The unsatisfactory termination of the attempts to use the 24-hour period in studying the influence of the ingestion of food upon metabolism led to a rearrangement of the experimental plans and the substitution of shorter experiments in which the metabolism was studied during that section of the digestive cycle when the maximum digestive activity would normally be expected. The experimental period would thus begin at approximately 9 o'clock in the morning and continue for about 8 hours.

Critique Of 8-Hour Method

In the 8-hour periods it was easier for the subject to follow a prescribed routine, such as sitting quietly in a chair without major muscular movement, than in the 24-hour periods. Furthermore, the uncertainty as to the length of time the subject slept was usually eliminated, as the experiments were made in the daytime, when the men were for the most part awake. Theoretically the 8-hour period experimental plan would have been very satisfactory if it had been possible for the subjects to sleep throughout the period and thus provide an ideal condition for measuring the quiescent metabolism. With many people there is a tendency to sleep after eating; our subjects, however, were nearly all young, many of them being college students; sleep after eating was therefore not a common experience; hence a uniformity in sleep could not be accepted as certain.

These experiments were subdivided into 2-hour periods and in the Boston experiments into periods even shorter. There was therefore opportunity to secure information as to the time relations of the increase in the metabolism and possibly the maximum effect of the food.

Many of the disadvantages found with the 24-hour plan apply, also, to the 8-hour method. The possible errors in the measurements are the same as with the longer periods, especially with the large chamber at Wesleyan University, but with the shorter periods they assume more importance, since there is less opportunity for compensation and the total amounts are smaller. Furthermore, with the protein-rich diets, the total effect of the ingestion of food is not obtained, as it is unquestionably true that the stimulus frequently continues longer than 8 hours. During the 8-hour period only one or two meals can be given; the daily routine, with period of sitting or lying after food, must therefore be sacrificed.

Finally, if we use as a basis of comparison the metabolism determined in an 8-hour period without food, as was done in all of the Mid-dletown experiments considered in this section and in some of the Boston experiments, we must still rely upon the determination of the base-line on one day and the observation of the food period on a subsequent day. In the 24-hour experimental plan the periods usually succeeded one another without interval, with the subject under careful surveillance the entire time and with like muscular activity throughout the days compared. In comparison experiments with an 8-hour basal unit, a period of some 8 to 14 hours, and sometimes one or more weeks, may intervene between the fasting and food measurements. During this time the subject is not under supervision; the activity and possibly the diet are therefore not known. The influence of a previous diet, muscular activity, and psychical excitement is as yet too uncertain for us to assume with surety that the basal metabolism will be alike during the periods compared, for although there is a general agreement between experiments made in this way, the katabolism does not remain exactly the same from day to day, either with or without food. Even when we average the results of a large number of fasting experiments and deduct this average from the results obtained after food to find the increase in the metabolism due to the ingestion of food, the errors introduced may amount to a considerable percentage of the whole increment.

With large increments this possible difference in conditions would of course have less significance than with small increments. Consequently, while both the 24-hour and the 8-hour basal units permit reasonably satisfactory measurements of large increments in the metabolism as a result of the taking of food, they are open to very serious objection when used for the measurement of small increments.

Discussion Of Results Of 8-Hour Experiments Without Food

The 8-hour experiments were made with the respiration calorimeters at Wesleyan University, Middletown, Connecticut, and the Nutrition Laboratory, Boston, Massachusetts. Inasmuch as the Boston 8-hour experiments differed somewhat in plan from those made in Middle-town, the two groups of experiments will be discussed separately. Those carried out at Wesleyan University will first be considered, not only because they are first in chronological order, but also because the apparatus used and general technique, other than duration, were like those of the 24-hour experiments which have just been discussed.

In the collection of the data a number of basal values were secured, ranging over a period of several weeks or months, and not infrequently a year or more. Hence this collection of basal data has special importance as indicating the possibility of seasonal variation, and furthermore as showing the probable trend of the metabolism on any given day when the metabolism was measured in 4 to 6 consecutive 2-hour periods. The basal metabolism experiments only will be considered in this section; the results of the experiments with food will be given later in the discussion of the effect upon the metabolism of different classes of foods.