456 PRINCIPLES OF ANIMAL NUTRITION. 
gains on the basal and on the augmented ration, then, we may deter- 
mine what proportion of the metabolizable energy of the added 
food was stored in the gain of tissue. In other words, we may 
determine two points on the line ‘SU in the figure on p. 446, 
thereby determining the line if it is a straight line. 
If the added metabolizable energy of the larger ration were de- 
rived solely from the material added, the result would show the utili- 
zation of the energy of that material. As we have seen, however, in 
connection with the discussion of the metabolizable energy of the food 
in Chapter X, this is rarely if ever the case with herbivorous animals. 
The difference in metabolizable energy between two rations usually 
includes, in addition to the real metabolizable energy of the added 
food, differences in the digestibility of the original ration and in 
the losses in urine and methane. Accordingly, we are here con- 
fronted with the same alternative as before, viz., whether to attempt 
to eliminate these secondary effects and base our computations 
on the real metabolizable energy of the feeding-stuff under experi- 
ment or to take the apparent metabolizable energy as representing 
the actual amount of energy contributed to the metabolism of the 
body. In the one case, if successful, we shall obtain a result which 
will be physiologically correct but which when applied in practice 
will require modification for the secondary effects just mentioned. 
In the other case we shall have a summary expression including 
all these results, but with the disadvantage of being more or less 
empirical in its nature. Hither method has its advantages and 
disadvantages. In the present case we shall use the apparent 
metabolizable energy of feeding-stuffs as computed on pp. 285-297 
and in Tables I-VI of the Appendix as the basis of computation. 
This does not, of course, affect the absolute amount of energy 
utilized from a unit weight of the material, but only the percentage 
calculated upon the metabolizable energy. 
Sources of Uncertainty in Computation.—While the computation 
of the energy utilized from feeding-stuffs in the manner just indi- 
cated is in principle very simple, certain complications arise in its 
execution from the impossibility of securing exactly comparable 
conditions of experiment. Two of these in particular require 
consideration here. 
Differences in Organic Matter Consumed.—As was noted in the 
