resistance to be overcome being that of the ergometer itself. Under these conditions the work would be simply the internal friction of the legs. The unequal balancing of the legs discussed by Berg, du Bois-Reymond, and Zuntz, with consequent muscular work in raising the legs when rotating the pedals, should be taken into account in these experiments also.

Measurements were made only with subject M. A. M., ergometer I being used in two experiments and ergometer II in three experiments. The results are brought together in table 114. The speed varied from 60 to 98 revolutions per minute, but the results obtained with the two machines are singularly concordant. The increase in the oxygen consumption during the muscular activity ranged from 90 per cent to 308 per cent, increasing regularly with the speed. The marked influence of changes in speed is noted when it is seen that the increase in the oxygen consumption is more than doubled in the change from 60 to 84 revolutions per minute, while at 98 revolutions per minute the metabolism was practically four times that of rest. All of the values for the lying metabolism were secured on the day of the riding experiment, and hence it was not necessary to use an average value as in some of the other tables.

TABLE 114.-Metabolism in experiments without food with subject lying on couch compared with the metabolism of the subject in no-load experiments on ergometer without motor. (Subject M. A. M.)

Accompanying the increased metabolism we find a similar increase in the pulse-rate which usually proceeds more or less regularly with the increment in the speed; the one exception is the unusually high pulse-rate found in the first period of February 21. It should, however, be stated that this record is the result of but a single observation. The few records of the respiration-rate again show slight, if any, change in the rate.

It is obvious, therefore, that riding on an ergometer with no resistance is accompanied by a considerable increase in the muscular work, an increase very much greater than would be expected from the small amount of muscular activity incidental to overcoming the very slight friction of the machine. Here again it should be stated that in all of these experiments the subject found it extremely trying to ride without load, and continually complained of inability to hold himself at any given speed, his tendency being to "spurt,"

Date.

followed by an effort to restrain himself. A superficial inspection of the tension on the sprocket-chain bore out his assertion that there were varying pressures put upon the pedals during the tests. In view of the unquestionable inhibition on the part of the subject and the abnormal amount of internal muscular work, the same difficulty is found in using these values as a baseline as was found with the values obtained in experiments when the subject rode without load on a motor-driven ergometer. These abnormalities should be continually borne in mind in making subsequent comparisons. INCREMENT IN METABOLISM IN NO-LOAD EXPERIMENTS DUE TO A CHANGE FROM A MOTOR-DRIVEN ERGOMETER TO RIDING WITHOUT MOTOR.

As has already been pointed out, when the subject is riding on the ergometer without load, the legs are not in equilibrium and the actual work of rotation produces a considerable increase in metabolism. Furthermore, when the ergometer is driven by a motor, while theoretically, at least, the subject remains inert on the seat of the apparatus and does no work, there is without doubt a tendency to resist the work of the motor in driving the machine. Four experiments were made with the motor-driven ergometer, and these have been compared in table 115 with two experiments in which the motor was not used, the results being given on the basis of the rate of speed. As would be expected, when the subject rode without motor there was a greater metabolism than when he rode with motor, the average increase in the oxygen consumption being not far from 30 per cent. There was likewise a small increase in the pulse-rate.

TABLE 115.-Metabolism in no-load experiments without food with ergometer driven by motor compared with metabolism in no-load experiments without motor. (Subject M. A. M.; ergometer II.)

1 There was but one record of pulse in each no-load period on this day.
One record of pulse.

That the increase in the metabolism incidental to riding without motor when compared to riding with motor was not greater is astonishing, to say the least, and bears out the conviction of all the observers that with this subject, riding without a reasonable load was a very complicated muscular process. In fact, it may seriously be questioned whether the values obtained for riding without load, both with and without motor, can properly be taken for a base-line in computing the efficiency. On the other hand, the increase in the metabolism without the motor as compared with that with the motor was approximately uniform, which implies a regularity in the muscular action

that is quite inconsistent with the theory that there were large compensatory movements. To anyone who watched the sprocket-chain during all of these tests it was, however, very obvious that it was impossible for this subject to sit quietly upon the motor-driven machine and allow his legs to rotate freely without attempting to stop them or to increase the speed of the machine. Likewise, in the no-load experiments without the motor, it was impossible for him to maintain an even rate, and there was unquestionably a tendency to "back-pedal" at times.

GROSS AND NET EFFICIENCY OF SUBJECTS IN RESPIRATION EXPERIMENTS WITH THE BICYCLE ERGOMETER.

In conformity with the custom of practically all of the earlier writers, we have computed the gross efficiency of our subjects, using simply the relationship between the heat equivalent of external muscular work and the total heat output; the results are given in table 116. The values for the net efficiency are likewise given, together with the data for the oxygen intake and heat output per minute during both the work and the rest periods, with the heat equivalent of muscular work.

The values for the different factors have been arranged, first, in the order of subjects, and second, on the basis of ascending values for the heat equivalent of external muscular work. The reason for this particular arrangement will appear in examining the figures in the table, since the gross efficiency percentages as here shown indicate invariably an increasing gross efficiency with an increased load. Thus, in the experiment with E. P. C., in which the load was very light and the effective work equivalent to but 0.44 calorie per minute, the gross efficiency was 11.3 per cent. When the load was approximately double, and the effective work 0.93 calorie, the efficiency was increased to 16.1 per cent, while with a load approximately triple, and the effective work 1.41 calories, the efficiency was 19.9 per cent.

The reason for this increase in the gross efficiency with an increasing load is easily seen when we consider the fact that a considerable amount of heat is required to maintain the body over and above that used for effective muscular work. When the subject lies quietly upon a couch, the heat output is only that required for body maintenance, this being an approximately constant value in all experiments. With a light load the maintenance requirement forms a large proportion of the total energy output, so that the deduction of the resting metabolism has a pronounced effect. As the load is increased, however, there is no increase in the maintenance requirement, since it is an approximately constant value; consequently, with increasing loads, the extra heat output is due solely to the external muscular activity, and the percentage for the gross efficiency is accordingly increased. As an illustration, a comparison may be made with a boiler in a boiler-house where sufficient steam must be maintained to keep the boiler and the boiler-room warm. If in addition a small draft is made upon the supply of steam for running a small engine, the output of mechanical energy from the engine bears a relatively small proportion to the total heat output from the boiler, and the efficiency of the plant as a whole is very low. With increasing drafts upon the supply of steam, the percentage of the basal requirement for keeping the boilers and the engine-room warm becomes less and less, and the percentage of gross efficiency is correspondingly increased.

TABLE 116.-Gross and net efficiency in respiration experiments with the bicycle ergometer.1 [Basal values obtained with subject lying on couch.]

1 Ergometer I was used for all experiments to January 19, 1912, inclusive; ergometer II was used in all experiments subsequent to that date.

2 Determined in respiration experiment of April 6, 1912. The corresponding respiratory quotient was 0.81. 3 Assumed from results obtained with this subject in respiration experiments made Jan. 4, to April 25, 1911. The corresponding respiratory quotient was 0.84. (See table 35.)

4 Determined in respiration experiment of April 4, 1912.

5 Determined in a series of respiration experiments made Jan. 9 to 13, 1912. The corresponding respiratory quotient was 0.86.

6 Determined in respiration experiments of Feb. 12 and April 6, 1912. Corresponding respiratory quotient 0.86. 7 Determined in respiration experiment of March 14, 1912. The corresponding respiratory quotient was 0.84. 8 This figure is a general average of results obtained with this subject while lying on the couch before work. (See table 91.)

1 This figure is a general average of results obtained with this subject while lying on the couch before work. (See table 91.)