240 TRANSACTIONS LIVERPOOL BIOLOGICAL SOCIETY. 



a single day, and they do not by any means give an 

 adequate idea of the quantitative variation among 

 individual catches. Thus on September 10th surface nets 

 I and II contained 250 and 550 respectively, while two 

 days later the corresponding numbers were 13,495,500 and 

 16,300,500; on April 8th two hauls of the Nansen net 

 gave respectively 198,000 and 3,739,000, and many other 

 such cases could be quoted. 



Such differences as above cited are due in the main 

 to the great abundance of some single organism, generally 

 Rhizosolenia semispina, Chaetoceros contortum, C. clebile 

 or Thalassiosira nordenskioldii. Thus of the two enormous 

 surface nettings of September 12th given above, 

 Rhizosolenia semispina accounts for thirteen millions and 

 sixteen millions of the organisms respectively ; again the 

 high average (477,664) of 32 hauls made on September 

 20th is traceable largely to the influence of four of the 

 catches amounting together to 13,230,150, of which 

 13,085,000 were Rhizosolenia semispina. 



Besides, however, these great fluctuating changes it 

 will be seen that there is a more regular seasonal change. 

 This is brought out more clearly by the diagram (p. 269). 

 Owing to the frequency of the spring and autumn hauls 

 it is possible to take 3-daily averages from March 27th to 

 April 26th, and again from August 20th to September 

 30th, but it should be noted that while the spring catches 

 and those of the middle (August 23rd to September 19th) 

 of the autumn period were made with several kinds of nets 

 outside the Bay, together with surface nets w T ithin the 

 Bay, those on other occasions were made only by the 

 surface nets within the Bay ; at other times than the 

 above two periods the nettings numbered from three to 

 four per month. 



The curve shows two humps, a well-marked one in 



