Methods Sodium and potassium concentrations for serum, sweat andurine were determined on a direct reading flame photometer (Domingo and Klyne, 1949), or by a direct reading EEL flame photometer. Chloride determinations, serum pH, carbon dioxide content, total body water, total body chloride, were all measured by methods set out in a previous communication (Cheek, 1957). To determine total chloride the volume of distribution of bromide was measured and a micro diffusion method was used as described previously (Cheek and West, 1955). Here the diffusion units of Obrink (1955) were used (obtained from Rudolf Grave Company, Stockholm, Sweden). These units obviate the need for fixative and for subsequent prolonged washings of the units and increase the accuracy of the method. The variation between duplicate determinations of serum bromide was less than 1%. Before the calculation of the extracellular volume index from the bromide space, 10% of the injected bromide was deducted from the total injected, to allow for the entrance of 8% of injected bromide into the red blood cells and for the fact that the volume of distribution of bromide is 1-2% greater than the chloride space. This correction has been applied previously (Cheek, 1954). Recent work indicates that in the rat almost all the intracellular chloride of the body is in red cells; hence it may be that the corrected Cl or Br space in man is a good approximation to true extracellularvolume. Sweat was collected following a pharmacological stimulus similar to a method described (Mauer and West, 1956). Urinary 17-ketosteroid excretion, and corticosteroid excretion were assayed by accepted procedures (Medical Research Council Committee on Clinical Endocrinology, 1951; Reddy, 1954). To assess urea clearance, urinewas collected over a 24-hour period accordingto the technique described by Landis, Elsom, Bott and Shiels (1935).

Case Report FC, a male infant, aged 3 months, was the first-born of parents who had been previously in good health. The mother was admitted to the matemity hospital during the last two months of pregnancybecause of an excessive gain in weight, hypertension and albuminuria. Labour was induced and the baby was born two weeks prematurely, weighing 5i lb. Labour was not difficult and delivery was uneventful. The infant was breast fed for two and a half months but failed to gainsatisfactorily. A dried milk mixture was triedtemporarily but at the age of 3 months he was brought to the Royal Children's Hospital because of poor feeding, unsatisfactory weight gain and cyanotic attacks. The cyanotic attacks had been noticed during the previous month, and were not associated with feeding. They were not accompanied 252