﻿INTRODUCTION. 



There is no more striking instance of the difficulty of interpreting 

 fossil remains by a close comparison with the skeletons of existing 

 animals, than that presented by the Palaeozoic Fishes. When 

 the first fragments of Coccostean plates from the Lower Old Eed 

 Sandstone of Caithness were noticed by Sedgwick and Murchi- 

 son sixty years ago, nothing more closely similar among existing 

 animals could be found than the dermal plates of the mud-tortoises. 

 Trionyx was accordingly entered in the list of Caithness fossils \ 

 Nearly eight years later, the Eussian geologist Kutorga 2 , when 

 attempting to interpret fragmentary teeth and dermal plates from 

 the corresponding formations of Livonia, was led to name a long 

 series of mud-tortoises, lizards, and Ichthyosauri from that country, 

 giving good figures and detailed descriptions of the evidence upon 

 which the restoration of so remarkable and unexpected a fauna was 

 based. Even when such entirely erroneous impressions were re- 

 moved by the discovery of more satisfactory specimens, and when 

 the far-reaching researches of the ichthyologist, Louis Agassiz, had 

 shown that all these remains pertained to fish-like organisms no 

 longer existing, the same tendency to interpret the past by a rigorous 

 comparison with the present everywhere prevailed, and the frequent 

 result was a distortion of the facts of structure in the fossils to con- 

 form to arrangements observed in the present fish-fauna. Not only 

 was Hugh Miller induced, by Agassiz's researches, to compare in 

 detail the skulls of some of the Old Eed genera with that of the 

 living cod-fish 3 , but this recent gadoid was actually used by Agassiz 



1 Trans. Geol. Soc. [2] vol. iii. (1829), p. 144, pi. xvi. fig. 6. 



2 S. Kutorga, ' Zweiter Beitrag zur Geognosie und Palaontologie Dorpat's,' 

 1837. 



3 H. Miller, 'Footprints of the Creator,' (1849), p. 48. 



