126 



Proof of Van' 'fc Hoff's Constant. 



Thus, by this plan, the constant 1'87 or 1/88 appears very 

 satisfactorily in the case of sugar and urea, and in the case of 

 alcohol the constant is only about 1 J per cent, too small. The 

 general run of the values of /3 in the cases of cane-sugar, urea, 

 and alcohol, shows that in the more concentrated solutions the 

 values obtained for the molecular lowering of the freezing- 

 point are greater than 1*87, and that it is only as the solutions 

 become more dilute that the constant 1*87 becomes apparent. 

 Now Van't HofF's laws were only deduced for very dilute 

 solutions, for " ideal " solutions. 



This year I repeated the investigations of cane-sugar, 

 alcohol, &c. under more favourable conditions, where no closed 

 ice-cap but ice-bands or network of ice are formed round the 

 Hg bulb of the yoo o° thermometer, and the convergence tem- 

 perature was under the freezing-point temperature. The 

 difference in the capsules is conditioned by the differing of the 

 ice-crystal, which is brought into the over-cold liquid to crys- 

 tallize it. When the convergence temperature is under the 

 freezing-temperature, the correct freezing-point of water, as 

 well as of solutions, is directly obtained with the ice-cap. [A 

 consideration of the processes which take place at the freezing- 

 temperature, when the convergence temperature is above 

 and under the freezing-temperature, is of great interest and 

 will form the subject of a separate communication.] Thus I 

 have shown by this third method, that Van't HofFs constant 

 holds good in dilute solutions. Table I', gives the experimental 

 data obtained with the xoVo"° an( ^ Too° thermometer for cane- 

 sugar, alcohol, and from these data Table IV. is deduced. The 

 Van't HofF constant appears very clearly, and is in the 

 most dilute solution only about 1^ per cent, less than 1/87. 

 Table III', gives the molecular depressions, starting from the 

 most dilute solutions, for cane-sugar &c. 



Table III'. 







1/1000° Thermometer. 



1/100° Thermometer. 



2-1. 

 3-1. 



a corr. 







a. 



ft 



a. 



/3. 



0003985 

 0011087 



00073 

 00221 



1SS-2 

 189-1 



00009 



(59) 



0218 



1731 



(1481) 



186-5 



4-1. 



0-018727 



0-0346 1S4-7 



0339 



181 



5-1. 



0*028437 



00521 1SJ-J 



00519 



1825 

 &c. 



