116 THE ENUMERATION OF BACTERIA IN MILK 



Results reported by the Chicago Department of Health 2 

 on the relative toxicity of raw and pasteurised milk also confirm 

 this hypothesis. 



After this consideration of the " raison d'etre " of the bac- 

 terial enumeration, the methods by which this is accomplished 

 will now be treated in detail. These may be divided into two 

 groups: (a) plating methods and (6) direct microscopical 

 methods. The former are based upon the ability of the indi- 

 vidual organisms to reproduce at such a rate upon the medium 

 employed as to produce a visible colony within the period of 

 incubation, and the latter upon suitable preparation for direct 

 enumeration under high magnification. 



Until within the last few years the former method was the one 

 usually employed, and as it is still in universal use, it will be 

 convenient to discuss it first. 



Plain nutrient gelatine prepared with fresh beef infusion 

 was first used with the plate method for the enumeration of 

 bacteria in milk and still enjoys considerable repute with many 

 workers for this purpose, the colonies being usually counted 

 after four to five days incubation at 20 to 22 C. In late 

 years, however, and especially in America, this method has 

 largely been supplanted by the substitution of agar for gelatine 

 and the incubation period reduced to forty-eight hours at blood 

 heat. Although the agar medium does not produce as many 

 visible colonies within the incubation period as the gelatine one, 

 it possesses certain advantages which more than offset this 

 drawback. In routine work it is very desirable that results 

 should be obtained in the shortest possible time, and in this 

 respect the agar medium is decidedly preferable as it reduces 

 the time required by 60 per cent. If necessary the colonies 

 may be counted after twenty-four hours incubation, but the 

 results so obtained do not exhibit the sharp contrasts given by 

 the longer period. Some of the author's results are given in 

 Table XLVIIL 3 



The average of the ratio of the forty-eight hour count to the 

 twenty-four hour C9\int is 3.4, but if the abnormal value of 



