102 



Part III. — Twenty-sixth Annual Report 



Measurements Represented as Percentages of the Length of the Fish. 

 Range of Variation — continued. 



The data set out in these tables are not of direct value from the point 

 of view of specific discrimination. But a detailed examination of them will 

 reveal wherein lie diagnostic characters. 



One species may be readily separated on the first examination of the 

 fish by some prominent distinguishing mark, as, for example, the black 

 area on the side of seglefinus. But in another the formulation of a 

 specific description is difficult, not always because the fish resembles its 

 neighbours closely, but owing to the difficulty of expressing the difference. 

 And that obtains, even although the two species may be quite easily 

 separated, when compared side by si^e. An accurate and detailed 

 description of a single fish will not serve for a specific description in every 

 case. In some instances it might do so, but not in the genus Gadus. 



A perfect specific description would be sufficient to enable one to 

 diagnose a fish by itself, without having recourse to direct comparison 

 with another fish. But that is not altogether necessary, since it is usually 

 possible to make use of pictures of some or all of the species. It must, 

 however, be comprehensive enough to admit of the diagnosis of a damaged 

 fish. This can only be assured when the scheme of classification is an 

 extended one, working along various lines, by each of which the species 

 may be reached, or at least found in a reduced group. A fish may be 

 quite normal although deprived of some of its so-called specific characters. 



