480 JOURNAL OF COSMETIC SCIENCE hair tresses, but the ethanol itself did not substantially affect the hair-tress luster one hour after treatment (Figure 3 ). Figure 3 shows that increased luster was detected by SAMBA even at 0.1 % phenyl trimethicone and that there was a good differentiation between 0.1 % phenyl trimethicone-treated tresses and untreated tresses. The luster of the treated hair tresses exhibited a concentration-dependent increase, with treatment of 2% phenyl trimethicone giving the maximal increase in luster, about 31 % (Figure 3). The increase in luster was due to an increase in the specular reflection (S), a decrease in the diffuse reflection (D), and a reduction in the width of the specular peak at half-height (W112 ) after phenyl trimethicone treatment (data not shown). FIBER SURFACE EXAMINATION To confirm that these increases in luster were caused by the deposition of phenyl trimethicone, we examined the surface of the hair fibers for thin film formation using AFM. Typical topography of an untreated hair fiber is shown in Figure 4A and C. After treatment with 0.5% phenyl trimethicone, a thin film was observed on the hair fiber (Figure 4B). After application of 2% phenyl trimethicone onto the hair fiber, a thicker film was formed (Figure 4D). We determined that roughness was reduced by 6.7% after 0.5% phenyl trimethicone treatment and by 19.7% after 2% phenyl trimethicone treatment (Table I). AFM studies provide convincing evidence that phenyl trimethicone increased hair fiber luster by deposition as a film on the hair surface. 40 ..-----------------------------------, ** 30 ···································································································· 10 ................................ , ............................................. Base 0.1% 0.5% 1% 2% Phenly trimethicone concentration Figure 3. Luster increase by phenyl trimethicone in oriental hair tresses as measured by SAMBA. The luster of the treated hair tresses exhibited a concentration-dependent increase. Data are expressed as mean ± SD. **p 0.0001 compared with base.

EVALUATION OF LUSTER IN HAIR 481 Before_0.5% After_0.05% Before_2% After_2% Figure 4. AFM topography images of hair fibers. (A,C) Untreated hair fiber. (B) Hair fiber treated with 0.5% phenyl trimethicone. (D) Hair fiber treated with 2% phenyl trirnethicone. CONSUMERS' SUBJECTIVE SHINE ASSESSMENT To determine whether consumers' perception accurately correlated with the instrumen­ tal luster value from SAMBA, we carried out panel tests using untrained panelists. First, we asked the panelists to rank the luster of untreated hair tresses on a scale ranging from 1 (most luster) to 5 (least luster) to identify different degrees of shine. The panelists could not distinguish differences in shine among the untreated hair tresses (data nor shown). Then, we treated the hair with phenyl trimethicone at the same concentrations we used for the SAMBA analysis. Panelists compared the degree of shine at various phenyl trimethicone concentrations. The rank test produced a score in which lower