72 MAINE STATE COLLEGE 



The writer has for the past few months devoted what limited 

 time could be spared from other station work to studying methods 

 with the idea of especially adopting some one for testing cream. 

 As a result of this investigation the method devised by Dr. 

 Babcock of the "Wisconsin Experiment Station was selected, it 

 being more simple and rapid than any that had appeared at the 

 time of beginning this work. 



Since that time the Beimling method has been brought out and 

 by some is considered equally good, but the apparatus is a little 

 more expensive. The Babcock method is fully explained 

 in Bulletin No. 24 of Wisconsin Experiment Station, there- 

 fore it is only briefly described here. The milk is measured into 

 a test bottle with a graduated neck, and about an equal amount of 

 sulphuric acid is added, which dissolves all solids except the fat. 

 The bottle is then at once whirled in a centrifugal machine to collect 

 the fat on the surface of the liquid. Hot water is then added to 

 raise the fat into the graduated part of the neck of the bottle, 

 where the percentage can be read. The bottle used in the milk 

 test as devised by Dr. Babcock is graduated to estimate no 

 higher than 10 per cent, of fat. and as cream usually contains 

 much more than that, it is evident this bottle cannot be used, if 

 the same quantity (18 grams) is taken for the test. 



To take a smaller quantity increases the liability to error and 

 impairs the accuracy of the results. A pipette one third the size 

 used for milk, made to deliver about six grains of cream, was first 

 tried, but so much care was necessary to obtain good results the 

 idea was abandoned. In reading a long column of fat. unless one 

 is expert and very careful to keep the temperature constant, he is 

 liable to make an error of 0.2 or 0.3 per cent., and when only 

 six grams of cream are used the reading must be multiplied by 

 three, which increases the error three-fold. Dr. Babcock recom- 

 mends using three bottles, dividing the pipette full into three 

 parts, putting one third into each bottle and adding the three read- 

 ings. This method involves so much work that it would not be 

 pratical in testing a large number of samples. To overcome these 

 difficulties the writer used a bottle like No. 2. graduated to read 

 from 0.2 to 25 per cent, of fat, the bulb on the neck holding 10 per 

 cent. TTiih this bottle cream containing not more than 25 per 

 cent, of fat can be as easily and accurately tested as milk, and 

 as cream raised in deep cans submerged in cold water very seldom 

 contains much over 20 per cent, these bottles are sufficiently large 



