1902.] OSTEOLOGY OF THE FALCONIFOKMES. 293 



hypei^apophyses are particularly well developed. The 2iid to 4tli 

 vertebrae bear hypapophyses : beyond these they are replaced by 

 catapophyses, which near the cervico- thoracic vertebra again 

 give place to hypapophyses. These catapophyses never coalesce 

 to form a carotid canal. In many of the larger Accipitres the 

 cervical vertebrae from say the 5th to the 8th have the neviral 

 plate deeply excised posteriorly, so that the postzygapophyses 

 appear, each as an articular facet at the exti-emity of a long beam. 

 The ventral surface of the centrum is either flattened or slightly 

 grooved. 



There is a very close similarity between the cervical vertebrae 

 of the Accipitres and the Striges : so close is this resemblance 

 that the one is hardly distinguishable from the other. The chief 

 differences appear to be in the fact that the cervical ribs of the 

 Striges are relatively slightly longer and the catapophyses some- 

 what more sharply defined. The pleurosteal lamella is also 

 somewhat more band-like and sharply defined. The hyper- 

 apophyses of the axis are abruptly truncated in the Owl, tubercular 

 in the Accipitres. For the rest, the differences are not greater 

 than those which normally obtain between species or genera, at 

 most. 



The cervici-thoracic vertebrae may be two or three in number. 



The thoracic vertebrae, as ah'eady remarked, are free save in 

 the Falconidse and Polyborida?. When fi'ee, they may be dis- 

 tinguished from those of the Stiigidee in that the neural spines 

 are relatively lower, wider antero-posterioi-ly, and not markedly 

 inclined forward. Hypapophyses in the Strigidae do not extend 

 beyond the second vertebi-a : in the Accipitres with free vertebrae 

 to the third. In the Strigidfe there may be a large pneumatic 

 aperture between the aiticular surfaces for the capitulum and 

 tuberculum, and a second caudad of this. As in the Accipitres 

 so in the Striges, each transverse process sends forward and 

 backward a long slender spike from its extreme lateral border ; 

 each spike overlaps similar spikes from the vertebra next in front 

 and behind it. The neural spines may also send backwards from 

 the upper border a pair of short spike-like processes, to embrace 

 the neural spine immediately behind it. 



In Serpentarius the thoracic vei'tebrae, from the 2nd to the 5th, 

 are pierced by a large pneumatic foramen, opening at the base of 

 the neural spine immediately behind the anterior zygapophysis. 

 This leads into an extensive chamber, excavated out of the 

 vertebral tissue and extending down to the spinal cord, being 

 separated therefrom only by a thin plate of bone. Other 

 pnevimatic apei'tures pierce the lateral walls of the neural tube, 

 and the centrum below this. 



In the majority of the Accipitres, the pneumatic apertures of 

 the thoracic vertebrae are restricted to a single opening at the 

 base, and caudad of the transverse process and immediately in 

 front of the base of the postzygapophysis. 



In the Cathartae the apertures are three in number, and lie on 



