THE INFLUENCE OF FOOD UPON MILK. 15 
\ 
caused us to investigate. With the Babcock we found 
the duplicate samples in the evening run very close, never 
varying over .4 of 1 per cent., while those taken the next 
morning varied from J to 2 per cent. The explanation 
is that the cream rises, sometimes dries on top, and fre¬ 
quently is sour, when it is impossible to mix and secure a 
fair sample. At times, without the knowledge of the 
operator, his pipette will draw in a clot of cream, while 
again from the same* vessel its mouth is surrounded by 
the poorest of milk, containing almost no butter fat. 
We find, from repeated analyses with the Babcock, that 
after sampling, the milk may stand in the test bottles 
until it is sour and coagulated, without the results being 
changed. 
From a study of the tabulation we learn that the 
*/ 
quality of milk was quite perceptibly influenced by the 
change of food given these cows. It will be also observed 
that in every case by the Babcock analysis, the wheat 
bran produced the best results, and that the gravimetric 
analysis exhibited two cases as good or better, with the 
other two but slightly lower. It must be remembered 
that in the two cases which showed a lower per cent, 
when the cows were on bran, the samples were from the 
two longest in milk ; and, further, that if there is any 
advantage from this fact, it was given the oat chop ration, 
which was fed first. Three of the cows lost in yield of 
milk, which might he due to some extent to the same 
cause, but more likely to natural fluctuations or the con- 
dition of the weather at that time. This is the more 
likely, since there is sufficient evidence extant, that bran 
causes a better flow of milk than oats. While they 
gained in weight on the oat chop, each lost a few pounds 
on the bran ration. The difference in either case could 
have been caused bv the difference in water drank, at a 
i j J 
single time. With these suggestions, we leave the con- 
