JOURNAL OF COSMETIC SCIENCE 436 CONCLUSIONS The whole of the experimental data show how it is possible to follow satisfactorily the variation of the pigment composition obtained from the reactions under a study carried out at different pH values by colorimetric analyses. In fact, if the colorimetric data can be considered a good descriptor of the formed pigment, our hypothesis of using the refl ec- tance as an innovative parameter for the prevision of the color yielded by such reactions can be asserted. Furthermore, the obtained data show that the color formed on the wool may be considered a good expression of the whole of the chromatic features of the major components present in the reaction solution. The results agree with the hypothesis that the wool fi ber does not interact chemically with the colored molecules. The weathering experiments show that, for all the couplers, the color decay, near to the destruction of the colored molecules, displays color alterations similar to those shown by the same colored molecules on an inert support like the silica gel used for the TLC. Since the color fastness on the wool is greater than that on TLC plates, we can assume that the colored molecules reside in the interior of the wool fi ber. The data seem to exclude any chemical interaction between the support and the dye formation, and as a consequence, wool seems to be an ideal substrate for a preliminary study of the application of colorim- etry as a new base of inspection in color prediction. ACKNOWLEDGMENTS This work was carried out with the fi nancial support of Tocco Magico (Italy). REFERENCES (1) J. F. Corbett, Benzoquinones imines. Part I. p-Phenylendiamine-ferricyanide and p-aminophenol-ferri- cyanide redox systems, J. Chem. Soc. B, 207–212 (1969). (2) J. F. Corbett, Benzoquinones imines. Part IV. Mechanism and kinetics of the formation of Bandrowski’s base, J. Chem. Soc. B, 818–822 (1969). (3) J. F. Corbett, Benzoquine imines. Part VI. Mechanism and kinetics of reaction of p-benzoquinone di- imine with m-phenylendiamines, J. Chem. Soc. B, 827–835 (1969). (4) J. F. Corbett, Benzoquinone imines. Part VII. The mechanism and kinetics of reaction of p-benzoqui- none di-imine with monohydric phenols and the ultraviolet, infrared, and nuclear magnetic resonance spectra of the resulting indoanilines, J. Chem. Soc. B, 1418–1426 (1970). (5) J. F. Corbett, Benzoquinone imines. Part IX. Mechanism and kinetics of reaction of p-benzo quinone di-imine with m-aminophenols, J. Chem. Soc. (Perkin II), 539–480 (1972). (6) J. F. Corbett, The role of meta difunctional benzene derivatives in oxidative hair dyeing. I. Reaction with p-diamines, J. Soc. Cosmet. Chem., 24, 103–134 (1973). (7) J. F. Corbett, Chemistry of hair colorant processes—Science as an aid to formulation and development, J. Soc. Cosmet. Chem., 35, 297–310 (1984). (8) J. S. Mukund, W. S. Tolgyesi, and A. D. Britt, Cooxidation of p-phenylendiamine and resorcinol in hair dyes, J. Soc. Cosmet. Chem., 23, 853–861 (1972). (9) R. J. Crawford and C. R. Robbins, A replacement for Rubine dye for detecting cationics on keratin, J. Soc. Cosmet. Chem., 31, 273–278 (1980). (10) M. Dolinsky, C. H. Wilson, H. H. Wisneski, and F. X. Demers, Oxidation products of p-phenylendi- amine in hair dyes, J. Soc. Cosmet. Chem., 19, 411–422 (1968).

J. Cosmet. Sci., 60, 437–465 (July/August 2009) 437 The cell membrane complex: Three related but different cellular cohesion components of mammalian hair fi bers CLARENCE ROBBINS, 12425 Lake Ridge Circle, Clermont, FL 34711. Accepted for publication February 11, 2009. Synopsis The structure, chemistry and physical properties of the cell membrane complex (CMC) of keratin fi bers are reviewed, highlighting differences in the three types of CMC. Starting with Rogers’ initial description of the CMC in animal hairs, several important developments have occurred that will be described, adding new de- tails to this important structure in mammalian hair fi bers. These developments show that essentially all of the covalently bound fatty acids of the beta layers are in the cuticle and exist as monolayers. The beta layers of the cortex are bilayers that are not covalently bonded but are attached by ionic and polar linkages on one side to the cortical cell membranes and on the other side to the delta layer. The delta layer between cortical cells consists of fi ve sublayers its proteins are clearly different from the delta layer that exists between cuticle cells. The cell membranes of cuticle cells are also markedly different from the cell membranes of cortical cells. Models with supporting evidence are presented for the three different types of cell membrane complex: cuticle–cuticle CMC, cuticle–cortex CMC, and cortex–cortex CMC. INTRODUCTION GENERAL STRUCTURE OF THE CMC The cell membrane complex (CMC) consists of cell membranes and adhesive material that binds the cuticle and cortical cells together in keratin fi bers. G. E. Rogers from his seminal high-resolution transmission electron microscope (TEM) studies (1,2) provided evidence for the current structure of the CMC, consisting of a central delta layer approxi- mately 15-nm thick sandwiched by two lipid layers called beta layers each about 5-nm thick see Figure 1, adapted from Fraser et al. (3). Questions still exist about the relative thickness and composition of the beta layers be- tween cuticle cells versus the beta layers of cortical cells (see Figures 2,3) and between the upper beta layer versus the lower beta layer of cuticle cells (see Figure 2). Although most authors quote the thicknesses of the beta layers between 2.5 (6) and 5.0 nm, 6.0 nm has also been cited (11). Swift (7) in his review of the human hair cuticle describes in detail the diffi culty of obtaining accurate measurements of the beta layers in the high-resolution TEM, and his explanation clarifi es the uncertainty that exists in ascribing monolayers or bilayers to these lipid strata on the basis of TEM measurements alone.