For the individual men of Squad A, the average reaction times for the whole series, from October 28 to January 27, range from 186 σ (Gul) to 249 σ (Bro). Spe shows an average, 266 σ, which is still higher. He did not, however, have so much opportunity for practice. All his reaction records show abnormally long times. The average variability for the individual subjects ranges from 18 to 29 per cent. The average reaction time for the whole experiment for the ten men is 217 σ, with an average individual variability of 20.5 per cent. With Squad B the general average for the 3 normal experiments is 224 σ, with like variability of 20.5 per cent. The average for the 2 low-diet experiments is 206 σ, with a variability of 19 per cent. The averages and variabilities for the individual subjects are not particularly noteworthy. There were no normal eye-reaction experiments for Squad A. The total number of these measurements was further reduced by the treadmill experiments on January 6 and 28 with Squad B, and on February 3 with Squad A, and the standard electrocardiograms on December 20 with Squad A, as on all four dates the morning psychological program had to be omitted.1

1Dodge and Benedict, Carnegie Inst. Wash. Pub. No. 232, 1915, p. 89.

Table 168. - Squad B - The Eye-Reaction Time And Its Variability

[M. in σ, S. D. in σ, and C. in per cent].

The available data for Squads A and B are plotted in figure 112. The two upper curves are for the variability and they show about 20 per cent for both squads in the first two experiments. In the third experiment with each squad there was a definite increase which, with Squad A, was to about 25 per cent. As seen in table 167, several men in Squad A showed unusually high variabilities in that experiment, as, for example: 34.5, 34.6, and 37 for Kon, Pec, and Tom. These values are high enough to account for this fluctuation. With Squad B the percentage for the third experiment was 21.7. This is not a large lThe fragmentary data for Squad A on November 11 are due to a technical difficulty in the apparatus; the subjects were in no way to blame enough fluctuation to be particularly noteworthy. Near the end of the experiment the eye reaction time decreased, and the variability was also smaller. The curves for the general averages are at the bottom of the figure. Squad B shows a rather consistent practice effect from the beginning to end, with the exception of a slight decrease in the third experiment, which corresponds with what has been noted above for the variability. There was no lengthening of reaction time associated with the period of food reduction for this squad. Squad A shows no significant change in the second experiment over their first one. The data for the second experiment are very fragmentary; indeed, they might have been omitted, since on account of the faulty action of the shutter (see S in figure 30), many of the shorter reactions failed to appear on the records. There was improvement up to December 9 and slightly lengthened reactions during January. The increase following the shortest reaction-time average of December 9 was approximately 5 per cent.

For comparison with the general averages, we may give attention to the short reactions made by the various subjects. Short reactions, which are at the same time true reactions, would seem theoretically to measure the real process involved better than the average of all reactions, many of which are unavoidably complicated by fluctuations in attention and other disturbing conditions, objective as well as subjective. The 5 shortest reactions made by each subject at each experiment were averaged (see tables 169 and 170) and the averages of these figures for the 10 men of Squads A and B used to plot the curves shown in the middle of figure 112. The initial values are 172 σ and 182 σ for Squads A and B, respectively. The similar value for the normal series of 1917 previously referred to was 193 σ. The curve for Squad B shows a gradual and uninterrupted practice effect from the value 182 σ to the final value of 160 σ. Squad A show3 a practice effect approximately equal or parallel to that of Squad B, up to and including December 9. In the last 2 experiments there was a definite, accumulative decline. This agrees very well with the findings for the general averages of the same squad on the same dates. It therefore appears that the eye-reaction time according to the technique here employed was not changed in any significant way by the reduced diet. According to the present standards our averages and the variability values with the fluctuations in these are well within normal limits.

Fig. 112. - Eye-reaction time and its variability.

Solid lines represent Squad A and broken lines Squad B. The two curves at the bottom represent average reaction time; the two in the middle are for the five shortest reactions made by each subject at each experiment; the two curves at the top show the coefficients of variability.

Table 169. - Squad A - Average Of Shortest Eye Reactions

[Values in σ].

Table 170. - Squad B - Average Of Shortest Eye Reactions. (Values In σ]