Sabre-like canines have evolved several times in mammals, not only among carnivorans, but also
among other eutherians and even metatherians. The changes in skull morphology and non-canine dentition
are remarkably similar in this odd mix of carnivorous mammals. These changes has been observed in
arctocyonids such as Anacodon, creodonts such as Apataelurus and Machaeorides, nimravids, such as
Eusmilus and Hoplophoneus, barbourofelines such as Barbourofelis and Sansanosmilus and in many
machairodonts, e.g., Smilodon. Among the metatherian borhyaenids the genus Thylacosmilus evolved some
of the most extreme sabres of all carnivores. The sabres themselves differ among genera, but two general
morphotypes were present. Dirk-toothed, with long slightly and crenulated canines, and scimitar-toothed
with shorter and more strongly serrated canines. Whether short or long sabres, the skull morphology
changed in numerous ways. Ubiquitous changes included reduction in dentition, enlargement of the
carnassials, reduction of the coronoid process and reduced width of the zygomatic arches, among others.
An often overlooked feature of many sabretooths are the presence of carnassial rotation, creodonts as well
as felids, in which the teeth in the lower tooth-row rotate to minimize inter-carnassial distance. The
presence of attrition (thegosis) on the carnassials but not on the sabres, is undoubtedly linked to the rotation
of the lower carnassials. The functional morphology of this unusual trait is discussed.

The virtual Eusthenopteron: inside the head of a Devonian lobe-fin with CT
Computerised Tomography (CT) is a powerful tool for visualising the internal structure of objects
in three dimensions using serial X-ray "slices", and has in recent years found increasing use in the study of
fossils. Compared to a grinding series it is incomparably quicker, non-destructive, and allows �weightless�
computer reconstruction of the objects in virtual space.

In collaboration with the University of Texas and Parc de Miguasha, we are reconstructing the
skull of the Devonian sarcopterygian Eusthenopteron from a CT scan series. The aim is to complement
Jarvik�s (1980) landmark study of the skull of this fish by using a different technique and a much larger
individual. Our results largely confirm Jarvik�s interpretation, with the exception of a few differences such
as a more posteriorly positioned external nostril, differently proportioned nasal capsules and a large
fontanelle in the roof of the otoccipital. However, we find that the palatoquadrate, hyomandibular and
braincase have been subtly disarticulated from each other in Jarvik�s reconstructions, in ways that obscure
the precise functional relationships between these structures. This is probably due to their having been
modelled individually in wax, and only then repositioned into (approximate) life positions with the help of
metal armatures. Our CT-based skull model thus allows us to take a fresh look at the cranial mechanics of
Eusthenopteron, and to more accurately infer the position of unpreserved but important structures such as
the spiracular cleft.

Arthrodires (and placoderms in general) show a great deal of morphological diversity by the Late
Devonian. Much of this diversity is seen in the skull and jaws, indicating a variety of possible feeding
niches, from durophagous bottom dwellers to piscivorous cruisers. Current phylogenies place the
Placodermi as sister-group to all other gnathostomes. The organization and function of placoderm feeding
systems is vital to understanding early jaw evolution. The goal of this study is to understand the
relationship of Late Devonian arthrodire skull diversity to ecology and feeding ability.
Functionally relevant morphological and mechanical data were collected from nine placoderms
from the Gogo Formation, western Australia. These skulls were analyzed using a dynamic computer model
based on linkage mechanics that calculates skull kinetics and mechanical feeding metrics in fossil
vertebrates. This model has previously produced original results for a North American species,
Dunkleosteus terreli. Here it is used to compare several arthrodire species to each other and to modern
groups. Results reveal a range of kinematic transfer coefficients (KT) and mechanical advantage values.
Arthrodire KT values (3.2-3.4) are generally higher then a subset of modern labrid oral linkages (0.45-1.5),
but near the high end of the range of hyoid linkages (0.07-4.7), indicating that they may have utilized
suction. The high range of mechanical advantage measurements (0.1-0.68) indicates a great deal of
diversity in feeding ability. Such results lend strong support to the possibility of obtaining quantified hypotheses of ecological diversity within a phylogenetically remote, and long extinct basal gnathostome
taxon.

