diapophysis of
atlas centrum
atlas centrum
proably fused
atlantal intercentrum
Fig. 9. Atlas-axis complex of an unidentified specimen (BMNH 35302), possibly representing Brachypterygius, from the Upper
Greensand. A) anterior and B) left lateral views. Scale equals 50 mm.
There is some variation in the development of the rib
facets on the atlas^axis complex. The diapophyses and
parapophyses are both well developed in the axis vertebra
of the two specimens described here (Figs. 4B, 5B), but they
are less clearly demarcated for the atlas. Describing the
situation for Ophthalmosaurus, Kirton (1983) noted that the
rib facets are poorly defined for both the atlas and the axis,
with much individual variation, and also variation between
left and right sides. The dorsal surfaces of the atlas
and axis are very similar, and are like those of the succeeding
vertebrae. Centrally is a depressed area, forming the
floor of the neural canal, flanked on either side by a well-
defined facet for articulation with a pedicel of the neural
arch (Figs. 4D, 5A-B). The two halves of the neural arch of
the atlas in BMNH R6697 are separate from one another,
without any development of a neural spine. Although they
are not united, they are fused with the underlying left and
right sides of the neural arch of the axis (Fig. 6A-B). Viewed
from behind, the neural arch of the axis has a discrete pair
of posterior zygapophyses (Fig. 6C). However, the anterior
zygapophyseal surface of the succeeding neural arch appears
to be an unpaired structure (Fig. 6D). The same
seems to be generally true of the succeeding vertebrae,
their articulations being through unpaired zygapophyses.
However, it is likely that there is some variation in this, as
indicated by a String of vertebrae (BMNH R1201) from
fairly far back - probably from the posterior dorsal series
- in which the (posterior) zygapophyses are widely separated
(Fig. 7).
The atlas-axis complex in Ophthalmosaurus, from the
Oxford Clay (Callovian-Oxfordian), is similar to that of the
Early Jurassic Ichthyosaurus. There is a well-marked suture
between the centra of the atlas and axis, and there is a large
facet for articulation with the atlantal intercentrum, but
there does not appear to be a separate axis intercentrum
(Kirton, 1983). In one well-preserved specimen (BMNH
32701), which is probably referable to Ophthalmosaurus, the
atlas and axis are both drawn out ventrally into low prominences,
as in the Lower Lias material (Fig. 8). There is a
large, triangular facet, presumably for articulation with the
atlantal intercentrum (Fig. 8A, E), but the intercentrum has
not been preserved. In contrast, there is no facet for articulation
with the axial intercentrum. Since there is a well-
marked ventral suture between atlas and axis centra, it is
clear that the axial intercentrum is not fused with them
(contrast Fig. 4A with Fig. 8E).
Two unidentified specimens (BMNH 35302 and 35415)
from the Upper Greensand (Aptian-Albian), which may
represent Brachypterygius, show complete fusion between
the atlas and axis, with no evidence of a suture between the
two elements (Fig. 9). The anteroventral margin of the atlas
centrum lacks an articular facet for the atlantal intercentrum
(Fig. 9A). The margin comes to a blunt point, reminiscent
of the shape of the anterior aspect of the atlas of
Ichthyosaurus when the atlantal intercentrum is in place
(compare Figs. 5D and 9A). This suggests that the atlantal
intercentrum is fused with the atlas in these specimens.
The atlas-axis complex is closely similar in Platypterygius
(BMNH 10894). Therefore it appears that there has been a
change in the atlas-axis complex during the latter stages of
ichthyosaurian evolution, with the loss of the suture between
the atlas and axis, fusion of the atlantal intercentrum
with the centrum of the atlas, and loss of the axial intercentrum.
Remaining Presacral Vertebrae
Several trends occur in the vertebrae in passing from the
cervical to the sacral regions: the centra tend to increase in
size, especially in dorsoventral height; the diapophyses,
which are confluent with the facets for the neural arches
anteriorly, become more ventrally placed, thereby losing
their contact with these facets; the parapophyses become
situated more ventrally; if the diapophyses and parapophyses
were confluent anteriorly, they become separate;
the diapophysis eventually becomes reduced, leaving a
Fig. 10. Lateral view of vertebral column, starting from third cervical (top left), to just beyond apical region (bottom right), showing
some of the vertebral trends discussed in the text. Drawn from Ichthyosaurus communis (BMNH R1162). Scale equals 100 mm.
single rib facet - this may occur well before, immediately
before, or at the level of the sacrum (Figs. 10-11). The neural
spines increase in height rapidly on the first few vertebrae,
remain fairly steady, and then decrease in height. There is
also a change in their general proportions, the spines being
more slender anteriorly, more broad posteriorly. They all
project posterodorsally. Although the neural arches are
typically not fused to the centra, fusion does sometimes
occur, as in a rather well-preserved specimen of Ophthalmosaurus
(BMNH R873). An interesting feature is that the
distal ends of the neural spines are deeply excavated, suggesting
a cartilaginous extension. This feature can be clearly
seen in the acid-prepared material from Dorset (Ichthyosaurus
communis) and was also recorded for Ophthalmosaurus
(Andrews, 1910).
As noted earlier, these are general trends, and there is
much variation, both individually and bilaterally. It is also
likely that there are taxonomic differences. However, as
pointed out elsewhere (McGowan, 1994C), vertebral features
should be used with much caution to characterize
taxa. This is largely because so few comparative data are
available to determine whether a given vertebral feature is
constant for a given species. For example, Kirton (1983)
noted for Ophthalmosaurus that the diapophysis and para-
pophysis become confluent at about the level of vertebra
42 or 43, which marks the position of the pelvic girdle, and
therefore the end of the presacral series. This feature can be
Fig. 11. Lateral views of vertebral centra of Ophthalmosaurus
(BMNH R2180), showing how the anterior caudals (B) are relatively
thinner (antero-posteriorly) than the precaudals (A). Notice
the single rib facet (probably representing the parapophysis) in
the caudal centrum. Scale equals 50 mm.