8 ON THE SYSTEMATIC POSITION OF HESPERORNIS. 
The fifth cervical vertebrae are equally similar, so far as the known characters 
go, but the neural spine and the pleurapophyses are damaged in the specimens of 
(aN NY | 
lg = \ \ 
SY; 
Fig. 10. Firra cERVICAL VERTEBRA OF COLYMBUS. 
Hesperornis. They differ on precisely the same lines as the last from the same 
vertebra in the Ostrich or the Rhea. 
The dorsal vertebrae of Hesperornis agree equally well with those of Colymbus, 
except that in the first dorsal the long hemal process or hypapophysis is single 
instead of forked: the hemal processes are broken in the remaining dorsal vertebrae 
ficured by Marsu. No hemal processes, or very rudimentary ones, exist in the 
Ratite. In other respects, in the shape of the articular surfaces of the centra, the 
position of the articular processes, the long transverse processes, and the long, high, 
and compressed neural spines, the agreement between Colymbus and Hesperornis is 
complete. 
The ribs are nine in number in Hesperornis, the same number as in Colymbus : 
in Hesperornis the three first, in Colymbus one only, is without attachment to a 
sternal rib. They are in both birds flatter and broader than in the Ostrich, and the 
first is fairly well developed, instead of being very minute. The uncinate processes 
afford better characters. In the Ostrich and Rhea they are small and three in 
number only: in the other living Ratites they are rudimentary or absent. In 
Hesperornis and in Colymbus they are very large, flat, and long, directed upwards 
at an acute angle, and are borne by six ribs in both genera. In Hesperornis they 
are free and unanchylosed with the ribs; such is the case in four out of the six ribs 
in my most adult specimen of Colymbus. 
