It seems evident, therefore, that it will be more desirable to present the data as a whole first and then draw from the average figures such deductions as are permissible. Accordingly, in table 92, are presented the average quotients found before, during, and after work. Such quotients as were obtained in later periods will be found in the footnotes. No indication is given in the table as to how many periods are included in the averages nor of the actual amount of work done, since the data are intended to present only a general picture of the results."

TABLE 92.-Respiratory quotients in respiration experiments without food before, during, and after work.

1 The work-periods on this day were begun about 11 hours after the conclusion of the periods lying before work. The subject rode without load and without motor for 1 hours in the interval.

2 Only one period lying before work was run on this day.

3 The respiratory quotient during the third hour lying after work was 0.80; during the fourth hour, 0.77; and during the fifth hour, 0.80.

4 The respiratory quotient during the third hour lying after work was 0.79. Another work-period followed for which the respiratory quotient was 0.81. In the three-quarters of an hour subsequent to this last work-period, the average respiratory quotient was 0.77 with the subject lying on the couch.

5 The work-periods on this day were begun about 21 hours after the conclusion of the periods lying before work. During the interval the subject sat inactive on the ergometer for two periods; he then rode for about 1 hour without load on the motor-driven ergometer.

6 Other work-periods were then run for which the average respiratory quotient was 0.90. In the half-hour following the latter work-periods, the respiratory quotient was 0.81 with the subject lying on the couch. 7 Other work-periods were then run for which the average respiratory quotient was 0.94. In the threequarters of an hour following the latter work-periods, the average respiratory quotient was 0.95 with the subject lying on the couch.

8 Other work-periods were then run for which the average respiratory quotient was 0.93. In the three-quarters of an hour following the latter work-periods, the average respiratory quotient was 0.88 with the subject lying on the couch.

9 Other work-periods were then run for which the average respiratory quotient was 0.87. In the 45 minutes following the latter work-periods, the average respiratory quotient was 0.74 with the subject lying on the couch.

10 The respiratory quotient during the third hour lying after work was 0.82; during the fourth hour, 0.79; during the fifth hour, 0.75; and during the sixth hour, 0.81.

In order to bring out in the sharpest contrast the relationships of the periods before work, during work, and after work, two separate averages

a The data are given in greater detail in tables 93 and 136.

have been drawn, first, for the 34 days on which such values were obtained, and second, for the 16 days on which the after-periods were of such length that a subdivision could be made of the values obtained for the first and second hours after work, respectively, this last average necessarily including only data secured in the experiments with the subject M. A. M. In the grand average for 34 days, it will be seen that the respiratory quotient of 0.85 was raised during work to 0.88, falling after work to 0.78. An inspection of the data shows that this rise and fall was present in the majority of the experiments, although in several of the earlier experiments there was a fall in the respiratory quotient during work, with a further lowering-very rarely a rise -during the period after work. The greater number of the experiments included in the average for the 16 days were made in the latter part of the experimental year, for, with the exception of the experiment of December 21, there were none prior to January 9. This average also shows a slight, though definite, rise in the quotient during the work, with a subsequent marked fall in both periods after work.

From the research as a whole, therefore, the significant conclusions can be drawn that there was a distinct, though numerically small, increase in the respiratory quotient during work, with a very appreciable decrease in the quotient following work, especially in the first hour of rest. This rise and fall may be interpreted in various ways. The rise in the quotient during work would distinctly imply that a larger proportion of carbohydrates was burned during this period than before or after work, especially as the rise in the quotient could not have been produced by an excessive disintegration of protein, since practically all of the quotients obtained in the period before work were above 0.8. It also seems probable that the marked fall and general continued lower level of the respiratory quotient after work indicates the combustion of a larger proportion of fat than was the case in the period before work, the lower quotients showing that the supply of carbohydrates in the body had been materially depleted. The increases in the respiratory quotient during the work-period as compared with the period before work was but small, namely, from 0.85 to 0.88 and from 0.84 to 0.86— an increment that might almost be considered within the limit of experimental error when the great differences in the physiological condition of the man are taken into account. Of themselves, therefore, these increases do not indicate a selective combustion of carbohydrates during muscular work, since it is possible that the combustion of carbohydrates in the period after work was proportionally the same as in the period preceding work, and that the low quotients in the last rest-periods were simply due to a natural depletion of the storage of carbohydrates during muscular work.

Furthermore, a close inspection of the data in table 92 shows that there are variations in the respiratory quotients obtained with the professional bicycle rider, M. A. M., between the earlier and the later experiments. To bring this out more clearly, the quotients may be averaged in the following way: For the 20 comparable days from December 15 to January 31 the average quotients for this subject were 0.85 before work, 0.85 during work, and 0.78 after work, while for the 12 comparable days from February 1 to April 16 the average quotients were 0.85 before work, 0.91 during work, and 0.78 after work. Dividing the experimental year with this subject

into halves, we found that the average respiratory quotients for the 16 comparable days from December 15 to January 19 were 0.84 before work, 0.84 during work, and 0.77 after work, while for the 16 comparable days from January 24 to April 16 the quotients were 0.85 before work, 0.90 during work, and 0.78 after work. Apparently in the later experiments there was an increase in the respiratory quotient during the working period which was not observed in the earlier experiments, this probably being due to the fact that the excessive muscular work was performed in the last part of the research.

