488 Dr. Harry C. Jones on the 



to be completely free from dextrose. The specimen from 

 Berlin -was purified by alcohol, and was in the form of a 

 powder. That from Halle consisted of small crystals. The 

 latter sample was on the whole to be preferred, and although 

 measurement of solutions prepared from each were made, only 

 those results obtained from the Halle sugar have been given 

 here. 



From the table of results it will be seen that the molecular 

 lowering at 0*003 n is about 2*36. This decreases with in- 

 crease in concentration to about 0'l?z, and then increases 

 again as the solution becomes more and more concentrated. 



This increase in the molecular lowering with increase in 

 concentration for the more concentrated solutions of cane- 

 sugar was found by Arrhenius *. His method was, however, 

 not sufficiently accurate to enable him to employ solutions 

 dilute enough to detect the corresponding increase in this 

 region. 



Raoult f , using an improved method, has found that the 

 molecular lowering of cane-sugar increases with increase in 

 dilution from about 0"1 n, and also increases with the concen- 

 tration in the more concentrated solutions. He did not use 

 solutions more dilute than 0'02 n. These determinations of 

 Eaoult are probably the most accurate hitherto made for 

 cane-sugar solutions. 



I am entirely unable to understand the results of Loomis J, 

 who has found the smallest molecular lowering for the most 

 dilute solutions employed (0*01 ?i), and that this increased 

 throughout with the concentration. He has also found similar 

 results with other organic compounds. 



Since the results with cane-sugar were of an unsuspected 

 character, it was determined to test the last apparent possi- 

 bility of experimental error being their cause. It was barely 

 possible that some considerable change in the bulb of the large 

 thermometer had taken place and escaped notice. To test 

 this it was decided to remeasure a series of solutions of sodium 

 chloride. The results of these measurements agreed with the 

 results obtained several months before, in every case to within 

 a few ten-thousandths of a degree. 



The facts with cane-sugar are then : 1st, the molecular 

 lowering increases from a certain concentration (about 0*1 n) 

 with increase in dilution, aud also increases from this same 

 concentration with increase in concentration. 



A possible explanation of the latter fact has been furnished 



* Arrhenius, Zeits.phys. Chem. ii. p. 491. 



t Eaoult, ibid. ix. p. 343. 



X Loomis, Ber. d. deutsch. Chem. Gesell. xxvi. p. 797. 



