26 EXPERIMENTAL DAIRY BACTERIOLOGY 



substance, fictitious accuracy should be avoided in the expres- 

 sion of the results. In reporting the results of the examina- 

 tion of ordinary market milks, the Committee on Standard 

 Methods of Milk Analysis of the American Public Health 

 Association has made the following recommendation : 



From 50,000 to 500,000 bacteria per cubic centimeter 

 report results to nearest 50,000. 



From 500,000 to 1,000,000, to nearest 100,000. 



From 1,000,000 to 2,000,000, to nearest 200,000. 



From 2,000,000 to 5,000,000, to nearest 500,000. 



Above 5,000,000, to nearest 1,000,000. 



The error with present methods, even with very careful 

 work, amounts to 10-15 per cent, so it is evident that the 

 above scheme is sufficiently accurate. 



In reporting the results of very clean milk (certified and 

 sanitary milk) the exact figures should be given, neglecting 

 the first significant figure, as 510 bacteria per cubic centi- 

 meter instead of 514. 



Recapitulation of preparation of plate cultures. The pro- 

 cedure for the examination of a sample of milk from which 

 plates are to be made containing 1/100, 1/1000, and 1/10,000 

 cc. of milk is as follows. Shake sample twenty-five times. 



Mark two 99-cc. water blanks 1 and 3 respectively, and 

 mark one 9-cc. water blank 2. 



With a sterile pipette transfer to blank No. 1, 1 cc. of milk, 

 and return pipette to milk. Replace the cotton plug in the 

 water flask and shake well. No harm will be done if the 

 cotton becomes wet. 



With another sterile pipette, transfer 1 cc. from flask No. 1 

 to No. 2, and also 1 cc. from No. 1 to No. 3. Return pipette 

 to No. 1. The cotton plug need not be replaced. Mix Nos. 2 

 and 3, and in each place a sterile pipette. 



