ILLINOIS STATE DAIRYMEN'S ASSOCIATIOiN 153 



test bottles should be given an occasional bath in some alkali 

 or grease solvent solution. 



Persons who desire to make accurate tests wil find it 

 profitable to provide themselves with a small copper tank, 

 which can be filled with some weak alkaline solution in which 

 the test bottles can be put to soak after they have been cleaned 

 with hot water, leaving them completely covered with this 

 liquid until they are to be used another time. 



This liquid should be warmed, and if ,the tank is pro- 

 vided with a small faucet at the bottom the liquid can be 

 drawn off without the trouble of fishing for them in the tank. 

 The writer has been able to clean test bottles in a very satis- 

 factory manner by adding about a tablespoonful of ''savogran'^ 

 to about two gallons of water and then soaking the test bottles 

 in this hot solution. Sal soda, gold dust or Lewis lye are 

 about as efficient for this purpose as "savogran," but the 

 cleansing properties of any of these substances are increased 

 by warming the liquid. It is probably unnecessary to state 

 that the test bottles should be rinsed with hot water after 

 they are taken from this bath and before they are used for 

 testing milk. 



The black stains that sometimes stick to the inside of 

 test bottles that have been used for some time can often be 

 removed with a little muriatic acid. 



In our daily tests of separators the student takes three 

 samples, one of whole milk, cream and skim milk. One 

 pipette is generally used for measuring each of these samples 

 into the test bottles. It sometimes happens that a very rich 

 skim milk test is reported because the whole milk or cream 

 is measured into the test bottles before the skim milk is 

 measured out. A part of the fat from the cream sample 

 adheres to the pipette and is naturally found in the skim milk 

 test. This high test of the skim/milk is not the fault of the 

 separator, but of the operator of the test. 



The skim milk sample should be measured into its test 

 bottle before the others are taken, in order to give the sep- 

 arator full credit for the work it has done. 



The proper strength of the acid for testing milk can gen- 

 erally be determined by the color of the fat which separates 



