340 JOURNAL OF THE SOCIETY OF COSMETIC CHEMISTS An alternate way to obtain the moduli is from a plot of (y vs. ½, where one has • = E'½ + E" (½02 - ½2) 1/2 (6) and the shape of the curve is that shown in Figure 3 fore o = •o = 1 and8 = 60 ø . The elastic modulus is given by •(e = eo) E' = (7) •0 where it should be noted that the slope of the major axis of the ellipse is not identical with E'. The loss modulus is given by E" = E'e'/(eo 2 - e'2) •/' (8) or by area of the loop E" = (9) 2 qT •0 where ½' is the displacement at zero force. Equations (7) and (8) are easily used with the computer to find maxima, minima, and crossing points. RESULTS AND DISCUSSION The elastic modulus as measured by the GBE is quite sensitive to water, as is shown in Figure 4. Water was applied from a squirt bottle with the probe attached. The loop o (force) Oo •o (displacement) Figure 3. Force vs. displacement (g0 = ½o, • = 60ø) for linear response.

MECHANICAL PROPERTIES OF SKIN 341 Untreated Grams E' = 110 gm/cm 3 E" =30 gm/cm Water-treated E' =45 gm/cm E" = 13 gm/cm ' MM -.75 -.5 -.25 5 .5 .75 -1 Figure 4. Effect of water on viscoelasticity of human skin (back of hand) in vivo (force vs. displacement) at 20% relative humidity. shape reached equilibrium within minutes. The loss modulus is also lowered by water and the ratio E' to E" remains essentially constant. The outer layers of the stratum corneum contribute significantly to the mechanical properties of the skin, since the moduli can be reduced by almost 50% with four tape strips as shown in Figure 5 for a single individual. (Here the amount of skin removed was measured by Lowry (22) protein analysis). These elastic storage modulus results are similar to those of Christensen et al., who also observed a reduction in E' upon applying water and lotions to the skin or upon tape stripping (17). However, the loss modulus results presented here are in contra- diction with the behavior of their analogous parameter 0. They found 0 to increase, rather than decrease, upon water treatment, or upon making progressively smaller circumference circular incisions through the epidermis surrounding the probe site. We suggest this discrepancy may largely be a result of the differences in maximum probe displacement used in the two studies. Christensen et al. cite a typical skin displacement of 3-4 mm, which is up to five times greater than our maximum dis- placement of ---0.75 mm. We have found that larger displacements form nonelliptical stress-strain curves, which exhibit preconditioning and stress relaxation effects, all hall-