HAIR VOLUME AND TEXTURE 343 camera used to capture images was a Sony XC-77CE miniature CCD video camera module with an Olympus 50-mm F2 auto-macro lens. Hair body was evaluated as a ratio of the test tress area to the control area. An average of four images per tress (imaged at 90 ø to each other) was used for the assessment. This provides a "three-dimensional" measurement of hair body, similar to a human assess- ment. Four images (and not two) were necessary, since the tress was rotated manually (exact angles of rotation cannot be verified) and not by an automated device. These measurements can be made rapidly: three minutes or less per tress, including setup. PANELISTS' EVALUATIONS Evaluations were made by panelists who evaluated hair attributes (volume, texture, etc.) using their own personal experience criteria. For the most part, a panel of ten evaluators (five women and five men) was used. In some experiments this panel was supplemented by other assessors. "Visual" hair body was evaluated by panelists observing a tress from all directions. For "visual + feel" hair body assessments, panelists also touched the hair fibers while estimating body in the previously described manner. HAIR AND TREATMENTS Light brown European hair, dark brown Oriental hair, and black Afro-American hair, all purchased from DeMeo Brothers, New York, were employed in experiments de- scribed in this paper. Some of this hair was permed for experimental use utilizing commercially available home permanent wave kits. Tresses were made by weighing the hair (1 to 6 g for experiments reported here) and binding the root end with a rubber band. In all cases, before final treatment and after fashioning into tresses, hair was washed twice (one-minute wash one-minute rinse, 43øC) with 20% surfactant (ammo- nium lauryl sulfate) to ensure an initially clean fiber surface, and combed twenty strokes to make sure that tresses were of uniform combability. After drying, the tresses were weighed and hair was removed (if necessary) to provide tresses with the required quan- tity of hair for the experiment. For Experiments I and II, tresses of varying weights (and type) were used in Experiments III, IV, and V, 3.5 g of virgin European hair was utilized. "Temporary" surface treatment regimens were applied to hair following man- ufacturers' instructions. When only an active ingredient was used, treatment time was thirty seconds followed by a thirty-second rinse. Hair treatments for Experiments III and IV were as follows: ß A, a quaternized protein (2% laurdimonium hydrolyzed animal protein) that in- creased hair volume but did not appreciably alter hair feel ß B, a particulate treatment (2% barium sulfate) that both increased hair volume and provided a rough feel to hair ß C, a "heavy" hair conditioner, based on cetearyl alcohol, keratin and stearalkonium chloride, that left hair very soft and smooth to touch but did not enhance hair volume ß D, a surfactant (20% ammonium lauryl sulfate)•-clean hair with no post-shampoo treatment

344 JOURNAL OF THE SOCIETY OF COSMETIC CHEMISTS ß E, a post-shampoo product containing panthenol, aloe vera, proteins, and botanicals that claimed to increase hair body and thicken hair ß F, a hair styling gel (post-shampoo product) that thickens and stiffens hair, thus increasing hair body, bounce, and hold ß G, a hair pomade, based on petrolatum and mineral oil, that caused fibers to adhere closely, giving a greasy, matted look (and feel) to hair RESULTS AND DISCUSSION To test the assumption that hair body can be adequately assessed by measuring the volume the hair occupies, panelists were exposed to a set of seventeen tresses, widely varying in volume (Experiment I). Figure 1 shows tresses representative of the hair body range of this experiment. Untrained panelists were asked to give rank and visual ratings (1 to 10 scale) to the tresses for hair body on the basis of visual examination only. Tresses included those prepared from straight hair, clean hair, greased hair, and hair with a natural or permanent wave. Also included were five tresses assembled from black Afro- American hair (very curly hair). Using a planimeter, tress areas were outlined from photographs of the hair tresses, and paper replicas of them fashioned and weighed. This procedure assumes that the two- dimensional tress area is representative of the three-dimensional tress volume. A non- parametric test (Spearman rank correlation) was applied to the data, yielding a Rho value of 0.865, Z = 3.461 (p 0.001). Figure 2 summarizes these data and shows the correlation between visually assigned panelist hair body ratings and the paper weights of the (seventeen) tresses. The black Afro-American hair tresses caused confusion among panelists, resulting in extreme ranking positions. These tresses were judged to have very little hair body by some panelists and a lot of body by other panelists. When these data Figure I. Photograph of hair tresses representative of those used in hair body/volume experiment (Afro- American hair on far left).