272 Correspondence. 



of the greater part of the resemblances which he discovers between 

 objects the most apart from each other in general structure, seems to 

 be a general law of nature which has ruled that, in every group of 

 animals there should be a minor group more essentially carnivorous, 

 another minor group more essentially herbivorous, another more 

 aquatic or natatorial, and so on. These minor groups may also be 

 characterized by one being more essentially terrestrial, another more 

 essentially aerial, another more aquatic, another more amphibious, 

 and so on. These general principles are the occasion of resem- 

 blances between animals the most distinct in their structure, and 

 therefore I understand perfectly what Swainson means when he 

 speaks of a Rasorial type of fish ; yet surely it is an incorrect ex- 

 pression, for so far from fishes having been created on the models of 

 Rasores or Grallatores, for all that we know, birds may have been 

 created on Plagiostomous or Cyclostomous types. The general model 

 was undoubtedly one ; but why Swainson should assume this one model 

 to have been taken from birds I cannot divine, except that in Orni- 

 thology he is most at home. However, to return to the subject of 

 Cyprinidae, your arrangement of them shews another set of analogies, 

 which I also think very conspicuous, for instance, 



The Pceonominaeare the types of the family Cyprinidae. 



The Sarcoborinae represent the Esocidae. 



The Pcecilianae represent the Clupeidae, 



The Cobitinae represent the Salmonidse. 



The Placycarinae represent the Siluridae. 



You will perhaps say, that the Cobitinae ought to represent the Silu- 

 ridae; but the relation between the Cobitinae and Siluridae is one of 

 direct affinity, in which I perfectly agree with Swainson; and I have 

 accordingly made the Cyprinidae and Siluridae contiguous groups in 

 the table of CLUPEINA, given on the opposite page. 



When I can secure a safe private hand, I shall beg your acceptance of 

 a copy of the third part of the " Illustrations of the Geology of South 

 Africa." In the mean time I must refer you to a copy which I gave our 

 friend Dr. Cantor. In page 9 of that work, you will see a Table which is 

 in perfect accordance with your views of the value of the word genus ; 

 but not perhaps with your view of the word family ; nor is what I have 

 written above consistent with the view I have taken in that Table 

 of the value of the words genus and family. The truth is, what in the 

 foregoing part of this letter are called genera, are families and ought 

 to end in ides, as the peculiar designation of that rank of group; 

 but as these groups agreed wonderfully with the extent of the old 



