X CONTAMINATION OF MILK 179 



these investigators recommend glucose in preference to lactose 

 agar. 



The committee of the American Public Health Association 

 on standard milk methods adopted nutrient agar containing 

 1 per cent agar, and of a reaction of + l^o. They remark : 



Much work yet remains to be done on media; the above is 

 recomimended as giving the highest and most uniform counts as far 

 as our comparative work has extended, and, with but shght varia- 

 tion, is the medium in most common use. 



The proper dilution of the sample has already been con- 

 sidered, the other steps are briefly as follows : Tubes of 

 nutrient agar medium are melted, and cooled to a temperature 

 of about 42° 0. The diluted milk fractions are then added 

 by sterile pipette. The contents of the inoculated tubes are 

 thoroughly mixed, poured, and solidified in the ordinary way. 

 The plates are inverted and incubated at 37° C. The colonies 

 should be counted after 40 to 48 hours' incubation. Many 

 workers count after 24 hours' incubation, but usually the 

 •colony development is then very incomplete, and the number 

 of colonies not easily counted. The counting should be done 

 by the nuked eye. 



Media, temperature of incubation, period of incubation, 

 must all be accurately recorded, and it is absolutely im- 

 perative for comparative work to enumerate under precisely 

 ■similar conditions. 



At least three plates should be poured, and only those 

 plates enumerated which do not contain an excessive number 

 of bacteria. Convenient amounts of the milk to add to the 

 agar tubes are, for fresh byre milk O^Ol, 0'02, and 0-004 cc, 

 and for ordinary vended milk 0-002, 0-001, and 00002 c.c. 

 If, however, something is known about the milk, it will usually 

 be possible to obtain a closer and more satisfactory range of 

 ■dilutions. 



Several workers have attempted to estimate the number 

 of bacteria from an examination of the stained centrifugal 

 deposit of a fraction of milk. The estimate is a very rough 

 one, and no numerical figure, with any pretence to accuracy, 

 can be obtained, but it gives a useful idea of the number 

 present. Dealing with samples of doubtful origin, it is a 

 decided advantage to centrifugalise a small definite quantity 



