of the Fishery Board for Scotland. 



349 



characters which would enable them to locate readily any new specimen 

 which came to hand, a classification based on one character was for the 

 most part sufficient and satisfactory. The aim of the classifier was then 

 to find out that external character which most clearly marked off 

 one group consisting of many species and genera from another such 

 group, and it has already been mentioned that no wider-reaching or 

 more general external character could have been found for the separa- 

 tion of the Ileterosomata into groups than that chosen by GUnther, 

 namely, the size and structure of the mouth. Even this, however, 

 shows the weakness of a one-character classification, because the Soles 

 had to be left to one side as an ill-defined heterogeneous group con- 

 taining both small- and large-mouthed species. Other similar classifica- 

 tions suffer from similar defects. The best classification of the 

 Heterosomata hitherto — that, namely, of Jordan and his collaborators — 

 chooses the symmetrical or asymmetrical position of the ventral fins 

 with respect to the median ventral line of the body as the chief 

 character differentiating the Turbot- from the Halibut- and Plaice- 

 groups. But it has already been shown in this paper, and it will be 

 ref erred to again later when discussing the various groups, how this 

 external difference — viz., symmetry or asymmetry — depends upon 

 certain differences in the arrangement of the underlying structures, 

 and how there are various intermediate stages from the one arrange- 

 ment to the other ; so that a form, e.g. Citharus* may have the ventral 

 fins slightly asymmetrical and yet be more nearly related to the 

 symmetrical group of flat-fishes. Hence, even with this character, we 

 soon stumble upon difficulties. The Plaice- and Halibut-groups are cer- 

 tainly well marked off from .the Turbot- and all three from the Sole- 

 group by its means, but a difficulty arises when we wish to find the limit 

 between the Halibut- and Turbot-groups. All intermediate stages 

 from the symmetrical condition of the Halibut- to the asymmetrical of 

 the Turbot-group are present, and we should be quite at a loss in fixing 

 the limit between them, even though the representative species in the 

 intermediate stages are few in number, if we could not call in the aid 

 of another character. This further character is the form and structure 

 of the olfactory epithelium. By means of this character the Halibut- 

 group is even more sharply divided from the Turbot-group than by the 

 former, and it is found that the intermediate stages between these two 

 groups, though more nearly related to the Halibut than to the Turbot 

 in regard to the structure and position of the ventral fins, have the 

 same type of olfactory organ as the Turbot. The genera possessing the 

 combination of these two characters are consequently separated from 

 both the Halibut- and Turbot-groups and placed in an intermediate 

 sub-family. 



It might be thought that the structure of the olfactory organ was a 

 character of greater importance than the position of the ventral fins 

 for the purposes of classification. Such is not the case, however. There 

 are in reality only two types of olfactory organ — the one with the 

 sensory epithelium arranged linearly in longitudinal lamella?, the other 

 with this epithelium in the form of a rosette. The different forms 

 which are found] in the Soleidce are but varieties of the latter type. 

 Hence, if we choose this as the chief character, we shall he obliged to 

 classify the Heterosomata into two sharply-divided groups, each of which 

 contains a number of genera at some distance from their natural allies. 

 This will be displayed in more detail when the characters of the various 

 groups are considered with regard to their distribution. It may be 

 mentioned, further, that the sensory epithelium of the olfactory organ 

 *Even in Paralichthys, which has a close superficial resemblance to the Halibut, the 

 ventral fins are not exactly symmetrical, thus furnishing an additional reason for 

 grouping this genus with Citharus, vide pp. 354-356. 



