THE FOOD AS A SOURCE OF ENERGY. 283, 
other classes of herbivorous animals have been reported, although . 
partial data are available from experiments on horses and swine. 
MetHop or Sratinc Resutts.—The determination of the 
metabolizable energy of a given ration by experiments like the- 
above is, in principle, very simple, although requiring many appli- 
ances and much technical skill. When, however, we attempt to 
generalize the results much greater difficulties are encountered 
than in the cases previously considered. 
In investigations upon carnivora and upon man the metaboliz- 
able energy, as we have just seen, is usually computed upon the 
total nutrients of the food—that is; upon the total amounts of. 
protein, carbohydrates, and fat—the deduction for the loss of 
energy in the feces being included in the factors employed. This 
is possible because the amount of potential energy thus removed 
is small in itself and subject to relatively small variations on ordi- 
nary diet and also because the crude nutrients composing the food 
are largely chemical penIpOUnes which are at least fairly well 
known. 
The food of herbivora, on the contrary, is both more complex 
and less well known chemically and contains a much larger and very 
varying proportion of indigestible matter. As a consequence the 
feces, instead of being chiefly an excretory product, consist mainly 
of undigested food residues with but a small proportion of meta- 
bolic products, and contain a large and variable part of the total 
potential energy of the food. For all these reasons it seems likely 
that any attempt to compute general factors for the metab- 
olizable energy of the crude nutrients of feeding-stuffs similar to 
those of Rubner or Atwater for the nutrients of human foods would 
be confronted by almost insuperable difficulties. 
It was natural, then, to attempt to eliminate these difficulties 
by computing the results upon the digestible nutrients of the feed- 
ing-stuffs, but even here considerable difficulties arise. The di- 
gested nutrients, particularly in the case of coarse fodders, are far 
from being the pure protein, carbohydrates, and fats which our 
ordinary statements of composition and digestibility assume them 
to be. Furthermore, a considerable and a variable proportion of 
the waste.of proteid metabolism in the herbivora takes the form of 
hippuric acid, a body less completely oxidized than urea, and ac- 
