96 



FOODS AND FOOD ADULTERANTS. 



The hoininy chops used showed, on analysis, the following compo 

 sition : 



Two of the other samples were received April 27 and 30 from a thor- 

 oughbred Jersey, four years old, weight about 600 pounds, in milk since 

 July 1, 1885, giving at the time about 5 quarts daily. On the 29th of 

 April samples of milk were also treated from the same cow, but after 

 dilution the centrifugal separation, although more than usually diffi- 

 cult, did not require so long a time as on the occasion mentioned. 



There is nothing shown by the analysis, by the breed of cow, nor by 

 the food which gives any definite idea of the cause of the peculiarity in 

 these milks which does not permit a speedy separation. It certainly 

 is not the quantity of fat present, for other milks having the same, 

 more or less, amounts of fat separated without difficulty. In the ab- 

 sence of any further evidence on this point we can only attribute the 

 phenomenon to bovine idiosyncrasy. 



In all 90 samples were compared by the usual method of separation 

 and by the centrifugal. By the former method the mean percentage of 

 fat obtained was 4.01 and by the latter 3.88. It thus appears that the 

 numbers obtained by the centrifugal method must be increased by .13 

 in order to correspond to those of the old method. This discrepancy is 

 readily explained when it is remembered that by the centrifugal motion 

 the percentage of ether left in emulsion would naturally be less than 

 with the former process of separation. The ether-fat solution thus be- 

 comes more dilute and consequently has a lower specific gravity. When, 

 therefore, the percentage of fat in a milk determined areometrically, is 

 calculated by the tables given for the old method of separation, it should 

 be increased by .13 in order to represent the actual quantity present. 



I think it safe to conclude from the data which have been obtained : 



First, that the method of Soxhlet cannot be applied to the determi- 

 nation of fat in American milks, especially if they be from individual 

 animals. It works somewhat better on mixed milks from a large dairy, 

 but even in this case it is a rare thing to secure a p rompt separation 

 and in most cases the method would be very difficult of application. 



Second, that by the use of the centrifugal machine described a prompt 

 separation of the ether- fat solution can be obtained in all cases, even in 

 those in which after forty-eight hours no separation whatever takes 

 place by the usual method. 



Third, that the estimation of the fat in milk by Soxhlet's areometer 

 can only be accurately secured when standard volumes of aqueous ether 



