TWELFTH ANNUAL YEAR BOOK— PART VIII 381 



or aliquot portions of tlie amount the patron delivers. When the 

 samples are taken in aliquot portions and kept in jars with glass 

 stoppers, or jars sealed with cork so there will be no evaporation, 

 and kept in a cool place away from the rays of the sun in the hot 

 months and in the cold days kept where they will not freeze, the 

 composite sampling may be fairly correct. How many do those 

 things all the time? Now, unless these precautions are taken, how 

 can the sample represent the average richness of the various lots of 

 cream from which samples are taken? 



You will be much surprised, if you test every day, at the variations 

 to be found in the patron's cream at different deliveries. There are 

 variations in the richness from day to day, and as a result the amount 

 of cream will vary. One day he will have more pounds of cream 

 with a lower test than he did the day before with less cream. Many 

 factors enter into the variations in the amount of cream and in the 

 test which the farmer does not always have control of and as a 

 result he has richer cream one day than another. Now, unless the 

 buttermaker takes the sample in the correct proportion for these 

 variations in the richness and amount at each delivery, how can he 

 have a correct composite sample? 



We will now discuss the dipper sample. As an illustration we will 

 say Mr. Jones makes two separate deliveries of cream to our cream- 

 ery and we take two samples for testing at each time, one for every 

 day testing, and one for the composite. Today he delivers 200 pounds 

 of cream testing 20% which makes 40 pounds of butterfat. Tomor- 

 row he brings 100 pounds testing 40% which makes 40 pounds of. 

 butterfat. Testing this for every day work he has actually delivered 

 80 pounds of butterfat. Now we test the composite sample and find 

 the average test is 30%. He has delivered 300 pounds of cream 

 and with the 30% test he would be paid for 90 pounds of butterfat, 

 or the creamery would have to pay for a loss of 10 pounds of but- 

 terfat. You may reverse the illustration, and suppose that Mr. Jones 

 delivered 100 pounds of cream testing 20% making 20 pounds of 

 butterfat. The next day he delivers 200 pounds of crean™ testing 

 40% making 80 pounds of butterfat. In all he has 100 pounds of 

 butterfat. We will now test the composite sample and find the aver- 

 age test to be 30%, the same as in the preceding example, and he 

 has delivered the same amount of cream, 300 pounds. Thirty times 

 300 makes 9 pounds of butterfat the creamery has to pay for. In 

 this case the creamery wins and the patron loses 10 pounds of butter- 

 fat. If the patron only knew he was losing what a "howl would go 

 up from Denmark." Now, on the other hand, it is all right with 

 him when the loss comes on the creamery. 



This example may be somewhat overdrawn in the differences we 

 make in the tests from day to day, but they are possible The varia- 

 tions found from day to day though smaller mean just the same and 

 have the same evil effects, and show the results which are had by 

 the dipper method. The man who takes the sample with the dipper 

 takes the amount each day through force of habit whether there is a 



