THE UTILIZATION OF ENERGY. 527 
of more or less uncertain estimates and assumptions regarding the 
expenditure of energy in digestion and assimilation such as those 
discussed in Chapter XI, § 3. 
The second possible general method for the determination of 
the percentage utilization of the energy of the food in work pro- 
duction is that employed in the determination of the utilization in 
tissue production. Having brought the animal into equilibrium as 
regards gain or loss of tissue and amount of work done with a suit- 
able basal ration, the material to be tested is added and the work 
increased until equilibrium is again reached. The increase in the 
work performed compared with the energy of the material added 
would then give the percentage utilization of the latter. 
The accurate execution of this method would require the em- 
ployment of a respiration apparatus or a respiration-calorimeter 
for the exact determination of the equilibrium between food and 
work, while the skill of the experimenter would doubtless be taxed 
in the endeavor to so adjust food and work as to secure either no 
gain or loss of tissue or equality of gain or loss in the two periods 
to be compared. Indeed, it may safely be said that exact equality 
would, as a matter of fact, be reached rarely and by accident, and 
that as a rule it would be necessary to correct the observed results 
for small differences in this respect. To make such corrections 
accurately, however, requires, as we have seen in § 1 of this 
chapter, a knowledge of the net availability and percentage utili- 
zation of the food, and we are thus brought back to the necessity 
for more accurate knowledge upon fundamental points. 
The extensive investigations of Atwater & Benedict * upon man 
appear to be the only ones yet upon record in which the actual 
balance of matter and energy during rest has been quantitatively 
compared with that during the performance of a measured amount of 
work. Unfortunately, however, the gains and losses of energy by 
the bodies of the subjects in these experiments were relatively 
considerable, while the experiments thus far reported seem to 
afford no sufficient data for computing the net availability of the 
food for maintenance or its percentage utilization for the production 
of gain. Moreover, the authors appear to regard the measurements 
* U.S. Department Agr., Office of Experiment Stations, Bull. 109; Mem- 
oirs Nat. Acad. Sci., 8, 231. 
