In connection with the walking experiments the uniformity in the number of steps and in length of step while walking a given distance was studied by means of photographic records. (See p. 129). These data are in table 144. From the distance walked, length of period, and the number of steps as computed from the photographic records, the average number of steps taken per minute, length of step, and total number of steps per 100 meters are found. It is evident that if the mill is maintained at a definite speed, these last three values are functions of each other, and if the man shortens his step he must take more of them to cover a given distance.

By comparing the results of Squad B on January 6 and 28, it is seen that Ft* took longer strides and fewer steps per 100 meters on January 28 than on January 6, and that Sne had the same length of stride and number of steps per 100 meters on the two days. The record for How for January 28 is not available for comparison.

With these exceptions, the other members of the squad all shortened their stride and increased the number of steps per 100 meters on January 28. That this change in the length of stride is characteristic throughout the whole period, and that there is no marked alteration as the walking progressed, is shown by the figures given in the same table giving the length per step at the sixth, twelfth, and twenty-fourth minutes, as calculated from the photographic records and the distance traveled during that portion of the experiment when the record was made.

On January 6 Har showed an apparent change in stride following the sixth minute and on January 28 Sch apparently shortened his stride. ' Aside from these variations, there appear no marked changes except that the one-minute observations of January 6 indicate more irregularities than the other periods; this is probably due to the regulating of the mill, which, as previously stated, generally took a minute before adjustment to the desired speed of 70 meters per minute was secured. For this reason the figures in this column have not been included in the averages for this day.

As normal data for Squad A were not obtained, the data for this squad observed with low diet can only be compared as an average of 11 men with the average of the 11 men of Squad B. Such a comparison is of less significance than the comparison of the individual changes for each man, for one or two men with long legs might alter the average of the groups considerably. However, the two measurements W and R employed in measuring the body-surface area of the men on January 27 and February 2 (see pp. 234-237), show that the difference in length of leg was slight, the length for Squad B averaging 89.7 cm. and that for Squad A 88.7 cm. The averages of the two groups are accordingly compared in table 145 which shows the decrease in the average length of stride and the increase in the number of steps per minute and per meter for the three tests.

Table 144. - Record Of The Steps Per Minute, Length Of Stride, And Number Of Steps Per 100 Meters For Squads A And B On January 6, 28, And February 3

Table 144. - Record of the steps per minute, length of stride, and number of steps per 100 meters for Squads A and B on January 6, 28, and February 3, together with the length of step as computed from the photographic records on the sixth, twelfth, and twenty-fourth minutes of walking.

Squad B, Normal, Jan. 6, 1918.

Table 145. - Comparison Of Average Of Distance Walked, Steps Taken, And Length Of Step Of Squads B, Normal, B 20-Day, And A 120-Day During Walking In The Treadmill Chamber

How close may be the connection between these results and the lowering of the heat output per horizontal kilogrammeter is uncertain. That it may have some bearing is reasonable to suppose. Not a little of the actual work of forward progression consists of lifting the body on the toes as one leg is swung past the other. Benedict and Mursch-hauser1 published some measurements on this toe-lift and found that their Subject I, who weighed 73.1 kg., in walking at a rate of 75.9 meters per minute, lifted the body on the average 3.78 meters per minute, which was equivalent to 0.65 large calorie, or 23 per cent of the increase over the standing metabolism.

Jendrassik2 has shown that in horizontal walking the body is held longer in equilibrium on one foot, while in grade walking the equilibrium is maintained longer on both feet. It is possible that in these experiments, with a lowered nutritional level, the body endeavored to spare itself the effort of maintaining equilibrium on one foot and excessive toe-lift, and found in a shorter and quicker step a means to this end. It is, at least, an interesting coincidence that Fis and Sue, who showed no decrease in length of stride nor increase in number of steps per 100 meters on January 28 over January 6, both showed an increase in the heat output per horizontal kilogrammeter. The cases of Har and Liv, who also showed increased cost per horizontal kilogrammeter, are in contrast to the behavior of these two.

1 Benedict and Murochhauser, Carnegie Inst. Wash. Pub. No. 231, 1915, p. 80. 2 Jendriadk, Arch. f. Phya., 1904, Supp. Band, p. 287.