METHODS. Determination of coagulation time.-The measurement of the coagulation time was in every case undertaken at half blood-heat, orat a temperature closely approximating to this. Thecapillary coagulation-tubes employed were made from ordinary glass tubing drawn oat in the flame. They were calibrated by introducing into the wide upper end of the tube in each case 5 c. mm. of mercury, only those tubes being selected for use where this volume of mercury formed in the capillary stem a column exactly 5 cm. in length. In carrying out this calibration, the technique described by one of us in a recent number of the'Lancet 1 was followed. Some minor modifications were, however, introduced. The first of these was that, blowing out the tubes in the ordinary way over filter-paper, we adjudged coagulation to be complete as soon as a definite coagulum made its appearance, irrespectively of the fact that such a clot might not be adhering firmly to the walls of the capillary tube. 2 Another modification which we introduced was that in our later observations-and these include all those undertaken upon normal men-we made it a practice to fill in our seriesof coagulation-tubes from a succession of slight pricks made as they were required, instead of from one deep prick made at the outset. In this way we alto-gether avoid the fallaciously accelerated coagulation which