DAIRY MEETING. 157 



tester and I began to follow out Prof. Hills' instructions to the 

 letter. Those of you who have read his talk or have heard him, 

 will remember that he said the first cause of variation in the test 

 would be a variation in the conditions of the cow, — the tempera- 

 ment, the way she was handled. I handle my cows myself, and 

 it took me only a month to convince myself that although by a 

 sudden fright, or something of that sort, I could make one cow 

 vary very much in her test, taking a composite sample it could 

 only be reckoned in tenths. Another point the professor made 

 was in the separator getting out of order. I use a separator. I 

 unbalanced my bowl, to try that, and there I found that I got 

 large results, as I think, I got about 2.1 variation by a wabbly 

 bowl. I got as high as 2.7 variation by a change of speed from 

 42 to 52 revolutions. Those experiments were made in the win- 

 ter season. I did not get around to an important thing which 

 was brought out by a conference with our Dairy Commissioner 

 late this fall, but I had an opportunity two weeks ago to satisfy 

 myself that one of the principal troubles with me in the past has 

 been in the time of separating my milk. During the cold spell 

 we had last week I tried milking 25 cows and letting the milk 

 stand until they were all milked, then separating it ; and then the 

 next day I separated it as fast as the boy milked, and there was 

 a variation of 3 per cent in the test of that cream in the two 

 mornings. Both of them were cold mornings, so that particles 

 of ice formed in the lean-to. Now to go back to the point which 

 I wish to make and emphasize, and which I really hoped this 

 Association would take some measures on before it adjourned. 

 What interested me the most was the manner in which the 

 creamery man took the sample from which to test my cream. 

 The man who drives on my road has to drive in the neighbor- 

 hood of 20 miles, collecting cream. He arrives at my house, 

 in the summer season, at about 10 o'clock and then has to drive 

 3^ miles to make his train. In the winter season I think the 

 train arrangement gives him an hour longer, but he does not get 

 up quite so early, so is about the same length of time on the 

 road. I began asking him to save me a sample when he took the 

 sample for the factory. I found when I tested my sample and 

 got my test from the factory there was a variation. I do not 

 think it ever exceeded one per cent, but it was not business. I 

 then began taking a sample myself, and then letting him take a 



