No great significance should be attached to the pedometer readings without due consideration of the factors influencing them, as otherwise they might lead to false conclusions. As pointed out earlier, the pedometer actually records the up-and-down motion of the body and is supposed to be used exclusively for indicating the distance walked in horizontal forward progression. As previously stated, from the various activities of the men in these squads, we are certain that the pedometer readings may not be directly considered as so many miles walked, especially when we are attempting to attribute a quantitative energy value to the several readings. In other words, each unit recorded on the pedometer may by no means have the same calorific significance. Thus, in going upstairs, each step requires much more energy than a step in level walking, and yet it would be recorded as one unit by the pedometer. Indeed, with level walking a marked change in the character of the terrain would alter the calorific value of each pedometer unit. Walking on a smooth and level sidewalk would have one value; walking over slippery ice would have a value which would be the resultant of a shorter step and accompanying increase in the leg tension for balance. These pedometers were worn the entire day. The men were cautioned, when riding in automobiles or trolley cars, to be sure that the motion of the vehicle did not cause registration. These false records, however, probably play very little role in the series as a whole. It is perfectly conceivable that the activity indicated by a half hour of wrestling or calisthenic exercise in the gymnasium, with jumps up and down, is not at all comparable to the equivalent number of pedometer units registered while walking, hence, it must be recognized that the pedometer does not differentiate in the character of these various units.

1Four of the subjects, Kon, Mon, Moy, and Tom, show their highest walking record in the Thanksgiving reeess, and Vea his surprisingly high record during the Christmas vacation. As may be seen from his individual records in table 191, on December 31 Vea took an unusually long walk. The second highest record for Vea was during the Thanksgiving recess.

We must further consider the factors contributing to activity in general, as recorded by the pedometer. These may be summed up as follows: First, the novelty or the psychological effect is a stimulus to increased record. This would be expected to appear, if at all, in the first week of wearing the instrument. The further possibility of a psychological effect with the squads expressing itself in a desire to make a better showing at the start than a competitive squad should also be considered. A second factor would be weather conditions. With fair weather, such as that obtaining in early fall, walking would be more pleasant than in bad weather, such as might occur later in the season. During icy conditions there would be less tendency for walking but, as pointed out previously, there would follow a distinct change of gait, and possibly a shortening of step with a consequent more rapid registration of pedometer units. It should likewise be remembered that with this shortened step on an icy walk there would be a greater consumption of energy for walking the same distance. Third, with time available for walking so limited during the busy college year, relaxation and recreation would be first sought in walking, as was clearly shown by the records for the Thanksgiving and Christmas recesses. Fourth, the state of nutrition would also affect the inclination for walking.

As has been pointed out, there is a reasonable relationship between the quantity of walking registered by the pedometers and the net available energy in the diets for corresponding periods of time. And yet on close inspection we can not convince ourselves that this is of special significance, because several other factors must be taken into consideration, such as the psychological factor mentioned above and the time available for walking in the vacations. On the assumption, however, that to walk a horizontal mile requires 60 calories, it is very clear that the differences in miles walked at the different stages is by no means sufficiently large to in any wise account for the actual changes in the net energy consumption. On the other hand it is true that horizontal walking is the easiest part of a man's activity to maintain up to a normal standard amount under conditions of physical weakness and discomfort. A man can walk with comfort and pleasure when he would avoid running and exercise more strenuous than walking, unless prompted by necessity or some special motive. It seems probable, therefore, that if we had quantitative estimates for physical activity more strenuous than walking, these might follow the fluctuations found with the walking, and the variations in activity might conceivably be larger. It is not surely indicated that they would be large enough, however, to account for the energy differences from period to period of the experiment. Such correlations between activity and energy intake are not definitely proved in this research to be physiological necessities. The influence of mental attitude, ranging from depression, with much restricted diet, to the feeling of euphoria, with more liberal diet, would make for a similar correlation.

In interpreting the records of activity and in attempting correlations with diet, clear distinction must be made between those periods in the research that should more strictly be considered as transitional periods and those that are maintenance periods. With Squad A we have at the beginning of the experiment a distinctly transitional period. Immediately after the Thanksgiving recess we have a period of greatly reduced diet to overcome the excess eating during this vacation. A similar short transitional period appears in the early part of January. With Squad B the entire reduced diet period must be considered as a transitional period. Strictly speaking, the two periods of preeminently maintenance levels for Squad A are those about the middle of December and the latter part of January.

For purposes of analysis it is quite unfortunate that the transitional period in the early fall was coincidental with the period of fair weather, thus contributing towards considerable walking. After October 18 it is apparent from the figures in the last column of the table that there is no pronounced tendency for these subjects to reduce materially their physical activity, at least as indicated by the pedometer records. It should still be pointed out, however, that this table does not indicate the probable number of miles that these men would have shown in the week prior to dietetic restriction.