THE CHLORINE-HAIR INTERACTION 95 Table I Effect of Cosmetic Treatment/Sequence and Cycles of Chlorination on the Coefficient of Friction and Percent Stick of Brown Hair Pretreated Posttreated Cycles Treatment p• s* % Stick s p• s % Stick s 0 Control 0. 130 0.017 36.6 1. 140 Bleach control 0. 196 0.027 38.0 1.225 Dye control 0. 166 0.015 36.8 1.095 Perm control 0. 196 0.015 38.4 0.894 5 Control 0. 180 0.016 38.0 0.707 Bleach 0.230 0.031 42.0 1.871 Dye 0.214 0.023 42.6 0.548 Perm 0.218 0.013 42.2 1.643 10 Control 0.246 0.027 41.8 2. 168 Bleach 0.218 0.028 41.6 2.881 Dye 0.276 0.013 39.2 1.304 Perm 0.234 0.018 42.4 1. 140 15 Control 0.244 0.021 43.2 1.483 Bleach 0.290 0.023 41.4 2. 191 Dye 0.272 0.023 37.8 0.837 Perm 0.254 0.018 43.0 2.828 30 Control 0.278 0.026 46.8 1.304 Bleach 0.326 0.015 52.2 3.421 Dye 0.294 0.036 39.6 0.894 Perm 0.280 0.025 54.6 3.782 0.300 0.035 45.4 2.408 0.193 0.019 38.0 0.816 0.244 0.009 42.6 1.342 0.306 0.025 47.8 2.490 0. 174 0.009 43.2 1.304 0.276 0.011 47.8 4.324 0.326 0.056 49.6 4.393 0.226 0.027 42.4 1.817 0.292 0.019 51.4 4.506 0.262 0.025 44.6 2.608 0.200 0.012 40.0 1.225 0.268 0.017 45.3 2.500 * s = standard deviation for five observations. STATISTICAL ANALYSIS Statistical analysis computations were performed using various procedures of the statis- tical analysis system (SAS). The general linear model procedure and the analysis of variance procedure were used to analyze the sources of variation. Pairwise comparisons were used to determine differences between levels of a given source of variation. All tests of significance were made at the 95 percent level. RESULTS AND DISCUSSION The effect of cosmetic treatment sequence and number of one-hour cycles of chlorina- tion on average values of the coefficient of friction are given in Tables I and II. Treat- ment and cycle effects are evident, with the blond and brown hair showing similar trends. The cosmetic treatments alone significantly increased the coefficients of friction beyond that of the untreated controls (0 cycles). When the cosmetic treatments pre- ceded chlorination, coefficients of friction generally increased with increased cycles of chlorination. This continual increase was similar to that of the control samples. These results indicated a gradual softening of the surface of the fibers, first with the cosmetic treatments and then with continual chlorination.

96 JOURNAL OF THE SOCIETY OF COSMETIC CHEMISTS Table II Effect of Cosmetic Treatment/Sequence and Cycles of Chlorination on the Coefficient of Friction and Percent Stick of Blond Hair Pretreated Posttreated Cycles Treatment p• s* % Stick s p• s % Stick s 10 3O Control 0. 132 0.013 36.8 0.837 Bleach control 0. 194 0.015 39.2 0.837 Dye control 0. 184 0.040 36.8 1.304 Perm control 0.202 0.023 40.0 1.581 Control 0. 180 0.014 38.6 0.548 Bleach 0.266 0.026 43.2 2.280 Dye 0.222 0.008 39.6 1. 140 Perm 0. 194 0.013 38.2 1.304 Control 0.222 0.031 42.8 Bleach 0.268 0.024 40.2 Dye 0.280 0.065 39.0 Perm 0.214 0.021 43.2 Control 0.220 0.014 43.6 Bleach 0.310 0.027 41.8 Dye 0.254 0.019 37.0 Perm 0.246 0.013 44.6 0.254 0.019 44.0 0.236 0.184 0.015 43.4 0.894 0.248 0.013 44.0 1.732 3.033 2.387 0.256 0.025 43.6 2.702 1.581 0.220 0.022 42.0 2.236 1.095 0.264 0.013 49.2 5.263 1. 140 1.924 0.294 0.032 47.2 2.683 1.225 0.206 0.029 40.6 3.362 1.673 0.306 0.025 50.0 3.162 Control 0. 302 0.023 51.4 5. 597 Bleach 0.312 0.008 51.4 2.074 0.248 0.026 41.2 1.483 Dye 0.358 0.035 39.6 1.673 0.208 0.033 39.8 2. 168 Perm 0.312 0.023 46.2 3.701 0.272 0.016 48.2 3. 114 * s = standard deviation or five observations. When the cosmetic treatments followed chlorination, however, different trends were obtained. For the bleached and the permed samples, the average values of coefficient of friction increased with chlorination up to 15 cycles of treatment. These values were generally significantly higher than those of the control samples and either similar to or significantly higher than the values of the corresponding pretreated samples. For the postdyed samples, the values of coefficient of friction did not change much with in- creased chlorination and were consistently significantly lower than the corresponding values of the postbleached and postpermed samples. Similar results were seen for the percent stick parameter (Tables I and II). There was little difference between the percent stick values of the control samples and the pre- bleached and prepermed samples. Values generally increased with increased cycles of chlorination. In the postbleached and postpermed samples, percent stick increased up to 15 cycles of chlorination and then either decreased significantly (postbleached brown, postpermed) or remained relatively unchanged (blond, postpermed). In the dyed samples, the results of percent stick were inconclusive generally, there was little effect of cycles of chlorination or of treatment sequence. Examination of fibers in the scanning electron microscope showed little definite differ- ences in the morphology of the cosmetically pretreated and control fibers. The cosmetic