LIFE SCIENCES CONTRIBUTIONS
ROYAL ONTARIO MUSEUM

NUMBER 83

Evolutionary Trends in

Longipinnate Ichthyosaurs
with Particular Reference to
the Skull and Fore Fin

C. McGOWAN

RC QO M

LIFE SCIENCES CONTRIBUTIONS
ROYAL ONTARIO MUSEUM

NUMBER 83

c.McGowaAn Evolutionary Trends in
Longipinnate Ichthyosaurs
with Particular Reference to
the Skull and Fore Fin

Publication date: 20 December 1972

Suggested citation: Life Sci. Contr., R. Ont. Mus.

Editors:

ROYAL ONTARIO MUSEUM
PUBLICATIONS: IN- DIBE SCIENCES

The Royal Ontario Museum publishes three series in the Life Sciences:

LIFE SCIENCES CONTRIBUTIONS, a numbered series of original scientific publi-
cations, including monographic works.

LIFE SCIENCES OCCASIONAL PAPERS, a numbered series of Original scientific
publications, primarily short and usually of taxonomic
significance.

LIFE SCIENCES MISCELLANEOUS PUBLICATIONS, an unnumbered series of
publications of varied subject matter and format.

All manuscripts considered for publication are subject to the scrutiny and
editorial policies of the Life Sciences Editorial Board, and to review by
persons Outside the Museum staff who are authorities in the particular field
involved.

LIFE SCIENCES EDITORIAL BOARD, 1972-1973

D. BARR
J. R. TAMSITT

CHRISTOPHER MCGOWAN is Assistant Curator, Department of Vertebrate Paleontology,
Royal Ontario Museum, and Assistant Professor, Department of Zoology, University
of Toronto.

PRICE: $2.00

© The Royal Ontario Museum, 1972

100 Queen’s Park, Toronto, Canada

PRINTED AT HANSON & EDGAR LIMITED, KINGSTON, ONTARIO

Contents

Abstract, /

Introduction, 2

Materials and Methods, 7
The Genus Platypterygius, 15

The Evolution of Longipinnate Ichthyosaurs, 22
Evolution of the Longipinnate Skull, 22
Evolution of the Longipinnate Body, 27

Evolution of the Longipinnate Fore Fin, 30
Size Ranges in Longipinnates, 31
Conclusions, 33
Acknowledgments, 36

Literature Cited, 37

Digitized by the Internet Archive
in 2011 with funding from
University of Toronto

http://www.archive.org/details/evolutionarytren00mcgo - 4

Evolutionary Trends in
Longipinnate Ichthyosaurs
with Particular Reference to
the Skull and Fore Fin

Abstract

The history of ichthyosaurs is briefly traced from the Middle
Triassic to the Upper Cretaceous. Platypterygius Von Huene,
the only Cretaceous representative, is compared with other
longipinnates, including Cymbospondylus petrinus Leidy,
Middle Triassic; Proteosaurus platyodon (Conybeare),
Lower Liassic; and Stenopterygius quadriscissus (Quenstedt),
Upper Liassic.

Several evolutionary trends occur in the skull and fore fin
of ichthyosaurs (McGowan, 1972) and whereas some of
these trends are well advanced in Platypterygius, others are
retarded. The diameters of the orbit and internal sclerotic
ring are small in Platypterygius, as in Cymbospondylus Leidy,
whereas the orbit lies more posteriorly, and is in this respect
comparable with Stenopterygius Jaekel. Maxillary reduction
in Platypterygius is intermediate between Cymbospondylus
and Stenopterygius. The fore-fin trend towards an increase in
the. total number of digits is more advanced in Platypterygius
than in any other longipinnate ichthyosaur, and the fin is
unique for the presence of three pre-axial accessory digits.
Platypterygius also differs from other longipinnates in its
body proportions, the skull is relatively larger, the tail fin
much smaller.

An attempt to reconstruct the locomotor habits of Platyp-
terygius indicate that the fore fins were used for propulsion,
unlike most ichthyosaurs, and it is suggested that speed and
manoeuverability may have been inferior to Jurassic longi-
pinnates.

Longipinnate evolution is discussed, and an attempt is
made to quantify the rates of cranial evolution. Rates of
change were highest during the Liassic and lowest during
the long, post-Liassic interval. The remarkable homogeneity
of ichthyosaurian body form during the Cretaceous, exem-
plified by the single genus Platypterygius, supports the con-
clusion that the rate of evolutionary change was at its lowest
level. Loss of adaptive plasticity at a time when a new
competitor, the mosasaur, was rapidly evolving is suggested
as a possible cause of extinction.

Introduction

A recent investigation of Cretaceous ichthyosaurs (McGowan, in press,
Contr. Geol.) demonstrated a remarkable homogeneity within the known
material, all referrable to the single genus Platypterygius, which is a longi-
pinnate. The distinction between latipinnate and longipinnate ichthyosaurs
was discussed elsewhere (McGowan, 1972), but it would be useful to
summarize the main points.

Latipinnate ichthyosaurs have fore fins with four distal carpals, each
supporting a primary digit, whereas longipinnates have only three distal
carpals and therefore only three primary digits. The total digital count is
increased in both groups by the presence of accessory digits, and in lati-
pinnates by digital bifurcation also. The evolutionary trend towards an
increase in the total digital count is more rapid in latipinnates than in
longipinnates, so that latipinnates usually have higher counts than their
longipinnate contemporaries. The two groups can also be distinguished
on cranial characters, latipinnates having relatively larger orbital and
internal sclerotic ring diameters and smaller maxillae than do longipinnates.

Platypterygius was widely distributed geographically and geologically. It
has been found in England, Germany, south-western Russia, north-eastern
India, western United States, Argentina, north-central Queensland and
Western Australia and was probably worldwide in distribution. Ranging
from the Neocomian to the Cenomanian (to the Santonian if the Danda-
ragan Chalk is of the same age as the Gingin Chalk, both of Western
Australia; see Teichert and Matheson, 1944, and McWhae et al., 1958),
Platypterygius existed from the Lower Cretaceous to the Upper Cretaceous,
a duration of about 45 million years. It is noteworthy that Platypterygius,
the last of the ichthyosaurs, was a longipinnate, the last known latipinnate
(Ophthalmosaurus Seeley) having apparently become extinct during the
late Jurassic.

Some cranial trends seen during ichthyosaurian evolution were weakly
expressed in Platypterygius and had barely progressed beyond the stage
exemplified by the earliest known longipinnate, Cymbospondylus, Middle
Triassic, North America. Before discussing Platypterygius, a brief review
of the geological history of the Ichthyosauria is warranted.

Unfortunately, the fossil record of the ichthyosaurs is discontinuous,
and remains are known from six horizons: Middle Triassic, Upper Triassic,
Lower Liassic, Upper Liassic, Lower Cretaceous, and the boundary of the
Lower Cretaceous/Upper Cretaceous. The earliest material is from the
Middle Triassic of Nevada and Spitzbergen. The Nevada locality yielded
the earliest known longipinnate genus, Cymbospondylus, which is known
from several specimens, including some well-preserved skulls and a nearly
complete skeleton. Unfortunately the structure of the fore fin is incomplete
(Fig. 1A) but resembles that of Merriamia Boulenger (Fig. 1 £) a longi-
pinnate from the Upper Triassic of California (see Merriam, 1908).
Occurring at the same locality as Cymbospondylus was a partial skull of
an obscure reptile described by Merriam in 1906 as Omphalosaurus.

iY
G
0000.
000
000

S500

a
is)

00000

00000000
699099000900000 0000

Fig. 1. Fore fins of longipinnate ichthyosaurs. For ease of comparison each fin
is drawn as if it were a left fin in dorsal view. A. Cymbospondylus petrinus Middle
Triassic, Nevada, U.S. (after Merriam, 1908, p. 64); B. Shastasaurus osmonti
Upper Triassic, California, U.S. (after Merriam, 1908, p. 64); c. Delphinosaurus
perrini Upper Triassic, California, U.S. (after Merriam, 1908, p. 64); D. Toretoc-
nemus californicus Upper Triassic, California, U.S. (after Merriam, 1908, p. 128);
E. Merriamia zitteli Upper Triassic, California, U.S. (after Merriam, 1908, p. 64);
F. Eurhinosaurus longirostris Upper Liassic, Germany (from Hauff, 1953, pl. 21);
G. Stenopterygius quadriscissus Upper Liassic, Germany, BMNH R4086; H. Platyp-
terygius platydactylus Lower Cretaceous, Germany (after Broili, 1907, fig. 13).
Seale: ==-4. cm.

Characteristic of the genus are the blunt, button-like teeth set in several
rows. Merriam (1906) did not consider the material to be ichthyosaurian
and concluded that it was probably best to place it in a group of its own.
In a description of Middle Triassic ichthyosaurs from Spitzbergen, Wiman
(1910) described some jaw fragments that possessed teeth similar to
those of Omphalosaurus. Wiman tentatively referred these jaw fragments,
as well as other material, to a new ichthyosaurian genus, Pessopteryx
Wiman. Merriam (1911) criticised Wiman’s inclusion of these jaw
fragments in the Ichthyosauria. In 1916 Wiman agreed with Merriam

3

and concluded that the fragments should be referred to the genus Omphalo-
saurus. Thus it was recognised that Omphalosaurus was not an ichthyosaur
but merely formed part of the Middle Triassic faunas of Nevada and
Spitzbergen. Nonetheless, Omphalosaurus continues to be considered an
ichthyosaur, albeit incorrectly. Omphalosaurus has been compared with the
placodonts, rhynchosaurs, and with the plesiosaurs (Merriam, 1906) but
its true affinities are still not clear. In addition to Cymbospondylus, the
Middle Triassic locality of Nevada has yielded two more ichthyosaurian
species, Cymbospondylus (?) natans Merriam (1908) and Phalarodon
fraasi Merriam (1910).

