166 PRINCIPLES OF ANIMAL NUTRITION. 
the amount of proteids metabolized, was sufficient to account for 
the gain of fat... They therefore concluded that the carbohydrates 
simply protected these materials from oxidation and regarded the 
formation of fat from the former as improbable, being confirmed 
in this belief by the observation that the amount of fat produced 
was proportional to the proteids rather than to the carbohydrates 
of the food. : The apparent exceptions they regarded as due to a 
retention of undigested starch in the alimentary canal. In brief, 
Pettenkofer & Voit, while not denying that carbohydrates aid in 
the production of fat, regarded their action as an indirect one. 
It should be added that, contrary to the general impression, Voit 
did not absolutely deny the formation of fat from carbohydrates, 
but regarded it as improbable and unproved. Moreover, he came 
later to admit the truth-of the opposite view, and even furnished 
from his laboratory experinfental evidence in its support. 
At an earlier date Voit * had likewise made experiments on a 
milch cow, the result of which was that not only all the fat of the 
milk, but most of the milk-sugar as well, could be accounted for by 
the proteids and fat of the food. Voit also examined the numerous 
experiments of Dumas, Persoz, Boussingault, and others (p. 163) 
upon the origin of animal fat and satisfied himself that, although 
they undoubtedly showed, as their authors claimed, a formation 
of fat from other ingredients of the food, the amount produced 
could at least in the great majority of cases be accounted.for by 
the proteids of the latter. 7 
It is important to observe that the evidence supporting Voit’s 
view was negative evidence. The results could be explained on the 
hypothesis that the carbohydrates did not contribute to fat pro- 
duction, but while a large number of such results might render the 
hypothesis very probable, they could not demonstrate its truth. On 
the other. hand, even a single well-authenticated case in which the 
fat and proteids of the food did not suffice to account for the amount 
of fat formed in. the body. would suffice to establish the possibility 
of its formation from other materials. A few apparent cases-of this 
sort among earlier experiments Voit was able to explain plausibly, 
but there was one important exception, viz., the experiments of 
* Zeit. f. Biol., 5, 79-169. 
