
. ; a 
< CONSERVATION OF ENERGY IN LIVING ORGANISM. bi] 
than those for the succeeding days. Considering each experi- 
ment as a whole, and comparing the averages of the several 
experiments one with another, the range of variation is less. 
Here the net outgo varies from 103 calories below to 62 calories 
above the net income per day. Expressed in percentages of 
Pemincome, the range is from —4.1:to +2.9 percent. But in 
the average for the nine experiments the figures for the net 
‘income and outgo are practically the same, being 2,268 and 
2,259 calories, respectively. 
The discrepancies in the individual experiments with E. O. 
are the largest we have found. They occurred in the early 
experiments, which, however, were largely given up to the 
development of the experimental methods. In the later ex- 
periments, especially those with J. C. W., the variations were 
much smaller. The four widest variations in single days in 
the experiments included in the table above being from. +231 
to —258 calories, or from +4.6 to —4.7 per cent. and the four 
widest in the individual experiments, including more than one 
day from +157 to —96 calories or from +-3.1 to —2.1 per cent. 
of the net income in each case. 
The way in which the errors compensate each other is illus- 
trated by the averages for income and outgo in Table 18, these 
averages being obtained, not by averaging the averages of in- 
dividual experiments or groups of experiments, but by divid- 
ing the sum totals in each class by the corresponding numbers 
of days, as in the instance above cited. The sum totals from 
which the averages for the several classes of experiments in 
Table 18 were obtained are given below. ‘The experiments 
with J. C. W. were the latest and ‘therefore represent the ad- 
vantage of accumulated experience in the development of 
apparatus and method. 
In the average of the twelve rest experiments with ordinary 
food, covering 41 days, the sum total of the figures of daily 
income is 92,101 and for outgo 92,118 so that when the aver- 
ages per day are found by dividing these numbers by 41, omit- 
ting fractions of a calorie, the figures for the two are the same, 
namely, 2,246 calories. A similar identity is found in the work 
experiments with J. C. W., for in the average of all the 32 ex- 
periments, covering 107 days, with ordinary diet, the daily in- 
ers: 
come is 3,748 and the outgo 3,745 calories, making a difference 
of 3 calories or hardly o.1 per cent. of the whole. 