The type materia! of the species C. natans, which Merriam (1908)
only tentatively referred to the genus Cymbospondylus, consists of an
incomplete fore fin, which, Merriam noted, is somewhat similar to that of
the latipinnate genus Mixosaurus Baur. Wiman (1916) compared the fin
of C. natans with that of Mixosaurus nordenskioldii (Hulke), Middle
Triassic, Spitzbergen, and demonstrated a close similarity (see Wiman,
1916, figs. 1,2). From the above it seems that C. natans should be referred
to the genus Mixosaurus.

The type material of Phalarodon fraasi Merriam, Middle Triassic,
Nevada, consists of a small, partial skull bearing pronounced dimorphic
teeth. The anteriomost teeth are almost conical, whereas the crowns of the
posterior teeth are flat. Wiman (1916) compared P. fraasi with the Middle
Triassic Spitzbergen species M. nordenskioldii and found a close corres-
pondence between their dimorphic dentitions (compare Merriam, 1910,
pl. 40, and Wiman, 1910, pl. 5, figs. 9-13). Because possession of a
dimorphic dentition is characteristic of the genus Mixosaurus (see Merriam,
1908) I conclude that P. fraasi should be referred to that genus. Wiman
(1916) also noted that M. natans (Merriam) might represent the skull
of M. fraasi (Merriam) as they occurred together at the same horizon,
and this seems to be a reasonable conclusion. The Middle Triassic locality
of Nevada has thus yielded two well-known genera, the longipinnate genus
Cymbospondylus, and the latipinnate genus Mixosaurus.

The first description of ichthyosaurian material from Spitzbergen
(Middle Triassic) was given by Hulke (1873), who described two new
ichthyosaurian species, Ichthyosaurus nordenskiolditi and I. polaris. In
each the type materia! was fragmentary, but additional material was described
by Wiman (1910), who referred the first species to the genus Mixosaurus
and the second to a new genus, Pessosaurus. Wiman (1910) also described
a second new genus, Pessopteryx, to which he referred four new species,
P. artica, P. pinguis, P. minor, and P. nisseri. Whereas M. nordenskioldii
(Hulke) is represented by an amount of determinate material, Pessosaurus
polaris (Hulke) and the four species of Pessopteryx are poorly known.
Whether Pessopteryx and Pessosaurus are latipinnate or longipinnate can-
not be determined. In 1929 Wiman described a new ichthyosaur from
Spitzbergen which he referred to a new species, Grippia longirostris. More
adequate figures were given by Wiman in 1933 and it can be seen that
G. longirostris possessed the button-like teeth that are characteristic of
Mixosaurus (see Wiman, 1933, pl. 1). There seems little doubt that this

4

material should be referred to the genus Mixosaurus, and this suggestion
was made by von Huene in 1930 in a bibliographic reference (Neues
Jahrbuch, Min. Geol. Pal., Vol. 3, 1930, p. 264) to Wiman’s paper of
1929. Thus, recovered from the Middle Triassic of Spitzbergen are two
inadequately known genera, Pessosaurus and Pessopteryx, and two well-
known latipinnate species, Mixosaurus nordenskioldii (Hulke) and M.
longirostris (Wiman).

A number of well-preserved skeletons of the small latipinnate species
Mixosaurus cornalianus (Bassani) have been recovered from the Bassano
shales of Italy (Upper Triassic). M. cornalianus is similar to M. norden-
skioldii but for the absence of a post-axial accessory digit. M. atavus
(Quenstedt) from the Middle Triassic of Germany is a well-represented
species similar to M. cornalianus. Some ichthyosaurian vertebrae from the
Triassic of Timor, Indonesia, have been referred to as genus Mixosaurus
(see Broili, 1931, von Huene, 1936).

Four genera are known from the Upper Triassic of California, Merria-
mia Boulenger, Shastasaurus Merriam, Delphinosaurus Merriam, and
Toretocnemus Merriam, but the material is incomplete in every one.
Furthermore, in only Merriamia is the fore fin complete, so that our
knowledge of these ichthyosaurs is altogether fragmentary. Merriamia
possessed an unmistakable longipinnate fore fin and so did Delphino-
saurus (see Fig. 1). The fore fin of Toretocnemus appears to be similar
to that of Merriamia (compare Fig. 1D, E). It is unlikely that there were
more than three digits in Toretocnemus because the second digit lies
partly in line with the position once occupied by the second epipodial,
leaving only enough space to accommodate one more digit. The fore fin
of Toretocnemus is therefore concluded to be longipinnate. Although the
type material of Toretocnemus and Merriamia represents small individuals
(fore fin lengths of approximately 5 cm and 12 cm, respectively), Delphi-
nosaurus appears to be moderate in size (fore fin length in excess of 16
cm), and Shastasaurus (like Cymbospondylus) is extremely large (fore
fin length in excess of 40 cm). From the American Upper Triassic, then,
are four genera representing at least nine species. In numbers of specimens
and state of preservation, however, no horizon can compare with the
Lower Jurassic of Europe.

Most ichthyosaurs that have been found are from the Liassic deposits
of England and Germany. From the Lower Liassic of England are four
latipinnate species that belong to the genus Ichthyosaurus: I. communis
Conybeare, J. breviceps Owen, I. conybeari Lydekker, and I. tenuirostris
Conybeare. All four species are represented by complete skeletons. Occur-
ring with these latipinnates are at least three longipinnate species, including
Proteosaurus platyodon (Conybeare) and two new species of the genus
Proteosaurus that have not yet been described. Upper Liassic ichthyosaurs
are not frequently found in England and only one species, Stenopterygius
(?) acutirostris (Owen), is well known.

The Upper Liassic quarries of Holzmaden, Germany, have yielded
some of the best preserved ichthyosaurs. All appear without exception to

be longipinnate, and most are probably referrable to the genera Stenop-
terygilus Jaekel and Eurhinosaurus Abel. Among the better known species
are S. quadriscissus (Quenstedt), the most common, S. crassicostatus
(Fraas), S. megacephalus von Huene, S. hauffianus von Huene, S. mega-
lorhinus von Huene, and S. distinteger von Huene. Eurhinosaurus, distinct
from Stenopterygius, is characterised by a short lower jaw that does not
extend to the tip of the extremely long and tenuous snout. The fins are
long and narrow, and the angle of the tail bend exceeds that of all other
longipinnates. Only two species of this genus are known, E. longirostris
(Jaeger) and FE. hueni Swinton.

After the relative abundance of ichthyosaurs in the Upper Luiassic, little
material is seen again until the Late Jurassic, where the latipinnate genus
Ophthalmosaurus is found. Ophthalmosaurus, well represented in the
Oxford Clay of England, is moderate in size (about 4 metres in total body
Jength). Ophthalmosaurs are also found in the Upper Jurassic Sundance
formation of the western U.S. and are usually referred to the genus
Baptanodon, but further research into the group may well demonstrate the
inappropriateness of this allocation. The species described by Rusconi
(1948) from the Jurassic of Argentina, Ancanamunia mendozana is based
upon an incomplete skeleton, but the humerus is well preserved (Rusconi,
1948, fig. 65) and has three distal facets as in Ophthalmosaurus
(Andrews, 1910, fig. 36). Furthermore, the single phalanx figured
(Rusconi, 1948, fig. 69) has a rounded outline, like that of Ophthalmo-
saurus, and it seems certain that Rusconi’s species should be referred to
this genus.

Save for isolated teeth, vertebral centra, and other fragmentary material
little more of ichthyosaurian history is seen until early in the Cretaceous
when Platypterygius appeared. Platypterygius was a fairly large longi-
pinnate, with an adult body size of about 7 metres. The skull possessed
some characteristics that had barely progressed beyond the evolutionary
stage represented by Cymbospondylus of the Middle Triassic, and its fore
fins were unique (Fig. 1H). Pre-axial accessory digits numbered three,
giving rise to a total digital count of eight, which is unprecedented in the
longipinnate line. The fore fin was consequently broad, the base narrow,
and was primarily used for sculling. Platypterygius was examined in some
detail and will be compared with its predecessors in an attempt to gain an
insight into ichthyosaurian evolution.

Materials and Methods

Data were obtained by direct measurement of 22 longipinnates and by
indirect measurement from photographs of an additional 19. Of these
specimens, 10 were from the Middle and Upper Triassic of the western
U.S., 24 from the Lower Jurassic of Europe, and seven from the Lower
Cretaceous of Australia and the western U.S. The majority of specimens
were judged to be adult inasmuch as they were either close to the maxi-
mum size recorded for the particular taxon or they were large (skull
length exceeding one metre). Three of the specimens from the English
Lower Jurassic (Hettangian and Sinemurian) belong to a new species
of the genus Proteosaurus, and as this species has not yet been described,
it will be referred to throughout as Proteosaurus sp.*

All measurements were recorded in centimetres, using vernier calipers
where possible, and expressed to the first decimal place. When measure-
ments were taken from photographs, the estimated error (empiric) was
not in excess of 10%. Measurements taken from the skull are shown in
Fig. 2 and recorded in Table 1. Ratios derived from these measurements
are: Diameter of orbit to the length of jaw; internal diameter of sclerotic
ring to the diameter of orbit; length of external naris to the diameter of
orbit; distance between tip of snout and anterior border of external naris
to the length of jaw; and, distance between tip of snout and anterior tip of
maxilla to the length of jaw. Terminology used in describing post-cranial
structures is given in Fig. 3 together with those measurements recorded.
Note that the length of the fore fin is measured from the most distal point
of the humerus to the most distal phalanx.

3 1

Fig. 2. Diagram of an ichthyosaur skull showing those continuous variates that
were recorded. 1 — length of skull; 2 — distance between tip of snout and the
anterior border of the external naris; 3 — distance between tip of snout and the
anterior tip of the maxilla; 4 — length of external naris; 5 — diameter of orbit;
6 — internal diameter of sclerotic ring; 7 — length of snout; 8 — length of
jaw. Ang = angular; Dent = Dentary; Quad = quadrate; Sur = surangular.

