126 MUSCULAR WORK 
revolution averaged 71 per minute, and the net efficiency was 23 per cent. 
On February 29, with the same effective work per minute but with a lower 
magnetization of the armature and a much higher rate of revolution, i. e., 
an average of 104 per minute, the net efficiency fell to approximately 16 to 18 
per cent. It is interesting to note that the highest efficiencies found through- 
out the whole table are those of the group from March 12 to 14, in which the 
rate of revolution was approximately 70, the efficiencies on these dates 
being practically 24 to 25 per cent. On days when the muscular work was 
excessive, i. e., 2.2 calories or over, the rate of revolution was very high and 
the net efficiency almost invariably low. It is thus clear from an inspection 
of table 116 that some relationship exists between speed and efficiency, a 
relationship that will be subsequently considered. From this table, however, 
the general deduction may be made that the percentage of efficiency increases 
as a rule with the increased load, but that there are also definite indications 
of a marked influence of speed upon the efficiency. 
The values here found for gross and net efficiency are wholly in line with 
those previously reported by Benedict and Carpenter ° save that the extraor- 
dinarily high maximum value of 25.2 per cent of net efficiency with the sub- 
ject M. A. M. is not equaled by any of their subjects. On the other hand, the 
average net efficiency for all the experiments with M. A. M. in which the heat 
equivalent of muscular work was above 1.35 calories per minute was not far 
from 21 per cent, a net efficiency that compares very favorably with those 
found with the five other subjects as well as with the net efficiency in the 
experiments reported by Benedict and Carpenter. The results as a whole, 
therefore, indicate that the average net efficiency of the subjects used in this 
research, as computed by deducting the resting metabolism from that during 
work, is essentially 21 per cent; these later values thus confirm fully the 
experiments reported by Benedict and Carpenter in which ergometer I was 
used, and the heat was measured directly instead of being computed from the 
oxygen consumption and the respiratory quotient. While this substantiation 
of earlier results cannot be looked upon as a proof of the accuracy of indirect 
calorimetry as compared with direct calorimetry, nevertheless the uniformity 
in results furnishes strong support to the assumption that both the method 
of indirect calorimetry and the ergometer here employed give results that 
are not abnormal. 
Efficiency in Work Experiments Based upon Values Obtained in Experiments 
with the Subject Sitting on the Bicycle Ergometer. 
While it may seem that the logical method of computation would be to 
deduct from the total energy output during work the values obtained in ex- 
periments with the subject sitting upon the bicycle ergometer, it was pointed 
out in a discussion of the results secured in such experiments, that in almost 
every instance the subjects complained of considerable strain, and found the 
experiments very trying and tiresome; the belief was therefore expressed 
that the values obtained were abnormal, particularly those for the professional 
subject M. A. M. To indicate the influence of deducting the sitting values 
from the total metabolism, we have computed the efficiency on this basis 
See p. 110 of this report. 
