110 PRINCIPLES OF ANIMAL NUTRITION. 
a comparison of the income and outgo of carbon and nitrogen. 
Pfliiger,* however, has called attention to the fact that while Pet- 
tenkofer & Voit made direct determinations of the outgo of these 
elements, or at least of the principal factors of it, the income is not 
computed from actual analyses of the meat used, but upon the 
assumption of average composition. According to Pfliiger, not 
only are the possible variations from the average in individual 
experiments a serious source of error, but the average itself is 
erroneous, the percentage of carbon assumed in the meat being 
too high. Pettenkofer & Voit estimate the ratio of nitrogen to 
carbon in lean meat f as 1 : 3.684, while according to Pfliger it is 
not higher than 1 :3.28, and prqbably lower. Moreover, Petten- 
kofer & Voit failed to take due account of the fact that a part of 
the gain of carbon which they observed could be ascribed to the fat 
still contained in the prepared “lean” meat. .Another, although 
slight, source of error, according to Pfliiger, lies in the fact that 
the carbon in the urine was estimated from the amount of nitrogen 
found by analysis on the assumption of a ratio of 1:0.60, while it 
should be 1:0.67. 
Using the above corrections, Pfliger has recalculated twenty- 
four of the experiments by Pettenkofer & Voit, which have been 
generally accepted as demonstrating the formation of fat from pro- 
teids, with the results shown on the opposite page. 
In the great majority of cases the experiments as recalculated 
show a loss instead of a gain of fat, and in three of the four cases in 
which a gain still appears it is small in amount, and, as Pfliger 
believes, within the limits of experimental error. Naturally such 
calculations as the above can neither prove nor disprove the hypoth- 
esis that the proteids serve as a source of fat.. They simply show 
that the experiments which have served as the principal support 
for that hypothesis do not demonstrate what they were supposed 
to. The question turns largely upon the elementary composition 
of the meat used by Pettenkofer & Voit, which they failed to 
determine. It is manifestly impossible to repair this error now, 
* Arch, ges. Physiol., 51, 229. 
+ Including such fat as cannot be removed by mechanical means. 
{ Loc. cit., p. 267. The experiments which showed a loss of fat as origi- 
nally computed are omitted. 
