CHAP. iv. ORIGIN OF FAT OF MILK. 293 



There is, finally, abundant evidence for concluding that the carbon of 

 the newly-formed fat, equally in the milk of a cow as in the body of an 

 ox, may be supplied (1) from the carbohydrates of the food, or (2) from 

 the carbon surplus of the proteid food, or (3) from fats taken as food 

 which are not the natural constituents of the body-fat. 



If the butter fats of milk were derived directly from the food, and 

 transferred without change to the milk, it would be possible to control, 

 through the food, the quality and composition of the butter produced. 

 This was once regarded as possible because such plants as onions and 

 turnips impart, in a short time, characteristic odours to the milk of the 

 cows consuming them ; whilst the marked difference in colour and 

 flavour of butter made in the winter when the cows receive only dry 

 food, and in the summer when grazing, has been attributed to differences 

 in the fatty portion of the fodder. On the other hand, variations in 

 the composition of butter ascribed to breed and to individual 

 peculiarities of the cow, are so well defined and so constant, whatever 

 the character of the food, as to render a direct transfer of fats highly 

 improbable. The question of the essential oils of plants is rather 

 beside the point ; a peck of onions fed to a cow would scarcely impart 

 so much flavour to the milk as would a piece of onion directly added 

 to it. 



One way of elucidating the question is by carefully examining the 

 butter, and noting the changes in its composition which follow radical 

 changes in the food. Crude cotton-seed oil has some characteristic 

 properties by which its presence, when mixed with butter, even in small 

 quantities, may be detected. A cow whose butter had frequently been 

 tested was set apart at the New York Experiment Station, and cotton- 

 seed meal included in her food. The first cotton-seed meal was given 

 on April 18, half a pound, and the quantity gradually increased to 4 Ib. 

 on May 1. This amount of meal contained as much fat as the 

 remainder of the ration, and would therefore be likely to contribute an 

 appreciable amount of cotton-seed oil to the butter, if any direct transfer 

 does occur. Inasmuch as the soaps from crude cotton-seed oil have a 

 high viscosity, whilst those from butter fat are usually low in this 

 respect, the test of the viscometer was applied time after time, with the 

 result that, notwithstanding the change in ration, no appreciable 

 change in the viscosity occurred ; this was strong evidence that no 

 cotton-seed oil was transferred to the butter. Nor was any cotton-seed 

 oil detected in the butter when subjected to another delicate and reli- 

 able test for the presence of the former. In this case, for certain then, 

 there was no transfer of cotton-seed oil from the food to the milk. 



The same thing is shown by the constant and uniform difference 

 found in butters from different breeds of cows, and from individual cows 

 in the same herd all receiving the same food and attention. From such 

 data Dr. Babcock concludes that the composition of the butter fats is 

 practically constant for each individual cow under all circumstances of 

 normal feeding. As a matter of fact, quality of butter depends more 

 upon certain physical properties and flavours, which constitute scarcely 

 more than traces of the butter, and which as yet have not even been 



