2 24 Bulletin American Miisenni of Natural History. [Vol. X, 



It is possible, however, that Cope's type represents a Diplodocus. 



The mid- and posterior-dorsal vertebrae are so different that 

 Cope supposed that they represented different genera, viz.. Cavi- 

 ar asaurus and Amphicoelias. 



Fig. 3. Cainarasaurus. Estimated 5th caudal. Posterior dorsal vertebra. Am. Mus. 

 Coll. R. 222. Approximately j'g natural size. 



The sacrum is represented in this specimen by a single large 

 free sacral vertebra, and by the coalesced spines of three sacrals. 

 The significance of these facts will be pointed out later. 



The tail of Caniarasaiirus includes a far more powerful and 

 complex series of vertebrae than has been indicated in the figures 

 and descriptions of Marsh. 



Immediately behind the sacrum, the first caudal of Cainara- 

 saurus (Fig. 4) is an entirely unique vertebra, extending laterally 

 into two great transverse plates, which may have come in contact 

 at their extremities with the posterior portions of the ilia. These 

 plates are concave anteriorly. The pre- and post-zygapophyses 



