NATURAL SCIENCES OP PHILADELPHIA. 263 



common original stock a generalized gyrenccphalic type whose 

 form had alreadj^ been modified for continued aquatic life by the 

 reduction or atroph^^ of the hinder members, and that the existing 

 orders represent the extreme specialization in diverse directions 

 from such a common type, tban that the two forms have inde- 

 pendently diverged from two ambulatorial t^^Dcs. Even admit- 

 ting the utmost differences that have been urged between the two 

 forms, and the respective affinities of each to other forms, all such 

 are still so indisputably related, inter se, as alone to suggest 

 suspicions and doubts as to independent origin. Furtliermore, 

 it would appear more probable that such divergence had com- 

 menced from a type whose toes were separated and spreading, 

 than from one whose digits and corneous appendage had already 

 become specialized as hoofs ; the difficulties, indeed, incident to 

 the view of the development from a hoofed mammal of a mutilate 

 form, are so great and obvious that reasons stronger than any 

 yet given will doubtless be required before a general belief in 

 such development, or in the very intimate affiuit}^ of the Sireuians 

 and Pachyderms, prevails. 



It must of course be granted that the pisciform mammals are 

 the derivatives from quadrupedal types, for the evidence as to 

 the line of descent from the lowest quadrupedal mammal to the 

 higher, is unassailable; therefore, the terms embryonic' or syn- 

 thetic are inapplicable to the pisciform mammals ; on the contrary, 

 those mammals must be considered as in the highest degree de- 

 veloped and specialized t3q)es, and if the extent of specialization 

 in miy direction were a true test of rank, their claims to the first 

 rank would be indisputable. 



Character op Progenitor. 



As to the common progenitor of tlie Sirenians and Cetaceans, 

 the evidence is by no means as clear as would be desirable. We 

 have, in the Pinnipedia, a type that has been, in a liigh degree 

 (compared with ordinary quadrupeds), specialized for aquatic life. 

 The characters common to that type and the Fissipede carnivores, 



' Prof. Agassiz lias been able to see in tlie mutilate mammals " enibry- 

 onic" types, considering the Cetaceans to be "embryonic" carnivores, and 

 the Sirenians to be genuine " embryonic" types of Pachyderms. See Proc. 

 Boston Soc. Nat. Hist., iii. p. 309, 1850. 



