19011.] PUBLIC DOCUMENT — No. 31. 89 



The acid nniuber, obtained by heatino- a definite quantity 

 of the fat in alcohol, allowino- it to cool and titrating the solu- 

 tion with Icnlh normal alkali, represents the number of milli- 

 grams of potassium hydrate necessary to neutralize the free 

 acid in one gram of fat. The column headed free fatty acids 

 was calculated by the use of a mean molecular weight of 252. 



Inasmuch as the butter fat was prepared directly after the 

 butter was made, the small amount of acidity could not be due 

 to the action of bacteria, since butter fat offers no suitable 

 medium for bacterial development. It is not believed that the 

 acidity was due to a decomposition of the glycerides, as the 

 low temperature and exclusion of air would not be favorable to 

 such a transformation. Brown ^ has showai that fresh butter 

 fat has an acid nundier of .50; Lewkowitsch" likewise calls at- 

 tention to a similar condition. 



In the above table it will be seen that in the first two periods 

 both herds showed quite similar acid numbers, varying from 

 .32 to .38, equivalent to from .14 to .17 per cent, of acidity. In 

 the third period, in case of Herd I. the acid number increased 

 slightly to .48, and noticeably in case of Herd II. from an 

 average of .32 in the second period to 1.39. This change could 

 not have been due to the effect of long standing, otherwise the 

 butter fat from Herd I. would have shown a similar condition. 

 It may have been due partly to the direct entrance of the soy 

 bean oil ^ into the butter fat and partly to the disturbance as the 

 result of feeding the oil. 



The ether number, or difference between the saponification 

 and the acid numbers, represents the milligrams of potassium 

 hydrate required to neutralize the acids of the neutral fat. 

 Xaturally it varied with the saponification and acid numbers, 

 being lowest in the butter fat produced by Herd II. in the third 

 period. 



The percentage of glycerol in fat may be determined directly 

 according to the method w^orked out by Benedict and Zsig- 

 mondy^ and modified by Allen, ^ or in case of triglycerides it 



» Pennsylvania State College, report loco citato. 

 2 Tjewkowitsch, loco citato. 

 * The 8oy bean oil had an acid number of 1.27. 



< .Tournal Society Chemical Industry, 1885, p. 610; abs. Lewkowltsch, third edition. Vol. 

 I., p. -iK?. 

 6 Commercial Organic Analyses, second edition, IF., p. 290. 



