THE VERTEBRAL COLUMN. 



51 



as in the preceding dorsals. The transverse processes, which with the first dorsal began to 

 increase in length and to pass from the nearly horizontal position they occupy in the cervicals 

 to the more elevated positions which they assume in the median and posterior dorsals, are 

 now inclined upward at an angle of about 45° from the neural spines and rise nearly on a 

 level with the summits of the latter. 



DORSALS 4 TO 14. 



The structure of the succeeding dorsals is so similar that they may best be described 

 together by calling attention to those characters wherein they differ from one another anc 

 from the preceding dorsals. The capitular facet, which, as already noticed, suddenly shifted 

 to a more elevated position on dorsal 3, gradually continues to move upward on the succeeding 

 dorsals until in the fifth it is seen to be leaving the neural arch and moving out on the inferior 

 surface of the transverse process. This continues until in the posterior dorsals it occupies 

 a position midway between the apex and base of the transverse processes, which are a little 

 shorter and somewhat weaker in this region than in the middorsal series. The antero- 

 posterior diameter of the neural 

 spines is also somewhat greater in 

 the posterior and middorsal regions, 

 and the transverse processes increase 

 both in length and strength until 

 about the tenth, when they become 

 a little more slender and somewhat 

 shorter as we proceed posteriorly. 

 The neural arches are throughout 

 of about equal height. Every dor- 

 sal vertebra has two rib facets, a 

 capitular and tubercular, showing 

 that the ribs were double headed 

 throughout. In the middorsal re- 

 gion there are a number of long, 

 flat, ossified tendons lying on either 

 side of the neural spines near their 

 summits and closely applied to them, 

 as shown in fig. 48. 



In fig. 52 there is shown a de- 

 tached posterior dorsal of Triceratops 

 prorsus. As will be noticed by a 

 comparison of this figure with fig. 51 , the neural spine is much broader, the incisions of the neural 

 arch for the anterior and posterior zygapophyses are much deeper, the capitular rib facet has 

 moved far out on the inferior surface of the transverse process, the centrum has become pro- 

 portionally higher and narrower, the neural canal is smaller, and the neural arch is more 

 constricted. 



THE SACRUM. 



The sacrum is composed of ten coossified vertebras. The centra of the median sacrals are 

 somewhat compressed, but the centra of both the anterior and posterior sacrals are heavier 

 and more expanded. Although ten vertebras are firmly united by their centra to form the 

 sacrum, not all of these should be regarded as true sacrals. The first of these ten coossified 

 vertebras should be more properly regarded as a sacro-lumbar or dorso-sacral, and Marsh 

 considered the two anterior sacrals as such, while four or perhaps five of the posterior of these 

 coossified vertebras should be regarded as sacro-caudals, thus reducing the number of true 

 sacrals to four or five according to the number that have been eliminated as dorso-sacrals and 

 sacro-caudals. 



Fig. 52. — A, Lateral view of posterior dorsal of Triceratops prorsus, No. 4842, U. S. 

 National Museum, last dorsal of mounted skeleton; B, anterior view of same. 

 s, Neural spine; t, transverse process; h, capitular facet; z, anterior zygapophy- 

 ses; z' , posterior zygapophyses; n, neural canal; a, anterior; p, posterior. One- 

 eighth natural size. After Marsh. 



