252 JOURNAL OF THE SOCIETY OF COSMETIC CHEMISTS brush, and allowed to dry for five minutes. Three-tenths gram of the soap to be tested was dissolved in 3 cc. of sterile distilled water and added to the hand with the dried culture on it, rubbing up into a good lather on the palm by means of an- other sterile brush for a period of two minutes. The total lather was removed from the hand by means of the brush and dissolved in 500 cc. of sterile distilled water. One cubic centimeter of this solution was plated in agar and the number of bacteria present determined by counting the colonies that appeared in standard beef broth agar after incubating the same in a Petri dish for a period of forty-eight hours. The possibility of bacteriostatic effects here were exclu .ded by streak- ing the sterile plates containing the G-11 with a fresh staphylococcus culture and observing its normal growth. Under these conditions of testing we found: (A) Using ordi- nary soap for the washing, the plate count showed ll0 staphylococci and a few spore formers present (B) using soap of the same composition as in (A) but containing 2 per cent of Compound G-11, the plate was sterile (C) using the same amount of "Neko" soap (Parke, Davis 2% HgI2 soap), there were no staphylococci on the plate, but ap- proximately a dozen spore formers or "spreaders" were present. This was not the type of hand- washing experiment carried out in our remaining tests. All of our washing experiments were con- ducted in the following manner: 1. Ten sterile basins used. 2. Each contained 2000 c.c. freshly distilled H•.O at 23- 28øC. 3. Measured lengths of hands and forearms (14 inches from tip of middle finger). 4. Marked off with "wet-proof" tape. 5. By stop watch: 25 seconds to lather. 75 seconds to scrub (no ' brush). 20 seconds to rinse in basin. Figure 1.--,•/, washings, control soap used in all basins. Average of 25 subjects. B, washings, 2 per cent G-11 soap used in all basins. Average of 6 subjects.* . Subject moved to next basin and procedure repeated. Within two minutes after subject finished at each basin two 0.1-cc. samples removed and each placed in sterile Petri dish, 1S cc. of melted beef heart infusion hormone agar pH 7.2 at 45øC. added to each dish.

METHODS OF TESTING A GERMICIDE 253 8. Incubated forty-eight hours and then counted on Quebec colony counter. 1. In the first series of experi- ments-Fig. 1: A. Control soap was used to wash through all 10 basins on an average of 25 subjects. Figure 2.--,4, initial washings, control soap used in all basins. Average of 25 subjects. 2•, 2 per cent G-11 soap used in basins 4, 5, and 6. Control soap used in all other basins. Average of 16 subjects. C, 4 per cent G-11 soap used in basins 4, 5, and 6. Control soap used in all other basins. Average of 6 sub- jects. the curve when G-11 was used be- gan at a slightly lower level and re- mained slightly lower than the control curve, but the result was not striking. 2. In the second experiment-- Fig. 2: Figure 3.--,4, 70 per cent ethyl alcohol used in basin 5. Control soap used in all other basins. Average of 9 subjects. B, 70 per cent ethyl alcohol used in basin 5 for 50 seconds only, followed by washing with control for 5 seconds. Average of 6 subjects. C, 2 per cent G-11 soap used in basin 5. Control soap used in all other basins. Average of 9 sub- jects. B. Two per cent G-11 soap was used in all 10 basins on an average .of six subjects. This was done to determine the immediate killing effect of G-11. On ß counting the colonies cultured from the various basins it will be seen that * These tables (Figures 1-8) are reprinted by special permission of Surgery, Gynecology and Obstetrics. A. Control soap was used in all 10 basins for an average of 25 sub- jects. B. Two per cent G-11 soap was used in basins 4, 5, and 6. Control soap in all other basins on an average of 16 subjects. C. Four per cent G-11 soap was used in basins 4, 5, and 6. Control