Of the measurements at present available, the motor coordination for successive horizontal eye movements is one of the best indicators of the neuro-muscular condition.1 The finger movements, as a motor process, compare very favorably with the eye movements, but in the case of the latter it is not so possible, if indeed it is at all possible, for the subject to influence them voluntarily. The time required for making a horizontal sweep of the eye through 40° on the arc of vision is almost a neuro-muscular constant for the individual and for his physiological or neuro-muscular condition.

Eye-movement records were taken with 63 men in the normal series of 1917. They were all recorded from the right eye and with the left eye covered.2 The procedure was identical with that used in the low-diet research. Four series of records, i. e., two plates (for illustrative records, see fig. 52) were made with each subject, as in the present investigation. The number of eye movements left or right which are available for counts in the case of a particular subject is usually 20. Therefore, the average figure for any man on any date is usually drawn from this number of counted movements. The range of eye-movement speed shown by the men in the normal series of 1917 is tabulated in table 179. The units are 0.001 second (σ) and the column headings in the table are self-explanatory. Two subjects show speeds for movements to the left which average 76 and 79 σ, respectively. These fall in the group division 71 to 80. The slowest eye-movement times for left movement were the averages 141 and 148. The distribution between these high and low points is fairly characteristic of normal frequency. The mode is clearly at 101 to 110, 24 of the 63 subjects showing averages which came within this range. The total average for the left movement is 107.0 σ and the standard deviation for the series of 63 left movement averages is 13.9 σ, showing a coefficient of variability within the group of 13 per cent. The movements to the right fall within a narrower range than those for the left. The average fastest speed is 75, and the slowest 128. The distribution between these extremes is sensibly normal, the mode occurring at 91 to 100. The average for the whole series of 63 individual averages is 97.8 σ. The standard deviation is 11.7 σ and the coefficient of variability for movements right therefore equals 12 per cent.

1Dodge and Benedict, Carnegie Inst. Waah. Pub. No. 232. 1915, pp. 146 and 262. 2There were a very few exceptions to this latter statement due to errors in the procedure.

The measurements taken on the normal subjects during normal experiments by Dodge and Benedict in connection with their alcohol investigation gave the following results as averages of the first normal series for 7 subjects: Speed for movements to the left, 101 σ; for movements to the right 99 σ. There is fair agreement between these totally different groups of individuals.

With the subjects used by Dodge and Benedict, the eye movements were taken in the afternoon and evening. In the case of the normal series of 1917 the time of day was always between 7 and 10 p. m. and following a hearty supper. The measurements with the subjects on reduced diet were invariably taken in the morning and following the light breakfast at the Laboratory. The eye reactions and eye movements were successive measurements.

The average results for Squads A and B are given in tables 180 and 181. We have no normal records for Squad A. The first three dates, that is, November 4, 18, and December 16, for Squad B are normal in the sense that the diet had not then been reduced. November 4 (see table 181) shows an average for movements to the left of 95.1 σ, and for movements to the right of 88.9 σ. The time required for movements of 40° to the left, registration being from the right eye and with the left eye covered, is noted to be longer than similar movements to the right. As this is seen to be characteristic of the records for the low-diet research, it is prominent also in the averages of the normal series of 1917, and moreover in the normal records taken by Dodge and Benedict. The reason for this discrepancy between the time requirement for right and left movement can not be definitely assigned at this time. We have considerable data on the problem and it is under investigation at the Nutrition Laboratory. The averages for left and right are, as in the case of the normal data of Dodge and Benedict, somewhat below the values shown for the 63 men of the series of 1917. This may be a mere matter of chance, or on the other hand partly due to the time of day of taking the records. Undoubtedly sleepiness slows down the eye movements.1 The variability in the eye-movement records shown with Squad B, November 4, is slightly above 8 per cent as an average for the variability figures of the individual series. Individual subjects show variabilities which range from 4 per cent with Lon, left, to 12.7 per cent with How, left.

Table 179. - Range Of Eye-Movement Speeds Shown By A Group Of 68 Normal Men, Series Of 1917

[Records taken from the right eye and with the left eye covered].

Table 180. - Squad A - Eye Movement Speed And Its Variability

[L. and R. designate left and right movements, M. and S. D. in σ, and C. in per cent].

1From a totally different series of experiments we have records on one subject who, when the records were taken, was very sleepy and his eye movement time for 40° movements is in the range of 280 to 800; he started the series of movements at the signal from the operator, but the plate shows that he very soon closed his eyes and was asleep.