A balance experiment measuring potassium and sodium intake and excretion was carried out on Romney Marsh and Merino sheep. For either potassium or sodium the correlation between total intake and total excretion was high and no difference was found between the two breeds in the relative importance of the urinary and faecal routes of excretion. A mean value of 11% of the total potassium excretion and 12% of the total sodium excretion appeared in the faeces, with the respective remainders, 89 and 88%, being voided in the urine. The relative importance of cutaneous losses of the two cations in suint is discussed briefly. Neither total intake of potassium nor gain or loss in weight by the sheep could be shown to influence the relative importance of the two routes of potassium excretion investigated in this work. This was also true of sodium.

Haematocrit values, plasma osmolality and the plasma concentrations of sodium, potassium, chloride and insulin were measured in carotid arterial blood before, during and after intravenous infusion of NaCI (0· 5 moll-i) and KCI (0· 5 moll-i) at 2 ml min~l for 105 min into six conscious splenectomized sheep. Hypertonic NaCI infusion was associated with a fall in haematocrit of 1 ·30 ± 0·10 % (P < 0·001) and no consistent change in plasma insulin concentration occurred during this infusion.Hypertonic KCI infusion caused the haematocrit to increase by 1·70 ± 0·39 % (P < 0·001) and the plasma insulin concentration to increase by 60·0± 16·3 jlU ml-1 (P < 0·01). It was concluded that this increase in insulin concentration was caused by elevation of the plasma potassium concentration and was not due to coincident increases in plasma chloride concentration or osmolality. Shrinkage of the extracellular fluid volume during KCI infusion made no major contribution to the increase in insulin concentration which was probably the result of increased release from the pancreas.