ILLINOIS STATE DAIRYMEN^ ASSOCIATION. 209 



of milk added to the jars of series B was always the 

 fraction of the pounds of which it was a part, i. e., jar 

 C, patron No. 1, always received one-thousandth of the 

 number of pounds of milk he brought, and jar C, 

 patron No. 2, one five-hundredth; a constant propor- 

 tion was selected for each patron so that at the end of 

 the week the accumulated sample amounted to about 

 one pint. 



By this method there were collected three composite 

 samples of the milk brought by each of the twenty 

 patrons, also a test of each lot of milk used in making 

 up the composite samples. 



One point of the investigation was to observe how 

 well the test of the composite samples would agree 

 with the average of the seven daily tests of the milk 

 used to make up this sample. The results proved very 

 satisfactory. The milk in the jars of series A and B 

 had not soured at the end of the week, the cr,eam had 

 separated somewhat, but with a little care it was easily 

 mixed. This can be done by holding the jar right side 

 up and giving it a motion that causes the milk to 

 revolve in it as if on an axis which extends through the 

 center of the jar from top to bottom. This sort of a 

 motion cleans the cream from the sides of the jar and 

 mixes it again with the milk in a very satisfactory 

 way. Unless some such precaution is used to clean the 

 cream from the sides of the jar, a portion of it sticks 

 very tenaceously and vitiates the results of the test. 



In the jars of series C, to which no preservative had 

 been added, the milk of the composite samples was sour 

 and curdled; the whey and curd separated. The "pow- 

 dered lye" was added to all of these samples and by pour- 

 ing and stirring they soon became evenly mixed and thin 

 as new milk. The action of the lye on sour milk is 



