wo years ago we reported our results on T physicochemical characterization of more than one hundred albumin specimens that were isolated from the blood plasma of cancer patients.4 The dependence of the opticalrotatory power on pH was found to be the most suitable method for characterization. The levorotation values of the albumins were plotted against the PH of the solutions, and the resulting rotation curves showed a considerable variety. The albumins that were isolated from the blood of individuals with neoplastic diseases ordinarily possessed lower levorotation values than did normal albumins. Many of the "cancerous" albumins, moreover, showed peculiar maxima and minima in the rotation curves. Comparisons with albumins that were isolated from healthy persons or from those suffering from noncancerous diseases showed certain overlapping of the curves. This of course could be expected because of the following reasons: (1) the great variety of the cancerous and noncancerous diseases and conditions of the blood donors; (2) genetic inhomogeneity of the donors; and (3) unavoidable small variations in the methods applied for plasma fractionation and optical-rotation measurements. The albumin fractions used in the first years of our studies were 90 to 95 per cent pure. It was demonstrated by electrophorcsis and by carbohydrate analyses that the abnormalities in the optical rotation are de-From the University of Texas M.