enable six samples of cream, with duplicates, to be churned at 
- once. After the first churning, variations were made in the tem- 
_ perature of the cream churned to ascertain what range of tem- 
perature was suited to each cream for best results, and this range 
being ascertained it has been approximated in subsequent churn- 
ings. This is the only condition in which uniformity has not been 
adhered to. The comparatively small amount of cream churned 
renders it more sensitive to the influence of the temperature of 
the room, especially if the temperature of the room is much 
higher or lower than that of the cream in the churn, but effort is 
made to control the temperature of the room and hold it at about 
60° Fahr. This is not always easy to accomplish. 
Each churn was removed as soon as the butter granules were 
about the size of wheat kernels. When the churning of the day’s 
samples was finished, the buttermilk was drained from each 
butter, fine wire gauze being tied over the mouth of each churn to 
prevent the escape of small particles of butter. . The butters were 
then washed with cold water until the rinsings were clear, and, 
finally, the churns were wiped dry on the outside and then weighed 
on a balance sensitive to within 100 milligrams (about 1-283 of an 
ounce.) 
_ Sampling butter for analysis— The churns were then transferred 
to the laboratory, the butter was carefully melted, thoroughly 
shaken, in order to incorporate all the moisture on the inside of 
the churn, the duplicates were mixed, again thoroughly agitated, 
and a portion poured into a smaller wide-mouthed bottle and cooled 
to solidity as quickly as possible under a stream of cold water, 
with constant shaking. In no other way can a uniform sample 
of butter be easily secured for analysis. The samples thus 
prepared were then analyzed. 
The foregoing is the plan first adopted, but the following modi- 
fications have been found desirable: (1.) Prolongation of period 
of creaming. (2.) Making correction for loss in handling. 
(3.) Changing the amount of milk used for making butter. 
_ (4) Changing the manner of washing the butter. (5.) Changing 
method of ripening cream. 
(1.) Prolongation of period of creaming.—It was soon observed 
that the amount of fat lost in the skim milk was much larger 
than it should be, and, on special investigation, it was found that 
the twelve-hour creaming was, in most cases, responsible for the 
New York AGRICULTURAL EXPERIMENT STaTION. 199 
