10 
Bulletin 93 . 
neous, even though the three sheep give the same result, i. e., a 
negative one. The natural explanation would be to attribute it to 
some error, and as the result is common to the three sheep, the 
error, if any has been made, must be a fundamental one, and 
would seem to lie in the determination of the fat or ether extract 
in the hay itself. The hay used was first cutting hay, furnished 
by the Farm Department, probably cut when the plants were in 
early to half bloom, as it is our custom to cut the alfalfa when in 
this condition, though the analysis corresponds to much later cut 
hay. 
The results in the case of the ether extract being so remark¬ 
able, the analytical work, though already done in duplicate, was 
repeated in the case of the hay and the feces of sheep No. 3. The 
principal weakness in my data lies in the sample of hay, the com¬ 
position of which shows nothing unusual except a very small 
amount of fat, ether extract, which is even less than I have here¬ 
tofore found in the stems alone or in hay made from plants that 
were in full seed. 
This extremely low percentage of fat almost forbids the use of 
the coefficient obtained for it in this series of experiments. 
The crude protein, 13.12 per cent, is a shade low, and the 
crude fibre, 41.05 per cent, a little too high for prime, first cutting 
alfalfa hay. But they are so well within the range found for 
these constituents in alfalfa hay that they cannot justly be made 
the subject of adverse comment. The fat, however, being less 
than one-half the amount usually found in good alfalfa hay, is 
open to the gravest doubts. The feces of the sheep fed on this 
hay are, on the other hand, quite as rich in ether extract as the 
feces of other sheep fed with a much better alfalfa hay. The 
average ether extract found in the feces of this series of experi¬ 
ments, and being the average of fifteen determinations, is 3.10 per 
cent, while the average percentage of ether extract found in the 
feces of three other sheep, likewise based upon fifteen determi¬ 
nations, is 3.09 per cent. 
It would seem that if the feces of two sets of sheep, to which 
the same kind of hay had been fed, contained the same amount of 
ether extract (fat), we ought to find a corresponding agreement in 
the amounts contained in the hay feed, provided that the diges¬ 
tion processes had acted upon them in the same manner and de¬ 
gree ; but we do not find this to be the case, as will appear more 
fully in the statement of a subsequent series of experiments. I 
therefore feel it to be incumbent upon me either to reject this se¬ 
ries or to make a somewhat full record of the study made, which 
I shall do as briefly as possible. It would be easier to do the 
former and to use only such results as are in harmony with other 
experiments which are considered as altogether reliable, and the 
number of which add materially to their conclusiveness. 
