EVOLUTION AND PALAEONTOLOGY 259 



but it seems not impossible that these fins might have disap- 

 peared altogether and the posterior rays have extended for- 

 wards. With regard to the anterior position of the pelvic fins 

 in the Gadidae it is to be remembered that they are not directly 

 attached to the pectoral girdle, while in the Ccelacanthid;e 

 these fins show a marked approximation to the pectorals. The 

 lobes at the bases of the paired fins are much reduced in Ccela- 

 canthidai as compared with other fringe-finned ganoids. 

 Coelacanthidaa, e.g., Macropoma, extend to the Cretaceous, and 

 fossil forms of Gadida: begin to appear in the following period, 

 the Eocene. 



The most primitive of the spiny-finned Teleosteans (Acan- 

 thopterygii) are also the most ancient of the fossil forms belong- 

 ing to this Order, namely the Berycidae, which were already 

 abundant in the Chalk period. They are closely similar to the 

 ganoid Eugnathidae in most of their characters, only differing in 

 the presence of a few spines in the dorsal and ventral fins and 

 the attachment of the pelvic bones to the pectoral girdle. In 

 Beryx and Holocentrum the air-bladder even retains its com- 

 munication with the digestive tube. There can be little doubt 

 that these early examples of the spiny-finned type were evolved 

 in the sea, and that the comparatively few existing fresh-water 

 forms, such as the Perch, have ascended from the sea at some 

 later period. 



The above may seem a somewhat technical and condensed 

 discussion, beyond the easy comprehension of the average 

 reader. It is however impossible to state the most important 

 facts concerning fossil fishes, and the relative age of the geolo- 

 gical strata in which they occur, without the use of some tech- 

 nical terms. We will endeavour to give in simple language 

 the chief conclusions which our present knowledge enables us 

 to draw concerning the evolution of fishes. In the broadest 

 sense it may be said that there are two principal types of 

 structure in fishes, the shark type and the bony type. At the 

 present time fishes of these two types exist in large numbers 

 and in great variety side by side in the sea. We cannot 

 maintain that the differences of structure between sharks and 

 bony fishes correspond to equally marked differences in mode 

 of life : the shark and the tunny are both voracious monsters 

 living as far as we can see much in the same way, both 



