263 



From W. S. Macleay, Esq. A.M. F.L.S., dated Elizabeth Bay near Syd- 

 ney, N. S. W. 12th September, 1840. 



My dear Sir, — I cannot find terms to express my gratitude for your 

 kind letter of the 12th March last, and for the very valuable present 

 which it accompanied. I assure you, that your excellent work on 

 Cyprinidce has afforded me the greatest delight, and the more so, inas- 

 much as I am convinced natural arrangement is always best tested by 

 accurate analysis, and also inasmuch as I am not by any means satisfied 

 with Swainson's arrangement of Fishes. As from every thing Swainson 

 writes, there is information to be derived, so I assure you, his little 

 volume on Reptiles and Fishes has not been lost on me ; yet the perusal 

 of your Monograph on Indian Cyprinidse, has made me recur to my old 

 views on a subject, which our common friend, Dr. Cantor, may have told 

 you has long occupied my thought ; and although perhaps you will 

 deem these views not sufficiently worked out, and rather crude, I cannot 

 refrain from making you acquainted with them, in order that I may 

 have the benefit of comparing your general arrangement of Fishes with 

 my own. 



Fishes form a class of Vertebrata, which has never yet been satis- 

 factorily divided into orders. I do not think that Acanthopterygii and 

 Malacopterygii, for instance, are natural orders. In order therefore to 

 arrive at the first great and natural division of fishes, I think we must 

 commence by incontestable data, or at least by facts that are generally 

 agreed on. Such facts for instance I hold to be the three following, viz., 

 1. The near approach of fishes to Batrachian Amphibia, which with 

 Swainson I consider to be made by means of Lophius and Malthe. 2ndly. 

 The near approach of fishes to Cetaceous Mammalia, which with him also 

 I consider to take place by means of Selache and the viviparous sharks. 

 3rdly. As the grand character of fishes as a class is, their being the most 

 imperfect of Vertebrata, the most typical of fishes ought therefore to be 

 the most imperfect of them, I e. the farthest removed from the type 

 of Vertebrata. Such fishes are evidently the Cyclostomi of Cuvier, such 

 as Myxine, and other genera leading off to Annulosa. Though essentially 

 aberrant, as they relate to vertebrated animals, the Cyclostomous fishes 

 are typical as respects the circle of fishes. Now it is this circle of fishes 

 in which we have the three above data, namely the two aberrant orders 



