400 



Aqbioultubal Experiment Station, Ithaca, N. Y. 



It was hoped tliat tlie fermentation metliod could be applied; 

 but as results even on samples of pure glucose were unsatisfactory, 

 its use was given up. For all the determinations reported below, 

 the fruit was dried at 100 degrees, the dry residue was ix)wdei'ed, 

 and then extracted for the sugar by boiling water when this was 

 the extracting agent used, or by ninety per cent alcohol in the 

 apparatus used in this laboratory for continuous exti'action of fat 

 by ether. The aqueous extract was so dark colored that a volu- 

 metric determination by the Fehling solution could not be made; 

 the alcoholic extract was much less highly coloi'ed. This solvent 

 was used in the hope that it might extract less of other reducing 

 substances than sugar, than water would tahe up. The gravi- 

 metric method, in which the copper reduced fix)m the cuprous 

 oxide by hydrogen was weighed, gave very satisfactory results. 



The results obtained are given in the following table : 



The acidity was determined by means of a standard solution of 

 potassium hydix)xide, and calculated as malic acid. 



It seems certain that the volumetric results were too low. The 

 gravimetric results were obtained with the alcoholic extract; 

 dujjlicat/e results on the same extract agreed closely. It will be 

 noticed that one very low result on sugar in sample 6, was accom- 

 panied by a very high per cent of acid. 



G. 0. CALDWELL. 



