X . of certain JMrds of Cuba. ... 21 



to vary much less than in Mammalia; and consequently, to 

 merit in them more attention as a ground of division. Finally 

 therefore, we may conclude, that in that order of Birds where 

 this great principle of structure varies the most, there Nature 

 is — if I may use the expression — looking out for the structure of 

 some other class. 



This mode of reasoning is, I am fully aware, not without its 

 defect ; but if it can be admitted to possess any value, it follows, 

 that we must look among the Rasores and Natatores for the 

 outlets from the class. And as there cannot be the least doubt 

 of the Natatores* approaching to the Chelonian reptiles, we 

 must consequently look among the Rasores for the approach to 

 Mammalia. 



Now this agrees with theory, inasmuch as it is from the two 

 extremes of the three Aberrant groups that we should expect to 

 pass into the contiguous classes. 



But this question is of such extreme importance to zoology, 

 that it ought not to be dismissed slightly. I trust, therefore, 

 that I shall scarcely be deemed to trespass upon the time of the 

 Society, if I here attempt to investigate three subjects, which 

 have exercised the ingenuity of naturalists from the earliest 

 periods of their science, and which are as follows : 



1. The true analogies existing between the orders of Birds 

 and those of Mammalia. 



2. The connection of the various orders of Mammalia in their 

 own series of affinity t. 



3. The point of nearest approach made by Mammalia to 



Birds, 

 * Bonnet op this account divided the order into Aves Aquatics and Aves Amphibia. 

 See on this subject Horn Entomologicce, p. 0.63, et seq. 



t I ought here to acknowledge, that the subject of the natural division of the Mam- 

 malia has been taken up by a writer in the Annals of Philosophy for November 1826. 

 This gentleman not only shows much acquaintance with the class in detail, but has 

 ingeniously developed a number of natural approximations. Unfortunately, however, 



first 



