484 JOURNAL OF THE SOCIETY OF COSMETIC CHEMISTS .._.. 10 3 3:10 2 Nourishing C. ( 1 ) ß Nourmshlng C. ( 2 ) ß Nourishing C. ( 3 ) ß Cold C. (1) ß Cold C.(2) ß Cleansing C. ( 1 ) ß Cold C. ( 3 ) ß Nourishing C. ( 4 ) ß Nourishing C. (5) ß Cleansing C. ( 2 ) 10 2 10 3 10 4 Viscosity (cps) Figure 2. Hardness and viscosity of cosmetic creams Apparatus For Measuring Hardness and Viscosity Hardness was measured using a curd tension meter. In this instrument, a load is applied to the penetration rod which is fitted with a round disk (1.5- 10.0 mm in diameter) at the bottom. The hardness, expressed as grams per cm 2, is the value of the load just as the disk begins to penetrate into the sam- ple. Viscosity, expressed as cps, was measured with a Ferranti-Shirley Cone and Plate Viscometer at a shear rate of 1700 sec -• and a sweep time of 10 sec. Sensory Evaluation Sensory evaluation was carried out by trained panels of 10 subjects (5 males and 5 females each). The panel was instructed to arrange each series of sam- ples in their order according to their firmness or viscousness. This was repeat- ed 3 times by each individual. Evaluation of firmness was made by pressing the surface of the samples in the container with a finger, and viscousness was evaluated by taking out a small amount of the sample with the forefinger and applying it to the forearm. In the case of the standard, the spring was pressed down by pushing with a finger at the center of the cardboard and the resistance which is sensed when the spring is compressed 1 cm by a finger pressure was evaluated as "firm- ness," and "viscousness" of silicone oil was evaluated in a similar manner as in the case of the cream base substances.

SENSORY FIRMNESS AND VISCOUSNESS 485 Smnple Group Table II Correlation Coehqcients Sensory Firmness and Sensory Viscousness and Measured Hardness Measured Viscosity Average Correlation Average Correlation Hardness Coehqcient Sample Viscosity Coehqcient (g/cm •ø) (r) Group (cps) (r) A 8.8 x 10 • 0.812 a 7.9 x 10 • 0.660 B 8.5 x 10'-' 0.893 b 1.5 x 10 • 0.860 C 2.5 x 10 •ø 0.934 c 5.5 x 10-0 0.873 D 9.5 x 10 • 0.748 d 2.2 x 10 '0 0.831 E 3.4 x 10 • 0.580 e 9.0 x 10 • 0.905 ]•ESULTS Experiments on Cream Base Substances Correlation Coefficient Through Spearman's rank method, the correlation coefficient betxveen sen- sory firmness and measured hardness xvere given, and the same calculations xvere also made betxveen sensory viscousness and measured viscosity and are shoxvn in Table II. It xvas found that the correspondence of sensory firmness to measured hard- ness xvas shoxvn, on the whole, in a comparatively high hardness range hoxv- ever, in the extremely high hardness range the correlation betxveen the txvo xvas shoxvn to be small. On the other hand, the correspondence of sensory viscousness to measured viscosity xvas shoxvn to be essentially in a low vis- cosity range. Discrimination Threshold Differential threshold is usually obtained by using the paired comparison method, but the number of samples was not sufficient for carrying out the paired test in this experiment. Therefore, another method has been designed and employed. First, the panel members xvere asked to rank and arrange one series of samples composed of 6 or 7 samples according to the order of sen- sory firmness or viscousness. Next, every possible pair from the previously ranked data according to the sensory firmness or viscousness xvere compared in each pair, and xvas regarded as the correct judgment if it corresponded to the measured order. Accordingly, the percentage of correct judgments was calculated. All data were plotted xvith the percentage of correct judgments as the ordinate and the relative difference as the abscissa. The relative difference (A) is the proportion of the difference to a mean value of txvo samples ob- tained by instrumental measurement and is shoxvn as percentage.