94 FOODS AND FOOD ADULTERANTS. 



by the milk standing ou ice ; Soxblet thinks it is the result of deficiency 

 of fat; and others attribute it to differences in age and breed of the 

 cows. The resume which precedes shows that not only the actual vol- 

 ume of the ethereal solution, but also the time of the separation required, 

 has a serious disturbing influence on the specific gravity of the ether-fat 

 solution. 



Therefore, the method, in order to be of general application, must be 

 subjected to some radical modification. 



In this direction were the attempts to secure a more prompt separa- 

 tion by varying the amounts of caustic-potash solution employed. These 

 attempts, as the record has shown, were entirely unsuccessful. Even 

 if the different kinds of milk would permit a prompt separation by vary- 

 ing the quantities of alkali employed, the amount for each sample could 

 only be determined by numerous and tedious experiments. 



I, therefore, turned my attention in another direction. It seemed to 

 me that a centrifugal machine might be used to secure this separation, 

 and accordingly I had a castaway drug-mill, formerly used in the labo- 

 ratory, modified so as to serve for this purpose. The machine was so 

 arranged as to hold four separatory flasks and impart to them a high 

 speed of rotation. The form of the machine, with modifications made, 

 is shown in figure 2. 



At this point of my investigations this apparatus was finished and I 

 immediately subjected it to a trial. 1 



Four samples which had not separated at all at the end of three hours 

 were placed in the apparatus and whirled for ten minutes. At the end 

 of this time three of them had completely separated, and the fourth 

 nearly so. The apparatus was set in motion again for five minutes, at 

 the end of which time the separation of the fourth sample was accom- 

 plished. 



The number of revolutions per minute of the machine was about 350. 



It will be seen from the above that the very first trial of the machine 

 was completely successful, securing a perfect separation of the ether-fat 

 solution in a few moments in samples which previous trial, by the usual 

 method, had failed to separate in several hours. 



The next determinations were made on a sample of milk purchased 

 at the Department restaurant. 



Duplicate flasks were treated in the usual way to secure the separa- 

 tion, and only at the end of two and a half hours was enough clear solu- 

 tion obtained to get a reading: No. 1 gave 2.40 per cent, fat; No. 2, 

 2.30 per cent. fat. 



The first set of samples of the same milk separated by the centrifugal 

 gave the percentages following : No. 1, 5.52 per cent, fat; No. 2, 2.32 per 

 cent. fat. 



1 This apparatus was first described before the Chemical Society of Washington, 

 May, 1886, and next at the Buffalo meeting of the A. A. A. S., August, 1886. 



