The method generally employed for the evaluation of germicides is known as the phenol coefficient test. This test in one form or another is universally employed for the determination of the germicidal potency of compounds or preparations recommended for the destruction of bacteria. The most widely used modification in this country at the present time is the Ruehle and Brewer (1931), U. S. Food and Drug Administration (FDA) phenol coefficient method. The test was originally designed to be used for comparing the bactericidal potency of phenol with phenol-like compounds. However, in recent years the method has been used for the examination of compounds which are totally unlike phenol, with the result that such evaluations have been the cause of considerable confusion and of doubtful value. Another departure has been the freedom which some investigators have taken with the test itself, changing the procedure to favor the compounds under examination. Regardless of the many modifications of the original procedure which have been proposed, it is generally agreed that a phenol coefficient test is of very little or no value for determining the efficiency of a germicide intended for clinical application. A phenol coefficient attempts to compare the toxicity of a germicide for a given organism with that of phenol but gives no information as to its effect on living tissue. For example, a germicide with a high phenol coefficient and a proportionately high toxicity to living tissue would have no advantage over one with a low phenol coefficient and a proportionately low toxicity to living tissue. A new method was proposed in 1935 (Salle and Lazarus), and modified in 1938 (Salle, McOmie, Shechmeister, and Foord), for the evaluation of germicidal substances intended for clinical application (figure 1). The germicides were tested for their effect on living embryonic chick heart tissue fragments as well as for their ability to kill bacteria. Two tests were performed. In one test the bacteria were exposed to a series of dilutions of germicide for 10 min, then transfers were made to nutrient broth to determine the killing dilution. In the other test the tissue fragments were exposed to a series of dilutions of germicide for 10 min, then washed twice with saline, and finally embedded in rabbit plasma to determine the killing dilution. A number known as the Toxicity Index was calculated from the results, which was defined as the ratio of the highest dilution of germicide required to prevent growth of the tissue fragments in 10 min to the highest dilution required to kill the test organism in the same period of time and under similar conditions. Theoretically, an index greater than one means that the germicide is more toxic to the tissue fragments than to the bacteria. As the number increases in size, the germicide decreases in value. Conversely, an index smaller than one means that the germicide is more toxic to the test organism than to the embryonic tissue fragments. As the fraction decreases in magnitude, the germicide increases in value. The smaller the toxicity index the more nearly perfect the chemotherapeutic agent. A phenol coefficient is not determined, and no reference is made to it. There is no relation whatsoever between a phenol coefficient and a toxicity index. In the present investigation the tissue toxicity and the bactericidal tests have been combined in one operation making the procedure simulate actual conditions of germicide therapy as closely as possible.