SEA-FISHERIES LABORATORY. 313 



inclined to regard these percentages as the most valuable 

 indications of the relative occurrences. For example, in 

 July, Oceanic species occur on the average six days and 

 Neritic a little over four days ; but the six represents 

 nearly 86 per cent, of the possible occurrences in that 

 month, and the four is 58 per cent. It must not be 

 supposed that June, with 16 species showing 115 occur- 

 rences, is necessarily more constant than July with 16 

 species making 96 occurrences, since samples were taken 

 on nine occasions in June but only on seven in July. The 

 numbers for mean occurrence, seven and six, are there- 

 fore in this case not instructive, and the percentages 79 

 and 85 are, we consider, more correct. But as we give in 

 the tables all the figures bearing on the problem, any 

 possible false impression can be easily rectified. 



If we enquire further into the apparently marked 

 Oceanic character of December, with 56 per cent, of 

 Oceanic species in the table, we find that Metridia lucens 

 is present (only in December, November and January), 

 while the common Neritic forms Temora, Eucampia, 

 Asterionella, Chcetoceros sociale, Rhizosolenia, Noctiluca, 

 Pleurobrachia and Autolytus are all absent, and also the 

 less prevalent Podon and Evadne. The same explanation 

 can be given in the case of January, the month with the 

 next highest percentage of Oceanic species, and this 

 supports Gough's remark that Oceanic character is 

 largely due to absence of Neritic forms, not to any 

 unusual abundance of Oceanic species. One of the most 

 notable points brought out by this enquiry is the fairly 

 even numbers of Oceanic and Neritic species occurring 

 in each month, since the percentages in Table I range 

 only from 40 to about 56. 



Treating the records for '08 in the same way, we 

 find that the totals of Table I (the details of which need 

 not be printed in full) are, as shown on p. 314: — 



