0,04 JOURNAL OF THE SOCIETY OF COSMETIC CHEMISTS rigidity that the additional bonds formed (a{ter removal of the hygroscopic substances) impart to the structure. Due to the reduction in swelling capacity of the corneum imposed by this structural collapse, water uptake will also be less. This is also indicated by water-binding isotherms (1, 13) and differential thermal analysis ( 90 ). Indirect evidence for the cell membrane system rather than the keratin being the rheologieally active material in the stratum corncure is provided by the fact that isolated keratin fihns swell considerably in wate ", almost to the point of dissolution, and display no detectable elasticity. This behavior con- trasts with that of untreated and extracted eorneum, both of which have finite elastic modulus values in water (45). The concept that the mechanical strength of the corncure resides in the cell membrane system is supported by the fact that the cell membrane protein (110) has a much higher cystinc con- tent (x 3-4) than the keratin filaments (91) leading to more permanent (with regard to disruption by water) disulflde erosslinks in the membrane (4•). W•n B•m• Transepidermal water loss is a small fraction ooe the water loss from a com- parable area of water. The skin thus has the property of a water barrier, and the stratum corneum is generally recognized as representing the principal skin barrier to water loss as well as to entry of environmental noxious agents (6, 7, 60). Removal ooe the skin barrier by cellophane tape stripping will en- hance the absorption ooe almost any substance. The barrier possesses a resistance to the transport ooe water molecules, and it is irrelevant whether these water molecules are passing from the inside to the outside of the skin or vice versa. It is also irrelevant whether the water is in the liquid or in the gas phase on one or both sides ooe the skin (57). The vapor barrier properties ooe the stratum corneum are virtually mechanical and not dependent on living cells, functioning enzyme systems, intact cell mem- branes, or attributable to an "electrical double layer" ( 1, 7, 8, 12, 14, 88, 92). The location ooe this epidermal barrier has been the subject ooe considerable controversy. At one time it was thought that the barrier layer resided between the stratum granulosum and stratum corneum (92-94). Later it was placed in the lower portion of the corncure (1, 95, 96). However, at present, the avail- able data offer no evidence for the concept that different layers within the main bulk ooe the stratum corneum have different diffusivities (10). Water permeability experiments on isolated stratum corncure (8) indicate that the barrier to penetration is the entire stratum corneum ( 10, 15, 50, 54). Water diffusion rate can be principally used to evaluate the functional state ot • the skin barrier. It is probably the most sensitive measure of barrier intact- hess (97). Ioe the skin barrier is removed, water will evaporate from the skin surface at approximately the same rate as from a free •vatcr surface. Water

WATER AND THE SKIN 205 permeability is measured in terms of evaporative loss from an epidermal sur- face covering the aqueous phase. Most methods depend upon the use of two chambers, in one of which is placed a known quantity of the substance under study, separated from a second chamber by a piece of whole skin or separated epidermis. The amount of the substance recovered from the second chamber is measured (10, 60, 98-100). Barrier renewal is a complex process. After loss of the stratum eorneum, a temporary barrier is formed by rapid conversion of granular cells into para- keratotie cells. As the parakeratotie layer thickens water loss gradually creases. The temporary barrier persists until the regenerating epidermis be- comes capable of forming normally keratinized eelIs (gl,101). The reforma- tion of the barrier to increased sensible water loss is very rapid. Three days after the injury, the water vapor loss of the skin has been restored to such an extent that the difference between the water vapor loss of the damaged and normal skin has become very small. Total regeneration requires about two weeks (95, 101-103). There is limited knowledge of the composition of the barrier. However, it can be equated with that of the eorneum. The main cellular components are proteins, ]ipids, and water combined into an ordered structure. The eelIs are bound into a layered membrane which is ca. 15/• thick when dry but increases to ca. 48/• when fully hydrated (104). When hydrated, the stratum eorneum contains approximately 75% water, g0% protein, and 5% lipid. The sebaceous secretion, sebum, forms an irregular film on the surface of the skin, which has been estimated as averaging 0.4-4/• thick (73). Sebum does not contribute to any significant effect to the epidermal barrier. liemov- ing the sebum by mild swabbing with acetone or ether has no effect on the rate of water loss (105) and adding up to 10 times the normal amount (30 thick) has only a slight effect (106). The sebum apparently does not form a functionally complete lipid layer over the skin, leaving gaps which allow ac- cess to the underlying stratum eorneum (107). It is nonoeelusive and rela- tively porous due to the presence of branched-chain aliphatie compounds (87). However, the oil does help retain the water on the surface of the skin. Frequent washing with soaps and detergents, which may dissolve the oil, can cause loss of the water-holding capacity of the epidermis. While the sur- face lipids offer little resistance to the passage of compounds, studies of re- moval of lipids from the cutaneous surface (•, 7, 16, 69, 88, 109) indicate that lipids participate in epidermal water function. Onken and Moyer (7) sho•ved that barrier function is restored when extracted lipids are returned to the skin. This suggests some possibility of marked variations in biological membrane permeability, dependent largely on the specific nature or distribu- t/on of the lipid contained in the cell membrane. Swcency and Downing (88 applied a number of solvents to the epidermal side of hairless mice and hu- man skin. All of the solvents increased the rate of diffusion of tritiated water,