104 MEMOIRS OF THE CARNEGIE MUSEUM 
the intercentra. There could not have been more than one retained and it may 
be possible that all have disappeared. ‘This indicates a higher degree of specializa- 
tion than has hitherto been found to exist among the Ichthyopterygia. The Trias- 
sic form Shastasawrus,”” Merriam has five intercentra while in Ichthyosawrus there 
are but three. 
The paired neural arch is the only process supported by the atlas. They extend 
upward and backward overlapping the lateral basal portion of the spine of the axis 
and never uniting to form a spine. The left half of this arch was found displaced 
though in an excellent state of preservation. Viewed from the side (fig. 12, b), it is 
Fig. 12. Left half of paired neural arch of atlas of Baptanodon discus (No. 878). One half vatural size. «a, seen 
anteriorly ; b, seen laterally ; c, view of distal or articular end. 
Fic. 18. Superior view of an anterior cervical of Baptanodon discus (No. 608). One half natural size. a, anterior 
d and « the combined dia- and neurapophysial surfaces ; n.c., neural canal ; p, posterior. 
Fig. 14. Anterior view of anterior cervical of Baptanodon discus (No. 603). One half natural size. d and 2, the 
coalesced diapophysis and neurapophysial surfaces ; n.c., neural canal ; p, parapophysis. 
an irregular spear-shaped compressed bone that articulated with the centrum by an 
expanded subtriangular articular end (fig. 12, c). An anterior view (fig. 12, a) 
shows the sinuous curves made necessary for it to fit closely to the process of the 
axis. ‘The function of these bones appears to be the same as the corresponding ele- 
ments of I. longifrons as described by Owen, which they resemble somewhat in shape 
and position. This paired neural arch does not develop zygapophyses. The pedicels 
of the axis as they rise from the surface of the centrum converge and meet above the 
neural canal to form a well-developed neural spine, the first of the vertebral series. 
There are no indications of prezygapophyses on the spine of the axis belonging to 
No. 878, though the process of B. marshi shows such articulating surfaces. ‘The pos- 
terior portion of the spine is broken away so that its extent antero-posteriorly can- 
not be determined accurately from our specimens, and it would appear after an 
37 Merriam, J. C., ‘‘ Triassic Ichthyopterygia from California and Nevada,’’ Univ. of Cal. Pub., Vol. 3, No. 4, p. 75. 
