242 THE BOOK OF BUTTER 



composite sample test 3.95 per cent, and the composite 

 sample taken with the Scoville sampler tested 4.00 per 

 cent. Thus the daily test was again slightly higher than 

 the others, but it was well within the limit of error in read- 

 ing. These small differences might be explained in the 

 following way : There is a natural tendency to read a 

 test high in the same way that it is natural to give a small 

 over-weight or slight over-measure when weighing or 

 measuring butter or any other article. As long as there 

 were seven-, fourteen-, and thirty-day periods respectively, 

 there would be seven, fourteen, and thirty times as many 

 readings to make in the daily as in the composite samples. 

 Thus there might easily be a difference of the above 

 amounts of fat due to this one reason. 



222. Testing whole milk. The following are the steps 

 for testing whole milk for f at : 1 . Agitate the sample 

 thoroughly in order to obtain a representative portion. 



2. With the regular milk pipette, measure out 17.6 

 c.c. of milk and put it into the test-bottle. 



3. With an acid measure add the sulfuric acid (com- 

 mercial H 2 SO 4 , Sp. G. 1.82 to 1.83) to the milk in the 

 test-bottle, being careful to hold the bottle in such a way 

 that the acid will run down the sides of the neck and 

 the bowl rather than to allow it to go directly into the 

 milk. When the acid is added in this way it will go to the 

 bottom of the bottle and rise under the milk; thus the 

 fat is not so likely to be scorched as when the acid is 

 poured immediately into the milk. The test-bottle should 

 be revolved slowly when the acid is being added in order 

 to carry down any milk that may adhere to the neck of 

 the bottle. The exact amount of acid to add depends 

 largely on the temperature of the acid and of the milk. 

 If the temperatures are high, less acid is necessary than 



