188 ODONTORNITHES. 



judge from Jurassic Mammals and Reptiles, the next classes above and 

 below Birds, the avian forms of that period would still be birds, although 

 with even stronger reptilian features. For the primal forms of the bird- 

 type, we must evidently look to the Pakeozoic ; and in the rich land fauna 

 of our American Permian we may yet hope to find j the remains of both 

 Birds and Mammals. , 



The genera Arclmopteryx, Hesperomis, and Ichthyornis, each jwssessed 

 certain generalized characters not shared by the others. These characters 

 were undoubtedly united in some earlier form, and this fact gives us a hint 

 as to what the more primitive forms must have been, and suggests the 

 prominent features of the ancestral type. 



In the generalized form to which we must look back for the ancestral 

 {ype of the class of Birds, we should therefore expect to find the following 

 characters : 

 , \ (1.) Teeth, in. grooves. 



(2.) Vertebras biconcave. 



(3.) Metacarpal and carpal bones free. 



(4.) Sternum without a keel. 



(5.) Sacrum composed of two vertebras. 



(6.) Bones of the pelvis separate. 



(7,) Tail longer than the body. . . 

 c (8.) Metatarsal and tarsal bones free. 



(9.) Four or more toes, directed forward. 

 (10.) Feathers rudimentary or imperfect. 



These various characters may indeed have been combined in an 

 animal that was more reptile than bird; but such a form would be on the 

 road toward the Birds, rather than on the ancestral line of either Dinosaurs 

 or Pterodactyles, as feathers were not a character of these groups. With 

 this exception, all of the characters named belong to the generalized 

 Sauropsid, from which both birds and the known Dinosaurs may well 

 have descended. An essential character in this ancestral type would be a 

 free quadrate bone, since this is a universal feature in Birds, and only 

 partially retained in the Dinosaurs now known, v 



