We use the term "renal threshold for hemoglobin" to indicate the smallest amount of hemoglobin which given intravenously will effect the appearance of recognizable hemoglobin in the urine. The initial renal threshold level for dog hemoglobin is established by the methods employed at an average value of 155 mg. hemoglobin per kilo body weight with maximal values of 210 and minimal of 124. Repeated daily injections of hemoglobin will depress this initial renal threshold level on the average 46 per cent with maximal values of 110 and minimal values of 60 mg. hemoglobin per kilo body weight. This minimal or depression threshold is relatively constant if the injections are continued. Rest periods without injections cause a return of the renal threshold for hemoglobin toward the initial threshold levels—recovery threshold level. Injections of hemoglobin below the initial threshold level but above the minimal or depression threshold will eventually reduce the renal threshold for hemoglobin to its depression threshold level. We believe the depression threshold or minimal renal threshold level due to repeated hemoglobin injections is a little above the glomerular threshold which we assume is the base line threshold for hemoglobin. Our reasons for this belief in the glomerular threshold are given above and in the other papers of this series.

The minimal or depression renal threshold for dog hemoglobin is not modified by moderate doses of mercuric chloride. This type of renal injury involves the epithelium of the convoluted tubules but the glomeruli escape. We are unable to explain our findings if we assume that the tubular epithelium takes an active part in the passage of dog hemoglobin from the blood into the urine. The evidence points toward the glomerular tuft as responsible for the passage of the hemoglobin from the blood plasma into the tubules. The glomerular tuft establishes the true hemoglobin threshold under these conditions. If the convoluted tubules are normal, we note that hemoglobin is taken into the epithelium and this explains the high initial renal threshold. With repeated hemoglobin injections this tubular epithelium becomes stuffed with hemoglobin pigment fractions and can absorb no more, which explains the minimal or depression threshold. Further injury of this tubular epithelium with mercury causes no change in this minimal renal threshold, unless we produce actual tubular obstruction.

Frequent injections of superthreshold amounts of dog hemoglobin will cause deposits of pigment material in the renal tubular epithelium. When this has happened this dog will survive minimal lethal doses of mercuric chloride with little evidence of renal injury. In fact some dogs will tolerate twice the minimal lethal dose without severe reaction. There is no evidence that continued injections of dog hemoglobin in these amounts will cause any injury or functional disability of the kidney. Rest periods will effect a disappearance of this pigment in the renal tubules.