50 W. G. SMITH I be taken not merely of these variations, but of the fact that the writing point of the lever in moving up and down de- scribes not a straight line but the arc of a circle. In measur- ing the reaction curves I have not felt justified as a rule in doing more than dividing (with the help of lens, etc.) the space occupied by the hundredth of a second into halves, quarters or thirds. The results have all been calculated in thousandths of a second, but as the unit of measurement in the majority of the experiments was the hundredth of a second and as I do not wish to lay any stress on the more minute differences in the measurements, I have expressed all the data in terms of hundredths of a second. Table I. includes data from three subjects who showed considerable constancy in reacting according to the antagon- istic form. One subject reacted in this mode almost invari- ably : the other two, who are not so regular, were tested again after an interval of a year and gave similar results on both occasions. I was able in addition to carry out deter- minations on these three persons by means of the chrono- scope. The graphic and chronoscopic determinations were, as far as practicable, carried out at the same time, and afford an opportunity of comparing the data supplied by the two methods. In all the experiments of this table the mode of reaction was sensorial, attention being concentrated on the sound. On comparing the lengths of sensorial and muscular reaction time by means of the chronoscope, I found that they were practically the same with the subjects B and C. The subject A, on the other hand, found the muscular reac- tion more difficult, and as a fact it was considerably longer. 1 Hence it seemed simplest to adopt the sensorial type of reaction as the basis of these experiments. Table I. Subject. n. chron. mv. n. st-r. m v. r-ap.
A 57 20 03 27 18 06 05 01 B 38 17 02 25 -14 02 04 02 C 42 18 02 74 14 03 04 01 n = number of experiments: chron. = reaction time by chronoscope: mo; = mean variation : st-r= interval between stimulus and rise of curve : r-ap = interval between rise and apex of curve. 1 As this case is of some interest in relation to the discussion on re- action types, it may be mentioned that the average length of the muscular reactions (40 expts.) was '26 sec., m v *05 sec. It was instructive to
