In a general review of the subject of anhydremia published in 1923, Marriott (1) brought out several of the phenomena which had been shown to accompany severe clinical dehydration: an increase in the proportion of the total solids and in the dried weight of a unit volume of blood, an increase in its specific gravity, in red cell count and hemoglobin, and in the concentration of the serum proteins. He showed also, however, that when deprivation of fluid, either from lack of intake, from vomiting, or from diarrhea, persisted over a long period of time, a lowering of total solids, red count, and hemoglobin ensued, which was attributed to breakdown of body protein with the liberation of bound water. All of these phenomena might occur at a time when the plasma volume was subnormal, and consequently an investigation of the concentration of these substances in a unit volume of blood was shown to be of limited value in the study of the water economy as a whole in dehydration. Although these considerations emphasized the fatuity of placing reliance on the indirect evidence afforded by changes in the concentration of blood constituents and the necessity for direct investigation of blood and plasma volume changes in dehydration, such direct determinations have been singularly infrequent. Marriott mentioned some in a general way, but did not fully report his data.A number of studies made since 1923-for example, those of Hartmann (2) and of Hamilton, Kajdi and Meeker (3)-on the serum or plasma concentration of electrolytes in dehydration accompanying the infections and digestive disorders of infants have shown that a lowering of the total electrolyte concentration is, in fact, rather more com-333 BLOOD VOLUME AND ELECTROLYTES mon than an increase; and even when dehydration was brought about experimentally by reduction of the water intake, Schiff, Bayer, and Fukuyama (4) were unable to change the concentration of serum total fixed base beyond the limit of error of the method used. If, then, a lowering of serum electrolyte concentration of this kind, or even if no change in concentration occurs, one must abandon the concept that loss of water alone from the blood is the primary event in the development of dehydration, since this would only act to bring about a higher concentration of electrolytes. This actual lowering of serum electrolyte concentration in dehydration appeared to Hartmann (2) of sufficient importance to cause him to postulate that loss of electrolyte from the circulation might occur to an extent that required the body to retain nonprotein nitrogen constituents for their osmotic action so that the normal osmotic pressure of the plasma might be maintained. If water alone had been lost to excess, no such difficulty would have existed, for the osmotic pressure of the plasma would by that measure be increased. The lowering of serum electrolyte concentration could theoretically be brought about by dilution of the blood, though in view of the loss of body weight and the obvious loss of water from the body as a whole, ...