Lizard teeth share many structural and ontogenetic features with other vertebrates. The primitive
condition is pleurodonty, where the teeth are attached to the inner surfaces of the jaw bones and are
regularly replaced. Teeth develop from crown to root, and enamel has a complexly sculptured surface while
it is being laid down that is occasionally retained. Much of the internal tissue of the tooth is resorbed before
it is shed. Acrodonty, where teeth are firmly attached to the crest of the jaws and replacement is greatly
reduced, has evolved at least twice.

Crown shapes vary and sometimes show a strong correlation with particular dietary components, so
they can potentially be used to infer them. For example, plant eaters and forms that eat hard prey both often
have characteristic tooth morphologies. Patterns of wear are also informative about diet and sometimes
habitat as well. In herbivores, phytoliths and other abrasive plant components may produce characteristic
scratching, and excavations at the bases of their teeth may result from dental caries, perhaps promoted by
sugars in the diet. Ant-feeders sometimes show pitting of the enamel surface, probably caused by the
formic acid their prey contains. Some wear patterns can be mimicked by simple experimental procedures.

All of the four major radiations of terrestrial mammals on Madagascar (primates, carnivorans,
rodents and insectivorans) are very diverse yet, with few exceptions, DNA sequence data support
monophyly of the island representatives of each order. Malagasy insectivorans are composed primarily of
the Tenrecidae, a diverse group consisting of ca. 20 species in 8 genera. Based in part on a new technique
for extracting information from museum specimens, and on a morphological dataset that samples fossils,
this presentation addresses whether or not the pattern of island monophyly best describes the radiation of
Malagasy tenrecs relative to their African mainland relatives. A case has been made using morphological
data that semi-aquatic Malagasy and mainland African tenrecs form a clade with each other to the
exclusion of other tenrecs. This relationship is not supported by sequence data, or by a combined
morphology-DNA dataset. Also addressed is the quality and quantity of the morphological data necessary
to have confidence in phylogenetic hypotheses of fossil tenrecs.

Neogene Siwalik deposits of northern Pakistan span much of the Neogene and document local and
global environmental changes in fluvial sediments and vertebrate faunas. Fieldwork since 1975 has
documented over 1000 fossil localities and 50,000 fossil specimens. While fossils represent a broad range
of terrestrial and freshwater vertebrates, mammalian remains dominate the record. Episodes of
immigration signify changing intercontinental connections. Late Miocene changes in vegetation and mammalian faunas indicate climatic change toward greater seasonal aridity, documented by stable carbon
isotopes in palaeosols and mammalian enamel.

This setting is ideal for testing models of biotic turnover. For the late Miocene interval of climatic
change, changes in mammalian faunas (appearances, disappearances, and ecophenotypic and evolutionary
changes within lineages) were analyzed in relation to two models of turnover. The first model emphasizes
habitat fidelity and environmental sorting of species according to their habitat preferences. This model
predicts a replacement of resident species with immigrant species adapted to more arid vegetation. The
second model emphasizes demographic properties: species with larger populations and geographical ranges
should persist longer and vice versa. This model predicts persistence of dominant resident species through
the interval of vegetation change and dietary accommodation to arid vegetation. Evolutionary changes in
dietary adaptations are also expected.

We evaluated these models with data about biostratigraphic ranges, stable-isotopic composition
and microwear of mammalian teeth, and relative abundance of 40 species from the Siwalik sequence
between 10.7 and 5.7 Ma. While some predictions of both models are upheld, the model of environmental
sorting is a better fit to the Siwalik record.

Palaeobiology of Pietraroia fishes as inferred from phosphorus determination in coprolites
Phosphorus is an important element for many essential processes in the body of all living
organisms, as it is involved in the metabolism of fat, carbohydrate and protein, and in energy metabolism.
For these reasons, phosphorus is very widely distributed in both plants and animals, and thus in their foods.
When a fish ingests its prey, most of the phosphorus ingested finally ends up in excrement, since the
amount required by the fish is negligible as compared to the ingested amount. Here we show that, by
combining a method to quantify the phosphorus in fish coprolites, using a set of equations widely used in
field marine ecology and some very basic assumptions, it is possible to assess the kind of prey ingested and
its maximal dimensions. Though a rough one, it is also possible to obtain an estimate for the composition of
the food web at the level of fishes in the ancient seas of Pietraroja.

