336 



Part III. — Eighteenth Annual Report 



the separation of species. Of somewhat greater value are the relative 

 dimensions which are usually given of different parts of the body. 

 These are also subject to great variation, and are in many cases useless 

 and misleading. Yet if those dimensions are chosen which express the 

 " form " of the species, and sufficient examples are measured, one can 

 readily see that they should be of great value. For example, the 

 " form " of the Plaice, Flounder, and Halibut cannot be mistaken for 

 that of a Turbot, and neither for that of a Lemon-dab or a Sole. Apart 

 from the general " form " the dimensions of parts, e.g. of the lengths of 

 tins or of the mandible, the breadth of the interorbital space, and so on, 

 are of no more than specific value. 



The indefinite nature of the external form of the flat- fish is shared 

 by various other characters which constitute the external appearance of 

 the animal. They seem at first sight to be definite, even generic 

 characters, and thus worthy of being ranked as important in a system 

 of classification, but yet when examined more closely they seem capable 

 of indefinite and indeterminable variation. One of these characters is 

 the position which the eyes occupy, whether on the left or right side of 

 the animal. By some writers (40) this character has been ranked as of 

 generic importance, but most systematists have given it a subordinate 

 rank. A great deal might be said for both sides. When we find that 

 a character varies even within the limits of a species, as this one does for 

 many species both of the European and American fauna, we are apt to 

 regard it as useless for classificatory purposes. And yet if we omit these 

 variable species we find that this character is constant and generic. 

 Further, if we disregard what are evidently abnormalities, such as 

 Turbots with arrested migration of the eye, we find that the variable 

 species are transitional forms of some kind or other. This character is 

 therefore employed as of great importance in the present classification. 

 It aids to bring out the affinities of the different groups, and it will be 

 shown how even the variable forms, when properly classified, serve the 

 same purpose. 



Another character of the same description is found in the position of 

 the anterior extremity of the dorsal fin on the head. This was employed 

 as generic by Giinther and previous writers, but discarded by more 

 recent systematists. Taken by itself, this character suffers from our 

 inability to give it a clear definition, but when considered in its relation 

 to the position of the nasal organ of the blind side, as will be done later, 

 it is one of the most important characters. The reason for this is not 

 far to seek. It is certain that the fiat-fishes are descended from some 

 type of round -fish, and in the latter the dorsal fin does not extend 

 forward over the head. Hence the different degrees we find in the 

 forward extension of this fin should mark the different stages in 

 specialisation of the various groups of flat-fishes. Such, as we pass from 

 the Halibut-group through the Plaice- and Turbot-groups to the Soles, 

 is indeed the case, and the only difficulty in the way of employing this 

 character is our inability to define properly the position of the anterior 

 extremity of the fin. By employing the position of the nasal organ we 

 obtain a character which is more easily expressed, although both char- 

 acters, as will be shown later, must be taken together in measuring- 

 affinities. 



The size of tit e mouth was another character considered by Giinther 

 as of even greater importance than the position of the anterior 

 extremity of the dorsal fin, and which has also been degraded from 

 this high importance by the American writers. It has, indeed, far 

 less to recommend it than the former character. The Turbots 

 and Halibut, for example, are grouped together by Giinther under 



