14 DAIRY CHEMISTRY 



does not impair the usefulness of the test, because 

 frequently the losses in skim milk and buttermilk 

 are greater than this, and so far as the whole milk is 

 concerned the method is perfectly reliable. When 

 the Babcock test shows only a trace of fat in the 

 skim milk or buttermilk, the losses are very small. 



15. Sampling Milk. Milk should be thoroughly 

 mixed before sampling. The milk as it comes from 

 the cow or when it has been standing is not in a 

 condition to sample until it has been thoroughly 

 mixed, either by pouring from one pail to another 

 or by stirring with a long-handled dipper. Milk 

 brought to the creamery in cans also requires thor- 

 ough mixing before sampling. This is best accom- 

 plished by the use of a long-handled dipper. Before 

 sampling, the milk should be weighed. 



The milk should not be measured into the test 

 bottles when it is either hot or cold. At a high 

 temperature the milk is expanded and may contain 

 an abnormal amount of dissolved air ; while at a low 

 temperature the milk may be unduly contracted. A 

 temperature of 70 to 80 is the most suitable for 

 measuring milk. If the milk has been standing for 

 some time in the sample bottle, it is necessary to mix 

 it thoroughly before measuring with tne pipette. 

 To do this, turn the milk from the sample bottle into 

 another bottle or dish, pouring it down the side to 

 prevent the formation of foam. In case the sample 

 bottle has been standing until the fat has separated 

 into a layer of cream, the bottle may be placed in a 



