TESTIiNG FOR MOISTUEE. 209 



than 1.0 per cent fat, simply because of the effect of 

 this one source of error which is introduced when 

 weighing the butter for testing. 



Another point in testing butter for fat by the 

 Babcock test, that will have a great effect on the 

 final results, is the temperature of the fat column 

 when it is measured. The extent to which the fat 

 column will change with the increase or decrease in 

 temperature is shown by the following figures : 



The expansion of fat by a change of 40 degrees 

 in temperature amounts to .07 per cent- fat in milk 

 testing 5 per cent fat; that is, if a milk testing 5 

 per cent fat is read at 110° F. the reading will be 

 5.07 per cent fat when the temperature is raised to 

 140° P. 



This same effect of expansion on the test of butter 

 containing 80 per cent fat will be 16 times more, or 

 .07X16 equals 1.12 per cent fat, which will be the 

 difference in the reading of the butter fat test at 

 110° and 140° F. Or, if the fat test of a sample of 

 butter when read at 110 is 80 per cent fat, the same 

 sample read at 140° F. will be 81.12 per cent fat. 



All the other errors of manipulation in testing 

 butter by the Babcock test, such as determining the 

 point of reading the top and bottom of the fat col- 

 umn, loss of moisture and of the butter when it is 

 being weighed into the test bottles, and other me- 

 chanical losses, are at least five times as great for 

 a fat as for a water determination in butter, and 20 

 times as great for butter as for milk testing 4 per 

 cent fat. 



