MILK TESTING 



19 



two large and three small divisions, b = 2.6. 7.6 

 2.6 = 5.0, the per cent of fat in the milk. 



Do not read from other points than a and 6, other- 

 wise the results will be too low. The bottles are 

 made to read in 

 just this way. 

 The reading 

 should be done 

 before the fat 

 cool sand streaks 

 down the sides. 

 In case a num- 

 ber of readings 

 are to be made, 

 the test bottles 

 should be set in 

 a pan of hot 

 water, or hot 

 water can be 

 run into the pan 

 of the machine, 

 to prevent the 

 bottles from 

 cooling. Dividers can be used as indicated in the 

 figure, but should be spread from to 5. 



19. Calibration of Test Bottles. All test bottles 

 should be rejected when inaccuracy of the divisions 

 can be detected with the eye. In the most careful 

 work they should be calibrated with mercury : 27.18 

 gm. of clean metallic mercury should just fill the 



FIG. 7. Reading the test. 