Length of external

naris/diameter of

orbit

Distance between tip
of snout and anterior

boundary of external

naris/length of jaw

0Ss'0 $90
970 89°0
Lv'0 89°0
1S’°0 89'0
-" 89°0
Sv 0 $90
Iv'0 89'0
cS'0 69°0
0Ss'0 69°0
vr 0 SL0
vv'0 OL'0
aos =
se) 1 1) 3S Ps
oO Paste SB 2)
Oo ~eo ao
(9°) OS uy <= =p
gg a5
s a6 S
Ass

sinesoAy}Yyo1 ayeuurdisuo, wo

= £70
= ae vc 0
ca 070
in 170
ae cc0
— 810
8c 0 1c 0
a £70
ca 170
vv 0 810
OF'O 1c 0
ee 0 cl'0
L70 €l0

Internal diameter
of sclerotic ring/
diameter of orbit
Diameter of
orbit/length
of jaw

Bjep jeueiy — T ATaVi

496 ‘Id
€S6l Jgney
ye ‘jd
€S6I JyneH

4¢ ‘Id
eS6l jgney
4op ‘Id
€S6l JyneH
1qp ‘Jd
eC6l JyneH
OO€Sa HNWA
t9IZf WNO

O8I¢ WoU
LSSL WOO
9807 HNWA
v6LI WOU

Icpe MN
4€SVTA VNO

Specimen number

snssiasiaponb
snisk{sajdouajs

SNUDIIAIUD * J

S1]D41SnD
snisksajddjv]d

Species

‘

‘sydvisojoyd Wolj uaye} s}uawaInseayy |

a

‘paquiosap aA JOU smunvsoIjotg JO soisads MoU YW x

Or 0 sv 0 1¢°0 $s'0 = 910 0$66 dWwon
Lt 0 6r'0 9¢°0 790 (EG) St0 £166 dwon snut4sjad
sn] kpuodsoqu&ky
tr 0 6r 0 6£°0 c9'0 ae 0c 0 LI¢ad HNING
0S '0 8r'0 Ir0 c9'0 a3 0c'0 YI JINS
99°0 bro 6£°0 6$'0 9c 0 070 IL6Eb HNWA ."ds snanosojjoig
850 6F 0 a 790 S70 VEY) ip9SPl HNWE
19°0 6r 0 ers 790 67 0 S10 iSSTIYM HNING uoposin}d
SHANDSOIJOLg
8£°0 9¢5°0 960 69'0 HlaO weeT ‘Id
€C6l JJNVAY SNUIysO]DSatu “¢
OF'0 ESO sr'0 $90 Or'0 070 iez] ‘jd
€S6I JJNCH supoydaavsaiu “¢
= = a 99°0 sr'0 17'0 egy ‘Id
€C6l JJneH Lada UISIP “F
a rs'0 9r'0 99°0 aig 170 iqel ‘Id
ES6l FINCH SNYISOIISSvII “S
6r'0 6F0 re0 r9'0 6r'0 L10 461 ‘Id
€S6l JNeH ({) Siisounov “¢
a = ss 590 a 170 4411 ‘Id
tSol JyneH
a a . 09'0 =< 970 yey] yd

tSol JJneH

Pre

’ ‘fle xura| tail length
eee) 6
Ie . “— So

7, fa
4a,
tail bend "On “yy

SS

Fig. 3. Generalised body of an ichthyosaur showing characters referred to in the
text.

A graphic method was used to trace the evolutionary changes occurring
in the skull during the geological history of longipinnates. First, all cranial
data for a given taxon were collected, and mean values for the individual
ratios were calculated. A straight line representation of the skull of each
taxon was then produced in unit length, and the positions of the orbit,
internal aperture of the sclerotic ring, external narial aperture, and the
anterior tip of the maxilla were marked upon it (Fig. 4). Having obtained
a series of straight lines, one for each taxon, they were transferred to a
time scale, and the corresponding cranial foci were joined by straight
lines (Fig. 5). There are three possible criticisms of this method. First,
the resulting graph was constructed from only a few points. Second, when
a species was represented by only one individual, bias could result by
comparing specimens of different ages. That the graph is constructed
from relatively few points reflects the small sample size, but there are
some checks that can be made to assess its validity, and these will be
discussed later. Discrepancies resulting from allometric growth can
largely be dismissed. An extensive investigation into allometric growth in
the latipinnate species Ichthyosaurus communis and I. breviceps showed
that intracranial growth, for the most part, was isometric (McGowan
1969), and the same also seems to be true for the longipinnate Stenop-
terygius quadriscissus. Furthermore, the two specimens representing Platyp-

B S-SA N—N M

Fig. 4. Generalised ichthyosaur skull showing significant cranial variates converted
to linear foci. A — anterior boundary of orbit; B — posterior boundary of orbit;
S-S — internal diameter of sclerotic ring; N-N — length of external narial aper-
ture; M — anterior tip of maxilla.

10

Time scale
x 10° years

species B

species A

10 20 30 40 50 60 70 80 90 100
4% SKULL LENGTH

Fig. 5. Straight-line representation of two hypothetical species, A and B, trans-

ferred to a time scale. A — anterior boundary of orbit; B — posterior boundary
of orbit; N-N — length of external narial aperture; M — anterior tip of
maxilla; 1, 2 and 3 — graphs obtained by joining the foci representing respecti-

vely the orbital centre, narial centre and the anterior tip of the maxilla. Gradient
of graph 3 is y/x.

terygius australis and P. americanus and the two representing Cymbospon-
dylus petrinus are all obviously mature specimens. A third, more subtle
criticism of this technique is that with interspecific variation in the size
of the adult skull, the observed changes in orbital and narial sizes could
reflect the attainment of optimum organ size in the larger species. Thus,
suppose the optimum size of the eye (i.e., that size beyond which an
increment does not affect its efficiency) is such as to give an orbital
diameter of, say, 10 cm. Clearly, an orbital diameter of 10 cm in species A
(Fig. 6), the skull of which is 100 cm long, will appear to be relatively
smaller than in species B, the adult size of which is 50 cm. If species A
were descended from species B, it could be understood that the eye
became reduced during evolution, an erroneous conclusion. This problem
can be overcome by making comparisons with other species of similar
adult skull size and will be discussed later.

BI

(

1OO7CM:

Fig. 6. Comparison of skulls of generalised ichthyosaurs drawn to same scale
(upper pair) and to same size (lower pair).

Certain trends occurred during the evolutionary history of the skull that
are immediately apparent from the graphic method outlined above. These
trends are discussed below, but it is appropriate to define here some of
the terms used and to describe how the rates of these changes can be
quantified. The orbit and external nares tended to become more posterior
in position, resulting in a reduction of the post-orbital portion of the
skull. Accompanying these trends was a reduction in relative size of the
maxilla, manifest in a posteriorward retreat of the anterior tip of the
maxilla. These three trends are referred to in the text as orbital recession,
narial recession, and maxillary recession. Changes also occurred in the
relative diameter of the orbit, internal diameter of the sclerotic ring,
and in the length of the external narial aperture.

eZ.

The rates of orbital recession, narial recession and maxillary recession
between two given geological horizons were measured graphically (Fig. 5,
see also Figs. 13 and 14). Lines were drawn connecting points representing
the centres of the orbit, the centres of the external nares, and the anterior
tips of the maxillae (Fig. 5, nos. 1 - 3). Gradients of these three lines
were then measured between the various geological horizons. The gradient
of a given line between two given horizons has the units of time/per cent
recession. Thus, the reciprocal of the gradient has the units of per cent
recession/unit time and is therefore a measure of the actual rate of
recession of the character in question. For example, in Fig. 5 the gradient
of line 3 is y/x, so that the rate of maxillary recession, the reciprocal of
this gradient, is given by x/y and has the units of per cent recession per
10° years.

The rates of change in the diameter of the orbit, internal diameter
of the sclerotic ring, and length of the external narial aperture between
two given horizons are obtained directly. Thus, to compare the rate of
change in the diameter of the orbit between, say, the Middle Triassic and
Lowermost Jurassic, the difference between orbital diameters of ichthyo-
saurs at these two horizons is taken and divided by the time interval
between them (about 30 x 10° years). The units of the rate of change are
per cent change/10°® years, but, since changes in size of cranial characters
are more rapid than recession rates, it was more convenient to express size
changes in units of 4 x 10° years, the approximate separation in time
between Lower and Upper Liassic.

A graphic method similar to that used in tracing the changes in the
skull was used to investigate the evolutionary changes occurring in body
shape throughout geological time. Because growth of the skull relative
to the body is not isometric (see McGowan, 1969) and because there is a
tendency for the angle of the tail bend to increase with size, it is im-
portant to select adults for this treatment. First, all relevant data for the
shape of the body were collected for each taxon. A straight line representa-
tion of the body of each taxon was then produced (Fig. 7), the trunk
made equal to unit length. Representations thus produced were then
transferred to a time scale (see Fig. 17).

<< O7FRWUE = fa bond
aS
IR ones

ist. cervical

vertebra TRUNK pelvis THA

SKULL

Fig. 7. Conversion of significant variates of the ichthyosaurian body to linear
foci. L — length of fore fin; 1 — length of hind fin. The lengths of the fins are
shown in the diagram by the lengths of the black bars.

13

When a given evolutionary trend, e.g., orbital recession, has barely
progressed in a given taxon, that taxon is said to be primitive with
respect to that particular character. When the trend has progressed con-
siderably the taxon is said to be advanced with respect to that particular
character.

