TESTS OF SEPARATORS. 

 J. M. Bartlett. 



These tests were not undertaken with the idea of making an 

 extensive investigation of separators, but to compare, prin- 

 cipally for our own satisfaction, the three hand machines that 

 are now in the dairy building; two of which, the Empire and 

 United States, have been placed there free of cost for use in the 

 dairy school. The DeLaval Company also supplied us with one 

 of the Humming Bird machines for that purpose. 



The machines used, sizes and capacities are shown in the 

 table. The operator could detect no material difference in the 

 ease of running of the United States No. 5 or Baby No. 2. The 

 Empire being a larger machine of course ran harder. He, how- 

 ever, preferred the United States to the others on account of the 

 simplicity of the bowl, it requiring less time to set it up, clean, 

 etc., 



If in handling the DeLaval discs they are arranged on a wire, 

 the difference in time and work required to clean and set them 

 up more than that required for one of the simpler bowls, is not 

 so much as one would think. 



In the first tests that were made the bowl washings were 

 tested, to estimate losses in that direction, but as the amount 

 of loss was about the same for each machine and the percentage 

 was proportional to the amount of milk run through, this factor 

 was disregarded in the experiment. 



The determination of fat in the skimmed milk was made in 

 most cases both by the gravimetric and the Babcock volu- 

 metric methods. 



The Babcock method, the writer finds, is not reliable when the 

 fat runs below .05 or .06 per cent, even when the B. & W. 

 skimmed milk bottle is used. In several cases, not given in the 

 table, where the Babcock method showed no fat or only a trace, 

 the gravimetric showed .06 to .08 per cent. There are a great 



