E. S. Biggs — Brachiosaurus altithorax. 305 



elusion is borne out by the presence of four ankylosed verte- 

 brae in the sacrum, the subquaclrate form of the coracoid, the 

 relatively short tibia, the Morosaur-like thoracic vertebrae and 

 the expanded blade of the scapula. Unfortunately there are 

 parts of two skeletons represented in the type of C. supremus, 

 which accounts for Cope's describing it as having twenty or 

 more thoracic vertebrae. 



There are reasons both for and against considering Atlanto- 

 saurus as a synonym of Camarasaurus. The size of the type 

 specimens is almost identical, as is shown by the length of their 

 respective femora. The similarity between the sacra holds so 

 far as the number and size of the vertebrae are concerned. The 

 one difference in evidence lies between the hollow centra 

 described by Marsh and the solid structure of the same as 

 stoutly maintained by Cope. The ischium of Camarasaurus is 

 not known ; that figured by Marsh is far from the Morosauroid 

 type. Neither is the pituitary canal described in the skull of 

 Atlantosaurus consistent with the characters which one would 

 expect in this genus. The writer was in error in stating in a 

 recent note on this form* that the type specimens of both genera 

 came from the same locality. That of Camarasaurus was col- 

 lected near Canon City, Colorado, while the type of Atlanto- 

 saurus came from near Morrison, one hundred miles farther 

 north. 



Apatosaurus is clearly distinguishable from Camarasaurus 

 by the narrow blade of the scapula if not by the doubtfully 

 constant character of three coalesced vertebrae in the sacrum. 

 With Atlantosaurus the sacrum forms the sole basis of com- 

 parison. If it be conceded that the primitive dinosaurian 

 sacrum is made up of three coalesced vertebrae, we may fairly 

 assume that the Apatosaur type represents merely a young 

 specimen of Brontosaurus. The size of the specimen, the 

 straight blade of the scapula, the imperfectly ossified border of 

 both scapula and coracoid and the character of the -dorsal' ver- 

 tebrae all bear out this conclusion. 



Amphicoelias is unique in the length and slenderness of its 

 femur. The biconcave type of caudal centrum is common to 

 a number of Sauropod genera ; but that this type of vertebra 

 persisted throughout the thoracic series may well be questioned. 

 The closest affinities of this genus seem to be with Dijplodocus. 



In view of this complexity of uncertain relationships, little 

 can be said of the affinities of Brachiosaurus. As the humerus 

 in the most nearly allied genera is not known, the fore leg 

 offers no basis of comparison other than the coracoid. This 

 bone differs from that of the type Camarasaurus supjremus in 

 * Field Columbian Mus. Pub. Geol. Ser., vol. i, No. 10. 



