JOURNAL OF THE SOCIETY OF COSMETIC CHEMISTS jects, who had used 2 per cent G-11 soap for a period of one week. The resultant curve showed that the Neko soap curve varied very little from the control soap curve and did not compare at all to that of G-11 soap. place to a small degree only the effect: of 2 per cent G-11 soap ap- parently is accumulative. 8. In this experiment the regular surgical scrub-up technique was compared to a modified scrub-up technique using G-11 soap (Fig. 8): Figure 7.--xt, initial washings, control soap used in all basins. Average of 6 subjects. B, control soap used in all basins. Average of 6 subjects who had used 2 per cent neko soap for a period of one week. C, control soap used in all basins. Average of 3 subjects who had used 2 per cent G-11 soap for a period of one week. In general, therefore, the above experiments showed that three 2- minute washings or even ten 2- minute washings with 2 per cent G-11 soap had slight effect in de- pressing the bacterial flora of skin as compared with the great reduction noted after one week's regular use of the same soap. Instantaneous killing of bacteria obviously took Figure 8.--/Y, subjects washed according to a routine surgical wash-up technique which included the use of iodine and alcohol after Basin 4. B, results of washings after subjects had used 2 per cent G-11 soap for one week. Here the surgical wash-up technique was modified by substituting control soap for hospital soap and omitting iodine and alcohol. Forty-five minutes' time elapsed between Basins 4 and 5. During this period subjects wore sterile rubber gloves and manipulated surgical instruments. A. The subjects washed accord- ing to the following procedure: Basin Z. The subject scooped up hospital soap and washed his hands only for thirty seconds. He then rinsed in the basin for twenty seconds.

METHODS OF TESTING A GERMICIDE 257 Basin 2. The subject picked up a sterile brush, wet it with hospital soap, and then washed the skin test areas of one limb, successively scrubbing the backs and fronts of each digit, the palms and wrists, the knuckles, and the forearms for a period of three minutes and rinsed twenty seconds. Basin 3. The subject repeated • the performance of Basin 2 on the other limb, again washing three minutes and rinsing twenty seconds. At this point the subject put aside the brush, laying it in a sterile Petri dish cover and cleansed his nails with the point of a nail file and an orange stick. No time limit was put on this action. Basin •. All the tests areas of both limbs were scrubbed with the brush, soap being used as needed for a period of four minutes, itevoting half of the time to each limb. The subject rinsed for' twenty seconds and discarded the brush. The ends of the digits to the distal interphalangeal joints were dipped into half strength tincture of iodine and withdrawn at once. The hands were briefly rinsed in 75 per cent (by volume) ethyl alcohol which was slightly tinged with iodine, and the arms were raised to allow the excess alcohol to run down the forearms. The subject wiped his hands and arms with a sterile towel, pulled on sterile muslin sleeves, powdered his hands with sterile talc and put on sterile rubber gloves. The subject manipulated sur- gical instruments for a period of forty-five minutes. The rubber gloves were aseptically removed and, the subject washed in Basins 5, 6 and 7, two minutes in each without a brush and using a bar of control soap. B. The subjects had used 2 per cent G-11 soap for one week. Here the surgical scrub-up technique was modified by substituting control soap for hospital soap and omitting iodine and alcohol. Forty-five minutes' time elapsed between Basins 4 and 5 during which the subjects wore sterile rubber gloves and manipulated surgical instru- ments. It was shown that an individual using 2 per cent G-11 soap regularly had a lower resident count after two minutes of washing than an individ- ual who had washed for twenty minutes with ordinary soap. In the surgical scrub-up the additional use of alcohol and iodine did not reduce the count or hold it as well as had the regular use of 2 per cent G-11 soap with only a two-minute wash- ing. We did not confirm Price's observation that the bacterial count was doubled in approximately forty to fifty minutes under the rubber gloves. It is possible that had our subjects actually been operating with the additional manipulation, tension, and sweating that the re- sults might have been different. The •uestion of Film Formation. Miller and his associates had shown by experiments that washing the hands with the use of a germicidal cationic detergent such as Zephi- ran chloride left the skin appar- ently free from bacteria, but subse-