lOB Rmldl Sfiidlcs ill Fishes 



In trying to characterise these populations, various quantitative 

 features are employed, both such as are determinable by counting (e.g. 

 the number of vertebrae or of fin rays), and such as involve measure- 

 ment (as for instance the size and shape of the head). We have thus 

 in the foruier case to deal with integrated, in the latter with graduated 

 variates. 



Where two populations have been found, by statistical examination 

 of a great number of specimens, to differ in the mean of one or more 

 characters, it is customary, in fishery biology, to say that each belongs 

 to its own race of the species in question. The word " race " is, how- 

 ever, not employed by all writers; some prefer the term "local forms" 

 or " families," while others again use all three indiscriminately. 



This uncertainty in the terminology itself serves to indicate the in- 

 completeness of our knowledge as to the true nature of races in fishes. 

 We do not even know whether they are genotypically determined or 

 merely phenotypieal phenomena. And a genetic analy.sis is still wanting 

 in this sphere. The reason for this lack of knowledge must be sought, 

 partly in the great experimental difficulties attending investigation of 

 our mai'ine food fishes, partly also in the fact that fishery biologists most 

 frequently regard their problems as solved when once they have shown 

 that two or more populations actually are distinct and biologically inde- 

 pendent. 



The question then is still before us : What is the cause of the racial 

 differences found >. Why, for instance, should the race nf herrings living 

 on the east coast of Scotland have a higher avei-age number of vertebrae 

 than that — or those — of the Baltic, ami why have the Baltic herrings 

 again a higher figure for this character than the herrings of the White 

 Sea i. Or how is it that the plaice {Plearonectes platessa) in the North 

 Sea manage to develop a gi-eater number of raj's in the anal fin than the 

 " plaice of the Baltic waters 1 



The generally accepted view among fishery biologists is that racial 

 differences arise from the influence of differing external conditions under 

 which the races in question live, such as for instance salinity of the 

 water, its temperature, etc. It is thought that the average values for 

 the differing qualities, i.e. the racial characters, are actually dependent 

 upon this or that temperature, salinity, or the like, which happens to be 

 peculiar to the water in question. Certain writers, such as Heineke, 

 imagine these external conditions as exerting a directly determinative 

 effect ; others are more inclined to regard them as acting indirectly, 

 through a process of selection. In this latter case, the variates not 



