﻿INTRODUCTION. IX 



an essential character in the diagnosis of subclasses of Fishes. In 

 short, the terms " Ganoid " and " Teleostean," while convenient for 

 use in alluding to well-defined hony-scaled types and modern bony 

 fishes respectively, can no longer be employed as means of precise 

 scientific expression. 



At the same time, however, that modern research has led to these 

 difficulties, the combined results of comparative anatomy and palae- 

 ontology have suggested an alternative classification, which seems to 

 express all the more important facts at present known. It is to be 

 expected that any subdivision of a class into " orders " or " subclasses " 

 will gradually become less cogent as the earlier types are more 

 fully revealed ; but when all discoveries tend to prove that these 

 subdivisions are divergent phyla, meeting only in remote antiquity, 

 an approximately natural classification seems to have been attained. 

 Among fishes, for example, it is now well known that, at least since 

 Lower Devonian times, there have been two distinct plans of cranial 

 structure, between which no definitely intermediate forms occur. 

 As pointed out both by Stannius x and Huxley 2 , the upper segments 

 of the mandibular and hyoid arches are directly fused with the 

 chondrocranium in Chimcera, Protojzterus, and their allies ; while they 

 are loosely articulated, the upper segment of the hyoid arch forming 

 a movable suspensorium, in all the Elasmobranchs and the so-called 

 Ganoidei and Teleostei. These types of cranial structure are termed 

 respectively the " autostylic " and " hyostylic " 3 . It is now generally 

 admitted that the first division passes through some of the early 

 Dipnoan fishes into the Amphibia, and thus into terrestrial Verte- 

 brates ; while it seems equally clear that the extreme specialization 

 of the second division has resulted in the modern types of fishes — the 

 vertebrates most completely adapted to an aquatic existence. 



Again, it will be observed that in the earliest known Palaeozoic fish- 

 fauna there are representatives both of the autostylic and hyost3 ? lic 

 types on the same primitive biological level, so far as the develop- 

 ment of the appendicular skeleton and the axial skeleton of the trunk 

 are concerned, but yet differing in the nature of the exoskeleton. 

 Some families exhibit mere ' k placoid " dermal calcifications, traversed 

 by delicate branching nutritive canals, these isolated plates not 

 uniting even in the region of the branchial apparatus to form any 

 covering of the clefts ; other families are well encased in dermal and 



J H. Stannius, « Handbuch der Zootomie— Fische,' (1846), pp. 18, 32. 



2 T. H. Huxley, ' Elements of Comparative Anatomy' (1864), pp. 195, 209. 



3 T. H. Huxley, Proc. Zool. Soe. 1876, pp. 40, 41. 



