For a true comparison of the results of the different experiments, it is necessary to compare only those in which approximately the same amounts of nitrogen were ingested and with experimental periods of approximately the same length. In many instances the total effect of the food had by no means ceased at the end of the period of observation. In other experiments it was evident that the full effect of the food ingestion was obtained, inasmuch as increments of less than 0.5 gram, as well as negative values, were found in the later periods. It was hoped that some information might be obtained as to the relationship between the amount of nitrogen ingested and the increments in the metabolism. Generally speaking, the larger amounts of meat produced the larger increments. This may not hold true, however, when different individuals are used for subjects, as may be seen by a comparison of the experiment with A. W. W., April 6, 1907, in which 755 grams of meat were ingested, with that with J. R., December 4, 1908, in which 418 grams were taken. Although both of the experiments continued for 8 hours, the increment in heat production was slightly more in the second experiment than in the first, but the increment in oxygen consumption and carbon-dioxide production in the experiment with J. R. was about 70 per cent of that in the experiment with A. W. W. It will thus be seen that marked irregularities occur in all these experiments, and no constancy was found in comparisons with different individuals and rarely in comparisons for the same individual.

It was also hoped that some light could be obtained as to the influence of animal protein as compared with that of vegetable protein. A superficial examination of the data in table 230 shows no material difference in the two classes of proteins in their influence upon the metabolism, but here again the comparisons are complicated by the fact that the experiments are made with different individuals and with different experimental plans. With the purest protein substance used (glidine), the experiments in which the largest amount was given, i. e., 9.70 grams of nitrogen, gave duplicate values for the same subject which were only reasonably satisfactory. Comparing these values with those obtained with beefsteak or with beefsteak combined with potato chips or bread, in which essentially the same amount of nitrogen was ingested, we find that the average values with glidine are slightly higher than those for beefsteak, although even with the same amount of nitrogen the values with beefsteak vary widely. The gluten bread experiments, while complicated by a relatively small amount of animal protein in the form of skim milk, show increments comparable with those obtained with beefsteak. One must conclude, therefore, that these experiments, defective though they are, indicate that there is no clearly defined difference between animal and vegetable proteins in their influence upon the metabolism.

The series of respiration experiments which are summarized in table 215 were made subsequent to most of the calorimeter experiments included in table 230 and were designed to throw more light upon the quantitative relationships. Varying amounts of beefsteak were taken in these experiments, although in none was so large an amount eaten as in the first two calorimeter experiments given in table 230. The heat production in the periods subsequent to the taking of the food invariably exceeded the basal value. Usually the experiments were not continued sufficiently long to include the total effect of the food, so that the basal value would again be reached; consequently the increases recorded in the last two columns of table 215 frequently represent incomplete increments. Most of the experiments did not extend over a period longer than 6 hours, although in one case the observations were continued over a period of nearly 12 hours. The irregularity in the effect upon different individuals of the ingestion of the same amount of nitrogen is strikingly shown in the percentage increase above the basal value, these figures being given in the last column of the table. While theoretically we should expect to find continually decreasing values for these percentages, as the experiments are arranged in the table in the order of decreasing amounts of beefsteak eaten, this is not actually the case.

Making due allowance for the fact that the time over which the experiments were continued varies somewhat, it is still clear that there is no uniform relationship between the amount of nitrogen ingested and the actual increase above the basal metabolism. Whether such a relationship could have been established if the experiments had been continued until the effect of food had completely ceased would appear, from the data obtained, extremely improbable. Experiments of this length are very tiresome for both subject and observer; nevertheless such experiments should ultimately be made. For the present, therefore, we can only reiterate the deductions made from the results of the calorimeter experiments to the effect that while the ingestion of protein in almost any amount invariably produces an increase over the basal metabolism which may be 25 per cent for several hours and for short periods may rise to 45 per cent (see tables 198 to 229), no definite mathematical relationship between the amount of protein ingested and the increment in the total metabolism can be noted from these values. It is probable that in any study of these results it should be remembered that these subjects were unlike in body-weight and in active mass of protoplasmic tissue.