SummaryChanges in specific gravity and hæmoglobin content of human arterial blood during water diuresis *The density of whole arterial blood and arterial plasma was studied in 30 patients, by the technic of Lindeström-Lang. On the same blood specimens the hæmoglobin was measured with the Pulfrich photometer.One hour after ingestion of one liter of water, there is a decrease in density of blood and arterial plasma and also drop of hæmoglobin content.The difference between the density of the blood and that of water at the same temperature decreases about 2,5 %. The corresponding change in density is 3,5 % in the plasma. The reduction in hæmoglobin content is about the same as that of the density of the whole blood (2,73 % as mean of 30 experiments).As, after ingestion of water, the density of the plasma and whole blood drops, there is no doubt that both are diluted. But, if one wishes to express the dilution in terms of increased volume and to relate it to the drop in hæmoglobin content, one has to consider several hypotheses.The first possibility is the blood being diluted by pure water without gaining or losing any protein or red cells. Such conditions being granted, it is easy to calculate that, one hour after ingestion of one liter of water, 100 cc. of arterial blood should contain 2,61 cc. of excess water, distributed in proportion of one fourth in the red cells and three fourths in the plasma.This should be in perfect agreement with the observed changes in density of whole blood and plasma and also with the drop in hæmoglobin content. But in order to prove the concordance between the inference and the facts, one ought to measure by direct observation an increase of 1,39 % in the volume of the red cells, which is hardly feasible by the methods available. Indeed, this would correspond to a very small variation in the diameter of the red cells and to a difference of about 1/200 in the hæmatocrit reading.Other hypotheses might be suggested concerning the mechanism of blood dilution. One might assume that the fluid entering the blood is not pure water but a solution of salts and proteins, or postulate a removal of red cells from the blood after ingestion of water. Such hypotheses are not incompatible with the experimental results. However, they seem to be less probable that the first one and are not easier to prove by direct observation.* The original paper will be published in«Revue Belge des Sciences Médicales»(Louvain).