The nitrogen output has a special interest in this connection when considered as an index of the level of the protein katabolism. The data for the nitrogen output for most of these subjects with presumably low metabolism have been collected in table 2, which shows that Prof. C. has a nitrogen excretion per kilogram per hour of 8.3 mg. This is somewhat higher than the average found for 10 subjects, including Prof. C. and H. F., and reported by Benedict and Joslin,2 the average for all subjects being 6.85 mg. On the other hand, both Dr. H. and H. F. show a low nitrogen excretion, the average of two experiments with Dr. H. being 4.8 mg. and the three values found for H. F. being 5.6, 4.5, and 2.5 mg. No post-absorptive nitrogen figures are available for Dr. Kellogg. It is clear, however, that Prof. C. was excreting nearly twice as much nitrogen as the other subjects and hence was on a fairly high nitrogen level. The fact that his metabolism per kilogram of body-weight was so low - but 0.2 calorie higher than that of H. F. or 19.9 calories as compared with 19.7 calories - must in all probability be due solely to excess body fat. Both subjects were distinctly over-weight and showed low metabolism per kilogram of body-weight.

Table 2. - Nitrogen Excretion Of Subjects Studied For Low Metabolism. (Subjects Post-Absorptive.)

* Introduced for comparison.

1 Computed from table 4, Medico-Actuarial Mortality Inv., New York, 1912,1, p. 38, deducting 8 lbs. for clothing. 2 Benedict and Joslin, Carnegie Inst. Wash. Pub. No. 136, 1910, p. 196.

It thus appears that a critical examination of the metabolism data for these individuals with supposedly low metabolism does not show appreciable deviations from the normal. With a rare pathological case there may be justification for laying weight upon the metabolism of a single individual, but the Nutrition Laboratory has, in recent years, strongly opposed the use of such individual data for comparison. Furthermore we have strenuously objected to an undue use of a so-called "standard" figure for the metabolism, as we believe that individual variations may be so great as to render these individual comparisons practically valueless. Hence, while the evidence for both Mr. Fletcher and Dr. Kellogg, and particularly the latter, did not indicate a noticeably low metabolism due to dietetic habits, it seemed best to study the question with groups of individuals, since with these only can convincing data be obtained.

With the idea that a vegetarian diet, which might be assumed to be likewise a low-protein diet, would result in a low metabolism, a series of investigations was carried out with approved technique at the Battle Creek Sanitarium, by Benedict and Roth (Journ. Biol. Chem., 1915, 20, p. 231). Through the courtesy of Dr. Kellogg, a large number of men and women vegetarians were thus studied. It became clear to us at this time that the so-called group system of comparison was absolutely necessary, namely, that only individuals of like height and weight may properly be compared. Hence for comparison with the men and women vegetarians we selected a group of normals, i. e., non-vegetarians, of like weight and height. The vegetarians, even with a presumably low nitrogen output and a less stimulated plane of metabolism due to the lowered nitrogen metabolism, did not have a lower total metabolism than the individuals subsisting on a mixed diet.

In an analysis of the results obtained with some 150 individuals1 it was found not only that, strictly speaking, there is no constancy in the basal metabolism, but also that those instances in which the metabolism is low give no indication of a general picture of unusually low metabolism due to other than well-known causes. In consideration of the fact that sex, age, muscular training, and body composition (i. e., proportion of inert body-fat and active protoplasmic tissue), height in individuals of the same weight and age, sleep, and the after effect of exercise, all have an influence upon the basal metabolism, the conclusion was drawn that the basal metabolism of an individual is a function, first, of the total mass of active protoplasmic tissue and, second, of the stimulus to cellular activity existing at the time the measurement of the metabolism was made. It was furthermore maintained that "apparently at present no law can be laid down that will cover both of these important variables in the basal metabolism of an individual."1

1 Benedict, Journ. Biol. Chem., 1915, 20, p. 263.