PROCEEDINGS 



OF 



THE GEOLOGICAL SOCIETY OF LONDON. 



Vol. n. 1834—1835. No. 37. 



November 5, 1834. — The Society assembled this evening for the 

 Session. 



Hugh Edwin Strickland, Esq., of Craycombe House, near Eves- 

 ham ; Sir William Molesworth, Bart,, Member of Parliament for the 

 Eastern Division of Cornwall, Pencarrow, Cornwall ; and Herbert 

 Mayo, Esq., of Cheshunt, Herts, were elected Fellows of this 

 Society. 



A paper was read "On a new Classification of Fishes, and on the 

 Geological Distribution of Fossil Fishes," by Prof. Agassiz, of Neu- 

 chatel. 



The author begins by observing that the state of the science of 

 Ichthyology had obliged him to undertake an examination of recent 

 fishes for the sake of comparing them with the fossil species, and in 

 doing so that he had arrived at a classification, in general, differing 

 considerably from the various arrangements previously adopted by 

 naturalists. One of the essential characters of fishes is, to have their 

 skin covered with scales of a peculiar form and structure. This co- 

 vering, which protects the animal without, is in direct relation with 

 its internal organization, and Dr. Agassiz has found that by an at- 

 tentive examination of the scales, fishes may be divided into more 

 natural orders than had hitherto been adopted. In this manner he 

 has established four orders, which bear some relation to the divisions 

 of Artedi and Cuvier ; but one of which, hitherto completely misun- 

 derstood, is almost exclusively composed of genera, whose species 

 are found in only the most ancient strata in the crust of our globe. 

 These four orders are, the Placoidinns, which comprehend the car- 

 tilaginous fishes of Cuvier, with the exception of the sturgeon ; the 

 Ganoidians, which comprehend above fifty extinct genera, and to 

 which we must refer the Plectognaths, Syngnaths, and Acipensers j 

 thirdly, the Ctenoidians, which are the Acanthopterygians of Cuvier 

 and Artedi, with the exception, however, of those which have smooth 

 scales, and with the addition of the Pleuronectes. Lastly, the Cy- 

 clo'idians, which are principally Malacopterygians, but which com- 

 prehend, besides, all those families excluded from the Acanthoptery- 

 gians of Cuvier, and from which we must take the Pleuronectes 

 placed among the Ctenoidians. 



If we estimate the number of fishes now known to amount to about 

 eight thousand species, we may state that more than three fourths of 

 this number belong to two only of the above-mentioned orders ; 

 namely, the Cycloidians and Ctenoidians, whose presence has not 



VOL. II. K 