Abbreviations for institutions referred to in the text are as follows:
BMNH, British Museum (Natural History), London; sMc, Sedgwick Mu-
seum, University of Cambridge; UcMp, Museum of Paleontology, Univer-
sity of California, Berkeley; Uw, University of Wyoming, Laramie; OUM,
Oxford University Museum; QMA, Queensland Museum, Australia.

Fig. 8. Comparison of Cymbospondylus and Platypterygius. a. Cymbospondylus
petrinus UCMP 9913, Middle Triassic, Nevada, U.S.; B. Platypterygius americanus
UW 2421, Lower Cretaceous/Upper Cretaceous boundary, Wyoming, U.S.; c. P.
americanus UW 2421, right fore fin, ventral view; D. P. australis QMA F3348 (and
F3388), right fore fin, dorsal view, Lower Cretaceous, Queensland, Australia.
Scale = 10 cm.

14

The Genus Platypterygius «riz. s)

The systematics of Platypterygius and the structure of the fore fin were dis-
cussed in detail elsewhere (McGowan, in press, Contr. Geol.). For a detailed
description of the skull, the reader is referred to Romer (1968). Addi-
tional information may be found in Nace (1939, 1941) and Broili (1907).
Further description of Platypterygius here would be superfluous, but it
would be useful to compare it with certain other longipinnate genera.

ge}

Comparison of the skull cf Platypterygius with those of
Cymbospondylus, Proteosaurus, and Stenopterygius (Fig. 9)

The skull of Platypterygius is more slender than that of Cymbospon-
dylus and has a lower crown. The orbit is smaller in Platypterygius, but the
internal diameter of the sclerotic ring is about the same in both genera.
The orbit is set more posteriorly in Platypterygius, so that the snout is
correspondingly longer and the post-orbital segment shorter. The maxilla,
robust in Cymbospondylus, is slender in Platypterygius and does not
extend as far anteriorly. The size of the external narial aperture is similar
in both genera, and teeth are sparse.

Compared with the Lower Liassic genus Proteosaurus, the skull of
Platypterygius is more slender and has a much narrower lower jaw. The
orbit is smaller in Platypterygius and placed more posteriorly, but the in-
ternal diameter of the sclerotic ring is comparable in the two. The narial
aperture of Platypterygius is smaller, and the maxilla is more slender and
does not extend so far forward as in Proteosaurus.

When Platypterygius is compared with the Upper Liassic genus Stenop-
terygius, the contrast is striking. Although both skulls are slender, with
tenuous upper and lower jaws, Platypterygius has a cranial depth and
orbital diameter only half that of Stenopterygius. The internal diameter
of the sclerotic ring is also considerably smaller in Platypterygius, as is the
length of the external narial aperture. Although the maxilla is slender in
both genera, it extends farther forward in Platypterygius.

In short, Platypterygius has a slender, low-crowned skull and an orbit
that is slightly smaller even than that of Cymbospondylus. The position of
the orbit is more posterior than in any other genus, except Stenopterygius
(in which orbital position is similar); and the maxilla is slender and
extends more anteriorly than in Stenopterygius but not as far forward as
in either Cymbospondylus or Proteosaurus.

Comparison of fore fins of Platypterygius and Stenopterygius

(Fig. 10)

The fore fins of these two genera, although superficially dissimilar, have
several features in common. The fins of both taxa possess a distal carpal
series of three elements (numbers 1-3 in Fig. 10), three primary digits,
and two post-axial accessory digits. Platypterygius, however, possesses
three pre-axial accessory digits. Furthermore, in Platypterygius the distal
phalanges are not much smaller than the proximal ones and appear to
remain in contact with one another. The fore fin of Platypterygius, there-
fore, is much broader and more robust than in Stenopterygius. Although
not apparent in Fig. 10, the humerus of Platypterygius has prominent
dorsal and ventral trochanters (for the insertion of the extensor and flexor
muscles) that are not found in Stenopterygius (see Fig. 8).

16

Stenopterygius quadriscissus

Cymbospondylus petrinus

Fig. 9. Comparison of the skulls of longipinnate ichthyosaurs. Stenopterygius

quadriscissus ROM 3180, Upper Liassic (Lower Toarcian) Germany; Proteosaurus
platyodon BMNH R1158, Lower Liassic (Hettangian-Sinemurian), England; Platyp-
terygius americanus UW 2421, Lower Cretaceous (Upper Albian), Wyoming, U.S.;

Cymbospondylus petrinus UCMP 9913, Middle Triassic, Nevada, U.S. Scale = 20
cm.

LF

Fig. 10. Comparison of the fore fins of Stenopterygius and Platypterygius.
A. Stenopterygius quadriscissus BMNH R4086, right fore fin, dorsal view (figure
has been laterally inverted and therefore appears as a left fin in dorsal view);
B. Platypetrygius platydactylus left fore fin, dorsal view (after Broili, 1907,
fig: 13). Scale <4 cm:

Comparison of body proportions cf Platypterygius with these of
Proteosaurus and Stenopterygius (Fig. 11)

Difficulties arise when body proportions are compared because of the
problem of differential growth, but, provided only adult individuals are
selected, bias is at least reduced. Choice could not be exercised in the
selection of Platypterygius, however, as only one complete specimen has
been described (P. platydactylus Broili 1907, which was destroyed during
World War 11); but from its large size, it was probably adult. In Proteo-
saurus and Stenopterygius it was possible to select the largest individuals,
which were obviously adult. Proteosaurus and Stenopterygius have similar
body proportions but differ from Platypterygius. The head of Platypterygius
is considerably larger, the angle of the tail bend greater, and the post-
flexural portion of the tail is considerably shorter than in the other two
genera. Fore fins are proportionally the same size in all three genera but
are somewhat wider and more rounded in Platypterygius. The length of the
body anterior to the tail bend is longer in Platypterygius.

18

Platypterygius platydactylus

Proteosaurus platyodon

Fig. 11. Diagrammatic comparison of body proportions in Stenopterygius qua-
driscissus, adapted from Hauff (1953, pl. 9b); Platypterygius platydactylus,
adapted from Broili (1907, pl. 12); and Proteosaurus platyodon, BMNH 2003.
Scale = 50 cm.

Summary

From all comparisons, Platypterygius emerges as a mosaic of primitive
and advanced characters. In orbital and narial recession Platypterygius is
more advanced than Cymbospondylus or Proteosaurus and is comparble
to Stenopterygius. But in the size of the orbit, Platypterygius is a8 primitive
as Cymbospondylus, whereas in the reduction of the maxilla it is inter-
mediate between Cymbospondylus and Stenopterygius. Although the ex-
treme angle of the tail bend in Platypterygius is not without precedence
(the angle of the bend in Eurhinosaurus is slightly greater), the shortness
of the post-flexural section of the tail is probably unprecedented. It is the
unique structure of the fore fin, however, that distinguished Platypterygius
from all other ichthyosaurs. A reconstruction of Platypterygius platy-
dactylus is given in Fig. 12 and compared with that of Stenopterygius
quadriscissus.

19

(is. WS Ae es

Fig. 12. Comparison of body form in Lower Jurassic and Cretaceous longipin-
nate ichthyosaurs. A. Reconstruction of Platypterygius, based upon the type speci-
men of P. platydactylus (Broili, 1907, pl. 12), the tvpe specimen of P. hercynicus
(Kuhn, 1946, pl. 3), P. americanus (UW 2421) and P. australis (QMA F3348);
B. body outline of Stenoptervgius quadriscissus (adapted from Hauff, 1953,
pla 7a). Seale = 50cm.

Discussion

Platypterygius, a moderately large ichthyosaur that attained a maximum
length of about seven metres, was larger than most Liassic longipinnates
but smaller than the Middle Triassic longipinnate Cymbospondylus. The
angle of the tail bend was large, and the post-flexural portion of the tail
short. As the ichthyosaurian tail was externally symmetrical, its overall
size was proportional to the length of the post-flexural segment. The tail
of Platypterygius was thus relatively small when compared with that of a
Liassic longipinnate (Fig. 12). While swimming, the forward propulsive
thrust of the ichthyosaurian tail was accompanied by a downward thrust
produced by the epicaudal lobe. The relative magnitude of the epicaudal
downward thrust was inversely proportional to the angle of the tail bend,
whereas the forward thrust produced by the entire tail was a direct
function of its size (McGowan, in preparation). As the tail of Platyp-
terygius was small and steeply angled, it must have produced relatively
smaller downward and forward propulsive thrusts than the tail of Stenop-
terygius. The dorsal and ventral trochanters of the humerus of Platyptery-

20

gius were enormously developed and indicative of an extensive musculature.
The femur similarly possessed large dorsal and ventral trochanters, and it
seems clear that the fore fins (probably also the hind fins) contributed to
the forward propulsive thrust, thus augmenting the power output of the
tail.

Functionally there are two types of fore fins, a broad-based structure
with a low aspect ratio used as a hydroplane, and a narrow-based
structure with a high aspect ratio probably used mainly for sculling
(McGowan, in press, Cont. Geol.). Although the hydroplane, by far the
most common, is found in longipinnates and latipinnates, the narrow-
based fin is found only in longipinnates (with the possible exception of
Ophthalmosaurus) and is uncommon. Platypterygius had a fairly low
aspect ratio fin, and although it may be paradoxical that a low aspect
ratio structure should be used for sculling, parallels exist in cetaceans and
in some aquatic chelonians. Thus /nia geoffrensis, the Amazon dolphin,
has a broad but narrow-based fin that is frequently used for sculling
(McGowan, personal observation). An extensive attachment area for the
muscles of the fore fin must have been present in Platypterygius, but the
ichthyosaurian pectoral girdle is not extensively developed, especially
when compared with plesiosaurs, which used their paddles almost exclu-
sively for producing thrust (see Tarlo, 1958). But there is evidence that
the pectoral girdle was a fairly stout structure and that the scapula was
large and heavy (Nace, 1939, 1941; McGowan, in press, Cont. Geol.).
Moreover, in some vertebrates certain flexor muscles originate outside the
pectoral girdle, and the ribs could have provided an extensive attachment
area for these muscles in Platypteryegius.

