28 Station Bulletin 372 



mately 48 hours. Wet grass dried more slowly. In fact, it often molded un- 

 less it was turned during the drying process. 



After each batch was dried, it was run through a hammermill and 

 bagged for feeding. A total of about one ton of material was obtained. This 

 is now being fed to cows experimentally and under supervision. 



F. S. Prince, P. T. Blood 



DAIRYING 



Relationship of the Composition of Milk to 

 Methods Used for Determining Adulteration 



A study was made of the relationship between milk composition and the 

 results of tests used to detect added water. One hundred-and-thirty individual 

 milk samples were analyzed for fat (both Mojonnier and Babcock methods), 

 total solids, and solids-not-fat. The added water tests which were carried 

 out on the same samples included the determination of the freezing point de- 

 pression, refractive index of the acetic and copper sera, and ash of the acetic 

 serum. Some of the samples were chosen because of known low composi- 

 tion. 



The following table summarizes, in the form of group averages, the re- 

 sults obtained: 



Refractometer Reading 

 Number of Per cent fat Per cent Per cent Acetic Copper Per cent 



samples Babcock Mojonnier S. N. F. T. S. serum serum Ash serum 



All samples showed a normal freezing point depression, indicating that 

 none of the samples were adulterated with water. Ten of the samples had 

 refractometer readings indicating added water. Acetic acid reading below 

 .39 and copper serum reading below 36 indicated added water, according to 

 Standard Methods. These 10 samples averaged 38.9 for the acetic serum 

 reading and 35.7 for the copper serum reading. Acetic serum ash averaged 

 0.793, which did not indicate added water. The average composition of 

 these ten samples was: fat (by Babcock) 2.88 per cent; fat (by Mojonnier) 

 2.90 per cent; S.N.F., 7.66 per cent; and total solids, 10.56 per cent. The 

 refractometer readings on another group of 12 samples, with a S.N.F. aver- 

 age of 7.86 per cent, did not indicate added water. 



A group of 14 samples, all having a copper serum reading between 36.0 

 and 36.9, had an average solids-not-fat content of 8.26 per cent again indi- 

 cating that low-solids milk is likely to show a low refractometer reading. On- 

 ly two samples had a serum ash content that would indicate added water; but 

 the refractometer readings on these two samples were in the range for nor- 

 mal milk. 



