1898.] Osborn, Additional Characters of Camarasauriis. 23 I 



in the great majority of cases very much disturbed and confused 

 before embedding. An instance in point is the loss of all the 

 anterior caudals in Marsh's otherwise excellent B. excelsus 

 specimen. 



Let us suppose 

 that the three early 

 united vertebrae (of 

 the more primitive 

 Cetiosaur sacrum) 

 were found alone, 

 the others having 

 been detached and 

 lost, we would then 

 have the genus Apa- 

 tosauriis ; if one of 

 the freer vertebrae 



had united we Fig. 10. Camarasanrns siiprciiiKs type. Anterior dorsal 



vertebra. After Cope. Approximately je natural size. Com- 

 WOUld have the ge- pare Morosaurus, Marsh, 1897, fig. 31, p. 181. 



nus Atlantosauriis 



or Caniarasaurus ; if both of the freer vertebras had united we 



would liave Brontosaunis. 



4. Anterior Dorsal Vertebra of the Cetiosaurs are of 



Avian Type. 



Anterior dorsals without median spines have been described 

 by Cope in Caniarasaurus (Fig. lo) and by Marsh in Apatosaurus 

 (Fig. 9), Morosaurus and Diplodocus. This condition seems so 

 general as to constitute almost an ordinal character of the Cetio- 

 saurs. 



These vertebrae are, moreover, remarkable in resembling those 

 of certain Struthious birds such as Dronuriis (Fig. i), in the 

 absence of median spines, in the elevation of lateral spinous 

 processes above the zygapophyses, and in the abrupt development 

 of a median spine upon the third or fourth-dorsal. 



It is also not improbable that the abrupt transition from verte- 

 brae without median spines to a vertebra with a strong median 

 spine {Dromceiis, D.3) is paralleled in the blunt spine of the 

 Caniarasaurus dorsal (Fig. 11), \vhicli may well re[)resent I' 3 or 