In conclusion, Platypterygius used its fore fins mainly for propulsion,
augmenting the thrust of the tail. Even so, the use of the fins as hydro-
planes is not precluded, and they probably did have a dual role. The fore
fin is less efficient at producing thrust than the tail (every other stroke
of a fin is a recovery stroke, whereas every stroke of a tail is a power
stroke), and as Platypterygius probably obtained much of its propulsive
thrust from its fore fins, swimming efficiency may have been inferior to
that of its Liassic predecessors.

In Platypterygius the orbit and internal diameter of the sclerotic ring
are both small when compared with Liassic longipinnates, possibly
indicative of an eye that functioned under high light intensities. Perhaps
Platypterygius spent much of its time at, or near the surface, and did not
dive deeply in search of food. As in the Liassic longipinnates, the long
snout and sharp teeth suggest a diet of fish and cephalopods.

21

=

The Evolution of Longipinnate
Ichthyosaurs

The evolutionary sequence of the longipinnates was determined by exa-
mining and comparing the most adequately known species throughout
geological time. Unfortunately, the sequence had to be constructed from
only six species, representing but four geological horizons: Cymbospon-
dylus petrinus, Middle Triassic, Nevada; Proteosaurus platyodon, Lower
Liassic, England; Proteosaurus sp.* (a new species not yet described),
Lower Liassic, England; Stenopteryvgius quadriscissus, Upper Liassic, Ger-
many; Platypterygius australis, uppermost Lower Cretaceous, Queensland,
Australia; and Platypterygius americanus, boundary between Lower and
Upper Cretaceous, Wyoming, U.S. Data extracted from the material were
used to investigate the evolutionary changes taking place in the skull,
body proportions, and the fore fin.

Evolution of the Longipinnate Skull

Results of the graphic treatment (outlined above in Materials and Methods)
are shown in Figs. 13 and 14. When the method was discussed, it was

10° years
90

100

UPPER
CRETAC ECS

0 Platypterygius americanus

Uillisise:

thf, ~

Vibe

120 NS
NS
LOWER NV
CRETACEOUS NN
130 N
NN
140 N
\\
160 \\
N
JURASSIC \
\
‘
\N Stenopterygius quadriscissus
‘STS IKSSICuan es Ne SS Proteosaurus platyodon
LIASSIC 180 —
\\' Proteosaurus sp. *
UPPER \\
TRIASSIC 190 \\
N
200 N
NN
\\’
NN Cymbospondylus_ petrinus
MIDDLE 210 eene aay a
TRIASSIC Biever A mae

10) 20M LSOm 140 50 G0 1708 CONN GOMNHOO
% SKULL LENGTH

Fig. 13. Changes in significant cranial characters in the evolution of species of

longipinnate ichthyosaurs. A — anterior boundary of orbit; B — posterior
boundary of orbit; S-S — internal diameter of sclerotic ring; N-N — length of
external narial aperture; M — anterior tip of maxilla; 1, 2 and 3 — graphs

obtained by joining the foci representing respectively the orbital centre, narial
centre, and the anterior tip of the maxilla.

Ze

shown that problems could arise out of variations in adult skull size. The
problem arises here because Platypterygius americanus had an adult
skull more than twice as long as that of Stenopterygius quadriscissus. Was
there an actual reduction in the size of the orbit, sclerotic, and narial aper-
tures from the Liassic to the Cretaceous, or merely a proportional reduc-
tion, which would be less meaningful functionally? There was in fact a
reduction in size as well as in proportion, as is demonstrated by comparing
P. americanus with species of Stenopterygius from the Upper Liassic of
comparable adult skull size. Thus, if S. disinteger (Hauff, 1953, pl. 18a)
and S. acutirostris (?) (Hauff, 1953, pl. 19a), with skull lengths of 97 cm
and 153 cm, respectively, are compared with Platypterygius americanus
(UW 2421), skull length 122 cm, and if skulls of all three are reduced to
unit size, it is seen that the orbital, sclerotic, and narial apertures of the
two Upper Liassic taxa are significantly larger. An actual decrease in the
finite sizes of the orbital, sclerotic, and narial apertures thus occurred in
ichthyosaurs from Liassic to Cretaceous times.

10° years
90

100

UPPER -
CRETACEOUS

Platypterygius

LOWER
CRETACEOUS ;

JURASSIC

Stenopterygius
Proteosaurus

LIASSIC 480

UPPER
TRIASSIC 190

Cymbospondylus

MIDDLE 210

N—N
TRIASSIC M

naris

1Ofen 120). 230" 840" 4 50! 760: 70 60-90 + 100
% SKULL LENGTH

Fig. 14. Changes in significant cranial characters in the evolution of genera of

longipinnate ichthyosaurs. A — anterior boundary of orbit; B — posterior boun-
dary of orbit; S-“S — internal diameter of sclerotic ring; N-N — length of exter-
nal narial aperture; M — anterior tip of maxilla; 1, 2 and 3 — graphs obtained

by joining foci representing respectively the orbital centre, narial centre, and
anterior tip of the maxilla.

a3

Leaving aside the well-marked hiatus at the Lower Jurassic, shortly to be
discussed, several generalisations may be made from data given in Figs.
13 and 14. Throughout the evolution of the ichthyosaurs, there was a well-
marked tendency for the orbit and external narial aperture to move poste-
riorly. Also migrating posteriorly was the anterior boundary of the maxilla,
resulting from a reduction in the length of the maxilla. Accompanying
these changes in relative position of the orbit and external naris were
changes in their relative sizes.

From the Middle Triassic to the Lower Liassic (early Jurassic), the
diameter of the orbit increased by about 25%, with a corresponding
increase of about 20% in the internal diameter of the sclerotic ring.
From the Lower Liassic to the Upper Liassic the diameter of the orbit
was constant, but the internal diameter of the sclerotic ring increased
by about 40%. From the Upper Liassic to the Cretaceous, the orbital and
internal sclerotic diameters decreased by about 40%, both becoming quite
small. The size of the external narial aperture increased by more than 60%
from the Middle Triassic to the Lower Liassic, but from the Lower Liassic
to the Cretaceous, there was an almost steady decrease of about 30%.

POSSIBLE SIGNIFICANCE OF CRANIAL CHANGES

The posterior migration of the eye and external narial aperture during the
evolutionary history of the ichthyosaurs increased the relative length of the
snout. As the snout is the dentulous portion of the skull, the increase in
length may have been correlated with increasing the food-gathering
capacity of the head. That the increase in snout length was correlated with
improving the streamlining of the body is also possible. As the eyes and
olfactory organs migrated, the optic and olfactory lobes of the brain also
moved posteriorly; consequently, the brain case became relatively shorter,
but the significance of this is not known. The reduction in size of the maxilla
does not appear to have had any discernable functional significance. The
margin of the upper jaw was formed from the premaxilla and maxilla, but the
contribution of the latter was minor. As ichthyosaurs evolved, the maxilla
became progressively reduced, a tendency reflected in the posterior shift
of the anterior tip of the maxilla (Figs. 13, 14).

Changes in orbital and internal sclerotic ring diameters are treated
together because of their close relationship. Before discussing the signifi-
cance of these changes, however, the respective functions of these struc-
tures need consideration. The orbit is the receptacle of the eye and is
directly proportional to the size of the eye. But the size of the eye is not
proportional to the size of the pupil, and it is the size of the pupil that
gives information on how the eye was used. Direct evidence of pupil size
can be obtained from the sclerotic apparatus. Underwood (1970) discussed
the significance of the presence of the sclerotic sulcus in supporting the
cornea in those recent reptiles that use their eyes above and beneath the
surface of the water, and McGowan (in press, Bull. Br. Mus.) demon-
strated that such a structure was absent from the sclerotic ring of Liassic
latipinnates. A sulcus was also absent from the Liassic longipinnates that I
examined and from the longipinnates Cymbospondylus petrinus (UCMP

24

—------ ~~

9913) and Platypterygius americanus (UW 2421). I conclude that a
sulcus was probably absent from all ichthyosaurs, and that they used
their eyes exclusively for underwater vision.

Except during the Liassic, changes in the diameters of the orbital
and sclerotic apertures were of the same magnitude. During the Liassic
the orbital diameter remained constant, whereas the size of the sclerotic
aperture increased greatly. Consequently the size of the pupil increased
without any increment in the size of the eye. Perhaps Upper Liassic
longipinnates inhabited deeper waters where light intensities were lower,
or perhaps the turbidity of the sea was greater than it had been pre-
viously. From the late Liassic onward, orbital and sclerotic apertures
gradually decreased in size at approximately equal rates. The size of the
eye and of the pupil were thus decreasing, again possibly indicating
changing habits or environments. Certain features of the post-cranial
skeleton of Platypterygius, discussed above, suggest that the animal did
not dive deeply, which could account for the relative small size of the
sclerotic aperture.

