MEMOIRS OF THE NATIONAL ACADEMY OF SCIENCES. 



303 



fact that the most of the hydrogen in organic combination in the food La eliminated, not in 

 organic combination in the feces and urine, but in the form of water in the urine or respiratory 

 products. In other words, the figures in column / apparently represent hydrogen gained by the 

 body in organic compounds, but for the most part actually represent hydrogen given off as water. 

 The total gain or loss of hydrogen for the experiment is calculated by adding together the i^'dro- 

 gen apparently lost a~ water, column/', and the hydrogen in organic combination apparently 

 gained, column I. This total gain or loss of hydrogen is shown in column n. There was in this 

 experiment a gain of hydrogen on the first day and a loss on the three following days, making 

 an average lo^s for the experiment of 20.8 grams per day. 



It should be said, however, that the determinations of water and consequently of hydrogen 

 are less satisfactory than those of nitrogen, carbon and energy. 



Table XVI. — !,:<■■, m. ,<„,/ outgo "f watt r and hydrogen. Metabolism i xperinu ni .V... i j. 



Estimated gains and losses of body protein and fat. — From the data summarized in Tables 

 XV and XVI we may compute the gain or loss of protein, fat, and water on the successive days 

 of the experiment. These computations are shown in Table XVII. If nitrogen is gained or lost, a 

 corresponding gain or loss of protein is assumed. Protein compounds are here assumed to contain 

 on the average 16 per cent of nitrogen, 53 per cent of carbon, and 7 per cent of hydrogen. Accord- 

 ingly, the gain or loss of protein is computed by multiplying the gain or loss of nitrogen by 6.25, 

 and is shown in column o. Whatever protein is gained or lost must, by the above assumption, 

 contain 53 per cent of carbon and 7 per cent of hydrogen. The amounts of carbon and hydrogen in 

 the protein gained or lost in the successive days of this experiment, as thus computed, are shown 

 in columns d and h. The algebraic difference between the total carbon gained or lost and that in 

 the protein gained or lost gives the amount of carbon gained or lost in other compounds, namely, 

 fat. glycogen, etc. It is probable that the amount of glycogen in the body at the time of rising, 

 7 a. m.. does not differ greatly from day to day, and the assumption is here made that all of the 

 gain or loss of carbon above that in the protein gained or lost comes from change in the amount 

 of body fat. It is assumed that average body fat contains 76.5 per cent carbon a and the amount 

 of fat gained or lost is consequently computed by dividing the values in column e by .765, as is 

 shown in column/'. Assuming, as before, that there has been no change in the body content of 

 glycogen, the algebraic difference between the total hydrogen gained or lost and that in the protein 

 and fat gained or lost is assumed to represent the hydrogen gained or lost in the form of water. 



a Determinations of the percentage of carbon in body fat made in this laboratory by F. G. Benedict and E. 

 Osterberg in 1900, published in vol. 4 of the American Journal of Physiology, page 74, average 76.08 per cent. The 

 value 0.761 was therefore used instead of 0.765 in computations of fat gained or lost in later experiments, beginning 

 with No. 26 



