Test of Cream Separators. 163 



the gravimetric analysis will show between . i and . 15 of i per cent, 

 of fat in the skim milk. 



The samples of skim milk were taken in all cases from the 

 mixC'l skim milk of the entire run, and not caught directly from 

 the skim milk outlet of the machine at any period of the run. 



The runs were short particularly for the larger capacity ma- 

 chines and in making short runs the error in, computing the ca- 

 pacity per hour is largely increased for the reason that unavoidably 

 part of the milk at the last end runs through slower, so that in 

 interpreting the tables a certain allowance should be made in the 

 capacity per hour, particularly for the Alpha No, i and the 

 Sharpies' machines. 



The milk used was, m all of the tests between January 30th 

 and March 23d inclusive, thd milk furnished to the Dairy School 

 by farmers. This milk was not always in the best condition for 

 handling. The milk of the evening and following morning was 

 shipped together about twenty miles by rail, and reached the 

 Dairy Building in the afternoon and was not worked up until the 

 following day, so that all of it was twenty-four hours old and some 

 thirty-six hours old before being worked. Moreover, on many 

 occasions it was received at the Dairy Building in a frozen condi- 

 tion. In all of the tests made after March 24th the milk of the 

 University herd was used and it was separated immediately after 

 milking, so that we had in these tests milk in its most favorable 

 and unfavorable conditions. 



Mr. Goodrich, Mr. Saunders and Mr. Cavanaugh did a large 

 part of the work here reported and much of the credit of it is due 

 to them. 



In the tables following are given the data concerning the vari- 

 ous runs of the different machines. The tables are given in the 

 order of the list of the machines given at the beginning of the 

 Bulletin and any peculiarity of the running of the machines is 

 mentioned in connection with them. 



