466 JOURNAL OF COSMETIC SCIENCE Table III Screening Experiment Showing Efficacy of Complex Over Systems Used as Controls % Mending % Mending after after Durability Formula type treatment combing index 2.00% neutralized PVM/MA copolymer 90 15 0.17 2.00% Polyquaternium-28 75 40 0.53 2% Complex 0.20 neutralized PVM/MA copolymer and 1.80 polyquaternium-28 90 75 0.83 1.00% neutralized PVM/MA copolymer plus 1.00% polyquaternium-28 85 20 0.24 the same results were obtained. This provided evidence for two things. First, that the results were reproducible giving validity to the test method. Second, the complex could be formulated into a system without loss of efficacy. Results are portrayed in Figure 11. In both cases the durability indices of the mending after combing was over 70%. As mentioned in the procedure there are 100 fibers tested for each treatment, twenty fibers tagged on one tress times five tresses. The data illustrated in Figure 12 shows the variability of mending for each tress for both before and after combing on the full formulation. The average and standard deviation of the mending results after combing of the five tresses is 67.6 ± 12.6%. Comparing this mending data to a control formula containing just the thickening agent as can be seen in Figure 13 shows the significance of the difference. From this it can be concluded that the complex is responsible for the high mending scores. 100 80 � 60 .5 "'C C 40 20 0 Stereomicroscopy Results of Mending Efficacy Primary Splits, Complete+partial Mending -.-------7"'n"'l'"""7----� Complex alone, 149A before com bing §I In a full formula, 61 B 68 67.6 after combing Figure 11. Total percent mending of complex alone and in a framework formula both before and after combing.

100 90 80 C 70 =s C 60 G) :Ii 50 C 40 G) 30 G) 20 10 0 2006 TRI/PRINCETON CONFERENCE Stereomicroscopy Results of Mending Efficacy Primary Splits, complete+partial mending 95 467 Ave. = 93.3% +/- tress 1 tress 2 tress 3 tress 4 tress 5 �before com bing � after com l: - insj 2.4% ➔ average of 100 fibers, 67.6%+/-12.6% mending after combing Figure 12. Variance of mending efficacy for five tresses used in one experiment. Complexes based on different weight ratios of cationic to anionic polymer were tested for split end mending efficacy. It was discovered that although there were microgels formed by the interaction of Polyquaternium-28 and PVM/MA Copolymer, an optimal amount of repair was evident when the two polymers were combined when their charge ratios were 1:1 as can be seen in Figure 14. Split end mending efficacy is reduced as the polymer ratio deviates from the one to one charge ratio. This is further evidence in showing the importance of the polyelectrolyte complex in mending split ends. The advantage of the developed method is to determine the fate of the split ends through various treatments that test the durability of the mend. Since the tagged split end fiber is always part of a tress it can be treated in such a way that it is exposed to realistic treatments. In the previous case the tress was subjected to a controlled amount of combing to test the durability of the mend. Other treatment regimens can be conceived to test durability in a realistic fashion. A test was devised to determine the effect of multiple treatments of the formula with the complex and its removability with sham- poo. One cycle consisted of treatment with the formula, drying, combing, and then washing. Split end mending was assessed after treatment before combing, after treat- ment after combing, and finally after washing for one, two and five cycles of the treatment schedule. Results of the test are illustrated in Figure 15. First it can be observed that mending is high after treatment but before combing for treatment cycle one through five. Durability to combing stress over the multiple treatment cycles stays the same and is approximately 60%. Durability to washing is low and decreases with continued use. It was observed during the mending tabulation that the split ends became more severe with continued washing which explains the lowering of the percent mend-