CLASSIFICATION OF ANIMALS. 195 



repeat, in the same class, the characteristic changes which are pecu- 

 liar to the whole family, to require any further argument to show that 

 Palmipedes are not, necessarily, a natural division ; and though we 

 may fail for the present in rearranging the families of this class into 

 natural orders, I trust after these remarks, more importance will yet 

 be attached, and more attention paid in future, to the fact that Pal- 

 mipedes, as they are now characterized, have very different types of 

 wings and bills. I have, for my own part, been strongly impressed 

 with the resemblance which exists between gulls and frigate birds, 

 and the birds of prey, of the hawk and vulture families, in which the 

 toes are by no means so completely distinct as they are among other 

 birds. And, far from considering birds of prey as the highest family 

 among birds, I would only consider them as highest in the series 

 which includes simultaneously Procellaridoe and Laridse. Whether 

 the family of pelicans belongs to this group or not, I am not prepared 

 to say ; but, at all events, the fact of their preserving their four toes 

 in one continuous web shows them to rank lowest among birds. 



Again, among reptiles there will no longer be a foundation for 

 any arrangement resting merely upon impressions ; thus the terres- 

 trial turtles will stand higher than the freshwater, and these again 

 higher than the marine ; and among Batrachians, which are best 

 known in their embryology, we can already arrange all the genera 

 in natural series, taking the metamorphosis of the higher as a 

 scale, and placing all full-grown forms in successive order, accord- 

 ing to their greater or less resemblance to these transient states. 

 Even the relative position of toads and frogs may be settled with 

 as much internal evidence as any other question of rank in wider 

 limits, merely upon the difference of their feet. 



In my researches upon fossil fishes I have on several occa- 

 sions alluded to the resemblance Avhich we notice between the 

 early stages of growth in fishes, and the lower forms of their families 

 in the full-grown state, and also to a similar resemblance between the 

 embryonic forms and the earliest representatives of that class in the 

 oldest geological epochs ; an analogy which is so close, that it involves 

 another most important principle, viz., that the order of succession in 

 time, of the geological types, agrees with the gradual changes which 

 the animals of our day undergo during their metamorphosis, thus 



