lateral keels, and these first appear at about vertebra 50.
The bone is thickened where these keels meet the anterior
and posterior edges of the centrum, forming what she
interpreted as facets for hemal arches. These facets become
more prominent by about vertebra 61, though the point is
made that the ventral keels are not pronounced in every
individual. One of the specimens in which they are said to
be well developed is BMNH R2180. Although pairs of low
ridges can be seen in some of these vertebrae, they would
not be described as keels nor are the hemal facets noted by
Kirton obvious features. There are similar slight ridges on
the ventral surfaces of some of the caudal vertebrae of
BMNH 481 (a series of large vertebral centra, probably
referable to Temnodontosaurus), but these cannot be described
as prominent, or even very obvious. However,
there are some slight thickenings on the anterior and posterior
edges of the centrum, where the low ridges intersect
them, but, again, these are neither prominent nor very
obvious. Kirton (1983) acknowledged that although the
facets she identifies as being for hemal arches are found
throughout most of the caudal region in Qphthalmosaurus,
the arches themselves have not been found.
The major problem in trying to establish the presence
of chevrons is distinguishing them from ribs. The ribs
diminish in size fairly rapidly in the caudal region, becoming
reduced to small, stout, bony rods, well before the apex
of the tailbend. The other problem is that small elements,
including the caudal ribs, often are not preserved. Examination
of ichthyosaurs from the Lower Lias in the Natural
History Museum, London, revealed little evidence of chevron
bones. However, one specimen, BMNH 14563, proved
remarkably informative.
According to L ydekker (1889A: 58), BMNH 14563,
which is part of the H awkins collection, is said to be “probably
from Lyme Regis.” However, this partial skeleton has
more the appearance of coming from Street (Somersetshire),
and thus could be latest Triassic or earliest Jurassic
in age (McGowan, 1989A). The caudal region is remarkably
well preserved, with ribs continuing all the way back
to vertebra 77, which is at the start of the apical region (Fig.
17A,C). The ribs are short and broad, and steadily decrease
in size posteriorly. The corresponding (unpaired) facets for
their articulation with the centra remain throughout. The
diminutive last rib is accompanied by a minute bony rod,
lying immediately ventral to the centrum (Fig. 17C). Similar
single rods occur in association with the next several
centra. However, at vertebra 95, there is a pair of rods, and
paired rods occur elsewhere too (Fig. 17B). To verify
whether these rods were indeed paired, counts were made
of the number of rods in a given number of segments.
Sixteen or seventeen rods were counted between vertebrae
90 and 100, and up to 26 between vertebrae 90 and 105.
There is little doubt that the rods are paired. They are not
displaced ribs because they co-exist with ribs for several
segments. Furthermore, whereas the ribs decrease in length
posteriorly, the rods increase in length. There can be no
doubt that the rods represent the unfused chevrons. The
largest chevrons occur at about the middle of the post-
apical region. Diminishing in size posteriorly, they continue
almost to the tip of the tail.
Identifying Vertebrae
Isolated vertebral centra can readily be identified in terms
of their approximate position in the vertebral series. When
the diapophysis (dorsal rib facet) is confluent with the
facet for the neural arch, the vertebra lies well forward, and
is from the cervical or anterior dorsal regions. A heart-
shaped anterior or posterior aspect confirms that it lies
well forward, at about vertebra 3 or 4 (Fig. 18A-D). When
the diapophysis and parapophysis (which may or may not
be confluent) lie in the upper half of the centrum, the
vertebra is probably from the anterior half of the dorsal
series (Fig. 18E-F). Vertebrae in which the two facets lie in
the lower half of centrum are probably from the posterior
half of the dorsal series (Fig. 18G-H). When the diapophysis
and parapophysis are confluent and low down, the
vertebra is situated close to the sacrum (Fig. 18I-J). Such
vertebrae are typical of the caudal series, well anterior to
the tailbend. Vertebrae lacking rib facets, which are also
usually laterally compressed are probably postflexural,
unless they are wedge-shaped and therefore apical (see
Fig. 16).
Taxonomic Use of Vertebrae
With the notable exception of Ophthalmosaurus, vertebrae
are seldom seen in their entirety in associated skeletons.
Instead, vertebrae are usually exposed from one aspect,
most often in lateral view. Consequently, we have little
idea of the exact shapes of the vertebrae from the different
regions of the vertebral column for most species. Accordingly,
when isolated vertebrae are found, their appearance
is often so unfamiliar that they are ascribed to new taxa.
Owen (1840), for example, erected the new species Ichthyosaurus
trigonus for an isolated centrum, from the Kim-
meridge Clay of Wiltshire, which was remarkable for its
triangular appearance when viewed anteriorly. The specimen
was said to be broadest dorsally, the upper surface
forming a straight line. The sides were remarkable for their
straightness, converging ventrally at an angle of 70 degrees.
The specimen was not figured, but, based on the
description, it was probably a cervical vertebra of Ophthalmosaurus
(Fig. 19A). Incidentally, there are several triangular
centra in the collection of the BMNH that have been
incorrectly labelled as Ichthyosaurus trigonus. This is beprobably
diapophysis
diapophysis
Fig. 19. A) cervical centrum of Ophthalmosaurus (BMNH R2180). B) posterior dorsal centrum of Ophthalmosaurus (BMNH 24684 d),
redrawn from L ydekker (1889A : fig. 14) who erroneously referred it to Ichthyosaurus trigonus. C) posterior dorsal centrum of Ophthalmosaurus
(BMNH R2175). Scales each equal 40 mm.
cause the apex of the triangle is dorsal, rather than ventral
as described by Owen (1840). One of these (Fig. 19B) was
figured by L ydekker (1889A: fig. 14). Such vertebrae can be
identified as being from the posterior dorsal region of
Ophthalmosaurus (Fig. 19B-C). The shapes of the vertebrae
should clearly be used with great caution in taxonomy.
Vertebral counts, in contrast, are reliable taxonomic features.
Ribs
Ribs occur throughout most of the vertebral column, commencing
at the atlas, and, in some cases, continuing well
beyond the tailbend. The first few ribs are short, but they
increase in length rapidly, so that by about the fourth or
fifth rib, they are almost full length. As discussed earlier,
the vertebrae show a trend from having paired rib facets
throughout most of the presacral region, to having single
ones in the vicinity of the sacrum and beyond. Likewise,
Fig. 20. Transition from posterior dorsal to anterior caudal regions of BMNH 2013 (Ichthyosaurus communis). Notice how the ribs become
much shorter and markedly expanded distally. Scale equals 100 mm.