The double key, DK, when raised to touch the contact C, closed the tachistoscope circuit, which passed through the electromagnet, M, holding the shutters SS' of the tachistoscope closed, and thereby cutting off the image of light at I. The slightest

TABLE VIII

The Effect of Variation in Intensity on the Effect of Variation in the Preparatory Interval, in the Case of Reactions to Light with the Regular Procedure.

Subject Vs.

N, for each interval of each series 25. Total No. of reactions 3700.

downward movement of the key, DK, broke the tachistoscope circuit, and made the chronoscope circuit through the contact C', thus causing the shutters to fly open and the image to appear at I at the same moment that the chronoscope circuit was made. The time between the break at C and the make at C', however, was long enough to be measurable, and was found to be about 2 σ, already mentioned as a constant error.

With each of the four intensities reactions were taken with preparatory intervals varying from I to, usually, 28 secs. Twenty-five reactions were taken in succession with each interval, beginning with the shortest interval and going up to the longest and then stopping for that day. Each Series in Tables VIII and IX gives the results for one day's sitting. The intervals are given in secs. in the first line of the tables, headed Interval. The

TABLE IX

Same as Table VIII, but Subject Ww.

N, for each interval of each series 25. Total No. of reactions = 1675.

54 42

Average for all intensities.. 290

95 104 98 Average... 376 336 427 506 545 576 638 273 312 359 380 406 427

80

80

average reaction time and the mean variation in σ, are given in each case, the mean variation in italics under the corresponding average. The intensities are stated merely as bright, dim, quite dim and very dim. A more accurate idea of the intensity may be gained by a glance at the reaction time obtained with the 2 sec. interval. The results of Table VIII are shown in graphic form in Fig. 7, and those of Table IX, in Fig. 8.

200

12

16

20

28

24 The abscissae represent the duration of the preparatory interval in secs., and the ordinates the average reaction time in σ. Each curve represents the results obtained with one intensity of stimulus, the lowest curve corresponding to the greatest intensity and the uppermost curve to the weakest intensity.

The results presented in Tables VIII and IX and in Figs. 7 and 8 lead to a number of interesting conclusions. Studied merely for the effect of prolongation of the preparatory interval on the reaction time, they show in the case of each intensity the same general effect that has already been described as indicated by the results of Table I (Chapter II). In the case of each subject, with each intensity, the effect of prolongation of interval beyond the most favorable point is to produce a marked and regular increase in reaction time, and this increase occurs at first at a

more rapid rate than later. Since the curves with each intensity have the same general form, I have plotted in Fig. 9, separately for each subject, the average of the results obtained with all four intensities. This gives two curves, of a high degree of reliability, showing how interval affects reaction time, taking the average for four different intensities of stimulus. The curves for the two subjects are very similar to each other and also very similar to the curves of Fig. 2, Chapter I, which show the effect of interval on sound reactions for each of three subjects. The curves for each subject are in accordance with the formula, already arrived at in Chapter I, y=A+B. log x, in

The abscissae and ordinates have the same significance as in Fig. 7.

24

1

which y is the reaction time, r is the duration of the preparatory interval, and A and B are constants. In other words, the reaction, for any preparatory interval of 2 secs. or over, is equal to a constant plus the product of another constant into the logarithm of the interval. The deviations of the reaction times experimentally obtained from those calculated in accordance with the equation are shown, for each subject, in Table X. In the

FIGURE 9

Graphic Representation of the Average Reaction Time for All Intensities for Each of Two Subjects.

The abscissae and ordinates have the same significance as in Figs. 7 and 8. The uppermost curve is the curve obtained by combining the four curves of Fig. 7 and the lower curve that obtained by combining the four curves of Fig. 8.