204 GENERAL ORNITHOLOGY part ii 



Ichthyornis ; it is unknown in recent birds.' A centrum cupped in 

 front only is procodous ; one cupped only behind is opisthoccdous 

 (Gr. oTTicrOe, opisthe, behind). Such structure results in a ball-and- 

 socket jointing of vertebrae. In those vertebrae of birds in which 

 this arrangement obtains, it is always the posterior face of a centrum 

 which is cupped, the anterior one being balled ; such vertebrae are 

 therefore opisthoccelous. But in the freest vertebral articulation of 

 birds, that existing in the region of the neck, another modification 

 occurs. Both ends of each vertebra are saddle-shaped, i.e. concave 

 in one direction, convex in the other ; a condition which is called 

 heieroccelous (Gr. erepos, heteros, contrary). The concavo-convexity of 

 any one vertebra fits the reciprocal concavo-convexity of the next. 

 Anterior- faces of heteroccelous vertebrae are concave crosswise, up- 

 and-down convex ; posterior faces are the reverse ; consequently, 

 such vertebrae are procoelous in horizontal section, but in vertical 

 section opisthoccelous. The various physical characters of vertebrae 

 in different regions of the body, and their connections with and 

 relations to other parts of the body, have caused their division into 

 several sets, as cervical, dorsal, etc., which are best considered 

 separately. 



Cervical Vertebrse (Fig. 56, cv) are those of the necJc: all those 

 in front of the thorax or chest, which do not bear free pleura- 

 pophyses in adult life, or the free pleurapophyses of which, if any, 

 are not in two-jointed pieces and do not reach the breast-bone ; i.e. 

 have no haemapophyses. It is advisable, in birds, to draw this line 

 between cervical and succeeding vertebrae, no other being equally 

 practicable ; for, on the one hand, one, two, or more of the cervicals 

 (recognisable as such by their general conformation and free articu- 

 lation) may have long free ribs, movably articulated ; and all the 

 cervicals, excepting usually the first, or first and second, have short 

 pleurapophyses, ankylosed in adult life, but free in the embryo; 

 while, on the other hand, a vertebra, apparently dorsal by its con- 

 figuration and even its ankylosis with the dorsal series, may be 

 entirely cer^^cal in its pleurapophysial character.^ Thus, in Fig. 56, 

 of an owl's trunk, the bone which is apparently first dorsal, and is 

 so marked (dv), bears a free styliform " riblet " an inch long (c'), only 



^ Except to this statement, however, the oddly-massed pygostyle, which, in birds 

 where a terminal disk develops inferiorly, may be distinctly cupped at both ends, as 

 it is in a raven, for example. 



^ The case is very puzzling ; the more so because, viewing the whole scries of 

 birds, the ambiguous " cervico-dorsal," or two such equivocal vertebrffi, may lean in 

 different cases in opposite directions when the whole sum of characters is taken into 

 account. Therefore it may be best, as already said, to make the possession of a 

 jointed sternum-reaching rib the criterion of the first dorsal vertebra, even though an 

 antecedent one may have the physical characters of a dorsal, and be ankylosed with 

 the dorsal series. This is the view taken by Huxley, who says : " The first dorsal 

 vertebra is defined as such by the union of its ribs with the sternum by means of a 



