130 MANUAL OF MILK PRODUCTS 



is necessary because the curd, which contains the bulk of the 

 fat, separates out very quickly. 



The test is completed in the same manner as that of milk, 

 with the following modifications : 



Special attention should be given when the bottles are placed 

 into the tester. Test bottles in which the lower end of the 

 funnel-neck extends perpendicularly along the side of the bulb 

 to the bottom of the bottle, should be so placed that the funnel- 

 neck faces the center of the tester, otherwise the fat rises into 

 the funnel-neck. Test bottles in which the lower end of the 

 funnel-neck extends diagonally to the bottom of the bottle 

 should be so placed that the graduated neck faces the center of 

 the tester. This will prevent excessive breakage of this type 

 of bottles. The tester should run perfectly smooth in order to 

 prevent excessive breakage, as these bottles are of very delicate 

 construction. 



Use about 20 c.c. of acid instead of 17.5 c.c. and whirl the 

 bottles about twice as long as in the case of milk. 



The amount of fat in skim-milk and in buttermilk is, or should 

 be, so minute, the fat globules are so small, and the construc- 

 tion of the bottle is so crude, that it is difficult to secure very 

 accurate tests of skim-milk and buttermilk. 



The results of testing skim-milk and buttermilk should not 

 be relied on absolutely for accuracy, but they may serve as a 

 convenient guide, showing the operator whether these by- 

 products contain comparatively little or much fat. The above 

 suggestion is offered on the basis of the results of a vast num- 

 ber of skim-milk and buttermilk tests conducted by this de- 

 partment. In these tests the results were compared with the 

 chemical fat estimations. 



TESTING FROZEN MILK (Ross) 



Partly frozen milk should never be sampled for testing, 

 since a sample of such milk will not be representative. Such 



