Artificial Mothers'* Milk. 103 



for instance, four glass tubefuls have been drawn 

 from a vessel containing forty quarts, then from a 

 vessel containing but thirty quarts only three tube- 

 fuls should be drawn, or two from another vessel 

 containing only twenty quarts, etc. The test samples 

 are all collected in the same dish and the testing at 

 once performed. The temperature of milk for testing 

 should be 02 F., more or less. Two colateral tests 

 should be made of every sample to avoid errors. 

 Quite a number of methods and apparatus for testing 

 have been invented, the most accurate being probably 

 the Soxhlet ; for use in dairies, however, this method 

 is too complicated, and the best known tester in this 

 country is the Babcock. 



According to the instructions kindly furnished me 

 by Prof. S. M. Babcock, of the Wisconsin Agricultural 

 Experiment Station, the method of operating the test 

 is as follows : 



THE BABCOCK TEST. 



The estimation of fat in milk by this test is ac- 

 complished by adding to a definite quantity of milk, 

 in a graduated test bottle, an equal volume of com- 

 mercial sulphuric acid of a spgr. of 1.821.83. This 

 acid dissolves the casein, setting free the fat, which is 

 then completely separated from the liquid in the 

 bottle by whirling in a centrifugal machine. Hot 

 water is afterwards filled into the bottles to bring the 

 separated fat into the graduated neck, where the per 

 cent, is read directly from the scale. 



