A modification of the Stewart method, whereby heat elimination in calories per min. is measured, was found accurate for the detn. of blood flow in the finger tip. After release of sympathetic tone the blood flow per sq. cm. per min. was calculated from the formula, [image] , in which F was the blood flow in cc. per sq. cm. per min.; [DELTA]1, the rise in temp. of the calorimeter per min. after the finger was immersed; [DELTA]2, the fall in temp. of the calorimeter after removal of the finger tip; m, the vol. of water in the cup; e, the hydro-thermic equivalent of the cup, stirring device and thermometer bulb; s, the specific heat of the blood (approx. 1.0); a, the area of the immersed finger tip in sq. cms.; t3, the temp. of the arterial blood (mout temp.) and t4, the temp. of the venous blood (av. calorimeter temp.). The av. normal blood flow was 0.253 cc. per sq. cm. per min. [plus or minus] a standard deviation of 0.026 cc. The brachial-digital arithmetic mean arterial blood pressure gradients were also detd. after release of sympathetic tone. The avg. normal gradient was 23.1 mm. Hg [plus or minus] a standard deviation of 7.1 mm. In essential hypertension, aplastic anemia, hyperthyroidism, acromegaly and diverse chronic infections all without clubbing, the blood flows and pressure gradients were within the normal range. In simple symmetrical clubbing abnormally high blood flows were found, caused, in part at least, by increased digital-arterial pressures (decreased gradients). In hereditary clubbing and in hypertrophic osteoarthropathy the digital blood flows and blood pressures were within the normal range. The reasons for this are discussed.