128 MUSCULAR WORK 
the increase in efficiency as computed on this basis is 1.5 per cent greater 
than the net efficiency drawn from table 116. The experiment of November 
23 with the subject E. P. C. is probably the least liable to adverse criticism, 
and here the increase is approximately 0.5 per cent. Since we believe that 
the values obtained during the sitting experiments were too high owing to 
abnormal conditions, we must consider that the efficiency computed on this 
basis will always be a little higher than that which would be obtained if the 
sitting values were not vitiated by the extraneous muscular movements 
and discomfort of the subject. In any event it can be seen that the increase 
in the efficiency due to using the sitting values as the basis varies only from 
0.5 to 1.5 per cent. 
Efficiency in Work Experiments Based upon Values Obtained in No-load 
Experiments Without Motor. 
Although the increase in the metabolism incidental to changing from a 
lying position on the couch to sitting on the ergometer is large, when compared 
with the excessive heat-output during severe muscular work it plays a rela- 
tively slight role. On the other hand, the movement of the pedals with the 
feet, the involuntary throwing of the body from one side to the other in bal- 
ancing, and the unavoidable lateral motion of the head unquestionably 
produce extraneous muscular activity that is not measured directly upon the 
bicycle ergometer. Consequently it is desirable if possible to deduct from 
the total heat output the heat incidental to moving the pedals with the 
feet without the performance of external work. The no-load experiments 
(without the use of a motor to drive the pedals) were made in an attempt to 
secure such data. In discussing the no-load experiments (see p. 119) it was 
pointed out that the subject found himself uncomfortable, with a tendency 
either to "race" or else to "back pedal." Owing to this tendency, there was 
unquestionably considerable muscular activity, particularly noticeable in the 
alterations in the tension on the sprocket-chain. Notwithstanding the diffi- 
culty of using this base-line for a comparison, we have computed the results 
for a number of the experiments with the subject M. A. M., and give them 
in table 119. 
Ergometer I was used in all experiments up to January 19, 1912, inclu- 
sive, ergometer II being used subsequent to that date. As in previous tables, 
the data are, in general, arranged in the order of increasing load. Here again 
it will be seen that even on this basis the efficiency has a tendency to increase 
as the load increases, the highest values being found with a current of 1.5 
amperes when the heat equivalent of the external muscular work was 1.9 
calories or above. The values fluctuate considerably, much more than in 
some of the earlier experiments, the lowest value being 22.4 per cent on Jan- 
uary 19, and the highest value 41.0 per cent on February 26. The average 
efficiency for all of the experiments is 30 per cent. In considering this high 
efficiency, however, it is necessary to bear in mind that the values for riding 
the ergometer with no load are unquestionably higher than they should be. 
This we have pointed out clearly in the previous discussion of the results 
of these experiments. A further complication arises in that a change in 
speed produces a noticeable alteration in the oxygen consumption and con- 
sequently in the heat output. For this reason it has been difficult at times 
