200 FOOD INGESTION AND ENERGY TRANSFORMATIONS. 



It is very difficult to obtain a satisfactory method for computing 

 the percentage increment. Obviously the lengthening of the period 

 in which the measurements are made without an increment above the 

 basal metabolism simply increases the denominator of the fraction and 

 thus increases the basal value without affecting the increment. The 

 value of the increment in terms of per cent is thus decreased. In 

 table 123 it will be seen that the percentage for the greatest increment 

 above the basal value is in every instance considerably higher than the 

 percentage for the total increment given in the last column of table 124. 

 This is due in large part to the fact that the total increment in the heat 

 production took place during the first hour or tAvo, while in the subse- 

 quent hours the metabolism was essentially the same as the basal value. 



When the duration of the experiment is 12 hours instead of 4, 5, or 8 

 hours, as the case may be, the percentage total increment is naturally 

 greatly decreased. A striking illustration of this is shown by com- 

 paring the experiments with A. L. L., on March 30, 1906, and April 

 19, 1906, in w r hich essentially the same amounts of bananas and sugar 

 were given. The greatest increment was 34 per cent in one case and 

 31 per cent in the other; in both experiments this increment occurred 

 in some part of the first or second hour. The total increment above 

 the basal value was found to be 88 calories in one case and 72 calories 

 in the other. There was, however, a difference in the basal values of 

 the two experiments of over 300 calories, owing to the fact that in the 

 April experiment the length of the experimental period was 12 hours, 

 while in the March experiment it was only 8 hours. The percentage 

 increment in the March experiment was therefore nearly double that in 

 the April experiment, while in other experiments the values are fairly 

 good duplicates. 



The computation of the total increment above the basal value for ex- 

 periments in which food was ingested is justifiable. The computation 

 of the percentage of the increment is, however, open to serious criti- 

 cism, and it is difficult to see how such percentages can have real sig- 

 nificance. Yet they are frequently computed and reported in experi- 

 ments of this kind. Perhaps their greatest value for this discussion is 

 the fact that \vhile in these observations the experimental period 

 varied in length only from 3 to 12 hours, and usually from 4 to 8 hours, 

 it can readily be seen that were the remainder of the 24 hours added 

 to the experimental period, the percentage value would be greatly 

 decreased and, in fact, would nearly disappear. It can easily be under- 

 stood, therefore, why investigators employing the 24-hour period have 

 failed to note a material increase in the metabolism due to the ingestion 

 of carbohydrates, for although there is a distinct temporary increase, 

 which may at times reach 30 per cent or over, this increase, when com- 

 pared to the total 24-hour basal value, appears almost insignificant. 

 If, on the contrary, we are dealing with a substance which is delayed 



