MECHANICAL EFFICIENCY OF THE HUMAN BODY 131 
vations of the tension on the sprocket chain fully confirming his personal 
impression. Inasmuch as the values obtained were unquestionably higher 
than normal owing to this extraneous muscular activity, the attempt was 
made to obtain a more even revolution of the pedals by belting a motor 
to the rear wheel of the ergometer and driving the machine at varying rates 
of speed. In these experiments, which were made only with ergometer II, 
the subject endeavored to sit in a relaxed position, and allow the legs to move 
freely with the rotation of the pedals. The results are singularly indicative 
of an inability to relax completely, and show a continuous opposition to the 
work of the motor. Since the motor was used in relatively few no-load ex- 
periments and the rate of speed did not vary widely, the number of experi- 
ments which can be compared with the work experiments is much smaller 
than was used for the other comparisons. They have, however, been col- 
lected and presented in table 120. 
In table 120 the values are given chronologically in order to compare the 
results obtained with the different rates of speed. On nearly every day the 
subject rode at two rates of speed, but unfortunately the rate was on some 
days so great as to make a comparison with the work experiments impossible, 
since no riding experiments at these speeds were available. On the other hand, 
the comparisons given are sufficient to show clearly, first, that the average 
efficiency for all the experiments is not far from 27 per cent; and second, that 
there is a marked difference in the results with a change in the rate of revo- 
lution, the efficiency almost invariably decreasing with the high rate of 
revolution. 
Since it is believed that in the no-load experiments the values found 
were all too high, we must accept with considerable reserve the average 
value of 27 per cent for the efficiency as computed on this basis, and the fact 
should be clearly recognized that the percentage of efficiency was raised by 
many extraneous muscular movements, such as "back pedaling," which 
would have been absent in a work experiment. In this table, as in others, 
individual periods will be found that give abnormal results, but as before the 
results should be considered only as a general picture. Such a review of the 
results does not lead one to the conclusion so strikingly shown in the earlier 
tables, namely, that the efficiency increases with the load; on the contrary, 
the values appear to show the reverse. The values are, however, particularly 
significant in indicating the great variations due to alterations in the rate of 
speed. It will be noted that with one or two exceptions all the values of 29 
per cent or above were obtained with the revolutions per minute not greater 
than 92. The method of determining the base-line is, however, open to such 
severe criticism that values computed in this way must be taken with great 
reserve. 
Efficiency in Work Experiments Based upon Values Obtained with a Current 
of 0.5 Ampere. 
An inspection of the preceding tables shows at a glance the difficulties 
incidental to assuming a base-line other than lying, for the extraneous mus- 
cular activity incidental to sitting up or to riding with no load, either with or 
without motor, certainly calls for a greater metabolism than would be ex- 
pected when no external muscular work is done. Our reasons for believing 
