposed. It is now a lonir time ago since 

 Professor Ostwald told me that it is also his opinion that the method used 

 by Jones requires a further development, because the results obtained for 

 non-electrolytes b} 7 this method do not agree with the fundamental gene- 

 ralizations of the modern Theory of Solutions. The work of Lewis, which 

 is distinguished by great experimental accuracy, and by the minute and 

 judicious care with which the conditions necessary to secure the best 

 possible results are investigated, enables us now to submit many important 

 questions in very dilute solutions to a more accurate verification. 



t See ZeitscJir. ])hys. Chemie, xv. p. 365. 



