IN an earlier paper [Pickles, 1949] a modification of the plethysmographic method, which may be called baroplethysmography, was proposed as a solution of the problem of estimating the blood flow through the human mammary gland. It has since been shown by this method [Pickles, 1950[Pickles, , 1951 that the estimated mammary blood flow may vary in a number of ways, suggesting control by at least one vasomotor reflex and by the hormonal changes of the menstrual cycle, pregnancy, and the different phases of lactation. These results, and others obtained thermometrically [Pickles, 1952 a, b], show that mammary blood flow estimations on human subjects are of physiological and clinical value. It seems desirable therefore to review in detail the working of the baroplethysmographic instrument, and to attempt, on the basis of continued experience of its use, to assess its reliability for both mammary and other blood flow estimations.The original paper [1949] outlined the principle and the physical theory of the instrument. No direct evidence of its reliability could then be offered: human mammary blood flow could not be measured by any other method for comparison, and a technical difficulty which will be described vitiated attempts to apply the instrument to any part, such as a hand, where comparisons with a standard method could be made. This difficulty has now been largely overcome, and the present paper gives first an account of such comparative experiments on the hand.METHODS.The Plethysmograph.-The method to be described can be applied to any part that can be enclosed in a plethysmograph. It depends on certain characteristics of the venous drainage from the part, and it is therefore essential that the seal of the plethysmograph should not press on the veins. The seal has, moreover, to withstand changes of plethysmograph air.pressure up to + 3 mm. Hg, and this is the cause of the difficulty mentioned above in applying the instrument to the hand: the usual types of plethysmograph seal either interfere with the venous drainage to a slight extent, or are incapable of withstanding the necessary pressure changes. The difficulty does not arise, however, when the method is applied to the breast, since a plethysmograph can be con-175