SUMMARY1. The effect of acute alterations of plasma sodium concentration (PNa) on renal sodium excretion (UNaV) was investigated by three types of experiments on anaesthetized dogs:(a) A local increase in PNa at one kidney was produced by infusion of hypertonic saline directly into its artery while systemic levels of PNa were stabilized by haemodialysis.(b) Systemic levels of PNa were lowered by exchange transfusion of blood for an equal volume of salt-free dextran-in-dextrose solution. The results were contrasted with those observed after similar exchanges, but using dextran-in-saline solution.(c) The level of PN. was altered by varying the sodium concentration of a saline solution infused at a fixed rate either intravenously or into one renal artery.2. All three types of experiment suggest a dependence of UNaV on PNa. Analysis demonstrated that this relationship was not due to contemporary changes in: packed cell volume; plasma solids concentration; plasma potassium concentration; blood pressure or plasma hydrogen ion concentration. The distribution of these variables did not change with PNaexcept for plasma hydrogen ion concentration. Moreover, the relationship persisted when data were selected to exclude clearance periods in which the value for any variable had shifted past the group mean obtained before PNa was altered.3. The fall in UNaV at low levels of PNa could be attributed to a fall in glomerular filtration rate (GFR), but the progressive rise in UNaV seen as PNa exceeded 150 m-mole l-l occurred despite a fall in GFR and no apparent change in the mean filtered load of sodium.These results suggest that the increased sodium excretion accompanying raised levels of PNa is due to reduced tubular re-absorption of sodium.

SUMMARY1. The effects of acute changes in plasma Na concentration (PN.) on renal blood flow (RBF) and glomerular filtration rate (GFR) were studied in anaesthetized greyhounds. Saline was infused at a constant rate (0.1 ml. kg-1 min-') either into a renal artery or into a systemic vein. Plasma Na concentration was altered by varying the Na concentration of the infused saline from 0-154 to 0-077, 0-616 or 1-232 M.2. Blood pressure (B.P.), packed cell volume (PCV), concentration of plasma solids (PS) and the plasma concentration of H+ and K (PK) ions were measured but no attempt was made to contain their fluctuation.3. An infusion of hypertonic saline into a renal artery usually led to an ipsilateral increase in RBF for 5-15 min, followed by a progressive fall. Over-all, mean values of RBF fell with PNa throughout the range studied (120-190 m-mole l.-'). Glomerular filtration rate rose with PNa to reach maximal values at PNa levels of 140-160 m-mole 1.-i, but fell thereafter. The combined fall in RBF and GFR, without change in filtration fraction, at PNa values above 160 m-mole 1.-IL is consistent with an alteration in afferent arteriolar resistance. The fall in GFR despite a rise in RBF noted when PNa was reduced below 140 m-mole 1.-i requires an additional explanation.4. Renal blood flow was independent of PK; it was inversely related to [H+] and directly related to PS. Glomerular filtration rate was independent of PCV and PK. It was also inversely related to [H+] and directly related to PS up to a value of 6 g 100 g-' plasma, after which the relationship was reversed. These results suggest that the renal vascular responses to acute

SUMMARY1. The effect of changes in the haematocrit of blood perfusing the kidney on its intrarenal distribution was studied in dogs.2. Two types of preparations were employed. (i) In the isolated perfused kidney evidence is presented that flow in the autoregulating preparation represents predominantly cortical flow while flow in the 'low flow nonautoregulating' kidney reflects medullary flow. (ii) In the intact kidney renal blood flow rate and its intrarenal distribution was studied by the injection of 133Xe into the renal artery and measuring its clearance from the kidney by an external counter.3. In both types of preparation cortical flow was found to be independent of changes in P.C.V. but medullary flow varied inversely with haematocrit.4. A change in the haematocrit of the perfusing blood leads to alteration of its viscosity. It was argued that an increase in viscosity must lead to a reduction in the resistance of the cortical afferent arterioles but that medullary afferent arterioles were not able to respond in this manner.5. These findings demonstrate that changes in total body haematocrit cause a redistribution of blood flow between renal cortex and medulla.

1. Acute changes in haematocrit were produced by exchange transfusion of dextran-in-saline or packed red cells.2. There were no significant changes in glomerular filtration rate, blood pressure, central venous pressure, heart rate, respiratory rate, blood volume or extracellular fluid volume following the exchange transfusions.3. Urine volume increased after haemodilution but decreased after haemoconcentration.4. The diuresis after haemodilution occurred despite an infusion of ADH or alcohol. Thus it could not be attributed to a change in circulating ADH level.5. There were two types of diuresis. The ;water diuretic' response was characterized by an increase in free water clearance with a reduction in urinary sodium concentration; the ;sodium diuretic' response by an increase in urinary sodium concentration but no change, or a fall, in free water clearance.6. The results were related to changes in medullary osmotic gradient found by other workers to occur when medullary blood flow rate is altered.

No abstract