12 AMERICAN DAIRYING. 



Cream test. — The first test I applied to my 

 cows was the test for cream, and this was 

 twenty-five years ago. At that time it was 

 generally supposed that the percentage or depth 

 of cream that raised on samples of milk set in 

 glass test tubes or dishes of uniform size and 

 depth was a correct basis on which to judge of 

 the butter value of the different cows' milk. I 

 soon learned, as have hundreds of others in this 

 line of work, that this test was not reliable, as 

 I found that there was in some instances a dif- 

 ference of 25 per cent in the batter value of 

 milk that showed the same percentage of cream 

 in the glass tubes. 



The Connecticut Experiment Station found 

 cream furnished by different patrons of a cream- 

 ery who set their milk in deep or submerged 

 cans for twelve to twenty-four hours to contain 

 from 13.8 to 24.9 per cent of fat. I proved the 

 cream test to be unreliable by applying the 

 cream test and the churn test to the same milk. 

 The results of these comparisons caused me to 

 abandon the cream test and start anew with 

 the churn. 



Churn test. — This plan cost me much time 

 and labor, each cow's milk being set separately 

 and skimmed an d churned separately. For this 

 work I had a gang of six small churns of the 

 dasher pattern, with the dashers attached to a 



