Examination of shellfish 



261 



per c.c. of shell water. Partly because it does not do this 

 exactly, but also for simplicity and the avoidance of fractions, 

 the method of stating results as an arbitrary rating is 

 preferred by the committee. Practical experience with the 

 method also has appeared to justify this preference. 



Illustrations of the Application of the Method of Rating 

 Oysters for B. Coli. Sometimes results similar to the 

 following are obtained, that is, one or more oysters may 

 show positive results in small quantities of shell water, 

 while an equal number may show negative results in larger 

 quantities. In this case the next lower numerical value 

 shall be given to the positive results in the high dilutions, 

 and such positive results shall be considered as being trans- 

 ferred to a lower dilution gi\'ing negative results in another 

 oyster. This is done on the theory that inconsistent results, 

 mathematically considered, may follow natiu^ally from an 

 unequal distribution of the bacteria in the shell water. 

 This recession of the assigned values, however, shall not 

 be carried beyond the point where the nimaber of such 

 recessions is greater than the number of instances where 

 other oysters in the series failed to give positive B. coli 

 results. 



As examples of the method of obtaining the rating for 

 B. coli, the following illustrations are given. They repre- 

 sent results that may be met with in practice : 



CASE A. 



-RESULTS OF B. COLI TESTS IN DILUTIONS 

 INDICx\TED 



