CHAP. XIX.] FISHES. 467 



covered belong to the Ganoidei, a highly developed group which 

 has continued to exist down to our times, and of which the 

 sturgeon is the best known example. We may therefore be sure 

 that the Upper Silurian rocks in which these are found, although 

 so very far back in geological history, do not by any means lead 

 us to the time when the primitive fish-type appeared upon the 

 earth. In the Carboniferous and Permian formations numerous 

 remains of fishes are found, allied to the Lepidosteus or Gar-pike 

 of North America. The next group in order of appearance, are 

 the Plagiostomata, containing the existing Sharks and Eays. 

 Traces of these are found in the highest Silurian beds, and be- 

 come plentiful in the Devonian and Carboniferous formations 

 and in all succeeding ages, being especially abundant in Creta- 

 ceous and Eocene strata. The Holocephali appear first in the 

 Oolitic period, and are represented by the living Chimaridse. 

 The Dipnoi, to which belong the Lepidosiren and Ceratodus, are 

 believed to have existed in the Triassic period, from the evidence 

 of teeth almost identical with those of the existing Australian 

 fish. All the ancient fossil fishes belong to the above-mentioned 

 groups, and many of them have little resemblance to existing 

 forms. The Teleostean fishes, which form the great bulk of 

 those now living, cannot be traced back further than the Creta- 

 ceous period, while by far the larger number first appear 

 in the Tertiary beds. The Salmonidse, Scopelidse, Percidse, 

 Clupeidae, Scombresocidse, Mugilidse, and Siluridse, or forms 

 closely allied to them, are found in the Cretaceous formation. 

 In the Eocene beds we first meet with Squammipennes, Cypri- 

 nidse, Pleuronectidee, Characinidse, Mureenidse, Gadidse, Pedi- 

 culati, Syngnathidae, and Hippocampidse. 



Most of these fossils represent marine fishes, those of fresh- 

 water origin being rare, and of little importance as an aid in 

 determining the causes of the distribution of liAdng forms. To 

 understand this we must look to the various changes of the 

 land surface which have led to the existing distribution of all 

 the higher vertebrates, and to those special means of dispersal 

 which Mr. Darwin has shown to be possessed by all fresh-water 

 productions. 



