The limited number of procedures affording results in units of weight [Y. Kuno (l', G. E, Burch and collab. (2, a and b), J. S. Weiner (3)1 does not include any method which would be sufficiently convenient for extensive use under the ordinary conditions of a clinical laboratory. Quite recently E. Cohen (4) devised a quantitative method for determining the amount of palmar sweat by weighing a square of blotting paper before and 1) involves the risk of a weight loss through evaporation before the weighing is ended (4); 2) hardly eliminates the possibility of an admixture of fatty matter, or cellular and horny particles. In addition, the technic is still too inconvenient and time consuming for a great number of examinations within limited periods of time. In recent years we obtained accurate measurements by means of a special electrohygrometer, but its use was likewise hardly practicable for medical examinations on a large scale, since too much time was consumed by a single determination; moreover, the intricate construction of the apparatus necessitated frequent adjustments. The only procedure known to us for examining the degree of sweating which can be employed with relative ease, is the application of reagents which produce color changes in the presence of water. Their use also affords a most valuable means for direct examination of outpouring sweat, exclusive of insensible cutaneous perspiration. Until the present time, however, color indicators could be used only for approximations for which more or less arbitrary standards (5, 6) were chosen. Another handicap in the application of indicator-methods is the usually strong objection of both the subjects under test and those performing the examination, to the unavoidable discoloration of clothing, skin, etc., with staining products formed by the indicatorsubstances in use. Thus, the dark blue salt resulting from the reaction of ferric chloride with tannic acid in the presence of traces of water [method of J. J. Silverman and V. E. Powell (7)1 often causes irreparable damage to clothing and leaves stains on the skin which are removable only with difficulty. Similarly, though to a somewhat milder degree, the purple product of the starch-iodine reaction [methods of Minor (8), W. C. Randall (9), and othersl is apt to soil clothing. One of the greatest disadvantages of almost all the indicator-methods employed thus