Three genera of small ornithischian dinosaurs have been described from the Lower Shaximiao
Formation (Middle Jurassic: ?Bajocian) of Dashanpu, Sichuan Province, People�s Republic of China:
Yandusaurus multidens, Xiaosaurus dashanpensis and Agilisaurus louderbacki. The taxonomy of these
animals is extremely confused: numerous conflicting opinions exist regarding the validity of each species
and the possible synonymies that might exist between them and other ornithischian taxa from China and
elsewhere. Re-examination of the relevant material demonstrates that all three of the small ornithischians
from Dashanpu represent valid taxa, each of which can be adequately diagnosed. In addition, these species
can be distinguished from each other by many craniodental and postcranial character states. We show that
�Yandusaurus� multidens is not congeneric with either Y. hongheensis or any other taxon, and therefore
propose a new generic name for the reception of the species multidens. Consideration of the character states
present in �Yandusaurus� multidens, Agilisaurus and Xiaosaurus shows that these taxa are not primitive
euornithopods (�hypsilophodontids�), as had been suggested by most previous authors, but represent more
basal ornithischians. �Yandusaurus� multidens and Agilisaurus can be regarded as Genasaura incertae sedis,
whereas Xiaosaurus can only be recognised as Ornithischia incertae sedis on the basis of current data.

Supertrees represent a relatively new and controversial method of phylogenetic reconstruction. By
combining existing phylogenetic tree topologies (�source trees�), which can be based on any type of data,
the supertree approach allows the creation of accurate, fully comprehensive phylogenies of large clades of
organisms both quickly and efficiently. However, recent studies have highlighted the vulnerability of
supertrees to the inclusion of poor quality or duplicated data, and some have questioned the utility of the supertree approach as a whole (Gatesy et al., 2002). We show that application of an explicit protocol for
source tree collection (specifically intended to exclude poor source trees; Bininda-Emonds et al., 2004) can
result in supertrees that accurately reflect current phylogenetic evidence. We have used this protocol and a
novel variant of Matrix Representation with Parsimony (MRP) to produce a family-level supertree of
placental mammals, based on 725 source trees from 430 references. This phylogeny is the first to include
all 113 extant families, as well as two recently extinct families � Nesophontidae and Plesiorycteropodidae �
whose relationships remain obscure. The supertree is highly congruent with recent phylogenetic analyses,
but is considerably more comprehensive. In agreement with most current evidence (Springer et al., 2004),
the orders Lipotyphla and Artiodactyla are recovered as non-monophyletic, and extant placentals comprise
four superorders, the origins of which may be connected with tectonic plate movements. Morphological
convergence between these superorders is rife. Preliminary work suggests that other supertree methods
support a largely identical topology. We suggest that the supertree topology presented here is suitable for
use in studies that require a comprehensive higher-level placental phylogeny.

Dinosaurs have been named since 1824, and the rate of naming of new taxa followed a sigmoid
curve until 1980.!After that point there has been an explosive growth of new names.!Synonymy rates,
however, are enormous. Compared to 'normal' synonymy rates of about 20%, at least half of named
dinosaurian species are synonyms, and there is no sign that this rate of erroneous (?over-enthusiastic)
naming has slowed.!A new modelling approach uses such historical data to estimate the total number of
valid dinosaurian species preserved in the rocks (which may or may not correspond to the total number that
ever lived)

This paper describes a new genus and species of primitive wombat from the Riversleigh World
Heritage Property in northwestern Queensland. This taxon is found at two sites within the World Heritage
Property, both of which are estimated to be early or middle Miocene in age. The description of this new
taxon is based on two partial maxillae as well as isolated incisors and molars, and represents the smallest
described wombat to-date.
Previously, the early evolutionary record of wombats prior to the late Miocene was known from a
single described tooth of Rhizophascolonus crowcrofti from the Oligo-Miocene of central Australia. This new taxon is therefore critical to an understanding of the early evolution of this family. The dentition
displays the primitive condition of rooted molars and premolars with much of the cusp detail retained in at
least the early stages of wear. Details of the dentition of this taxon along with that of the late Miocene to
Recent taxon Warendja, suggest a trend towards the modification of a two-phase power stroke to that of a
single transverse movement of the jaw in modern wombats.

