JURASSIC BIRDS AND THEIR ALLIES. 313 



Having been engaged for several years in the investigation of 

 American Mesozoic birds, it became important for me to study the 

 European forms, and I have recently examined with some care the 

 three known specimens of Archceopteryx. I have also studied in the 

 Continental museums various fossil reptiles, including Compsognathus, 

 which promised to throw light on the early forms of birds. 



During my investigation of Archceopteryx, I observed several 

 characters of importance not previously determined, and 1 have 

 thought it might be appropriate to present them here. The more 

 important of these characters are as follows : 



1. The presence of true teeth, in position, in the skull. 



2. Vertebra? biconcave. 



3. A well-ossified, broad sternum. 



4. Three digits only in the manus, all with claws. 



5. Pelvic bones separate. 



6. The distal end of fibula in front of tibia. 



T. Metatarsals separate, or imperfectly united. 



These characters, taken in connection with the free metacarpals, 

 and long tail, previously described, show clearly that Ave have in 

 Archwopteryx a most remarkable form, which, if a bird, as I believe, 

 is certainly the most reptilian of birds. 



If now we examine these various characters in detail, their im- 

 portance will be apparent. 



The teeth actually in position in the skull appear to be in the pre- 

 maxillary, as they are below or in front of the nasal aperture. The 

 form of the teeth, both crown and root, is very similar to the teeth of 

 Hesperornis. The fact that some teeth are scattered about near the 

 jaw would suggest that they were implanted in a groove. No teeth 

 are known from the lower jaw, but they were probably present. 



The presacral vertebra? are all, or nearly all, biconcave, resembling 

 those of Ichthyornis in general form, but without the large lateral 

 foramina. There appear to be twenty-one presacral vertebra^, and the 

 same, or nearly the same, number of caudals. The sacral vertebra? 

 are fewer in number than in any known bird, those united together 

 not exceeding five, and probably less. 



The scapular arch strongly resembles that of modern birds. The 

 articulation of the scapula and coracoid, and the latter with the ster- 

 num is characteristic ; and the furculum is distinctly avian. The 

 sternum is a single broad plate, well ossified. It probably supported 

 a keel, but this is not exposed in the known specimens. 



In the wing itself the main interest centers in the manus and its 

 free metacarpals. In form and position these three bones are just 

 what may be seen in some young birds of to-day. This is an im- 

 portant point, as it has been claimed that the hand of ArchcBopteryx 

 is not at all avian, but reptilian. The bones of the reptile are indeed 

 there, but they have already received the stamp of the bird. 



