The percentages for the total increment above the basal value as computed for the mixed carbohydrates are likewise shown in table 124. The starch as ingested in the experiments with mixed carbohydrates was in three forms: first, in popcorn, which was dry and hence must undergo the process of imbibition in the stomach; second, in rice, which was cooked; and third, in the moist starch of bananas. The popcorn experiments were primarily designed to throw some light upon the ingestion of roughage in the diet and those with rice to give the effect of cooked starch. The effect of uncooked starch was studied with bananas, of which large amounts could be eaten with considerable ease. As carried out, however, the experimental method was somewhat faulty in that the bananas were given, in all but three experiments, with relatively large amounts of cane sugar; hence we have unquestionably a double influence upon the metabolism.

Total increments for bananas and sugar are frequently found of 17 to 24 per cent, showing very perceptibly the influence of the ingestion of this mixture of carbohydrates. No great stress should be laid upon these computations, owing to the irregularities in the length of the observations and the fact that frequently the metabolism returned to the basal value before the experiment ended. Nevertheless, the general picture shown for bananas and sugar is that of a very pronounced increase in heat production following their ingestion, which may rise in individual periods to a peak of 34 per cent, with a total increment above the basal value as high as 24 per cent and frequently 15 or more per cent, values which are considerably above those normally noted with pure carbohydrates. The effect following the ingestion of cane sugar is very pronounced; a considerable effect is likewise found with bananas. The high values obtained with the combined bananas and sugar point definitely to the conclusion that we have here an effect due to cane sugar which is superimposed upon the effect due to the large amount of carbohydrate taken simultaneously in the form of fruit.

The experiments with bananas without sugar gave results which are irregular; two showed a measurable increment, while in the other no increment was obtained. The two experiments with popcorn indicate a distinctly higher metabolism as a result of the ingestion of this material. But one experiment was made with rice, a fact which is to be regretted, since the slight increment noted, namely, 7 calories, should be confirmed. It is evident that our section of this research dealing with carbohydrates of a gross texture and the possible effect of roughage in the diet is altogether too limited for adequate discussion.

Respiration Experiments

As the research on the influence of the ingestion of food progressed, it became evident that measurements of the metabolism in short periods were essential, for many of the experiments indicated a somewhat rapid change in the character of the metabolism following the ingestion of carbohydrate. Experiments with periods of sufficiently short duration to show this rapid change were impracticable with an apparatus so large as the respiration calorimeter in Middletown. With the development and subsequent completion in the Nutrition Laboratory of the so-called "universal respiration apparatus"1 observations could readily be made in short periods with fairly satisfactory results. An extended series of such experiments was begun in the fall of 1910 and continued at intervals for several years. We are indebted to Mr. H. L. Higgins2 and Mr. L. E. Emmes for their kind cooperation, as the majority of the experiments made in 1911 were under their immediate supervision. The experiments previous to 1912 were made with the so-called "tension-equalizer" form of the respiration apparatus,3 which was later replaced by the spirometer type of the same apparatus. Both types of the apparatus have been carefully tested by one of us1 and their capacity for yielding accurate results has been proved.

1This apparatus is described in detail by Benedict, Deutsch. Arch. f. klin. Med., 1912, 107, pp. 156-200; also Carpenter, Carnegie Inst. Wash. Pub. No. 216, 1915, pp. 21-53.

2See, also, Higgins, Am. Journ. Physiol., 1916, 41, p. 258.

3Carpenter, Carnegie Inst. Wash. Pub. No. 216, 1915, p. 21, and Benedict, Am. Journ. Physiol.. 1909, 24, p. 345.