HOW INDIVIDUAL SELECTION IS MADE 141 



remaining in the tube is placed in a pint glass jar bearing the 

 name or number of the cow. These jars must be kept covered 

 tightly to prevent evaporation. If a sampling tube is not 

 at hand, a fairly accurate sample may be prepared by taking 

 equal quantities of the milk from each milking with a very 

 small dipper and placing in a jar provided for that purpose. 

 Some preservative is used to prevent the souring of the milk 

 before the time for testing. For this purpose formaline, 

 which may be purchased from any drug store, is one of the 

 best agents. Ten drops of this substance will be sufficient 

 to preserve half a pint of milk for several days. A small 

 amount of corrosive sublimate serves the same purpose. 

 This substance is a strong poison, and for this reason milk con- 

 taining it must be handled very carefully. A quantity as 

 large as a grain of wheat is ample to preserve a sample for 

 testing. It is well to use prepared tablets containing a color- 

 ing matter, or to add some common dye to prevent the milk 

 from being used for food by mistake. 



The samples when complete are to be tested by the Bab- 

 cock test. The reading of the test is per cent butter fat, 

 or pounds of pure fat per 100 pounds of milk. The total 

 yield of milk multiplied by the per cent of fat gives the total 

 fat yield. If the cream from this milk should be separated 

 and churned into butter, a larger quantity of butter would be 

 obtained than there was of butter fat. This results from the 

 incorporation of some curd, considerable water, and a small 

 amount of salt with the fat, the mixture being ordinary 

 butter. If proper methods are followed, the yield of butter 

 will exceed that of fat by about one sixth, which is known as 

 the overrun. In estimating the amount of butter from the 

 butter fat it is customary to add one sixth to the butter fat. 



