Fig. 9. Ground Plan of Reaction Box for Electrical Stimuli (Group 2). IC, interrupted circuits; CC, chronoscope circuit; X, key for making stimulus circuit and breaking chronoscope circuit; B, stimulus battery; S, string from reaction key to animal. Scale 1/2.

The reactions now under consideration were taken in sets of 24 in order to furnish evidence on the problem of fatigue. The stimulus was given at intervals of one minute, and the subject was moistened at intervals of ten minutes. To obtain 24 satisfactory reactions it was usually necessary to give from thirty to forty stimulations. Five animals, numbers 1, 2, 4, 5, and 6, served as subjects. They were green frogs whose size and sex were as follows:

Length.Weight.Sex.
Number 17.5cm.35 grams.Male.
Number 27.3"37"Male.
Number 48.2"50.4"Female?
Number 57.1"25"Female.
Number 67.8"42"Male.

For most of these frogs a one-cell stimulus was near the threshold, and consequently the reaction time is extremely variable. In Table X. an analysis of the reactions according to the number of repetitions of the stimulus requisite for a motor reaction has been made. Numbers 1 and 5 it will be noticed reacted most frequently to the first stimulus, and for them 48 satisfactory records were obtained; but in case of the others there were fewer responses to the first stimulus, and in the tabulation of series 1 (Table XI.) averages are given for less than the regular sets of 24 reactions each.

TABLE X.

ANALYSIS OF REACTIONS TO ONE-CELL STIMULUS.

Frog.Reactions to
first Stimulus.		To 2d.To 3d.To 4th.To 5th.More.Total No.
of Reactions.
1532100157
220125541258
43115102857
55111120166
64515631575
Totals,200551410727313

Table XI. is self-explanatory. In addition to the usual averages, there is given the average for each half of the sets, in order that the effect of fatigue may be noted. In general, for this series, the second half is in its average about one third longer than the first half. There is, therefore, marked evidence of tiring. The mean reaction time for this strength of stimulus is difficult to determine because of the extremely great variations. At one time a subject may react immediately, with a time of not over a fifth of a second, and at another it may hesitate for as much as a second or two before reacting, thus giving a time of unusual length. Just how many and which of these delayed responses should be included in a series for the obtaining of the mean reaction time to this particular stimulus is an extremely troublesome question. It is evident that the mode should be considered in this case rather than the mean, or at least that the mean should be gotten by reference to the mode. For example, although the reaction times for the one-cell stimulus vary all the way from 150σ to 1000σ or more, the great majority of them lie between 200σ and 400σ. The question is, how much deviation from the mode should be allowed? Frequently the inclusion of a single long reaction will lengthen the mean by 10σ or even 20σ. What is meant by the modal condition and the deviation therefrom is illustrated by the accompanying curve of a series of reaction times for the electric stimulus of group I.

__________________________________________________________________________
_8_|______________________________________________________________________
_7_|_____________________________________|________________________________
_6_|_____________________________________|________________________________
_5_|_____________________________________|________________________________
_4_|________________________________|____|____|___________________________
_3_|____________|___________________|____|____|___________________________
_2_|_______|____|____|_________|____|____|____|____|______________________
_1_|__|____|____|____|_________|____|____|____|____|____|____|____|____|__
100 110 120 130 140 150 160 170 180 190 200 210 220 230230

The column of figures at the left indicates the number of reactions; that below the base line gives the reaction times in classes separated by 10σ. Of thirty-one reactions, seven are here in the class 170σ. This is the model class, and the mean gotten by taking the average of 31 reactions is 162σ. If the mode had been taken to represent the usual reaction time in this case, there would have been no considerable error. But suppose now that in the series there had occurred a reaction of 800σ. Should it have been used in the determination of the mean? If so, it would have made it almost 30σ greater, thus removing it considerably from the mode. If not, on what grounds should it be discarded? The fact that widely varying results are gotten in any series of reactions, points, it would seem, not so much to the normal variability as to accidental differences in conditions; and the best explanation for isolated reactions available is that they are due to such disturbing factors as would decrease the strength of the stimulus or temporarily inhibit the response. During experimentation it was possible to detect many reactions which were unsatisfactory because of some defect in the method, but occasionally when everything appeared to be all right an exceptional result was gotten. There is the possibility of any or all such results being due to internal factors whose influence it should be one of the objects of reaction-time work to determine; but in view of the fact that there were very few of these questionable cases, and that in series I, for instance, the inclusion of two or three reactions which stood isolated by several tenths of a second from the mode would have given a mean so far from the modal condition that the results would not have been in any wise comparable with those of other series, those reactions which were entirely isolated from the mode and removed therefrom by 200σ have been omitted. In series I alone was this needful, for in the other series there was comparatively little irregularity.