300 



BOARD OF AGRICULTURE. 



[Jan. 



the temperature of the surrounding water at similar heights was 

 also recorded. In two cases parallel trials were made by the 

 Holstein method which has been in favor in that part of Germany, 

 and which consists essentially in setting the milk without pre- 

 liminary cooling in shallow pans, so that the milk is not more than 

 two inches deep, and keeping the temperature between 54° and 59°. 

 The following table gives the results of the first series: 



It is noticeable that in these experiments with the exception of 

 No. 3, the milk had a very uniform content of fat, 1 1 per cent, 

 being the range of difference. No. 3, was the milkmg on the 

 warmest day of October, when the average temperature was 65°, 

 and the maximum 71°. 



The result of the skimming in the first experiment compared 

 with the others shows a striking difference, and illustrates how 

 dangerous it is to draw conclusions from experiments with a sin- 

 gle lot of milk without proving that the milk was perfectly 

 normal. • 



In all these experiments from No. 2 on, the milk was not 

 normal. The cream was unusually slow in rising, and in the 

 dairy it was noticed that the milk soured very much sooner than 

 it had done before. This peculiarity was not due alone to the 

 warm weather, but also, as was proved by direct trial, to defect in 

 the food. 



The herd were given in their ration, chaff from grain which had 

 rusted badly, and during the milking, dust from this chaff settled 

 in the milk. 



