RESTORATION OF ICHTHYORNIS. 181 



To the evolutionist, who believes that birds are all closely connected 

 genetically, this difference in structure, at first sight, offers a most serious 

 difficulty ; since hitherto we have had no hint of a transformation from the 

 one fonn to the other. 



In the third cervical vertebra of IcMhyomis, however, we catch nature 

 in the act, as it were, of forming a new type ; by modifying one form of 

 vertebra into another. Following this clew, the connection between these 

 widely divergent types of structure soon becomes apparent ; and the 

 development of the modern style of avian vertebra from the fish-like, 

 biconcave form finds a ready solution. In the anterior articulation of 

 this vertebra of Ichtliyornis (figure 34), the surface looks downward and 

 forward, being inclined at an angle of nearly 60° with the axis of the 

 centrum. In vertical section, it is moderately convex, while transversely 

 it is strongly concave ; thus presenting a near approach to the saddle-like 

 articulation. None of the other known vertebrae of IcMhyornis possess this 

 character. 



This highly specialized feature occurs at the first bend of the neck, 

 and greatly facilitates motion in a vertical plane. If, now, we consider for 

 a moment that the dominant motion in the neck of a modem bird is in a 

 vertical plane, we see at once that anything that tends to facilitate this 

 motion would be an advantage, and that the motion itself would tend 

 directly to produce this modification. With biconcave vertebrae, the 

 flexure in any direction is dependent on the elasticity of the fibrous tissue 

 that connects them, as the edges of the cups do not slide over each other. 

 An increasing movement in the neck of Ichtliyornis in a vertical plane 

 would tend to deflect the upper and lower margins of the circular cup, 

 and to produce a vertical constriction, and at the same time to leave the 

 lateral margins projecting ; and this is precisely what we have in the third 

 vertebra of this genus. 



This modification of the vertebras would naturally appear first where 

 the neck had most motion, viz: in the anterior cervicals, and gradually 

 would be extended down the neck; and, on to the sacrum, if the same 

 flexure were continued. 



