The elephant, black and white rhinoceros and the hippopotamus inhabit similar country, feed off land vegetation and have superficially the same type of almost hairless skin. The hippopotamus, however, is unique in spending most of the daytime in water or mud and feeding almost entirely at night or in the early morning. A study has been made of the structure of the epidermis of these animals and the differences found can, broadly speaking, be explained as modifications that are in accord with the animal's mode of life and surroundings. A more detailed study of the hippopotamus's skin shows that its structure, while probably primarily one of many possible modifications for an aquatic environment, permits a rate of transepidermal water loss that is greater than that in other animals for which figures can be obtained. No certain biological function can be proposed for this rapid loss of water; it is suggested that it is brought about by evaporation from the exposed layers of PAS‐positive material which in this skin are continuous from the Malpighian layer to the surface of an unusually thin stratum corneum.

An adequate rate of evaporative water loss is considered es• sential for the maintenance of thermal balance in the elephant in warm climatic conditions. Histological studies have failed to reveal the existence of sweat glands in elephant skin. Trans• epidermal water•loss rate has been measured and shown to be sufficiently high for possible thermal needs. The structure of elephant skin and the behaviour of elephants are seen to con• tribute towards maintaining skin permeability and the necessary level of transepidermal water loss.

The carbonic anhydrase inhibitor, acetazoleamide (Diamox), causes striking and sustained falls in both intraocular (Becker, 1954) and cerebrospinal fluid (Tschirgi, Frost & Taylor, 1954) pressures. The falls in pressure have been ascribed to decreased rates of production of fluid and it has been estimated that the decrease in rate of formation of the ocular fluid must amount to some 65 % of the normal (Becker, 1955;Becker & Constant, 1955); if the rate of drip from the subarachnoid space is any true record of the rate of production of the cerebrospinal fluid, the fall in rate is also of this order (Kister, 1956).It was considered that such large changes in the rates of flow might be reflected in measurable decreases in the rates of turnover of 24Na in the fluids; moreover, since these fluids normally have different concentrations of ions and non-electrolytes from those in a dialysate of plasma, changes in rate of flow, and thus of the average sojourn in their respective cavities, might influence the steady-state distributions between fluids and blood plasma. In the present work, the influence of Diamox on the rates of turnover of 24Na in the two fluids, and on the steady-state distributions of sugar, chloride and bicarbonate between plasma and the fluids, has been examined in several species. Changes in some of these steady-state distributions have been found, and the problem has been discussed as to whether they are consequences of a primary decrease in rate of flow, or primary effects of the drug on the secretory mechanisms involved in the elaboration of the fluids. METHODS Kinetic 8tudie8. The technique employed for rabbits was that described earlier (Davson, 1955), a steady concentration of "Na in the plasma being maintained by continuous intravenous infusion, and the aqueous humour and cerebrospinal fluid being withdrawn at stated times after

The distribution of chloride between the aqueous humour and plasma shows characteristic variations according to the mammalian species examined (Davson, Matchett & Roberts, 1952). Thus in the guinea-pig the concentrations in the two fluids are such that there is a deficiency of the ion in the aqueous humour, compared with the concentration in a plasma-dialysate, the value of the ratio: Concentration in aqueous humour/Concentration in plasma (RAq) being 0 935, compared with the value of 1P02 for the ratio: concentration in dialysate/concentration in plasma (RDia1). In the horse, by contrast, the value of RAq is 1-14, indicating a large excess of chloride in the aqueous humour. In general, the variations in the distribution-ratio for the cerebrospinal fluid (RCf) are by no means so great, and in all the species examined there is an excess of chloride in this fluid (Davson, 1955): in consequence, the ocular and cerebrospinal fluids of the guinea-pig, for example, show a difference in chloride concentration of 21 %; in the horse, on the other hand, the ratios are very close, the difference being less than 4 %. The problem naturally arises as to whether the excesses of chloride are compensated by deficiencies of bicarbonate, and vice versa.In the present study the bicarbonate distributions have been determined in a number of species. Characteristic variations have, indeed, been observed, and an attempt has been made to relate these variations with the buffering requirements of the intraocular contents; since the buffering capacity of the aqueous humour depends not only on its bicarbonate concentration, but also on its rate of renewal, the latter factor, measured by the rate of turnover of 24Na, has been examined in a number of species. Wherever possible simultaneous studies have been carried out on the cerebrospinal fluid, as it is con-

No abstract