1 8 9 5- ORIGIN OF SPECIES AMONG FLAT-FISHES. 175 



An important specific difference between Z. punctatus and 

 Z. unimacnlatus is the concrescence of the pelvic and ventral fins in the 

 former and not in the latter. Yet the observations on the two 

 species show that this feature has nothing to do with the peculiar 

 habit described. No difference of habit has yet been detected 

 corresponding to this difference of structure. Nor have we yet any 

 evidence of the utility of the prolongation of the first dorsal ray in 

 Z. unimacnlatus. 



But these are not the only specific differences. There are others 

 which consist in differences of the degree to which the generic 

 characters are developed. In other words, the specific characters of 

 the three species form a series. Even the vomerine teeth and the 

 union of the pelvic and ventral fins are to be regarded as terms of such 

 a series, for the teeth on the vomer are, in Z. norvegicus, very small, in 

 Z. unimacnlatus absent, in Z. punctatus well developed, while in the 

 former two species the pelvic fins extend backwards so as to embrace 

 the anus and commencement of the anal fin, although they are not 

 united with it. The other characters, for example the accessory 

 finlets and the scales, are developed in different degrees in the three 

 species. The finlets are largest and the rays most branched in 

 Z. punctatus, in which each contains six rays, while in Z. unimacnlatus 

 and Z. norvegicus there are not more than four. The generic character 

 of the scales is that the exposed portion is short antero-posteriorly ; 

 there is a single row of spines of which the centre one is prolonged 

 and extends directly outward from the surface. In Z. punctatus this 

 character is developed to the maximum, the whole exposed portion of 

 the scales being bent nearly at right angles to the embedded portion ; 

 the projecting spines are not of one uniform length, but those of 

 maximum length occur at scattered points. It is this character of 

 the scales which gives the peculiar hirsute character to the skin of 

 Z. punctatus. It is developed also in Z. unimaculatns, but in Z. norvegicus 

 occurs to a much less degree. In the last species the exposed 

 portions of the scales are broader and lie flat to the body, but here 

 and there occurs one with a projecting central spine ; these are more 

 abundant on the head. 



If the generic characters were adapted to peculiar generic habits, 

 the specific differences in these characters might be similarly adapted 

 to differences in the habits. I have given evidence that the accessory 

 flaps of the fins beneath the tail are not necessary to the habit of 

 adhesion, which is produced by the movement of the posterior part of 

 the fins, and this is related to the characteristic shape of the body, 

 which depends on the breadth of the body and fins at the posterior 

 end. This, therefore, might have been selected, but on the other 

 hand the enlargement of the posterior part of the body might have 

 been produced by the movement. Young brill have been seen to 

 move the posterior parts of the fins when adhering to glass. The 

 peculiar character of the scales has, so far as the evidence goes, no 



