With diets consisting primarily of carbohydrates and fat no special indices are available as to the proportion of fat and carbohydrate burned in the body other than the relationship between the carbon-dioxide production and the oxygen consumption; but when protein enters into the katabolism, especially in excessive amounts, the nitrogen in the urine has commonly been taken as an index of the amount of protein katabolized. The intimate relationship between protein katabolism and heat production has been so pronounced as to lead writers to calculate quantitative relationships between heat production and the nitrogen excretion of the urine.

In the computation of the total energy transformation by means of the respiratory exchange, emphasis is laid for the most part upon the measurement of carbon-dioxide excretion and oxygen consumption, and heat production is computed from the calorific value of the carbon-dioxide or oxygen at the respiratory quotient actually measured. There are two methods for computing heat production from the calorific values for carbon dioxide and oxygen. In one no special attention is paid to the protein disintegration, on the general ground that usually about 15 per cent of the total energy is derived from protein metabolism. When a high degree of accuracy is desired, however, it is customary to compute from the respiratory exchange and the nitrogen in the urine the non-protein respiratory quotient, then compute the energy production due to the katabolism of the protein by multiplying the number of grams of nitrogen in the urine by a standard factor (26.51 calories). The remainder of the energy is then apportioned between fat and carbohydrate on the basis of the non-protein respiratory quotient. As previously stated (see page 203), this was not done in our computations of the energy from the gaseous exchange, as the non-protein quotient has relatively little significance, save in those experiments in which an excessive amount of protein was ingested.

By using the nitrogen excretion as an index of the protein katabolized, computing the total energy derived from protein and comparing it with the increment in the energy due to the ingestion of a protein food, certain relationships are made possible. This method of computation may be illustrated by using the results of the experiment with A. H. M. on April 5, 1907, in which 777 grams of beefsteak were taken by the subject. (See table 198, page 267.) The basal nitrogen excretion used was 1.06 grams per 2 hours (see table 28, page 80). The nitrogen excretion in the first 2-hour period following the ingestion of the food was 4 grams. The increment in the nitrogen excretion due to the ingestion of this large amount of a protein food was therefore 2.94 grams. As each gram of nitrogen in the urine represents a heat production from protein katabolized of 26.51 calories, the increment of 2.94 grams of nitrogen represents 78 calories of energy due to the increase in the amount of protein katabolized during this 2-hour period. Inasmuch as the total increment in heat production for the first period was but 31 calories, it is evident that at least 47 calories from the protein combustion took the place of energy originally derived from carbohydrate-fat combustion in a 2-hour period of the basal experiment. The total nitrogen excretion in the 8 hours of the experiment was 11.49 grams; the excess nitrogen excretion was therefore 7.25 grams, with an energy production of 192 calories due to the increase in the protein katabolized. The total increment in the heat production was but 136 calories; we may assume, therefore, that the replacement of basal energy derived from material other than protein was at least 56 calories.

The direct measurement of the protein disintegration from the nitrogen in the urine leads to the possibilities of further computation to determine the cause of the increase in the energy output following the ingestion of food. For example, when a protein food, such as beefsteak, is given in an experiment, we may compare the subsequent total increase in the metabolism (1) with the total energy of the food intake; (2) with the fuel value of the intake, thus obtaining the "cost of digestion"; (3) with that portion of the total energy or fuel value of the diet which is derived from protein alone; (4) with the total energy of the katabolized protein; or (5) with the increment in the heat production due to the increase in the amount of protein katabolized.

In the experiment with A. H. M. on April 5, 1907, the total effect of the ingestion of beefsteak was not obtained, as there was still a considerable increment in the metabolism even in the last period. We can not use the results, therefore, for an illustration of computing the specific dynamic action.1 An experiment better adapted for this purpose is that with the same subject on May 24,1907, in which the basal metabolism was obtained in the last period of the experiment and the total increment due to the ingestion of the beefsteak was therefore secured. (See table 200, page 269.) Following the usage of Rubner, the fuel value rather than the total energy of the diet may be used in the computation. The fuel value of the beefsteak eaten in this experiment was 644 calories, of which 70 per cent was derived from protein, or approximately 450 calories. The total increase in the heat production subsequent to the ingestion of the food was 70 calories. The total nitrogen excretion in the 8 hours of the experiment was 8.26 grams; as the basal nitrogen excretion which may be used for the same period is 4.24 grams, the excess nitrogen excretion due to the ingestion of the food was therefore 4.02 grams. This corresponds to an excess in the amount of protein katabolized (4.02 by 26.51) of approximately 107 calories as the result of an intake of 450 calories from protein.

A part of this increment of 70 calories may be properly ascribed to the influence of fat ingestion, since there was a considerable proportion of fat present in the beefsteak, but our evidence, as well as that of other investigators, indicates that the ingestion of fat has but a slight effect upon the metabolism and may probably be neglected in computations of this kind. Indeed, this was done in computing the values given in tables 249 and 250. The possibilities of differentiating between fat and protein in determining the influence upon the metabolism of the ingestion of a protein-fat diet should not, however, be lost sight of. It may be noted in this connection that Rubner carefully made such corrections in considering the influence of the protein-fat diets used in his experiments.

The experiments in our research with a predominatingly protein diet were not sufficiently extended or carried out with a sufficient degree of accuracy to justify a computation from their results of the so-called "specific dynamic action" of protein in the case of man. There is no question but that such a relationship exists between the increment in the protein katabolism and the increment in the heat production, but it may or may not be causal. Our experiments show that subsequent to the ingestion of a diet containing an excessive amount of protein there is prolonged and excessive heat production which continues for several hours. The nitrogen in the urine is likewise increased, although, as is seen from the foregoing discussion, the increase in heat production is not sufficient to account for the total excess protein katabolized.1 The fact should be recognized that this relationship is more apparent than real, for an increment in heat production is likewise found as the result of the ingestion of carbohydrates which is unaccompanied by material changes in the nitrogen excretion; one must therefore be cautious in associating too intimately the increase in the heat production with the increase in the amount of nitrogen excreted in the urine.

1Williams, Riche, and Lusk (Journ. Biol. Chem., 1912, 12, p. 352) have pointed out in an interesting manner the methods of computing the specific dynamic action, so called, from an increase in the protein katabolism.