72 JOURNAL OF THE SOCIETY OF COSMETIC CHEMISTS comparable results determined un- der similar experimental conditions .on days on which this preparation was not given to the subjects. Comparisons of the curve for per- spiratory odors in Chart VII with the curves in Charts IV and VI will afford convincing evidences of the failure of the chlorophyll deriva- tive either to retard or to inhibit the chemical and biological reactions which, on. the surface of the skin, re- sult in production of offensive odors. However, administration of the water-soluble chlorophyllins did ef- fect substantial reductions in the in- tensities of odor of which the source was an end product of chemical and biological reactions in the gastroin- testinal tract. Subsequent experi- mental studies have afforded indi- cations that chlorophyllins adminis- tered in a form which protects against precipitation by acid gastric contents and assures solubility of the chlorophyll derivatives in the intestinal contents, within certain limitations, may function as deo- dorizers and inhibitors of develop- ments of malodorous end products of metabolism of some species of bac- teria. However, these deodorant capacities of the chlorophyll deriva- tives are dependent upon contacts of the deodorizer with the sources of the odors. Hence, these effects of chlorophyll derivatives should be classified as manifestations of local rather than of systemic actions. In concluding this summary, it has been deemed advisable, and, it is hoped, profitable, to direct atten- tion to three basic principles which merit primary consideration in the planning and execution of in-vivo experimental studies of deodorants. First, the methods of use of a prod- uct under test should be made to conform qualitatively with actual use-conditions, in so far as the latter are reproducible in experimental procedures. From a quantitative standpoint, however, the reproduc- tions of use-conditions must be made so exacting that even the most ef- ficient deodorant would be incapable of exhibiting a 100 per cent perform- ance because a 100 per cent deodor- ant effect, under any set of experi- mental conditions, invariably leaves unanswered the question whether this result is representative of the maximum efficiency of the product under test. The second principle concerns the relative number of control experiments and of compa- rable experimental trials of a deodor- ant preparation. Lengths of con- trol periods and the number of ex- periments within these periods must be adequate for determinations of the day to day variations in develop- ments of perspiratory odors by any one subject and, also, the variations among subjects in respect to their capacities to produce odors under the conditions of the tests. An ex- perimental period should include two series of control experiments, one preceding and one following the test period. The number of experi- ments within the second control period need be sufficient only to demonstrate a return in the direc- tion of the mean result of the first control experiments.

TESTING DEODORANTS WITH CHLOROPHYLL AND DERIVATIVES 73 Of the three principles, the one which has the greatest potentiality of influencing, either constructively or destructively, the practical sig- nificances of experimental studies of deodorants relates to the criteria adopted in assembling groups of experimental subjects. A common practice adopted in clinical trials and in in-vivo experiments has been, over a period of several decades, the utilization of ambulatory clinic pa- tients or hospitalized cases as test subjects. Frequently, these groups have been supplemented by private patients reporting to physicians' offices at intervals of seven days or more. For experimental studies of deodorants, this procedure of select- ing subjects is wholly unsatisfactory and may be productive of misleading experimental data. The entire cutaneous areas of skins of all subjects should be free of functional and (•rganic disorders, in so) far as these abnormalities are recognizable by a qualified specialist in dermatology. It is a matter of common knowledge that among dill ferent individuals, marked varia- tions in rates of excretion of sweat may occur under identical conditions of thermal stimulations. Experi- ences, over a period of eighteen years, in conducting sweating tests, during control periods, by the method described in the discussions of experiments presented in Charts IV and V, have shown that adult female subjects may produce from 50 to 500 c.c. of perspiration dur- ing immersions in water at tem- peratures between 40 and 45øC. during periods of thirty minutes. Although in some cases, two or three months intervened between two series of tests, the daily volumes of perspiration collected from any one subject were surprisingly uni- form through two or more series of tests. Some subjects gave volumes from 50 to 100 c.c. these individu- als are referred to as the "light sweaters." The yields from the majority of subjects ranged from 150 to 300 c.c. of perspiration these women are called the "moderate sweaters." From a relatively small group of subjects, volumes of per- spiration collected in one test fell within the extremes of 325 to 500 c.c. these cases are classified as the "heavy sweaters." It has been the invariable policy of this laboratory, in experimental studies of antiper- spitants and deodorants, to include in a test of subjects all c)f the "heavy sweaters" who are available and then to complete the group with "moderate sweaters." Light sweat- ers are never used in studies of either deodorants or antiperspirants. In respect to either bacterial populations of perspiration or counts of viable microi3rganisms on their skins, variations among subjects are definitely greater than variations among volumes of perspiration. Extremes of variations, among counts of organisms, are from ap- proximately 2.00 millions to about 18.00 millions per 1 c.c. of perspira- tion. In general, the "moderate" and "heavy sweaters" show the higher and highest densities of bac- terial populations of perspiration.