THE MOST PRIMITIVE KNOWN REPTILE. 279 



but well-formed transverse process bearing a rib-facet distally. 

 The neural spine of this vertebra is destroyed. 



The restoration (text-fig. 7) is only uncertain in the shape 

 and size of the intercentrum between the odontoid and axial 

 centrum and of that element itself. The sutures between the 

 neural arch and centrum of the axis are obliterated. 



The 3rd and 4th vertebrae exactly i^esemble the 2nd, except that 

 the arches are wider and the transverse processes longer. The 

 neural spine is represented only by a little ridge. 



The 5th vertebra exactly resembles that in front of it except 

 that it has a low but very massive neural spine. 



The centra of the 3rd, 4th, and 5th vertebrae are short antero- 

 posteriorly and have a very well-marked keel, their sides being- 

 excavated into deep concavities on each side of the middle line. 

 The intercentrum between the 4th and 5th vertebrae is well 

 displayed, it is nearly as long as the centrum before it, and 

 laterally is carried out into well-marked processes for articulation 

 with the capitula of the ribs. 



The next vertebra, completely preserved is the 9th, from which 

 the series is complete and naturally articulated to the 10th 

 caudal. 



The 9th is a typical Seymouina vertebra with a small well- 

 rounded centri\m, considerably longer than that of the 5th. The 

 arch is enormously wide and massive, but the ends of the com- 

 paratively slender transverse processes project well bej'^ond the 

 extremities of the zygapophyses. The 9th vertebra has a low but 

 very massive spine whose summit is bifid. 



From this vertebra to the sacrum the sti'ucture of the vertebrae 

 remains very unifoi-m, the only changes being that the transverse 

 process gradually shortens, the spine becomes lower and more 

 slender, and the ventral surface of the centrum becomes flattened. 



The 10th vertebra is unique in that it is entirely devoid of the 

 slightest trace of a spine. I myself removed the matrix which 

 covered it, and the surface is still practically perfect. 



My skeleton is abnormal in that the 23rd vertebra is asym- 

 metrical, on the left side agreeing exactly with that in front, 

 whilst the right side of the neural arch is much elongated and 

 carries a sacral rib. 



The 24th vertebra, the sacral, is only exposed on the left side, 

 •where it carries a veiy massive sacral rib, applied to the inner 

 aspect of the ilium in the usual way. This vertebra has a rela- 

 tively high and massive spine. 



As Prof. Williston has already remarked, the first caudal difliers 

 markedly from all that precede it in its narrow and unexpanded 

 neural arch. It has a fairly long transverse process. From hei'e 

 to the sixth the caudals do not difier much, except that the spine 

 gets progressively more and more slender and leans more steeply 

 backward . 



The seventh and succeeding caudals bear no ribs and the 

 transverse process has disappeared from them ; their centra are 



20* 



