N... 10(i(;. OSTEOLOGY OF CAMrTOHAURlS—OILMORE. 273 



Detailed description. — The skull and jaws of the holotype are un- 

 known, and the complete neck with the first dorsal is all that is 

 preserved of the vertebral column of this individual. The cervicals, 

 posterior to the third, are all opisthocoelous, the second and third, 

 as in the other S2)ecies of the genus, being plano-concave. The 

 median ventral keel is Avider and heavier, and the centra are not so 

 deeply excavated laterally as in G. hrowni. The cervical region of 

 the holotj^pe of C, mecUus being still inclosed by the matrix, direct 

 comparisons with that species could not be made. 



The first dorsal, as in G. hrowni, is opisthocoelous and shows the 

 same transition of the parapophysis from the anterior end of the 

 centrum in the ninth cervical to a position well up on the side of 

 the neural arch on this vertebra. It is the first vertebra of the series 

 to bear a true spinous process — a short spine 11 mm. high, slightly 

 thickened and roughened at its upper extremity. The neuro-central 

 sutures are plainly shoAvn on all of the vertebrae preserved. 



The sacrum, No. 187T«, Yale University Museum (see Plate 13), 

 although considered by Marsh as pertaining to the holotype, in all 

 probability belongs to another individual. This is shown by a com- 

 parison of the measurements of the cervical and sacral vertebrae with 

 the homologous parts of Cat. No. 4282, U.S.N.M., the proper associa- 

 tion of the different elements of this skeleton being unquestionable. 

 Referring to the table of measurements on page 243, it will be 

 seen that the cervicals are smaller and the sacrals larger than 

 those of No. 4282. On account of the difference in proportions noted, 

 I am inclined to believe the sacrum represents a distinct individual, 

 and should, therefore, as it was included in the material first de- 

 scribed, be considered a paratype, to which the catalogue number 

 1877a has been given to distinguish it from the holotype. No. 1877. 



In the sacral region of No. 1877a there are five vertebra^, united 

 by suture, of which the joosterior four represent true sacrals, or those 

 which give support to sacral ribs. This interpretation excludes the 

 anterior vertebra of the series which may be regarded as sacro-dorsal 

 («, Plate 13), as shown by the weak diapophyses which carried a 

 single-headed rib. Marsh has described the peculiar peg-and-notch 

 articulation (see 3, fig. 37) of the centra of this region and makes • 

 it an important feature of the genus, but after a study of several 

 sacra I am inclined to believe too much stress has been laid on this 

 character, as it appears to be a variable one. For example, in the 

 sacrum of G. hrowni this peculiarity is only weakly developed be- 

 tween S3 and S4 (see fig. 17), and in G. nanus there is only the 

 suggestion of such an articulation. In the sacrals under discussion 

 this articulation is faintly shown between Sj and So, more pro- 

 nounced between So and S;j, and strongly developed between S3 and 

 Proc. N. M. vol. xxxvi — 09 18 



