The Colorimetric Determination of Glucose.

Several new descriptions of colorimetric methods for the determination of blood sugar have been described in recent years (for example, Polis and Sortwell, 1946; Somogyi, 1945; Nelson, 1944; Haslewood and Strookman, 1939). These procedures are all based on the use of alkaline copper reagents and on the production of molybdenum blue by the action of cuprous copper on complex molybdic acid solutions. With a view to finding the method or combination of methods which gives the most reliable results, particularly in the very low range of blood sugar, the following combinations of reagents were compared: (a) Harding's Schaffer-Hartmann copper reagent with the Folin-Wu phosphomolybdic acid; (b) Harding's copper reagent with the Nelson ar"senomolybdic acid; (c) Somogyi's copper reagent with the FolinWu phosphomolybdic acid, and (d) Somogyi's copper reagent with the Nelson arsenomolybdic acid. Much difficulty has occurred during the war years and subsequently in this country with colorimetric methods for blood sugar. Biochemists have repeatedly complained of the rapid fading of the blue colours. This fault has been more manifest because of the recent introduction of photo-electric instruments in which the rapid falling-off of the intensity of the blue colour has been noted by numerous workers. With Duboscq colorimeters the fading would not have been so readily noticed, since the rate of fading in standard and test is very similar. It has frequently been impossible during the last several years to obtain chemicals of Analytic Reagent quality, and this abnormal behaviour has been traced to the use of inferior chemical reagents. The ordinary or technical grade of sodium carbonate, for instance, contains considerable sodium bicarbonate as well as the sodiusil carbonate. The amount of this inferior reagent specified for the preparation of an alkaline copper solution contains insufficient sodium carbonate to produce the desired properties in the final solution, and an abnormal behaviour consequently results. Experimental