442 PRINCIPLES OF ANIMAL NUTRITION. 
the other hand, the amount of metabolizable energy required to 
produce it is computed on the basis of the results upon utilization 
obtained in other experiments, this larger amount being added to 
the energy of the excreta and the sum of the two subtracted from 
the potential energy of the food; that is the energy of digestion 
and assimilation above the maintenance ration is assumed to be 
waste energy. 
Computed in this way, and assuming further that the mainte- 
nance requirements of different animals are substantially propor- 
tional to the two-thirds powers of their live weights, the results are 
as follows: 
‘ Main- 
Live Stable 
No. of * tenance 
Animals. vee t, p as Require: 
Deg. C. Cals.’ 
Observed - 
Unfattened 2... cc. cece cee eee 7 632 15.2 13,470 
Hattened gem sca. serene ge wg snksere es 3 785 15.7 19,671 
Computed to same live weight - 
Unfattened oo. accede nee eae eed 7 800 15.2 15,760 
Fattened ............---. eee eee 3 800 15.7 19,920 
Kellner concludes from these figures that the maintenance re- 
quirements of fattened animals are greater per unit of surface than 
those of unfattened ones. 
These experiments, it is true, were on different animals and the 
individuality of the animal is an important factor in determining 
the maintenance requirement. The results on the seven unfattened 
animals, when computed to 600 kgs. live weight, show a range of 
1760 Cals., or 13.54 per cent. of the average, while the three results 
on fattened animals, computed to 800 kgs. live weight, show a 
range of 2420 Cals., or 12.16 per cent. of the average. Moreover, 
in making up the average of the unfattened animals, one animal 
was excluded on the ground that the results were probably abnor- 
mally high, but the same animal is subsequently included among 
the three fattened animals the results on which are averaged. 
Even after making all allowances for these facts, however, the 
results for the fattened animals are decidedly higher relatively 
Digitized by Microsoft® 
