May 23, 1941 1084 of either edema or erythema was considered to be the average of the measurements of their respective diameters. In this study the term" positive reaction" refers to a reaction with an area of edema (induration) whose average diameter measured 5 mm. or more at the time of the 48-hour reading. A reaction consisting of erythema only was considered negative, irrespective of size. The procedures for testing each group were essentially identical. An initial test employing one of the dilute solutions of tests 1 to 7 (table 1) was administered to each person. Increasingly larger doses of tuberculin were injected in those who had no reaction or a reaction consisting of an area of induration of less than 10 mm. It is to be noted that the criterion for elimination from further testing (10 mm. or more of edema at 48 hours) was different from that employed for the determination of positive reactions (5 mm. or more of edema at 48 hours). This was done to make certain that the smaller reactions were actually positive ones and n1ot artifacts. Definition of terms.-The testing of a person by the procedure outlined above determines the smallest dose of tuberculin PPD to which he will react. This dose may be defined as the" sensitivity level" of that individual. The application of this testing procedure to any population group determines the" sensitivity levels" of all of its members. For example, a certain number of the population will fall into the msost sensitive class (those reacting to the smallest dose), others will react to the next larger dose, and so on until the largest dose is reaclhed. The remainder will consist of those wlho do nlot react even to the largest dose. In order to determine the total number of persons who would react to any given dose if the entire population were tested with that dose alone, it is necessary to add the number of reactors to that dose and to all the smaller doses, since it may be assumed that those wiho react to any of the smaller doses would also react to a larger dose. For example, if test number 5 alone were applied to the entire population the number of reactors to that dose would be the number reacting to tests 1, 2, 3, 4, and 5. Therefore, in order to obtain the number of reactors to each of the doses if used alone the number reacting to the various test doses was accumulated as above and called the" accumulated number of positive reactors" to each test dose. When expressed as a percentage of the entire population tested, this accumulated number becomes the" accumulated percentage of positive reactors" to the various test doses of tuberculin. The plotting of the accumulated percentage of positive reactors against dosage results in a curve which may be defined as the" sensitivity curve" for the population under discussion. Thus, a population composed of a high percentage of persons hypersensitive to tuberculin would have a curve rising sharply for the smaller doses whereas a relatively insensitive population would