Classification of Teleostean Fishes. 



301 



the asymmetry which characterizes- them as a group. I am 

 fully convinced that if they do not actually form part of the 

 ancestral group out of which the flat-fishes were evolved, 

 they are very nearly related to them ; and it follows that the 

 Zeidge, of which our familiar John Dory is the best-known 

 representative, are the nearest known living allies of the 

 Pleuronectida3. The number of vertebrae in Amphistium is 



Restoration of Amphistium paradoxnm, Ag., from the Upper Eocene. 



rather in favour of than against such a view, since the least 

 specialized of living Pleuronectidse, Psettodes *, agrees with 

 it in this respect, all other forms of which the skeleton is 

 known having 28 or more. Although it is perfectly true that 

 in a general way the number of vertebrae has become reduced 

 in the course of evolution, this law certainly does not apply 

 to the particular groups, as seems to me proved by such series 

 of forms as we know in the Siluridte, Scombriformes, and 

 especially in this instance, where the increased number is 

 evidently related to the undulatory swimming movements 

 of these fishes. 



* In wliicli the eye of the blind side is not lateral, but on the dorsal 

 surface of the head, the dorsal fin does not extend on the head, the mouth 

 is large and symmetrical, and the pelvic bones and tins are placed as in a 

 normal Perciform. 



