18 
Bulletin 93. 
uble in the petroleum; subsequent attempts to separate the col¬ 
oring matters from these extracts, though very unsatisfactory in 
themselves, indicate that this assumption was not wholly justi¬ 
fied. The coefficient of digestion for the fat, petroleum extract, 
was negative, as in the case of the ether extract—showing over twice 
as much fat in the feces as was ingested with the hay; the negative 
coefficient for the ether extract being 111.67 and for the petro¬ 
leum 110.8. 
The next thing suggesting itself was that the excess of sub¬ 
stances extracted from the feces by the ether might be due to 
biliary products, and we sought for cholesterine and bile pigments. 
We did not obtain satisfactory crystallizations of cholesterine, but 
we did obtain a good Petenkofer reaction. This is hardly to be 
wondered at, as this substance occurs so generally distributed with¬ 
in the body. We obtained fairly good reactions for bile pig¬ 
ments, and were it not for the presence of other substances which 
might have produced the reactions observed, one would be justi¬ 
fied in asserting that they were present. As the matter stands, 
however, I am very doubtful about the actual presence of bile pig¬ 
ments, and I am very fully convinced that this class of products 
do not furnish the explanation for the excessive amount of extract 
in the feces. By excessive is here meant relative to the amount in 
the hay feed. 
We attempted to determine the chlorophyll in these extracts; 
the results were, as was to be foreseen, unsatisfactory, but indicated 
that from 30 to 35 per cent of the extract consists of chlorophyll 
and related substances. The petroleum extract was not colorless, 
but contained a considerable quantity of coloring matter. This 
coloring matter was also soluble in ether, for when the sample 
was first extracted with ether, and then with petroleum, the latter 
remained colorless. The large amount of coloring matter in al¬ 
falfa gave us trouble in other operations; for instance, we found it 
necessary to use lead and copper salts jointly in obtaining a col¬ 
orless solution from an alcoholic extract of alfalfa hay. 
The question of the coloring matters was not prosecuted fur¬ 
ther and was considered to this extent only because of their direct 
disturbing influence upon our fat determinations and indirectly 
upon some of our work due to the color imparted to the solutions, 
making it difficult to observe the reactions or to determine when 
the end had been reached. 
In all of this we have been unable to find any explanation of 
the fact that this series of experiments gives us no digestion coeffi¬ 
cient for the ether extract in alfalfa hay. I have canvassed all 
of the analytical difficulties which have occurred to me as pos¬ 
sibly being capable of furnishing even a suggestion of an ex¬ 
planation, the analysis of the hay and also those of the feces have 
been repeated several times by different operators and with great 
