282 
THE PHARMACEUTICAL JOURNAL AND TRANSACTIONS. 
[October 8,1870. 
oil the contrary to be tolerably well made out that 
fat is a product of the breaking up of albuminates, 
and Voit regards it as probable that the milk sugar 
in milk originates from the fat as the result of an 
oxidation, so that both fat and milk sugar would be 
derivatives of albuminates hi the food. 
But in my opinion the investigations conducted by 
Voit in reference to the origin of fat and milk sugar 
in the milk of cows, lead to entirely opposite conclu¬ 
sions, and it may not be uninteresting to examine 
somewhat closely the basis of his experiments and 
opinions. 
The most important argument adduced by Voit in 
favour of the probable formation of fat from albumi¬ 
nates is based upon some experiments made, together 
with Pettenkofer, by which Voit believes he has 
proved that, in the body of a dog fed on meat, fat was 
formed from flesh, or might have been so formed. 
Comparing the quantity of carbon in the meat- 
food taken, with the quantity given off as carbonic 
acid and in the other excretions, Voit found there 
was a deficiency of 3’8 grams carbon in the excre¬ 
tions, and, from considerations as to what may have 
become of that carbon, he is inclined to think that it 
was converted into fat and remained in that form in 
the body of the dog. 
Although the deficiency of 3'8 grams is but small, 
Voit, nevertheless, does not think it can be due to 
experimental error. 
Examining more closely the data employed in this 
calculation, it is, in the first place, noticeable that the 
daily excretion of urea varied between 100'41 and 
115'02 grams, while the ftecal excretion also varied 
between 18T and 53 - 6 grams. 
The quantity of carbon in the urine was calculated 
from the mean of ten experiments ; the quantity in 
the faeces from the mean of seven experiments ; but 
on the other hand, only three respiration experiments 
were taken into account for the determination of the 
carbon in the expired carbonic acid. 
It appears to me that this circumstance involves 
an error, though it may be a small one, inasmuch 
as a correct balance could be expected only when 
the determination of the carbon eliminated in the 
urine and feces referred to the same day on which 
the carbon in the expired carbonic acid was deter¬ 
mined ; but on that day the dog did not pass any 
feces, so that the numbers given could only re¬ 
present estimated quantities, and though they are as¬ 
tonishingly accurate when not differing more than 3 - 8 
grams in the carbon, they are not absolutely accu¬ 
rate, nor would any one acquainted with experiments 
of this kind regard them as sufficiently accurate to 
serve as a basis for a theory of fat production. 
But even if the accuracy of the deficit be not dis¬ 
puted, there is a further reason for refusing to attach 
any weight to Voit’s inference that the 3’8 grams of 
carbon wanting in the excretions had been converted 
into fat, because he forgot that the meat on which 
he fed the dog contained some portion of fat. In liis 
previous experiments with Bischoff, he says, “ the 
meat was good fresh cow-beef, always very carefully 
deprived of fat, bones, etc. Several analyses showed 
that it contained at the utmost only 1 per cent, 
of fat.”* 
I have myself determined the fat in a piece of 
very lean beef that had been selected by Professor 
* ‘ Die Gesetze der Ernahrung dee Fleisclifrcssers,’ p. 58. 
Bischoff for this purpose, and, by dissolving the meat 
with hydrochloric acid, I obtained -§■ per cent. fat. 
Then considering that in the experiments made by 
Voit and Pettenkofer, the dog was fed daily with 
1500 grams of meat, and assuming this to contain | 
per cent, of fat, the dog would receive 5 grams 
of fat daily in the meat; moreover, if there was 
really, as observed, a deficiency of 3’8 grams carbon, 
corresponding exactly to 5 grams of fat, it is in¬ 
finitely more probable that this quantity of fat, 
taken in the food, had simply remained in the dog’s 
body, than is the assumption that this 5 grams of fat 
was consumed in the respiration, while another 5 
grams was produced out of the albuminates of the 
meat. Such an assumption is indeed incompatible 
with the economic laws of the animal body. 
Besides this, Voit did not take into account the 
fact that in his earlier experiments together with 
Bischoff,* a dog fed with 1800 grams of meat (or 300 
grams more than in the case above referred to) lost, 
within seven days, 230 grams, or nearly half a pound 
of body-weight. 
This fact is as little in favour of the production of 
fat from flesh, as the more recent experiments. It 
may indeed be said that the loss of weight in an 
animal fed with meat does not at all disprove the 
production of fat from flesh, since the fat must dis¬ 
place a certain quantity of water, and thus give rise 
to loss of weight; but such an objection as this 
would not have any significance, unless the produc¬ 
tion of fat from flesh were indubitably proved, and 
that is not the case. 
In Voit’s investigation there is the same kind of 
error that Pasteur committed when, from the loss in 
his determination of the ammonia in fermented 
liquors, he inferred a positive fact, without knowing 
the source of that loss—thus adopting a procedure 
inadmissible in scientific investigation. 
As may be gathered from what has already been 
said, Voit’s experiments with the dog cannot any 
longer be admitted as arguments in the question as 
to the production of fat. 
In regard to Voit’s experiments with the milch 
cow, his exposition of them, just as in Thomson’s 
investigations, is based upon the erroneous idea that 
a diet rich in albuminates exercises an influence on 
the production of butter, and increases the quantity 
of it, while existing experience only goes to show 
that a strong diet (Kraftfutter) increases the yield 
of milk. 
The experiments made in this direction by Kuhn 
show that the composition of cow’s milk is very con¬ 
stant with different animals and kinds of food ; milk 
varies in the amount of water, but during long 
periods of observation there is scarcely any altera¬ 
tion in the relative proportions of its constituents. 
“ The greatest differences were observed in the fat, 
next to those in sugar. On the average of all the 
animals tins amounted to 0 09 per cent, in favour of 
food that was itself superior to the extent of about IT or 
18 per cent.” This is a remarkably small difference.! 
It is evident that it would only then be possible to 
speak of any influence of albuminates upon the yield 
of butter if, on the addition of albuminates to the 
fodder, the amount of butter in milk were sensiWy 
and permanently augmented; however, observation 
shows nothing beyond the fact that the yield of milk 
* Ibid., p. 79. 
f Landwirth. Versuchs-Station, Dr. Nobbe, 1869, xii. lot. 
