A STUDY OF RATIONS FED TO MILCH COWS. aoe 
the Station representative or the owner, and it is believed that 
the errors of judgment were not large. The weight was esti- 
mated to the nearest twenty-five pounds. 
Number of months since the last calf.—In most cases the time 
since the cow dropped her last calf was known and was recorded. 
Number of months till due to calf.—There was more or less 
uncertainty in regard to this point but in most cases it could be 
approximately ascertained. 
Daily milk yield during the tests.—The milk of each cow 
at each milking was weighed to the nearest tenth of a pound 
by the Station representative. 
Percentages and amounts of butter fat in the milk.—A sample 
of the milk of each cow, night and morning, was taken and 
from the combined sample a determination of the percentage of 
butter fat was made by the Babcock method. From the per- 
centages of butter fat and the weight of milk the daily yields of 
butter fat were computed. 
Kinds and weights of foods used.—During all the tests of the 
first year and the first tests on each herd of later years, the 
dairyman or farmer was requested to use as nearly as possible 
the same ration as he had been feeding immediately before the 
test began. ‘The quantity of food for each animal was weighed 
by the Station representative just before feeding. If any consid- 
erable portion of the food was left uneaten by the cows it was 
weighed and due allowance was made in estimating the amounts 
~ eaten daily. ; 
During the early part of the test samples of each feeding 
stuff used were taken and sent to the laboratory for analysis, in © 
order to determine the actual amounts of the different nutrients 
in the daily ration. Estimates were made of the amounts of 
digestible nutrients in the ration fed each herd, and a new 
ration was suggested which should furnish a considerably larger 
proportion of digestible protein, and have a correspondingly 
narrower nutritive ratio. In some cases the analyses of the 
feeds used in the first tests could not be made in season to be of 
use as a basis for calculating the new ration, and in other cases 
only the nitrogen (and protein) was determined at the time. 
Under such circumstances average figures for composition were 