The Ocean Sunfishes, family Molidae, include the most massive and some of the most bizarre of
the teleost fishes. The most conspicuous character of the Ocean Sunfishes is the punctuation of the body by
a deep, abbreviated, caudal fin-like structure extending vertically between the posterior ends of the dorsal
and anal fins, termed the clavus by Fraser Brunner. Homology of the clavus has been a matter of debate
since the first studies on molid anatomy in the early 1800's. Two hypotheses have been proposed: i) it is a
highly modified caudal fin; ii) it is formed by highly modified elements of the dorsal and anal fins. To
resolve this homology issue, we studied the ontogeny of the molid vertebral column and median fins and
compared it to that of a less morphologically derived gymnodont, a member of the family Tetraodontidae.
We show that in molids the chorda never flexes during development, that the claval rays form from the
posterior ends of the dorsal and anal fins toward the middle thus closing the gap inward, and that elements
of the molid clavus have an identical development and composition as the proximal-middle and distal
radials of the regular dorsal and anal fins. We thus conclude that the molid clavus is unequivocally formed
by modified elements of the dorsal and anal fin and that the caudal fin has been lost in molids.

Recent discoveries in northeastern Thailand are providing a better picture of the dinosaur
assemblage from the Khok Kruat Formation (dated as Aptian on the basis of sharks and palynomorphs).
Some elements are reminiscent of the better-known fauna from the older Sao Khua Formation; they include
a Phuwiangosaurus-like sauropod and a spinosaur (probably Siamosaurus). Broad-toothed sauropods,
which are known from the Sao Khua Formation, have not been found in the Khok Kruat Formation.
Dinosaurs present in the Khok Kruat Formation but so far absent from the Sao Khua Formation include
Psittacosaurus and an advanced iguanodontian reminiscent of Altirhinus. The Khok Kruat assemblage,
which also occurs in Laos, may be the result of a dispersal event introducing into Southeast Asia faunal
elements that apparently were not present there during earlier stages of the Cretaceous.

Ornithischia is a hugely diverse, abundant, and important clade of dinosaurs. Yet a full
understanding of ornithischian evolution is hampered by an absence of rigorously tested global phylogenies
for the clade. In order to reassess prevailing views of ornithischian phylogeny and evolution a �totalevidence�
approach was taken. Nine higher-level and 42 species-level taxa were selected, representing all
known valid ornithischians. The validity of all previously utilised characters was assessed, and direct
examination of taxa allowed the identification of further characters. Data was analysed using a range of
techniques including safe taxonomic reduction, reduced consensus methods and various measures of
support.

The general structure of ornithischian phylogeny is supported; however, positions of certain taxa
differ significantly from previous analyses. Heterodontosauridae forms the sister group to Genasaura and
may represent the basalmost ornithischians known, a position concordant with stratigraphy. Strong
evidence supports hypsilophodontid paraphyly, with some Middle Jurassic taxa (Agilisaurus and
�Yandusaurus� multidens) appearing as sister-taxa to Cerapoda (Marginocephalia + Ornithopoda).
Far greater congruence with stratigraphy is present than in previous phylogenetic hypotheses,
suggesting that the early ornithischian record may be better than generally believed. New time-constraints
are placed on evolutionary events. For example, previous hypotheses require a Late Triassic date for the
Ornithopoda-Marginocephalia split; this phylogeny predicts a Middle - Late Jurassic date for the same
event, obviating the need for an extensive marginocephalian ghost-range. This analysis highlights both the
necessity of rigorously testing the ornithischian tree and the areas in which efforts should be concentrated.

