THE UTILIZATION OF ENERGY. 539 
particularly in experiments at a trot, owing to the continual changes 
in the friction. Wolff believes that in his experiments, all made 
at a rather slow walk, the errors are less, but admits that they are 
sufficient to deprive his computations of utilization of all valve. 
Grandeau & LeClerc, however, were successful in improving 
the dynamometer, by the addition of an integrating apparatus,* so 
that its measurements of the total work were satisfactory, and this 
apparatus was.added to Wolff’s dynamometer in 1891. Before that 
date, therefore, Wolff’s experiments, while of great value in many 
other respects, afford no trustworthy direct data as to the utili- 
zation of the energy of the food for work production, although, as 
we have just seen, they afford some information on subsidiary 
points. From 1891, however, we may regard the measurements of 
the work done on the dynamometer as reasonably accurate. 
Corrections. -— Unfortunately, in the light of subsequent 
investigation, the same is not true of some of the other factors 
entering into the comparison, particularly the work of locomotion 
and the metabolizable energy of the food. 
In all his later experiments Wolff computes the work of hori- 
zontal locomotion per second by means of the formula ; (Fw, 
in which W equals the weight of the animal, g the force of gravity, 
and v the velocity per second. Zuntz’s experiments, however, 
appear to show that this formula gives too high results, the error 
increasing with the velocity, and Wolff + himself recognizes the truth 
of this for higher speeds. According to Zuntz’s determinations 
(p. 512), Kellner’s method of computation gives results agree- 
ing quite closely with those computed from his respiration experi- 
ments. Under the conditions of Wolff’s experiments this corre- 
sponds quite closely to 50,000 kgm. per 100 revolutions of the 
dynamometer, and in the comparisons which follow this amount 
has been substituted for that computed by Wolff, thus reducing 
materially the figures for the total work performed. 
Wolff estimates the metabolizable energy of the food, on the 
basis of Rubner’s results, by multiplying the digested fat by 2.4, 
adding the remaining digested nutrients, and reckoning the total 
* Ann. Sci. Agron., 1881, I, 464. 
{ Landw. Jahrb., 16, Supp. III, 119. 
