26 PROCEEDINGS OF THE NATIONAL MUSEUM vol. 66 



in length from the first to the last. In view of the above, all 

 allocations given below will remain tentative until a skeleton with 

 associated vertebrae is found. Inasmuch as the first six ribs on 

 the right side were preserved, it has been possible to determine 

 definitely the position in the column of at least three of these 

 dorsals. 



The first and second dorsals are missing, but the third (pi. 8, 

 fig. 1) is characterized by a rather long diapophysis, large oval 

 prezygapdphysial facets which lie in a slightly oblique plane, and 

 a narrow neural spine. The neural arch (pi. 7, fig. 3) is low, 

 relatively long, and on each side gives rise to a lateral process, the 

 diapophysis. The articular facet for the tuberculum of the third 

 rib is situated on the lower two-thirds of the outer face of the 

 diapophysis. The neck of the diapophysis is constricted dorso- 

 ventrally between the facet and the neural arch. The postzy- 

 gapophysial. facets (pi. 7, fig. 6) are large, elongate, and slope ob- 

 liquely inward. Both epiphyses are missing. 



The neurapophyses of the fifth dorsal (pi. 7, fig. 2) are not as 

 highly arched as the third, the diapophyses are shorter, and the 

 neural spine is longer. No posterior epiphysis (pi. 7, fig. 5) is 

 present and at least two-thirds of the anterior one is missing. The 

 prezvgapophysial facets are large and elongate, but they are 

 nearer together than those of the third, and the inward slope is 

 more oblique. The facet for the tuberculum of the rib (pi. 8, fig. 

 2) has shifted very slightly in position from that of the third, 

 but it is somewhat broader. The postzygapophysial facets are 

 nearly vertical in position. There is a trace of a median keel on 

 this vertebra. 



Unless the vertebra assumed to be the sixth (pi. 7, fig. 1) is ab- 

 normal, it will be difficult to explain certain features possessed by 

 it. The centrum (pi. 8, fig. 3) is too short to follow after the 

 fifth vertebra, but the neural spine is too high and long, and the 

 prezygapophysial facets are too close together to assign a more 

 anterior position to it. This vertebra may not belong to this por- 

 poise. The diapophyses (pi. 7, fig. 4) are much shorter than those 

 of either the fifth or the third, and the mesial dorsoventral con- 

 striction has largely disappeared. 



The neural spine of the seventh dorsal is broken off near the base 

 and the posterior epiphysis is missing. The prezygapophysial facets 

 (pi. 7, fig. 7) are strongly concave, oblique in position, with thin 

 raised inferior margins. The neural canal is more nearly circular 

 than in any of the preceding vertebrae and the neurapophyses are 

 slightly thinner. The diapophyses are short and their extremities 

 are occupied by kidney-shaped facets for tlie tubercula. In the an- 