Pietraroia plattenkalks are richly fossiliferous, fine-grained cherty limestones, from the Matese
Mountains - southern Apennines, Italy, and are well known for the exceptional state of preservation of the
fossils. Based on detailed taphonomical analyses, the fossil assemblage recognised in the Pietraroia
plattenkalk represents a taphocoenosis and an obruption deposit in the sense that most (if not all) of the
animals were transported to an accumulation site from other places. A likely scenario could be that the
sediments were deposited in a depressed area in which gravity currents carrying animals from above,
capturing both living and dead organisms, transported them into accumulation sites placed well below the
original place where they lived and/or died.

First considered as shallow lagoon deposits or as intra-platform small basin-fill, the Pietraroia
plattenkalk sequences, on the basis of sedimentological analyses and geometrical reconstructions, are here
interpreted as deposits from a submarine channel, the �Pietraroia Channel�, and that they document a major
transgressive event. Transgression was associated with the development of suboxic to anoxic conditions at
the seafloor that favoured fossil preservation and the deposition of coprolite-rich and bituminous layers
found within the plattenkalk sequences. A peculiar palaeogeographical and palaeotopographical setting,
strongly controlled by local tectonics, saw the contiguity of wide emergent areas with a relatively deepwater
channelised area where fossiliferous plattenkalk sequences were deposited.

The Alfred Leeds Collection is one of the major fossil marine reptile collections at the Natural
History Museum, London. It contains at least 300 specimens collected towards the end of the 19th Century
from the Middle Jurassic Oxford Clay around Peterborough. One of its more unusual treasures is BMNH
R2903, the first specimen to have been described as a fossilized 'reptile egg'. �The Sphere� declared that it
predated the discovery of dinosaur eggs in Mongolia (Anonymous, 1923). Then in 1950, W.!E. Swinton
featured the enigmatic fossil in an article for The Illustrated London News suggesting that it might be an
amphibious dinosaur egg. As an isolated curiosity, and given the unlikelihood of a fossil egg occurring
within the marine sediments of the Oxford Clay, it has been overlooked for decades and still lacks a
satisfactory identification. CT scanning work undertaken by the University of Glasgow in 1995, as part of a
comparative programme of scanning technologies used with fossil eggs, revealed the internal density
contrasts of the object, and appeared to show a discrete mass of components within. This CT scanning
work has recently been supplemented by SEM analysis of the surface of the object that indicated a
laminated structure to the outer crust or 'shell'. Although the thickness and structure of the 'shell' mitigates
against a dinosaurian/avian diagnosis it is still too early to exclude the possibility that it is in fact a
fossilised Middle Jurassic reptile egg. Further SEM studies may yet reveal the putative egg�s true identity.

The extant clouded leopard (Neofelis nebulosa) has long been regarded as an unusual, small (~ 15-
25 kg) pantherine and has caused problems in analyses of functional skull morphology and phylogeny.
Little is known of its ecology in the wild. Although famous for its hypertrophied canines, it is, nonetheless,
universally regarded as an unusual conical-toothed pantherine, with no morphological affinity to extinct
sabertoothed machairodontines. However, a whole suite of characters accompany the evolution of a
sabertooth morphology, and these have appeared convergently at least five times in the Tertiary in widely
different groups, such as borhyaenid marsupials, creodonts, nimravids and felids. No previous studies have
analysed these features in the clouded leopard and they have also been assumed to be absent in all extant
felids. However, the clouded leopard is found to possess a whole suite of skull features, formerly assumed
to be present exclusively in sabertooth predators, including the marked anterior angulation of the facial
region, narrowed zygomatic arches, lowered jaw joint and differences in the muscle mechanics of the
posterior jaw adductors. The clouded leopard differs from any extant form and may constitute a third extant
felid ecomorph (the others being the cheetah, Acinonyx jubatus, and a general pantherine ecomorph). Data
on its ecology in the wild are urgently required and scholars studying sabertooth functional morphology
should pay closer attention to the clouded leopard, rather than the large pantherines.

