THIRTY-SIXTH ANNUAL CONVENTION. 163 



facing the package. After the trier of butter is drawn and the 

 sample taken, the top two inches is replaced, thus leaving the sur- 

 face of the package in the same condition as if it had been ex- 

 amined for quality only. 



It is noted that the amount of brine forced to the surface 

 varies with the condition of the butter. When the butter is in a 

 very soft condition, little brine is forced out and at the same 

 time less water is seen on the trier. Storage or frozen butter 

 shows no visible water either on the top of the tub or on the trier 

 plug. Butter commonly found in the commission house did show 

 this free water. It seemed, therefore, reasonable to sample the 

 butter in that condition and at the same time carry on investiga- 

 tional work to ascertain the accuracy of this method of sampling. 



May lo, 1907, a churning of 242 pounds of butter was 

 printed in a Lusted printer. Before any of the butter was taken 

 from the churn, the surface was cut off, and a sample taken from 

 various places in the churn, and mixed in one sampling jar. This 

 sample contained 14.05 percent of water. From each tray of 25 

 pounds the center one pound print was removed and a portion 

 of it placed in a sampling jar. It was found that the sample 

 from tray i contained 13.25 percent of water and the analysis of 

 the other nine samples gave the following results : Sample No. 

 2, 13.41; No. 3, 13.05; No. 4, 13.38; No. 5, 13.21; No. 6, 13.01; 

 No. 7, 13.68; No. 8, 12.99; No. 9, 13.21, and sample No. 10, 

 13.25 percent. Average from the 10 samples taken from the one 

 pound prints of butter was 13.24 or 0.31 of one percent less 

 water than was found in the samples taken from the churn. The 

 butter made three days later was sampled in the same manner. 

 Samples taken from the churn contained 13.64 percent water. 

 The average percent of water, of the seven samples taken from 

 the butter after it was printed was 13.27. The highest water 

 content, 13.96 percent, was in sample from tray No. 5 and low- 

 est 12.71, from tray No. 4. 



May 15 all of the cream received was churned in two dif- 

 ferent churnings. The method of making and sampling the but- 

 ter was the same as in the two previous lots. The sample taken 



