THIRTY-SIXTH ANNUAL CONVENTION. 213 



The final average over-run is the same when based upon 

 butter fat churned and butter fat recovered, according to the per 

 cent of fat in samples taken from 320 tubs of butter representing 

 80 churnings. The over-run calculated from the per cent of fat 

 in the samples taken from the 80 churnings is 1.3 per cent higher. 

 This difference indicates that, either the reading of the Babcock 

 bottle was high or the churn sample gave a per cent of fat lower 

 than is actually in the butter. It is safe to assume that in pack- 

 ing the butter into tubs a small amount of water is forced out 

 which would cause a higher fat content in tub samples. The last 

 column in this Table indicates the per cent of over-run based 

 upon butter fat churned after deducting the butter fat lost in the 

 buttermilk. The actual over-run then is one-half of one per 

 cent higher than is possible according to butter fat analyses. 

 Considering the large number of samples of cream tested and 

 butter analyzed the results thus obtained are remarkably consis- 

 tent. 



