DRY FACIAL SKIN 173 in ethanol:water (70/30). One hour later, the site was sampled four times in succession and sebum excretion expressed as •g/cm2/min (5). STATISTICS Statistical analysis employed SAS (North Carolina), using a VAX 750 computer. Corre- lations among parameters were determined according to Kendall/Spearman procedures. Correlation coefficients (r) and statistical significance (p) have been calculated for all the experimental data. RESULTS The distribution of the dryness scores, on cheek and forehead, for the subjects in this investigation is summarized in Figure 1. For simplicity, we have divided the popula- tion into three groups, viz 1, 1 - 2 , 2 where 1, is marginal dryness. It is noteworthy that 70% of the subjects had appreciable dryness on the cheeks, while the figure for the forehead was only 10%. SC EXTENSIBILITY Figure 2 reveals a striking decrease in SC extensibility as the dryness score increases (r = -0.55, p 0.001). 40- CHEEK FOREHERD o r• m,m '-! Z 30 10 1 (1-2) DRYNESS SCORE

174 JOURNAL OF THE SOCIETY OF COSMETIC CHEMISTS 140 CHEEK ' BO • 60 Ld 40 X 20 o 1 (1-2) 2 nRYNESS SCORE Figure 2. Relationship between skin extensibility measured on cheek (arbitrary units) and dryness scores. ELECTRICAL CONDUCTANCE On both the cheek and the forehead, conductance was inversely related to the degree of dryness (r = -0.66 for cheek and -0.41 for forehead). While the correlations were relatively weak, especially on the forehead, the differences among the grades were sig- nificant for both areas (p 0.001 and p 0.002) (Figure 3). CORNEOCYTES The size of exfoliated corneocytes decreased with increasing dryness scores (Figure 4). The differences were significant (r = -0.32, p 0.05 on the cheek and r = -0.22, p 0.05 on the forehead). SER AND TEWL The SER was not found to correlate with the dry skin scores (r = 0.03, p = 0.85). A statistically significant trend (r = 0.28, p = 0.04) was observed with TEWL and the dry skin scores.