METHODS OF INVESTIGATION. 251 
olizing substantially proteids and fat, then, such as a fasting animal 
or one consuming only those two nutrients, a determination by any 
of the methods indicated in Chapter III of the amount of oxygen 
consumed will afford the basis for at least an approximately correct 
computation of the energy liberated during the same time, par- 
ticularly when, as is often the case, the proteid metabolism consti- 
tutes but a small proportion of the total metabolism. For the 
carbohydrates the figures are somewhat higher, and where these 
bodies constitute a considerable portion of the food the error will 
be more serious, but even then the results will be of value and 
especially will afford relatively correct figures for the heat produc- 
tion on the same diet at different times. 
The computation from the gaseous exchange of the amount of 
energy liberated assumes a more exact form in case it is desired 
to determine the increment arising from some change in the 
conditions of the experiment, notably from an increase in the 
muscular work done. In the latter case, as we have seen (Chap- 
ter VI), the increased metabolism is largely or wholly that of non- 
nitrogenous matter. Such being the case, we can compute in the 
manner indicated on p. 76 from the increments of carbon dioxide 
and oxygen caused by the work the proportion of each gas corre- 
sponding respectively to the oxidation of fat. and of carbohydrates, 
and from this it is easy to compute the corresponding amounts of 
energy. Thus, to take the example from Zuntz’s investigations 
there given, the increments of oxygen and of carbon dioxide pro- 
duced by the performance of 1 kgm. of work in the case of a dog 
were computed to be divided as follows: 
Oxygen Carbon Dioxide 
Consumed, Produced, 
¢.c. c.c. 
BY fats cance jada Guawe aan nee woe Rass 0.6939 0.4905 
“ carbohydrates ............0 0.2002 e eee 0.9765 0.9765 
Total s..ohGacSuce Serted cacao bes arses 1.6704 1.4670 
From this, using Zuntz’s factors and assuming that there was no 
change in the proteid metabolism, the total excess of energy liber- 
ated in the body during work over that metabolized during rest is 
computed as follows: 
