8 PROCEEDINGS OF THE AMERICAN ACADEMY. 



fourth-molar solution of crystallized chromic sulphate (170 grams in a 

 litre) was made ; and a portion of this was boiled in a Jena flask with 

 a reflux condenser and rapidly cooled. Thus were obtained a violet 

 and a green solution containing the same concentration of chromium, the 

 latter being always used very promptly after its preparation in order to 

 forestall as much as possible the reversion to the violet condition. 



The freezing point determinations were made in large test-tubes in a 

 Beckmann apparatus. Each tube contained 20.00 cubic centimeters of 

 the sugar solution, and 5.00 cubic centimeters of the chromic solution 

 or of another solution whose catalytic action was to be tested. In order 

 to determine the freezing point of the cane sugar solution another tube- 

 ful was prepared in which these last 5.00 cubic centimeters were of pure 

 water. 



The freezing points of these solutions were determined in the first 

 place immediately after mixing. They were then placed in a water tank 

 of fairly constant temperature (not far from 17°); and from time to time 

 further determinations of the freezing point were made. 



Oddly enough the green and violet solutions caused in the first place 

 the same depression of the freezing point of the sugar solution. It is 

 not wholly safe to conclude from this, however, that the average number 

 of molecules in a given bulk of the two solutions would be identical if the 

 sugar were absent ; for it is not impossible that a portion of the complex 

 substances existing in the green solution may combine with sucrose, as 

 we later found them to affect gelatine. 



In a short time the catalytic action of the two solutions was seen to be 

 very different. The green solution inverted the sugar rapidly, while the 

 violet solution had only a slight action. This is indicated by the follow- 

 ing table of data and results, most of which explains itself. The sixth 

 column labelled "acid" represents the freezing points of a 0.468 normal 

 solution of hydrochloric acid. Each of the temperatures represents an 

 average of at least four readings of the thermometer. The changes in 

 the zero point of the thermometer were registered by freezing-point deter- 

 minations of pure water ; these were essentially parallel with the changes 

 in the pure sugar solution and are not recorded in the table on the 

 opposite page. 



It is evident from this table that the green solution possessed at least 

 twelve times as great a catalyzing power as the violet, while the acid 

 catalyzed at first about 3.5 times as fast as the green. As the sucrose 

 diminishes in concentration the speed of the reaction diminishes from 

 this cause, and as time goes on, the green slowly reverts to the violet, 



