460 PRINCIPLES OF ANIMAL NUTRITION. 
first. because they include a correction for the differences in or- 
ganic matter consumed, and second, because the energy of the gain 
has been corrected for the amount of nitrogen retained in the body 
in the same manner as the energy of the urine (compare p. 285), 
viz., by deducting 7.45 Cals. per gram of nitrogen. In most cases 
the metabolizable energy is that already. computed in Tables I to 
VI of the Appendix, being based on actual calorimetric determina- 
tions in food and feces. In two instances (distinguished by being 
bracketed) the metabolizable energy has been computed by the 
writer from such data as are available.* 
In Table VII the final results are expressed as percentages of 
the metabolizable energy utilized. By combining them with the 
results contained in the six preceding tables of the Appendix they 
may likewise be expressed as percentages of the gross energy of 
the several materials and also as energy utilized per gram of total 
organic matter. The summary on pp. 461-2 contains the results 
expressed in all of these ways. 
EaruerR EXxprerimMents.—The earlier respiration experiments 
of Henneberg & Stohmann { on oxen, in 1865, while made in accord- 
ance with the experience then available, are now known to be de- 
fective in several respects. The respiratory products were deter- 
mined for twelve hours only, while the same authors subsequently 
showed that twenty-four hours was the minimum time necessary. 
The food consumed on the respiration day was less than the average 
for the whole experiment, but how much less does not appear, and 
finally the methods used for the determination of the hydrocarbon 
gases excreted have subsequently been shown to give too low re- 
sults. It seems useless therefore to enter into an elaborate com- 
putation of the results. In the later experiments of the same 
authors ¢ with sheep, these sources of inaccuracy were largely re- 
* The data used in these computations are as follows: 
For Ox IV the average results for Periods la and 1b have been com- 
puted on the assumption that the heat values of food and excreta per gram 
in Period 1b were the same as those determined in Period 1a. 
For Ox V the metabolizable energy in Period 3 has been computed by 
adding to that in Period 2a 3.345 Cals. for each gram of organic matter in 
the starch added, this being the metabolizable energy computed for the 
starch in Period 2a. 
+ Neue Beitrige, p. 287. } Loc. cit, p 68. 
