1902-3.] Bacterial Contamination of Milk. 487 



of this method of cleaning milk is similar to that of the gravel filters, and 

 according to Backhaus,^^ 95 per cent, of the mechanical impurities (hairs, 

 manure particles, etc.), are eliminated. The separator divides the milk 

 into three parts, the slime which adheres to the bowl of the machine, the 

 skim-milk and the cream. Several investigators have given us data of 

 the number of bacteria which are found in these three products. Thus 

 Popp and Becker^'' found the germ content per c.c. of the whole milk, to 

 be 72,954; of the cream of this milk, 58,275 ; the separator skim-milk, 

 21,735 ; and the separator slime, 43,891. 



Scheurlen^^ found in one litre of milk 2,050,000,000 of bacteria, and 

 after separation 1,700 in the 200 c.c. of cream, 560,000,000 in the 800 

 c.c. skim-milk, and 18,000,000 in the 6 c.c. of slime. 



Other investigators have also shown that centrifugation does not 

 decrease the number of bacteria in milk. Thus, Fjord and Fleisch- 

 mann^^ claim that centrifugal separation has little value as a means of 

 purification, and Conn^^ states that " milk after passing through a centri- 

 fuge, although it contains less gross impurities, shows more bacteria than 

 before. This is explained by the fact that masses are broken up, and 

 large numbers of bacteria liberated, " and again," the same writer says, 

 "centrifugal purification does not materially affect the bacteria, for there 

 seem to be about as many after treatment as before." 



Niederstadt^* obtained similar results, for he found that by the 

 centrifugal treatment of 300 litres of milk, about 130 grams of sediment 

 were obtained. The cream was richer in bacteria than the sediment. 

 The separator effected no purification of milk from bacteria, and 75 per 

 cent, of the bacteria went into the cream. 



Dunbar and Kister,^^ in an exhaustive series of experiments, found 

 in four instances fewer bacteria after separation, the average of these 

 four instances being as follows : — 



Raw Milk. Centrifuged Milk. 



446,000 per c.c. 146,000 per c.c. 



But in the remainder of the experiments, twenty-four in number, 

 more bacteria were found in the separated milk, the averages in this case 

 being : 



Raw Milk. Centrifuged Milk. 



1,400,000 per c.c. 2,200,000 per c.c. 



It would seem from these figures that the smaller the number of 

 bacteria present in the whole milk, the more efficient was the separator 



* 



m reducing their numbers. 



