of the Fishery Board for Scotland. 



351 



small group, such as the Siluridce and their nearest allies, only those 

 characters which were out of the common, e.g. the Weberian ossicles, 

 or those which were little employed and were thus little acted upon by the 

 strong hand of the environment, such as the bones of the pectoral arch, 

 could be relied upon in the least to show genetic affinities. Neverthe- 

 less, there is good reason for believing that the flat-fishes form a favoured 

 group. As a whole they are well marked off from all other groups 

 of the Teleostei, even from the Gadidse, to which they are supposed to 

 be most nearly related, and this sharp demarcation displays a fact 

 which is very evident throughout — viz., that the group is specialised 

 from the very beginning. The habits and habitat of the species 

 included under it are also peculiar and in certain ways quite uniform 

 throughout the group. We should expect, therefore, that this uniformity 

 will display itself in a certain amount of constancy in the characters, 

 and that if these latter change they will do so in accordance with 

 certain well-marked differences in the surrounding physical conditions, 

 or to greater specialisation of habits and habitat. The former element is 

 shown in the geographical distribution of the characters of the various 

 sub-groups or families ; the latter, in these characters also to a certain 

 extent, but more especially in those of the species within the separate 

 sub-groups or families. It follows from this that a classification of 

 the flat-fishes, if it is intended to represent the genetic affinities of 

 the various divisions, should display, along with the characters, their 

 distribution generally and their varying degrees of specialisation. This, 

 in truth, seems possible for the Heterosomata, and in the following 

 pages the endeavour is made to show in general outlines how such a 

 classification may be realised. The influence of convergency can also 

 be seen in the various groups, but only the more general and out- 

 standing examples need be mentioned. 



Classification of the Heterosomata. 



Families. 



a. Preopercular margin more or less distinct, not hidden by the skin 

 and scales of the head. 



Pleuronectidce. 



a. Hippoglossince (Halibut-tribe). 

 p. Pleuronectinai (Flounder-tribe). 

 y. Psettince ( = Bothinai) (Turbot-tribe). 

 aa. Preopercular margin aduate, hidden by the skin and scales of the 

 head. 



Soleidon. 



a. Achirince (American Soles). 

 fB. Soleince (European Soles). 

 y. Cynoglossino3. 



The above classification, with slight modifications, is taken from the 

 recent work of Jordan and Evermann (33). It is the most convenient 

 at the present time for grouping together the flat-fishes as a whole, but 

 one feels already that it is too definite and cramping to express well the 

 relationships of the various groups to one another. It is premature, 

 for much yet requires to be known, to endeavour to substitute this 

 scheme by another, but it is thought advisable to present the various 

 groups in the following loose arrangement, which has the double 

 advantage of displaying affinities, and at the same time of showing the 

 gaps in our knowledge. 



