DISCUSSION OF RESULTS. 77 



The results of all of the walking experiments with subject I are given 

 in abstract in table 13. Inasmuch as the experience of previous 

 investigators has shown that the variations in velocity play an impor- 

 tant role in the total energy transformations, we have recorded here 

 the distance per minute which the subject walked and have likewise 

 computed the number of horizontal kilogrammeters by multiplying 

 the distance per minute by the body-weight. From the kymograph 

 records it was possible to find the average height to which the body was 

 raised at each step by obtaining the average height of each recorded 

 movement; by multiplying this average height by the number of steps, 

 the total distance that the body was moved in a vertical direction could 

 be computed with reasonable accuracy, thus giving a component 

 for the analysis of the various factors involved in walking. The 

 measurements of the distance that the body was elevated per minute 

 are recorded in column d. 



The total heat-output was calculated from the gaseous exchange not 

 only for the walking periods but also for the standing periods preceding 

 the walking; these values are recorded in column e. The increase due 

 to walking was obtained by difference and the energy required to move 

 1 kilogram 1 meter in a forward direction, i. e., 1 horizontal kilo- 

 grammeter, is recorded in gram-calories in column g. These latter 

 values are obviously the most important result of this study. 



EXPERIMENTS WITHOUT FOOD. 



Considering first the values without food, it will be seen that the 

 heat-output per horizontal kilogrammeter varies from a minimum of 

 0.446 gram-calorie to a maximum of 0.637 gram-calorie, but the large 

 majority of the values lie close to the average value of 0.507 gram- 

 calorie. A general inspection of these results shows indications of a 

 periodicity or rhythm in the efficiency of this subject in walking. Thus, 

 low values are found on December 2 and 3, which are followed by a 

 group of high values continuing until December 22, while the last 

 three experimental days that the subject was without food give dis- 

 tinctly lower values than those found in the preceding group. There 

 is not, however, sufficient regularity in this rhythm to indicate the 

 effect of training of the subject or in the use of the treadmill. To be 

 sure, the first value, that of November 29, is considerably higher than 

 the average and it is probably fair to assume that the high value of 

 0.637 gram-calorie found on December 15 was more or less accidental. 

 On the other hand, we see no reason why these values should be rejected 

 and the figures are sufficiently numerous to make it immaterial whether 

 or not these or any other particular values are discarded so far as the 

 influence upon the general average is concerned. While it may appear 

 that the low values found on the last three days would imply the effect 

 of training, it should be stated at the outset that this subject was a 