INFLUENCE UPON THE RESPIRATORY QUOTIENT OF INCREASING AMOUNTS OF WORK.

It is obvious from the foregoing that for a better understanding of these respiratory quotients, it is necessary to analyze the results obtained for the periods before work, during work, and after work. In some experiments external muscular work to the extent of over 150 calories was performed. Obviously during these experiments there must have been heavier and more prolonged drafts upon the storage of body-material than in experiments in TABLE 93.-Respiratory quotients in respiration experiments without food, with accumulated amounts of work on the bicycle ergometer.

1 The respiratory quotients were determined during the last part of the period in which the work specified was done.

2 The respiratory quotient during the third hour lying after work was 0.80; during the fourth hour, 0.77; during the fifth hour, 0.80.

3 The respiratory quotient during the third hour lying after work was 0.79.

4 The respiratory quotient during the third hour lying after work was 0.82; during the fourth hour, 0.79; during the fifth hour, 0.75; during the sixth hour, 0.81.

which the external muscular work was but 35 calories. Consequently, in table 93 the respiratory quotients determined during muscular activity are classified according to the accumulated amount of work done, the values secured before and after the work-period also being included. The work each morning was practically continuous; the table therefore shows the quotients obtained at various times during the morning as the amount of external muscular work gradually accumulated.

The quotients given for the periods before work are usually the average of those obtained for two or three periods, and probably represent the true values, but those for the work-periods are individual determinations and hence are liable to the errors which may creep into the determination of both the carbon dioxide and the oxygen. While every precaution was taken to prevent error, it is pointed out that these individual values should be used with extreme caution, and deductions can only be drawn from the tabulated results as a whole.

The respiratory quotients before work demand no particular discussion at this point, but the quotients obtained during work, and particularly the relationship to the amount of work done, indicate that as the experimental year progressed the amount of work increased until in the latter part of the year practically all of the experiments were made with excessive amounts of muscular activity. An examination of the data shows that in the majority of the experiments the quotients tend to increase as the total amount of work increases on any particular day. This is not invariably the case, as in several of the experiments there is a marked fall in the quotient during the second work-period. Considering only the experiments with M. A. M., we find that on 13 days the quotients decreased in value while on 18 days they increased.

This tendency for the quotients to increase with the increased work has an important bearing upon the interpretation of the results. If, for example, the metabolism were of exactly the same character during muscular work as it was during the preceding rest-period, there would be a certain draft upon body-glycogen which, under the conditions of these experiments, namely, no food in the morning, would cause a rapid reduction in the storage of carbohydrates. Under these conditions we should normally expect to find an invariable decrease in the respiratory quotient as the work progressed, assuming that the character of the metabolism remains exactly that prior to the experiment. Since on the whole the results show an increase in the respiratory quotient rather than a decrease, this evidence would point strongly to the fact that the character of the katabolism during the muscular work does not remain the same as that prior to the work, but that there was an increased combustion of carbohydrate. This is furthermore accentuated by the fact that in all but a very few experiments there was a marked drop in the respiratory quotient during the period after the work was finished. This finding during the first hour after work is usually substantiated by observations made during the second hour, as well as in the relatively few instances in which the respiratory quotient was followed for several hours after work.

These figures then, taken as a whole, indicate strongly that during the period of severe muscular activity the character of the katabolism is altered by an increase in the amount of carbohydrates burned, the storage of carbo

hydrates in the body being drawn upon for the purpose. At the completion of the work-period the specific katabolism incidental to muscular work ceases, and the picture of a body subsisting upon a depleted storage of carbohydrate material is then indicated by the lower respiratory quotient.

INFLUENCE UPON THE RESPIRATORY QUOTIENT OF VARIATIONS IN THE AMOUNT OF CARBOHYDRATES IN THE DIET.

The analysis of the figures with the subject M. A. M. can be carried somewhat farther by isolating a few experiments which were made with a semi-controlled diet. On certain days the subject was fed in the laboratory and given a diet which was approximately controlled. During 6 days his diet contained 400 or more grams of carbohydrate a day, while on 6 other days it was noticeably deficient in carbohydrate material, although by no means carbohydrate-free. The results for these particular experiments are given with some detail in table 94.

TABLE 94.-Respiratory quotients in experiments without food following special diets. (Subject M. A. M.)

The average respiratory quotient during the preliminary rest-period with the carbohydrate-rich diet was considerably higher than that for the same period with the carbohydrate-low diet, i. e., 0.85 as against 0.79. During the periods with muscular work, there was a marked increase in the respiratory quotient with both diets which was somewhat greater with the carbohydrate-rich diet. The low quotient after work was also observed in both series of experiments, the lowest quotients being found with the diet poor in carbohydrates. This is fully in conformity with the view previously expressed that during muscular work there is an excessive combustion of carbohydrates so that in the subsequent period the katabolism is that of a body deficient in glycogen.

It is clear, therefore, that during the winter's experimentation there were variations in the storage of carbohydrates in the body, the relatively few days on which the diet was controlled giving a reasonably accurate picture of the conditions that would be expected to exist. It is interesting to note that even with the carbohydrate-poor diet, when the respiratory quotients indicated low amounts of glycogen in the body, there was nevertheless on the average a slight increase in the respiratory quotient during work,