When a measurement of serum protein fractions by the paperelectrophoretic method is made, it is assumed that the amount of dyestuff adsorbed onto serum protein of the paper is proportional to the content of the latter. This assumption, however, is not necessarily correct, since the amount of dye per unit of protein does not keep constant, if the concentration of adsorbed protein increases more than a definite one. This was shown by the present author's experiment. Accordingly, in order to get the true values, extinction readings must be corrected in regard to protein concentration by the calibration curve which was obtained here.
When the dyestuff of the paper was washed out by 0.5% acetic acid aqueous solution containing free chlorine in it, the adsorbed dyestuff of the protein as well as of the paper is simultaneously decolored to such an extent as to interfere with the extinction readings. Thus the percentage of each serum protein fraction will not indicate the true one. In order to prevent this effect by decoloration of the dyestuff, a drop of 5% sodium thiosulfate solution was added to 2 liters of acetic acid solution, the water of which was from city water supply. By this method the author obtained the satisfactory pattern of the serum protein fractions which indicated minimum standard errors.