A METHOD FOR THE PREPARATION OF BLOOD FILTRATES FOR THE DETERMINATION OF SUGAR

A few years ago the writer observed that tungstate filtrates of blood yield considerably higher reducing non-sugar values (nonfermentable reducing substances) when determined with an alkaline ferricyanide reagent than when copper is used as the oxidizing agent. The average reduction values, expressed in equivalents of glucose, were 21 mg. per cent with the modified Shaff er-Hartmann method and 40 mg. per cent with the ferricyanide reagent. As a consequence, the latter gave an average of 20 mg. per cent higher apparent sugar (total reduction) values than the Shaffer-Hartmann reagent (1). In seeming contradiction to this we encounter in the literature reports that the Hagedorn-Jensen method, with a ferricyanide reagent, yields results not higher than those obtained by several copper reduction methods (2,3). It would be entirely misleading, however, to compare copper reduction methods with the Hagedorn-Jensen procedure without taking into consideration that while in the former the sugar is generally determined in tungstate filtrates, the filtrates involved in the Hagedorn-Jensen method are prepared by a vastly different procedure, consisting in heat coagulation of the proteins in the presence of zinc hydroxide. With the amount of non-sugar reducing substances in such filtrates unknown, an adequate correlation of the sugar values is obviously impossible. Experiments recorded in Table I furnish information as to the quantity of reducing non-sugars in zinc hydroxide filtrates,’