10 



thus rendering a count of no value. The count after 24 hours 

 is shoAvn. In the North Training Wall the average number 

 of organisms per mussel is estimated from five plates after 

 24 hours incubation. Aiter 48 hours the increase in the 

 number of colonies teas very considerable, but these were all 

 microscopic ; consequently, it was decided not to count them ; 

 so the estimation shows the minimum organisms per mussel. 



Average numbers of organisms of different categories 

 found per mussel : — 



Acid and gas were produced in lactose, bile salt, litmus 

 broth in ^rixi^h part of a mussel, and acid only in gj^i^oth part 

 of a mussel in the case of the Church Scar sample. 



Acid and gas were similarly produced in g^oth part of a 

 mussel, but no reaction was observed in .^^o^li part in the case 

 of the Training Wall sample. 



Conclusions. 



Church Scar. — There is evidence of recent fsecal pollution 



borne out by topographical evidence, and particularly by the 



presence of Streptococci in relative abundance. The presence 



of acid forming bacilli was detected in .-;5;^oth part of a mussel. 



North Training Wall. — The pollution here appears to be 

 of a more diffuse nature. There are more lactose fermenting 

 bacilli and fewer non-lactose fermenting organisms than in 

 the Church Scar sample. Topographical evidence would 

 suggest that sewage bacilli have to undergo a longer period of 

 isolation from their normal habitat than those reaching 

 Church Scar. This may have resulted in a natural selection 



