PERMEABILITY OF EPIDERMIS 865 50,000 ---, 40,000 \ 30,000 • % 2o,oo o to,ooo o _ I - kp = 0.59x10 -• t 0 1 2 3 4 5 6 7 8 t (hours) Figure 9. Effect of DMSO on water permeability. Upper line is in the presence of DMSO, recovery line is after DMSO was removed, and lowest line is the original permeability of water .cates that contact with DMSO does not continue to alter the membrane structure appreciably beyond the initial effect of the solvation. This be- havior is consistent with the proposed mechanism for the mode of ac- tion of DMSO given above, i.e., that DMSO simply solyates the tissue. Unfortunately, for this theoretically satisfying picture, an appreciable part of the enhanced diffusion appears to be permanent and resolvation with water does not restore the barrier completely. For example, a permanent increase in water permeability is observed when DMS'O (100%) is held in contact with the tissue for 6 hours, then drained and replaced with water (Table IV, first series). Other investigators have reported that the effect of DMSO is re- versible (4, 6) this has occasionally been observed in our laboratory on particular samples, but it has not been observed consistently. Our efforts to pursue this intriguing aspect with a series of experiments aimed directly at verifying this point have been unsuccessful. Some lipids are extracted by DMSO and this could cause a permanent increase in permeability in the same way as that produced with nonpolar solvents. It is still possible that the mechanism proposed is essentially correct and the resolvation with water in the manner described after contact with DMSO and the accompanying extraction of the membrane is the major cause of damage. In the second and third series of experiments with DMSO (Table IV), care was taken to avoid mixing water and DMSO in the membrane.

866 ,JOURNAL OF THE SOCIETY OF COSMETIC CHEMISTS Table IV Effect of DMSO on the Permeability of H20 Initial Test Run Recovery k•, X 10 • k v X 10 ab k v X 10 sc Notes 0.53 22.6 0.44 27.7 0.33 27.2 0.37 24.2 0.55 21.4 0.41 27.7 0.53 26.6 0.42 23.9 0.53 26 0.59 32.2 0.67 28.1 0.51 35.2 0.67 24.8 0.50 26.2 0.5O 24.9 0.50 19.4 0.50 22.6 0.50 27.8 0.50 21.7 0.50 24.4 19.5 ) 14.3 19.6 20.6 16.7 I 20.2 I 19.3 I 12.8 J 10.9 ) 10.5 I 6.8 t 12.9 / 6.4 9.6] 5.8 I 7.7 / 4.85 8.7 I 7. DMSO run 280 min H20 run time 7.5 hours Butanol rinse before and after DMSO No HsO or butanol contact before DMSO Butanol rinse prior to HaO run • Initial runs are water permeabilities prior to DMSO contact (H20 donor) (average value = 0.50 X 10 -s cm/hr). b Test runs are water penneabilities in the presence of DMSO (DMSO 100% donor). •, Recovery runs are water perrneabilities after removal of DMSO (H20 donor). The hydrated membrane was rinsed with butanol (which does no measur- able damage) prior to adding DMSO and then after removing DMSO prior to rehydration. As shown in Table IV there was appreciably less permanent damage to the tissue with this procedure. There would be presumably still less permanent damage if extraction were minimized as in an in vivo application where the solvent is left to evaporate and any extracted substances are redeposited. Soaps and Detergents: the Anionic Surfactants We can distinguish still another kind of chemical disruption of the stratum corneum, namely, that produced by dilute soap or detergent solutions. Anionic surfactants appear to have an appreciable effect on the permeability of the skin even when applied from dilute aqueous solutions. Figure 10 shows the large increase in water permeability produced