NEW APPROACHES TO HAIR SPRAY EVALUATION 621 proves the coating on the hair fibers, presumably by reducing the resin solution's surface tension and viscosity. The purpose of studying coat- ability and finding ways to improve it was intended to develop hair sprays having superior curl retention at high relative humidity. Curl Retention o[ Hair Sprays Various methods have been evaluated to improve curl retention at high humidity without sacrificing natural appeaarnce, feel, or manage- ability. Initially, sprayed curls were evaluated in a chamber main- tained at 95-98% relative humidity. The curls were hung in front of a sheet of cardboard with graduated lines. The relative effectiveness of various formulations could be visually observed or photographed. Generally, three curls with the product being evaluated and three curls with a standard (usually considered best) were compared over a period of 3 hours. The pictures of the curls were studied to determine differ- ences in curl retention. Other pror•erties (luster, tack, feel, combability, and flakiness) were and still are evaluated subiectively. This method for determining curl retention was inadecluate because a sub'ective judg- ment was involved, a standard had to be run for each comparison, pic- tures minimized differences, and variations between curls st)rayed with the same product caused confusion. Other methods described in the literature (4, 7, 9) overcame some of these difficulties, but for one reason or another had limitations which made them not completely satisfactory. Therefore, we developed a method of measuring curl retention as ob- jectively as possible and applied statistics to validate our conclusions. The advantages of a statistical approach to improved curl retention are: it is reliable, conclusions can be statistically validated with a high de- gree of confidence, and a maximum amount of data is obtained from a •ninimum number of experiments. The method is readily applied to the optimization of formulations. Scanning Electron Microscope Studies Several laboratory formulations and commercial products were chosen for study. The laboratory formulations are given in Table I. Untreated European dark brown human hair* was made into 3-g switches and cleaned by the following procedure before initial use and for subsequent removal of hair spray films before reuse: wet hair, apply Castile shampoo, massage to high lather, rinse with warm water, towel • DeMeo Brothers, New York, N. Y.

622 JOURNAL OF THE SOCIETY OF COSMETIC CHEMISTS Table I Formulations for Studying Coatability Formulation 1 Resin D (50% soln) Acetulan a Perfume b Anhydrous denatured ethanol Freon 12/Freon 11 (50/50) propellant wt % 3.0 0.2 0.5 36.3 60.0 Formulation 3 PVP/VA copolymer (50% soln) Acetulan a Isopropyl myristate Anhydrous denatured ethanol Freon 12/Freon 11 (50/50) propellant Formulation 2 Resin G-1 (50% soln) AMPD c Acetulan • Isopropyl myristate Perfume b Anhydrous denatured ethanol Freon 12/Freon IlS (50/50) propellant Formulation 4 wt % 2.5 Resin D (50% soln) Acetulan" 0.1 Isopropyl myristate 0.1 Anhydrous denatured 37.3 ethanol Freon 12/Freon 11 60.0 (50/50) propellant wt % 40 01 01 01 O3 35 4 60.0 wt % 2.5 0.1 0.1 37.3 60.0 Formulation 5 wt % Resin R-1 (50% soln) 2.5 Isopropyl myristate 0.1 Acetulan" 0.1 Anhydrous denatured 37.3 ethanol Freon 12/Freon 11 60.0 (50/50) propellant • Amerchol, a Unit of CPC International, Inc., Edison, N.J. b GD-6763-1, Givaudan Corp., Clifton, N.J. ½ 2-Amino-2-methyl-l,3-propanediol, Commercial Solvents Corp., New York, N.Y. dry to remove excess moisture, apply creme rinse, massage well, and rinse with lukewarm water until clear. Hair to be evaluated by the SEM was air dried and combed before use. Coated hair was obtained by spraying the switch from a distance of 12 in. using five rapid back-and- forth passes on each side of the switch. The sprayed hair was air dried overnight. A portion of the sprayed hair was carefully placed over the sample mounts (ca. « in. in diameter) and glued at the ends on the underside. When secured, the excess hair extending over the ends of the sample mount was cut off. The sample mount with hair was placed