488 Transactions of the Canadian Institute. [Vol. VIL 



Eckles and Barnes^* have also investigated the purification of milk 

 by the centrifugal separator. They found a large proportion of the 

 bacteria removed by centrifuging, but no enhancement in keeping 

 quality. 



Russell in a private communication to the writer expresses his 

 opinion thus : — " I do not think clarification is worth the trouble, unless 

 the milk is exceptionally dirty.' 



At the suggestion of the Ontario Department of Agriculture, we 

 (my assistant, Dr. Streit and myself) have reinvestigated this subject. 



A power belt separator was used, run at the speed indicated by the 

 manufacturers. The milk came from farms in the vicinity, and was of 

 average quality, similar to the ordinary factory supply. About 150 

 pounds of this milk were thoroughly mixed in a sterilized can with a 

 sterilized stirrer. A half-pint sample of the milk was taken in a steri- 

 lized jar, the rest of the milk being put through a separator. The cream 

 and skim-milk were caught together in a sterilized can, and were again 

 thoroughly mixed with a sterilized stirrer, and another half-pint sample 

 of the clarified sample was taken. Both samples were immediately 

 carried to the laboratory, where suitable dilutions were made and plates 

 poured. 



The culture medium used was whey gelatine, with one per cent, of 

 peptone. The plates were kept at 20° C. and counted at the end of 

 forty-eight or seventy-two hours, depending on the size of the colonies. 

 In most cases the plates were counted by each of us independently, so 

 as to reduce the personal equation. 



Each result given in the table is the average of four plates, and thus 

 the gross average represents the numerical results obtained from 240 

 plates or analyses. 



