THE RELATIONS OF METABOLISM TO FOOD-SUPPLY. 139 
determinations of the fasting metabolism of the animals actually 
used for the experiments were made. By a re-computation of 
Munk’s experiments they obtain results varying but little from 
100 per cent. A computation from the average figures given on 
p. 136, assuming 3.4 per cent. of nitrogen in the meat and 0.51 
grams of nitrogen per day in the feces, shows that the minimum is 
probably less than 107 per cent. of the fasting nitrogen excretion. 
Much depends, however, upon whether we take as the unit of 
comparison the total nitrogen excretion or, like Voit & Korkunoff, 
eliminate that portion derived from the extractives. If we select 
the former, then it appears that with a liberal supply of carbohy- 
drates in the food the supply of proteids certainly need not exceed 
the fasting metabolism in order to maintain nitrogen equilibrium, 
and perhaps may be reduced materially below it. . 
Finally, it must be remembered that the fasting proteid meta- 
bolism itself is not a constant. In Chapter IV it was shown that 
as the store of fat in the body of a fasting animal becomes depleted 
the body proteids are drawn upon to an increasing extent to supply 
energy to the animal. It is not possible to show that the experi- 
mental results which have been cited are materially affected by this 
variability of the fasting proteid metabolism—indeed, it seems 
doubtful whether they are—but the fact that the demands of the 
organism for energy may affect the proteid metabolism is of itself 
sufficient to show that our uhit of comparison, while practically 
convenient and perhaps sufficiently accurate, is not invariable. 
Amount oF NOon-nriTRoGENous Nutrients REquirepD.—In 
most of the experiments which have been cited, the very low figures 
for the necessary proteid supply have been obtained by the em- 
ployment of an amount of non-nitrogenous nutrients materially 
in excess of the estimated requirements of the animal for energy, 
although in no case was this latter factor actually determined. 
Sivén,* however, experimenting upon himself with a diet equal 
in amount to that ordinarily required to maintain his weight, was 
able to gradually reduce the total nitrogen of his food to 4.52 grams 
and maintain nitrogen equilibrium. He did not determine his fast- 
ing metabolism, but the above figure, which is equivalent to 0.08 
gram of nitrogen per kilogram live weight, is lower than the low- 
* Skand. Arch. f. Physiol., 10, 91. 