The external narial aperture has two functions, respiration and olfaction.
McGowan (in press, Bull. Br. Mus.) suggested that the function of the
olfactory organ of Liassic latipinnates may have been to determine the
quality of the water rather than that of the inspired air, and that water
was probably continuously monitored while the animal was submerged. I
also tentatively suggested that the median-dorsal internasal foramen, a
structure seen in a number of latipinnate ichthyosaur skulls from the
Liassic, might have functioned in respiration. Moreover, I recently iden-
tified the same structure in some Liassic longipinnates (Proteosaurus platyo-
don ROM 7972, BMNH R1155, BMNH R215) and it could well be that this
foramen was consistently present, at least in Liassic ichthyosaurus. During
the interval between the early Triassic and the beginning of the Jurassic, the
size of the narial aperture increased, and was followed by a rapid reduction
in the early part of the Jurassic, a trend that continued, though at a reduced
rate, till late Cretaceous times (Figs. 13, 14). What was the functional signi-
ficance of these changes? From the standpoint of respiration it is reasonable
to assume that there was a certain minimum size of the narial aperture for
normal respiratory exchange. The narial aperture of Cymbospondylus was
smaller than in any pre-Cretaceous ichthyosaur but must have been ade-
quate for respiratory exchange. As Cymbospondylus was probably the
largest ichthyosaur in body mass, the increase in size of the narial aperture
after the Middle Triassic may not have been correlated with respiratory
function (not unless post-Triassic longipinnate ichthyosaurs were consi-
derably more active than Cymbospondylus, and required correspondingly
larger nares for the increased rate of respiratory exchange). The increment
in narial size from the Triassic to the beginning of the Jurassic may there-
fore have been correlated with increased olfactory acuity rather than with
an increased respiratory demand. The inverse relationship between the
size of the narial and sclerotic apertures during the Liassic is probably
significant. Could the Liassic have been a period of transition from an

25

olfactory to a visually-dominated sensory system? That the decrease in
narial aperture was correlated with the appearance of the internasal fora-
men is also possible.

RATES OF CHANGE IN CRANIAL EVOLUTION OF LONGIPINNATE ICHTHYOSAURS
Although the fossil record is far from complete, it is possible to obtain
some idea of the rates at which evolutionary processes occurred in the
longipinnate skull (see Materials and Methods), and results are shown in
Figs. 15 and 16. Three generalisations emerge from this study: (1)
During a given interval of time, the individual characters of the skull did
not change at the same rate. For example, during the interval from the
Triassic to the early Liassic the increase in narial length was about twice
as fast as the rate of increase in orbital diameter (Fig. 15). (2) The
direction of a given change sometimes reversed. Thus, the size of the
external naris increased during pre-Liassic times, then decreased (Fig.
15). Similarly, the trend toward an increase in orbital and internal
sclerotic diameters was reversed in post-Liassic times. (3) The most rapid
rate of change occurred during the Liassic, an interval of some 4 million
years, whereas the slowest rate of change occurred during the interval
from late Liassic to late Cretaceous, about 65 million years.

Different cranial characters not only evolved at different rates during
the geological history of the longipinnates, but a given structure was
sometimes acted upon by two different evolutionary forces (Figs. 15,

50

SCLEROTIC+

7./4x108 YEARS

EXT. NARES+

Xie z
Naees SCLEROTIC

Middle Triassic — Lower} Lower Liassic — Upper | Upper Liassic — Upper
Liassic Liassic Cretaceous

Fig. 15. Rates of cranial evolution with respect to percentage change in diameters
of the orbit and internal sclerotic ring, and in the length of the external narial
aperture, measured throughout each geological range. -+- = imcrease in Size;
— = decrease in size.

26

3-0 EXT. NARES MAXILLA

ORBIT MAXILLA ext. MAXILLA
EXT. NARES orpit NARES
Middle Triassic — Lower] Lower Liassic — Upper | Upper Liassic — Upper
Liassic Liassic | Cretaceous

Fig. 16. Rates of cranial evolution with respect to the recession of the orbit,
external narial aperture, and the anterior tip of the maxilla measured through-
out each geological range.

16). For example, the diameter of the orbit remained unchanged during
the Liassic interval but its rate of posterior recession was maximal.
Furthermore, structures considered to have a close functional relationship
did not always evolve at the same rate. Thus, the diameters of the orbit
and internal sclerotic ring changed at about the same rate during the
pre- and post-Liassic intervals but were dissimilar throughout the Liassic
(Fig. 15). The well-marked trends toward an increase in orbital and
internal sclerotic diameters, and in the size of the narial aperture,
reversed after the Liassic (Fig. 15), and by the Cretaceous the three
structures were relatively smaller than they had been even during the
Triassic. These changes undoubtedly represent an example of the
reversal of an evolutionary trend during a lineage. Not that such a
reversal is without precedent in ichthyosaurs, for it was shown that the
longipinnate fin, having undergone a reduction to four digits by the late
Triassic, increased to five or more digits by the Liassic (McGowan,
1972). The fore fin of Platypterygius, with a total of eight digits, represents
an extreme expression of this particular tendency. That the rate of cranial
evolution was fastest during the Liassic and slowest during the post-
Liassic interval has been documented and is probably indicative of an
overall slowing of the evolutionary process. This conclusion is corro-
borated by the homogeneity of body structure in Cretaceous ichthyosaurs.

Evolution of the Longipinnate Body

Because of the incompleteness of most of the skeletons and of the small
sample sizes, conclusions drawn from available data are only tentative.
Results are shown in Fig. 17, and five generalisations may be made:
(1) the skull increased in length relative to the body; (2) the angle of
the tail bend increased; (3) the pre-flexural segment of the tail decreased
in length; (4) the post-flexural segment of the tail decreased in length;
and (5) the length of the fore fins decreased. As only four species were

27

10° years
1st. cervical
vertebra

100 pelvis
UPPER SKULL TRUNK TAIL
CRETACEOUS (unit length)
aarG ue Platypterygius platydactylus
120
LOWER
CRETACEOUS
140
JURASSIC
160
Stenoplerygius quadriscissus

Proteosaurus platyodon

UPPER
TRIASSIC

MIDDLE
TRIASSIC

Cymbospondylus petrinus

220

Fig. 17. Changes in body proportions of longipinnate ichthyosaurs during evolu-
tion. Data based on conversion of variates to linear foci, with trunk lengths the
same. Black bars represent the icngth of fore and hind fins. Data for Cyimbos-
pondylus petrinus ere incomplete.

Platypterygius platvdactylus from Broili C907. -ple12).:
Stenopterygius quadriscissus from Hauff (1953, pl: 9b):
Proteosaurus platyodon BMNH 2913.

Cymbospondylus petrinus UCMP 9950.

studied, cross references were made to six other Upper Liassic longi-
pinnates for which data was available (Fig. 18). How well do available
data substantiate these generalisations?

1. Increase in length of skull

Skulls of the six other Upper Liassic longipinnates (Stenopterygius mega-
lorhinus, S. megacephalus, S. crassicostatus, S. disinteger, S. acutirostris (2),
and Eurhinosaurus longirostris, all from Hauff, 1953) were larger (relative
to the body) than that of Cymbospondylus petrinus. Furthermore, with the
exceptions of Stenopterygius quadriscissus and S. crassicostatus, the Upper
Liassic longipinnates had larger skulls than the Lower Liassic Proteosaurus
platyodon. The increase in relative skull length is thus well documented for
the Middle Triassic-Upper Liassic period but is less well substantiated during
the post-Liassic period. Thus, although skulls of S. guadriscissus, S. crasst-
costatus, and S. megalorhinus were smaller than that of Platypterygius
platydactylus, the reverse was true for Eurhinosaurus longirostris and S.
disinteger. Moreover, skulls of S. acutirostris (?) and S. megacephalus were
comparable in size with that of Platypterygius platydactylus. It is con-
cluded that the size of the skull increased relative to the body, and that
this trend was most marked during the Middle Triassic-Lower Liassic

268

SKUIEL 1st cervical TRUNK pelvis TAIL

verteb (unit length)
Stenopterygius megalorhinus 4
Stenopterygius megacephalus 32
Stenopterygius crassicostatus 42

S

Stenopterygius disinger \40

} x.
Eurhinosaurus longirostris \55 ~

\
Stenopterygius acutirostris N41

Fig. 18. Data of body prceportions for Upper Liassic longipinnates based on
conversion of variates to linear foci. Black bars represent the lengths of fore
and hind fins.

S. megalorhinis from Hauff (1953, pl. 13a).

5S. megacephalus from Hauff (1953, pl. 12a).

Ss chassicosiarus trony Hanif (1953. pl. 13b).

S. aisimeser irom Hautt (1953. pl. 18a).
Eurhinosaurus longirostris from Hauff (1953. pl. 21b).
S. acutirostris (7?) from Hauff (1953, pl. 19b).

period. From the Upper Liassic to the Upper Cretaceous the tendency
was apparent but less pronounced.

2. Increase in the angle of the tail bend

The angle of the tail bend was greater in Platyvpterygius platydactylus than
in any of the Upper Liassic longipinnates except Eurhinosaurus. Further-
more, with the exception of Steopteryvgius megalorhinus, angles were
greater in the Upper Liassic longipinnates than in the Lower Luiassic
species Proteosaurus platyodon. The conclusion is that there was an
incresse in the angle of the tail bend (a similar tendency occurred in the
latipinnates).

3. Reduction in the pre- and post-flexural segments of the tail

All Upper Liassic longipinnates had longer pre- and post-flexural tail
segments than did Platypterygius platydactylus and. with the exception of
Eurhinosaurus, had shorter post-flexural segments than Proteosaurus
platyodon. In the lengths of the pre-flexural segments, however, three of
the Upper Liassic longipinnates (S. disinteger, S. acutirostris(?), E. longi-
rostris) exceeded P. platyodon. J therefore conclude that there was a
general tendency for the tail segments to become shorter.

4. Reduction in fore fin

Although differences between the relative lengths of the fore fin of
Proteosaurus platyodon, Stenopterygius quadriscissus, and Platypterygius
platydactylus were not great, and although it is not certain whether the
fore fin of the holotype of P. platydactylus was complete (Broili, 1907,

29

fig. 13), the smallest fin was likely that of P. platydactylus. Furthermore,
the fore fin in each of the six Upper Liassic longipinnates was relatively
longer than in P. platydactylus. The trend toward a reduction in the
length of the fore fin therefore appears to be substantiated.

