ESTIMATION OF CARBOHYDRATES 221 



B. ESTIMATION BY REDUCTION OF COPPER SALTS. 



This method of estimating glucose is the one most frequently 

 used and is the one which is the most varied in manipulation. The 

 variations may be divided into the following groups : 



I. Complete reduction of cupric to cuprous salt. 

 Methods of Fehling-Soxhlet ; Pavy ; Gerrard ; Benedict. 



II. Incomplete reduction of cupric to cuprous salt 



(a) Gravimetric by the estimation of the precipitated cuprous oxide. 

 Methods of Maercher ; Allihn ; Kjeldahl ; Brown, Morris and 



Millar; Pfluger. 



(b] Volumetric. 



(i) Direct by the estimation of the precipitated cuprous oxide. 

 Methods of Mohr-Bertrand ; Caven and Hill ; Sidersky ; Bang's 

 second method. 



(ii) Indirect by the estimation of the residual cupric salt. 

 Methods of Lehmann-Maquenne ; Bang's first method. 



(i) Fehling-Soxhlet Method. 



Barreswil in 1844 was the first to use this property as a means of estimat- 

 ing glucose, his reagent consisting of an alkaline solution of neutral potassium 

 tartrate and copper sulphate. Fehling, in 1849 and 1858, established this re- 

 duction process of estimating glucose by showing that the ratio of glucose to 

 cupric oxide was 1:5. He used Rochelle salt, as suggested by Bodeker, in 

 place of neutral potassium tartrate. It was found later that the ratio of i : 5 

 was not exact and that it varied with the concentration, alkalinity and time of 

 boiling. These difficulties were overcome in 1880 by Soxhlet, who showed 

 that accurate results could be obtained if (i) the copper solution had always 

 the same concentration and (2) the sugar solution under examination had a 

 concentration of about i per cent. Soxhlet, following suggestions by Krause 

 and Staedeler and by Graeger, employed two solutions, which were mixed in 

 equal volumes immediately before use. Such solutions were found to keep well, 

 whereas, if mixed, the titer changed and the solution reduced itself on boiling. 



The method as described by Fehling with Soxhlet's modifications is the 

 simplest and the one most commonly used at the present time for a rapid 

 and very fairly accurate estimation of glucose. 



It is as follows : 



