EFFECT OF DIMETHVL SULFOXIDE 165 lawn, N.J. Bovine albumin (crystallized), lot 13 code BVO162, and beta lactoglobulin (3X crystallized from bovine milk), lot 33, were purchased from Pentex Incorporated, Kankakee, Ill. All other chem- icals were of highest purity available. The details of the picrate ion-guinea pig skin penetration studies are described in the preceding paper (1). Studies on the effect of DMSO on swelling (or expanding) soluble proteins were performed using essentially the methods, techniques, and calculations described by Schachman (9) and Van Holde and Sun (8). Relative viscosities were measured using a Cannon-Ubbelohde Semi- Micro No. 75 Dilution Viscometer. The viscometer was mounted in a water bath in such a manner that it could be removed at will and replaced with fair precision. The temperature of the water bath was maintained at 25 ø + 0.05øC. The flow times were measured manually using a Heuer stopwatch (Heuer Timer Corporation, New York, N.Y.) which was graduated in '} ½ second. Flow times for distilled water or KC1 were of the order of 125 seconds. Duplicate flow times were generally reproducible to better than 0.4 second. Because of the viscometer design, no kinetic energy correction was used (10). Protein solutions were prepared by diluting two equal aliquots of a stock protein solution. One aliquot was diluted with water and stock KC1 solution and the other was diluted with water, stock KC1 solution, and DMSO. The net result was that two solutions, equal in protein concentration and 0.1 M in KC1, were prepared. However, one solu- tion was 30% (v/v) in DMSO. No problem in protein solubility was encountered with this DMSO concentration. The protein concentra- tions of all solutions were of the order of 0.5%. The flow times and density measurements were done on the DMSO- protein solutions, while the pH and protein concentration determinations were done on the aqueous protein solutions. Protein concentrations were determined by evaporating aliquots of the aqueous protein solu- tions to dryness at 100-110øC for 18-24 hours and correcting for the residue weights. The pH of the aqueous protein solution was assumed to be the true pH for the identical solution which was 30% in DMSO. The difference in pH readings between the two solutions, as measured on a Beckman Zeromatic pH meter, ranged from approximately 0.7 pH unit at pH 4-5 to approximately 0.2 pH unit at pH 2-3. The pH of these solutions was lowered by adding equal volumes of 0.1_M HC1 to each protein solution. The 0.1_M HC1 that was added to the 30% DMSO protein solution was also 30% in DMSO.

166 JOURNAL OF THE SOCIETY OF COSMETIC CHEMISTS Relative viscosities were calculated from equation 1, •. = r•/r•o = (t/to) (p/po) where r•0, to, and p0 are the viscosity, flow time, and density of the reference liquid 7, t, and p are the corresponding terms for the protein solutions. The reduced viscosities were calculated by dividing the specific viscosity (r% = 1/r/rek-l. ) by the protein concentration. The density term P/Po Of equation 1 which generally is in the neighborhood of 1.002 (9) was measured by obtaining the weights of 1 ml of reference solution and of protein solution in weighing bottles. A 1 ml precision transfer pipet (accuracy within 0.006 ml certified by the N.B.S.) was employed for this purpose. The density of distilled water was also determined in this manner, and a density correction factor was calculated from known literature values. All densities were corrected to 25 øC. Studies on the effect of DMSO on swelling insoluble proteins (keratin) were performed using the technique of equilibrium liquid retention (11.) Experiments at room temperature and at pH 7.0 were performed on samples of brown, virgin Caucasian hair. Three tenths gram samples of hair were dried for four hours at 100-110øC, weighed, and then im- mersed in the appropriate solution (containing several drops of a nonionic wetting agent) until the uptake of liquid by the hair had reached an equilibrium value. The times necessary for equilibration varied, depending upon the treatment solvent. In pH 7.0 buffer it was 30-60 minutes and for DMSO solutions it was as long as 7 days. Equilibrium liquid retentions were measured by removing the hair samples from the treatment solutions and quickly placing them on wire supports in centrifuge tubes. The tubes were so designed that the hair samples rested 4.45 cm from the bottoms of the tubes. The tubes were capped with rubber stoppers and centrifuged for 10 minutes at approxi- mately 1000 g. The hair samples were then immediately transferred to capped weighing bottles and reweighed. The amount of liquid taken up or retained by the hair fibers was expressed as per cent of the dry weight. Swelling studies on individual hair fibers were performed by cutting them into 1 cm long pieces. Several of the pieces were immersed in pH 7.0 buffer on glass microscope slides for one hour, and a cover slip was then positioned. At the end of this time 60 readings of the fiber width were taken on a microscope (10 X magnification). The eyepiece micrometer had previously been calibrated with a 2 mm stage microm-