Evolution of the Longipinnate Fore Fin

The lack of knowledge of the structure of the fore fin of Triassic ichthyo-
saurs was mentioned previously, and although the structure of the fore
fin of Cymbospondylus petrinus (Middle Triassic) is not well known,
that of the Upper Triassic species Merriamia zittelli is (Merriam, 1903).
The fore fin of Merriamia consists of only three primary digits and a
remnant of a fourth (Fig. 1£). The proximal carpal series comprises
radiale, intermedium, and ulnare, and the distal carpal series contains
three elements. From available evidence the fins of Toretocnemus and
Delphinosaurus, both from the Upper Triassic, were probably similar
(Merriam, 1908). Thus, by the late Triassic the longipinnate fore fin
had undergone reduction from a previous pentadactyl condition (for a
discussion of fore fins see McGowan, 1972). Information on fore fins
is lacking until the early Liassic of England, where a few longipinnate
species are recorded. Again, the fore fin possesses only three primary
digits, but the individual elements have lost much of their earlier primi-
tive appearance. The radius and ulna are relatively short, and the carpals
and phalanges have become more oval. Several elements of the pre-axial
series, including the radius, have emarginate free borders, and the reten-
tion of this seemingly primitive character (Lydekker, 1889, p. 70) is
found throughout the Jurassic longipinnates.

By the late Liassic most longipinnates (Fig. 1F, G) had a digital count
of five, due to the possession of accessory digits. All had retained a distal
carpal series of three elements, and the total number of elements in each
digit had increased. Probably all of the five-digits fins were broad-based
(confirmed in several where the soft part of the fin is preserved as a
carbonaceous film) and used exclusively as hydroplanes. The three-digit
fins, on the other hand, tended to be long and narrow, as in S. acutiros-
tris(?) and E. longirostris (Fig. 1F), and were probably used as paddles.

Nothing more of fore fin evolution is known until the Cretaceous and
the appearance of Platypterygius. The acquisition of accessory digits had
gone further in Platypterygius than in any other longipinnate and, for the
first time in the known history of the Ichthyosauria, pre-axial accessory
digits exceeded one in number (Fig. 1H). Although the free borders of the
radius and succeeding radial elements were not emarginate, many bones
possessed small emarginations on their proximal and distal borders (see
Nace, 1939; Nace, 1941; McGowan, in press, Contr. Geol.). As noted
previously, the fin of Platyptervgius was considered to have been a broad
but narrow-based sculling structure. In the regular shapes of the phalan-
ges and for their orderly arrangement in vertical rows, the fin was striking.

30

Size Ranges in Longipinnates

Although the maximum attainable size in a given species cannot be pre-
dicted without a sufficiently large sample, taxa upon which the cranial
discussion was based are each represented by at least three specimens
and by many more in Upper Liassic species. Consequently, maximum
attainable sizes could be estimated for 14 longipinnate species, which are

Latipinnates

oaAaN OO & OW MY =| LOngipinnates

—_—
ais (©

ig @i@) ac)
|
|
|
1
N

O 20 40 60 60" 1007120 140 160
SKULL LENGTH cm

19. Approximate maximum sizes of some ichthyosaur species determined
from skull length. Latipinnates: A) Ophthalmosaurus; B) Ichthyosaurus cony-
beari; C) I. breviceps; D) I. communis; E) Mixosaurus cornalianus. Longipin-
nates: 1) Platypterygius australis; 2) P. americanus; 3) P. platydactylus; 4) Eur-
hinosaurus logirostris; 5) Stenopterygius megacephalus; 6) S$. megalorhinus;
7) S. crassicostatus; 8) S. disinteger; 9) S. acutirostris (2); 10) S. quadriscissus;
11) Proteosaurus sp.* (undescribed new species from the Lower Liassic of Eng-
land); 12) Proteosaurus sp. (a second undescribed new species); 13) Proteosaurus
platyodon; 14) Cymbospondylus petrinus.

Fig.

|

compared with five latipinnate species (Fig. 19). In general, longipinnates
always tended to be large and were much larger than their latipinnate
counterparts. Based only on skull lengths Cymbospondylus petrinus was
not the largest longipinnate, for its skull was only about half as long
relative to its body as that of Proteosaurus platyodon. But in total body
length, C. petrinus was probably the largest known ichthyosaur, reaching at
least 10 metres. P. platyodon probably did not greatly exceed 9 metres in
length. Except for Stenopterygius acutirostris (?), a veritable giant of nearly
8 metres (Hauff, 1953, pl. 19), the Upper Liassic longipinnates were
smaller than P. platyodon, usually less than 4 metres in length. Relative
to its body, the skull of Platypterygius platydactylus was slightly more than
twice as large as that of C. petrinus (see Fig. 17). Although the skull of
P. platydactylus was about the same size as that of C. petrinus, in total
body length P. platydactylus was probably only about 7 metres but, even
so, larger than any Upper Liassic longipinnate. Complete data from the
Upper Triassic longipinnates is wanting, but species of the genus Shasta-
saurus were probably the same size as C. petrinus, whereas the size of
Delphinosaurus perrini was comparable to that of Upper Liassic species,
e.g., Stenopterygius quadriscissus. Merriamia zittelli and Toretocnemus
californicus were both small species, probably only one metre long, and
were therefore the smallest of all longipinnates. It is conceivable, how-
ever, that the material referred to these two species represents immature
individuals.

I

Conclusions

When ichthyosaurs first appeared in the fossil record, in the Middle
Triassic, they were already highly adapted to an aquatic mode of life
and were represented by two distinct stocks, the latipinnates and the longi-
pinnates. Latipinnate ichthyosaurs always tended to be moderate in size,
gencrally not exceeding 4 metres in total body length, whereas the longi-
pinnate stock produced some giants, including Cymbospondylus petrinus
(Middle Triassic) the largest of ichthyosaurs with an estimated adult
body length of about 10 metres. From the available evidence it appears
that the longipinnates survived to the late Cretaceous, whereas the lati-
pinnates had become extinct by the late Jurassic (Fig. 20). The longi-
pinnates appear to have evolved most rapidly during the early part of the
Jurassic, but after this period the evolutionary process slowed to its lowest
rate. The changing rates of evolution during the history of the longi-
pinnates may have been correlated with changes in food availability. The

—+

ACOMOLS AMDO

=

Ichthyosoure®

O-NONS>—DA |O-HDHHSVDCeH

Z>-—-< DMV

Fig. 20. A phylogeny of the Ichthyosauria.

33

diet of Triassic ichthyosaurs is unknown, and they may have fed on
fishes and cephalopods (e.g., belemnites, ammonites, and squids, all of
which were becoming increasingly more abundant) and perhaps also on
other marine reptiles, e.g., placodonts. Fishes and cephalopods were not
as abundant during the Triassic as they were to become by the Liassic,
and the relatively low rate of evolution and paucity of ichthyosaurian re-
mains during the Triassic could be attributable to limited food resources.
Liassic ichthyosaurs fed on holostean fishes of the genera Lepidotes,
Dapedium, and Pholidophorus, as well as on belemnites and ammonites
(Pollard, 1968; McGowan, in press, Bull. Br. Mus.). The Liassic seas
abounded with these prey, and the Liassic was therefore a time of plenty
for ichthyosaurs. The rapid rate of evolution and the abundance of
ichthyosaur remains during the Liassic may therefore be attributable to
optimum environmental conditions. By Cretaceous times the marine fauna
had changed considerably, but there is no reason to suppose that the
seas were any less bountiful than they had been during the Liassic.
Teleosts had largely replaced holostean fishes, and the advanced sela-
chians had undergone an adaptive radiation. Ammonites and squids were
even more plentiful than they had been during the Liassic and, assuming
a diet of fishes, belemnites, ammonites, and squids, opportunities for
ichthyosaurs were probably no less during the Cretaceous. Why, then,
was the rate of evolution during the Cretaceous so low, and why were
ichthyosaurs relatively scarce? A failure of the fossil record might be a
plausible explanation, for Lower Cretaceous marine deposits are not
abundant. But the complete absence of ichthyosaurs from the rich
Niobrara chalk deposits of the Upper Cretaceous of the western U.S.
(Romer, 1966) and the remarkable homogeneity of all Cretaceous
ichthyosaurs thus far recovered (McGowan, in press, Contr. Geol.) give
convincing evidence that the group was indeed in decline, and an explana-
tion must be sought.

The slowing of evolutionary rates during the post-Liassic period may
be correlated with the prevalence of more constant environmental con-
ditions. By the Cretaceous selachians had completed much of their evo-
lution and were essentially similar to the present fauna. Ammonites,
belemnites, and squids still abounded, and populations were relatively
stable. But the bony fish fauna had not attained stability because the
teleosts were diversifying rapidly and replacing the holosteans. There are
therefore some grounds for concluding that a greater stability existed in
the marine environment during the Cretaceous. Although greater envi-
ronmental stability may have been a contributing factor to the slowing rate
of evolution, it was probably not the only one. Perhaps some genetic
changes occurred during the long post-Liassic interval, thereby reducing
variation. This suggestion is at present the most attractive one because
it leads to an hypothesis for the ultimate extinction of the ichthyosaurs
which will now be examined.

