CHEMISTRY. 107 



(4) The amouuts of different glucoses required for the reduction of 

 a definite amount of Febliug's solution are not the same, and in 

 order to accurately estimate the glucose content of a solution one 

 must have definite knowledge as to the kinds of glucoses in the solu 

 tion. If the inversion yields dextrose alone, the titration will give 

 quantitative results; but if it yields levulose, alone or mixed with other 

 glucoses, the quantitative determination is difficult. In this case the 

 inversion must be performed by heating for a short time with very 

 dilute acid, as in the case of cane sugar; by inversion with stronger 

 acid the levulose will be partially decomposed. 



The exact determination of the carbohydrates by the above piethod 

 in vegetable substances containing raffinose, stachyose, or lupeose is 

 practically impossible. Approximately accurate results may be obtained 

 by inverting a definite quantity of the vegetable extract and an equiva- 

 lent quantity of the pure carbohydrate under exactly the same condi- 

 tions, and then titrating both with Fehling's solution. Assuming 

 that the carbohydrate behaved the same in the pure solution is in the 

 vegetable extract, the amount of the carbohydrate may be approxi- 

 mately calculated from a comparison of the two results. 



The difficulties are nearly insurmountable when a vegetable extract 

 contains a number of carbohydrates which do not invert with like 

 rapidity, or one of which yields a glucose which resists the action of 

 acid only in a vslight degree. For instance, if an extract containing 

 cane sugar and maltose, or trehalose, was heated with an acid, the 

 inversion of the cane sugar would be completed before that of the 

 maltose or trehalose; and if the heating was continued until the latter 

 was inverted, the levulose from the cane sugar would be partially or 

 wholly decomposed. It would thus be impossible to determine the 

 total glucovse content from the solution resulting from inversion. Of 

 course, if the amount of maltose or trehalose present was small, the 

 error would be insignificant. 



If an extract contains several soluble or insoluble carbohydrates, 

 capable of inversion, error may result from the fact that inversion will 

 usually give a glucose mixture of varying reducing power, according 

 to the amount of different glucoses present. This error will be small, 

 as a rule, since the glucoses known at present do not differ very 

 greatly in reducing- power. 



On the analysis of the seeds of plants, E. Schulze {Chem. Ztrj., 

 18 {1894), No. 43, pp. 799-802).— In the case of both leguminous plants 

 and cereals, the husk, or outer covering of the seeds, and the kernel, 

 differ widely in nutritive value. Consequently the author believes an 

 analysis of these two parts separately gives a better idea of the com- 

 position than an analysis of the whole seed. This not only increases 

 the number of determinations, but also the labor of preparing- the 

 sample for analysis. 



