144 TRANSACTIONS LIVERPOOL BTOLOGICAL SOCIETY. 



therefore, chlorine in sea-water to the extent of 5 parts per 

 million does not interfere with the functioning of mussels — ■ 

 I did not expect that it would, since the chlorine throat washes 

 used are really stronger than this. It was, however, desirable 

 actually to make the trial. 



A tank with chlorinated sea- water, of concentration, 

 5 per million, was now prepared, and mussels were placed in 

 it. These mussels were part of the same sample used in making 

 the experiment of cleansing by a water circulation — they 

 contained on the average 7,400 sewage bacteria each. The 

 tank was allowed to stand in full daylight, though not in direct 

 full sunlight, and samples of the mussels were taken after one 

 and two days. 



First Sampling. Mussels in 5 per million chlorine sea-water 

 for one day. 



-. /mil (Plate 1, 17 red colonies. 

 l/50th 2 28 



mussel 1 „ 3,14 



Mean number of sewage bacteria per mussel = 985. 



Second Sampling. After two days. 



-. /™,i (Plate 1, 11 red colonies. 

 l/50th 2 u 



Mean number of sewage bacteria per mussel = 600. 



The mussels were not removed from the tank, but the 

 water was now run off and replaced by sea- water freshly dosed 

 with chlorine so as to make a strength of 5 per million. 

 This water was allowed to stand another day, when the mussels 

 were again examined. 

 Third Sampling. After 1 day, the water being changed. 



mussel ** plates, all of which were sterile. 



The mussels were therefore freed from sewage contamina- 

 tion, 



The experiment was repeated at Liverpool in July. Short 



