174 



MILK 



per cent, of the fat fails to appear in the column for this reason. 

 Van Slyke gives the following method to separate the fat more 

 completely from the mixture: "Use 20 c.c. sulphuric acid, whirl the 

 bottles at full speed for three to five minutes longer than usually, 

 and read the fat at a temperature of 130 to 140 F. Steam tur- 

 bine testers which keep the bottles hot give the best results." 



When carried out carefully the Babcock test gives results which 

 differ but slightly from results obtained by the ether extraction 

 method. The following figures given by Leach show the compar- 



Fig. 64. Test bottles used with skimmed milk and whey. Each division 

 of the scale represents 0.05 per cent, on the neck of the bottle at the right. 

 On the two bottles at the left each division represents 0.01 per cent. fat. 

 (Circular of Information No. 27, July, 1911, Univ. of Wis. Agri. Exp. Sta.) 



ative accuracy of fat determinations by the Babcock method with 

 the results obtained by the Adams-Soxhlet method: 



FAT DETERMINATIONS COMPARING THE BABCOCK METHOD WITH THE ADAMS- 

 SOXHLET METHOD 



Kind of milk. 



Whole milk 1 



2 



Watered milk 1...'. 



2 



Skimmed milk 1 



2... 



Adams-Soxhlet method, 

 per cent. fat. 

 4.27 

 4.28 

 2.70 

 2.74 

 0.16 

 0.14 



Babcock method, 

 per cent. fat. 

 4.30 

 4.35 

 2.70 

 2.80 

 0.15 

 0.15 



