20G MODERN BUTTER MAKING. 



it is safci- to make two tests of the same lot of but- 

 ler, as the result of two tests is a truer indication 

 of the actual moisture present. 



261. Testing butter for butter fat by the Babcock 

 test, and from results, computing the approximate 

 moisture content. Experience has taught that moist- 

 ure cannot be accurately computed by knowing the 

 approximate per cent of fat in butter. The per cent 

 of salt in butter varies too much for this to be a re- 

 liable test. It would be impossible to make an accu- 

 rate guess as to the per cent of salt in butter, and 

 to test for salt as well as for butter fat would take 

 more time than it would take to make a moisture 

 test. Even the use of the Babcock test for determin- 

 ing the per cent of butter fat in butter is not to be 

 recommended due to errors arising from various 

 sources. This is fully explained in the following by 

 Prof. E. H. Parrington, Ploard's Dairyman, July 1, 

 1910: 



"Among the various factors which make these re- 

 sults uncertain are the following: 



1. Taking a fair sample of butter. 



2. Accurate Aveighing of the butter into the test 

 bottle. 



.3. The high per cent of fat in butter which makes 

 errors in reading the length of the fat column due to 

 changes in temperature much greptcr than in milk. 



4. The difficulties in getting the fat free from 

 water when it is lucasured in the neck of the test 

 bottles. 



The effect of thes(! various errors of analysis is five 

 times as great on the fat as it is on the water deter- 