As far as can be established from the fossil record, Platypterygius was
the sole representative of the Ichthyosauria in Cretaceous times. There is
no evidence of any other genera, and species referred to Platypterygius

34

appear to be very similar. Adults probably reached body lengths of about 7
metres, and the propulsive thrust of the small tail was augmented by the
the fore fins (probably also the hind fins), perhaps a less efficient
swimming mechanism than that of Liassic longipinnates. Whereas a loss
of swimming efficiency may have been inconsequential during the early
Cretaceous, it may have become critical by late Cretaceous times when
ichthyosaurs were faced with increased competition from plesiosaurs and
new competition from mosasaurs. Plesiosaurs derived their propulsive
thrust from their fore and hind paddles, the extensive musculature of
which is reflected in the large size of the girdles. Mosasaurs, on the
other hand, derived their propulsive thrust from lateral undulations of the
elongate body and deep tail. (The paddles were almost certainly pressed
against the body during forward motion, as in extant swimming lizards,
and were used only when stationary.) Thus, both plesiosaurs and mosa-
saurs employed only one method to produce a forward thrust, while
Platypterygius practised a division of labour. Consequently, Platypterygius
may have been outclassed by its reptilian contemporaries in speed, ma-
noeuverability, or diving abilities. Nonetheless, caution should be exercised
when drawing conclusions on locomotory prowess from evidence of the
skeleton alone. The aquatic performance of the otter (Lutra), for example, is
incredible, but there is little or no skeletal evidence to indicate that the
otter is so agile. With reservations, then, it is concluded that Platypterygius
might have been less efficient in its locomotory abilities than the plesio-
saurs and mosasaurs.

Mosasaurs did not appear until late in the Cretaceous and then under-
went a rapid proliferation. At that time the ichthyosaurs were apparently
at an evolutionary standstill. Perhaps the reason for the eventual demise
of the ichthyosaurs should be sought in a loss of adaptive plasticity at a
time when a new competitor was undergoing adaptive radiation. Although
this hypothesis may satisfactorily explain the final extinction of the ichthyo-
saurs, the disappearance of the smaller longipinnates (e.g., Stenopterygius
quadriscissus) that had been so successful during the early Jurassic
remains unresolved. One would imagine that there has always been an
ecological niche for such animals, one today occupied by the dolphins
(Delphinoidea). Perhaps stenopterygians were extant during the Creta-
ceous but lived in environments different from those occupied by the
platypterygians and were excluded from the fossil record. A parallel exists
in the Upper Liassic; remains of latipinnates are apparently absent, even
though they are found in Lower Liassic and in Upper Jurassic deposits.
Only more Cretaceous material will resolve this point.

be)

Acknowledgments

I wish to thank Dr. Alan Charig and Mr. Cyril Walker of the British
Meseum (Natural History), and Dr. C. L. Forbes of the Sedgwick Mu-
seum, Cambridge, and Mr. J. M. Edmonds and Mr. H. P. Powell of the
Oxford University Museum for their generous help in making material
available for study. I thank Drs. M. Wade and A. Bartholomai of the
Queensland Museum for generously supplying me with material, photo-
graphs, and a considerable amount of data. I wish also to thank Drs. K.
N. Bell of the National Museum of Victoria and D. Rhodes of the Uni-
versity of Western Australia for supplying photographs and Drs. D. Mer-
rilees of the Western Australian Museum and A. Ritchie of the Australian
Museum, Sydney, for supplying information. For the loan of material and
warm hospitality, I thank Drs. S. P. Welles and J. T. Gregory of the
University of California, Berkeley, and Paul McGrew of the University of
Wyoming.

Thanks are also due to Miss Lynda Smith and Mrs. Judy Allan for
their continued assistance, Mrs. Sophie Poray for drawing the figures,
Miss E. Dowie and Mrs. P. Trunks of the RoM Library, and Mr. L.
Warren and staff of the ROM Department of Photography. I also thank
Dr. A. G. Edmund for reading the manuscript and for advice.

36

Literature Cited

ANDREWS, C. W.
1910 A descriptive catalogue of the marine reptiles of the Oxford clay.
Part 1. London, Printed by Order of the Trustees of the British Museum.
205 pp.
BROILI, F.
1907 Eine neuer Jchthyosaurus aus der norddeutschen Kreide. Palaeonto-
graphica, vol. 54, pp. 139-162.
1931 Mixosauridae von Timor. Wet. Meded. Dienst Mijnb. Ned.-Oost-Indié,
no; 17, pp. 3-10.
HAUFF, B.
1955 Das Holzmadenbuch. Ohringen, Verlag der Hohenlohe’ schen Buch-
landlung. 54 pp.
HUENE, F. VON
1936 Ichthyosaurierreste aus Timor. Zentbl. Miner. Geol. Paldont., abt. B,

pp. 327-334.
HULKE, J. W.
1873 Memorandum on some fossil vertebrate remains collected by the

Swedish expeditions to Spitzbergen in 1864 and 1868. Bih. K. Svenska
VetenskAkad. Handl., n.s., band 1, no. 9, pp. 2-11.
KUHN, O.

1946 Ein skelett von Jchthvosaurus (Platypterygius) hercynicus n. sp. aus dem

Aptium von gitter. Ber. Naturf. Ges. Bamberg. vol. 29, pp. 69-82.
LYDEKKER, R.

1889 Catalogue of fossil Reptilia and Amphibia in the British Museum
(Natural History). Part 2. Containing the orders Ichthyopterygia and
Sauropterygia. London, Printed by Order of the Trustees of the British
Museum. 307 pp.

MCGOWAN, C.

1969 The cranial morphology and interrelationships of the Lower Liassic
ichthyosaurs. Ph. D. thesis, University of London. 498 pp.

1972 The distinction between latipinnate and longipinnate ichthyosaurs. Life
Scie, Occ, Paps; R- Ont. Mus!, no. 20; pp. 1-12.

In The cranial morphology of the Lower Liassic latipinnate ichthyosaurs

press of England. Bull. Br. Mus. Nat. Hist. Geol.

In The systematics of Cretaceous ichthyosaurs with particular reference

press to the material from North America. Contr. Geol.

MCWAE, J. R. H., P. E. PLAYFORD, A. W. LINDNER, B. F. GLENISTER and B. £. BALME
1958 The stratigraphy of Western Australia. Carlton, Victoria, Melbourne
University Press on behalf of the Geological Society of Australia.
161 pp.
MERRIAM, J. C.
1903 New Ichthyosauria from the Upper Triassic of California. Univ. Calif.
Publis: Bull: Dep: Geol., vol. 3, no: 12, pp: 249-263.

1906 Preliminary note on a new marine reptile from the Middle Triassic of
Nevada. Univ. Cali.’ Publis: Bull, Dep. Geol.,-vol. 5, no: 5, pp- 75-79.
1908 Triassic Ichthyosauria, with special reference to the American forms.

Mem. Univ. Calif., vol. 1, no. 1, pp. 1-196.
1910 The skull and dentition of a primitive ichthyosaurian from the Middle
Triassic. Univ. Calif. Publs. Bull. Dep. Geol., vol. 5, no. 24, pp. 381-390.
MERRIAM, J. C. and H. C. BRYANT
1911 Notes on the dentition of Omphalosaurus. Univ. Calif. Publs. Bull. Dep.
Geol., vol. 6, no. 14, pp. 329-332.

hd

INAGE Rem le.

1939 A new ichthyosaur from the Upper Cretaceous Mowry Formation of
__ Wyoming. Am. J. Sci., vol. 237, no. 9, pp. 673-686.
1941 A new ichthyosaur from the Late Cretaceous, northeast Wyoming.

Am. J. Sci., vol. 239, no. 12, pp. 908-914.
POLLARD, J. E.
1968 The gastric contents of an ichthyosaur from the Lower Lias of Lyme
Regis, Dorset. Paleontology, vol. 11, pt. 3, pp. 376-388.
ROMER, A. S.
1966 Vertebrate Paleontology. 3d. ed. Chicago, University of Chicago Press.
468 pp.
1968 An ichthyosaur skull from the Cretaceous of Wyoming. Contr. Geol.
VOL 7, no. If pp. 27-41.

RUSCONI, C.
1948 Ictiosaurios del Jurasico de Mendoza (Argentina). Revta. Mus. Hist.
Nat. Mendoza, vol. 2, entregas 1-2, pp. 17-162.
TARLO, L. B.
1958 The scapula of Pliosaurus macromerus Phillips. Palaeontology. vol. 1}.

pt. 3, pp. 193-199:
TEICHERT, C., and R. S. MATHESON
1944 Upper Cretaceous ichthyosaurian and plesiosaurian remains from Wes-
tern “Australia, Aust: Ji. Sci5.vol. 6, no: 6,.pp. 167-170:
UNDERWOOD, G.
1970 The eye. Jn Gans, C. ed. Biology of the Reptilia. vol. 2. Morphology B.
London, Academic press, pp. 1-97.
WIMAN, C.
1910 Ichthyosaurier aus der Trias Spitzbergen. Bull. Geol. Instn. Univ.
Upsala, vol. 10, pp. 124-148.

1916 Notes on the marine Triassic reptile fauna of Spitzbergen. Univ. Calif.
Publs: Bully Dep; Geol, vol. 10; ne. 5, pp. 63-73:
1929 Eine neue marine Reptilien-Ordnung aus der Trias Spitzbergens. Bull.

Geol. Instn. Univ. Upsala, vol. 22, no. 5, pp. 183-196.
1933 Uber Grippia longirostris. Nova Acta R. Soc. Sci. Upsal., ser. 4, vol. 9,
no. 4, pp. 1-19.

38

7 = _ abt = 7 we 7 : i = “ 7 ~~) ee OV eee 4 a om
; J a | : >» ; 2
é - { ' ; = . an | rn t
3 ) - \ j
r fa j : ‘ ?
= 2 re & ‘ ; 7
oo x a ie ‘ ; 4 aaa
2
‘ ’
‘ i
*\ . ‘ 5 ; :
oA :
\ f : ~ ©€
a : : 7
/ a}
an h ¢ J
ad
al i
~ ns y — 7
- / a }
BAZ i)
@ % Ss
A 2 Vs,
‘ +
ih E S
2 = L '
: } = 7

x
'
t
'
-
~
—

: a at
— ao

a

‘
is
x
x
\
v4
{
¥
{
o,
.
'

pelts 5
=

.

oY «
}

> bes u c ‘ Pane i a : oF , ~* :

. . ees

“ pcr Ge A ae uae i , ex ead .
AT Oh iy eh, Sk On ee Plena NSS ie ee Oe ee