62 NATURE STUDIES. 



tions, however, were not teeth, but merely tooth-like 

 extensions of the horny covering of the beak. The 

 teeth of the Ichthyornis and Hesperornis, as is shown 

 by the smaller figure (showing a tooth, and, within it, 

 a tooth forming to take its place) were unmistakably 

 teeth. It does not take away from their dental cha- 

 racter that they were set in a groove in Hesperornis 

 and ArchceopteryXj instead of in separate sockets, as in 

 higher- toothed races and in Ichthyornis. 



It should be added that Professor Marsh has 

 examined the specimen of Archceopleryx in the British 

 Museum, and fully satisfied himself that it belongs to 

 the class of toothed birds. " The teeth seen on the 

 same slab with this specimen agree so closely with 

 the teeth of Hesperornis, that " he e ' identified them 

 at once as those of birds, and not fishes." 



He describes the leading characters of the ancestral 

 bird in the following terms : " In the generalised 

 form to which we must look for the ancestral type of 

 the class of birds, we should expect to find the following 1 

 characters: Teeth in grooves; vertebras biconcave" 

 (that is, the bones of the backbone shaped somewhat 

 as we see them in fish) ; " breastbone without a keel ; 

 tail longer than the body; bones of the hand and 

 wrist, as also those of the foot, free ; the bones of the 

 pelvis separate; the sacrum" (or hind bone of the 

 pelvis) " formed of two vertebras ; four or more toes 

 directed forward ; feathers rudimentary or imperfect." 



If we consider the circumstances under which, 

 according to the theory of evolution, the race of birds 



