Results of function tests show that symptoms of hydraemic nephritis (in spite of histological evidence) depend mainly on disease of the glomeruli, while in azotae-mic nephritis the functional failure is due chiefly to the tubules. The chief symptoms of hydraemic nephritis seem to result more or less directly from lowering of the colloid osmotic pressure of the plasma, probably itself due to loss of protein through damaged glomeruli. A simple method of estimating the degree of reduction is described, and has been applied in a number of cases. The colloid osmotic pressure of normal plasma is lowered to a disproportionate extent by dilution, and due allowance is made for this in considering the alteration of the plasma proteins in disease. There is no real hydraemia in so-called "hydraemic" nephritis, and the reduction of urine is due to passage of fluid into the tissues. It is a fallacy to regard an absolute increase in the globulin of the plasma as certain evidence of an extra-renal factor. In azotaemic nephritis the chief feature is a gradual failure of the kidneys to raise the molecular concentration of the urine above that of the plasma. The power of distinguishing between different waste products is not lost, and creatinine is still excreted better than urea; but all waste products are believed to diffuse back through the renal cells in the water which is reabsorbed from the tubules. Although urea is still concentrated, there is less chloride in the urine than in the plasma, and active reabsorption of chloride in the tubules may permit some concentration of the more harmful waste products even when the power of absorbing water against an osmotic resistance has been lost. Fluid in the diet is of theoretical value, and the use of saline diuretics is discouraged.