Gladbachus adentatus, from the Middle Devonian of southern Germany (Bergisch Gladbach), is
probably the most complete of only four early-to-mid Devonian sharks known from body fossils. The
single specimen consists of the dorsoventrally flattened forequarters of a reasonably well articulated fish,
including the braincase, jaws, hyoid and gill arches, and anterior vertebral column. Large areas of
squamation are intact, but the paired and median fins are largely absent. The challenge of this material is
whether it really represents a chondrichthyan, because many of the classic, defining, specialisations of the
group are missing: fin spines, simple placoid scales, replacement tooth series, and calcified cartilage with a
diagnostic, prismatic structure. Consequently, support for a relationship to crown-group chondrichthyans is
weak, but remains consistent with a stem-group position. If this is correct, then Gladbachus delivers a new
glimpse of diversity preceding the root of holocephalan (chimaeroid) and elasmobranch lineages, differing
sharply from that of standard models for primitive chondrichthyans, such as �Ctenacanthus� and
Cladoselache. The full consequences of this have yet to be explored, but it is already apparent that these
new data are incompatible with current scenarios of chondrichthyan jaw evolution, and recently postulated
similarities between early chondrichthyan and osteichthyan neurocrania.

Ever larger sauropod phylogenies have been produced, bringing with them the promise of better
resolution and support. However, despite increases in numbers of taxa and characters, and datasets being
merged to varying degrees, there is still much topological conflict. There are a number of possible reasons
for such conflict, including; choice and interpretation of characters; choice of taxa; interpretation of
homoplasy; amount of missing data; various forms of error. A comparison of the two most taxonomically
complete sauropod phylogenies (Wilson, 2002; Upchurch et al., 2004) revealed that all of these problems
are present to varying degrees.

The way in which, and the degree to which, each of the causes of conflict influences phylogeny
topology is largely unknown. The goal of this study was to use the debate over sauropod relationships as a
case study in the comparison of differing morphological phylogenetic analyses. The principle aims were:
quantification of the degree of congruence and conflict between the two phylogenies; assessment of the
causes of topological conflict and discussion of how such problems may be resolved; and evaluation of the
strategies available to the morphological phylogeneticist for data analysis and conflict investigation,
leading to methodological recommendations.

Results show that each of the phylogenetic hypotheses examined has topological areas that are well
supported and others that are poorly supported. A new phylogenetic hypothesis for the relationships of
sauropod dinosaurs is presented. Although the taxonomic focus is on the phylogeny of the Sauropoda,
many of the issues detailed are of wider concern in morphological phylogenetics.

The mammalian orbit, or eye-socket, is a highly plastic region of the skull. It comprises between
seven and nine bones, all of which vary widely in their contribution to this region between the different
mammalian orders and families. Yet, despite this plasticity, relatively little research has concentrated on the
orbital region. It is hypothesised that the structure of the mammalian orbit is principally influenced by the
forces generated by the jaw-closing musculature. In order to quantify the orbit, fourteen linear, angular and
area measurements were taken from 84 species of placental mammals using a Microscribe-3D digitiser.
The results were then analysed using Principal Components Analysis. The results of the multivariate
analysis on untransformed data showed a clear division of the mammalian taxa into temporalis-dominant
forms and masseter-dominant forms. This correlation between orbital structure and masticatory
musculature was reinforced by the results from the data transformed with respect to skull length, which
showed a separation of the taxa into the three specialised feeding types as proposed by Turnbull (1970):
i.e., �carnivore-shear�, �ungulate-grinding� and �rodent-gnawing�. Moreover, within the rodents, there was a
clear distinction between species in which the masseter is highly developed and those in which the
temporalis has more prominence. These results give a strong indication that, although orbital structure is in
part determined by the relative size and orientation of the orbits, the forces generated by the muscles of
mastication also have a large effect.

Finite Element Analysis (FEA) is a computer modelling technique, developed by mechanical
engineers, which can be used to test how structures respond to imposed loads. It has recently attracted
interest from biologists and palaeobiologists interested in skull and skeletal biomechanics, but the question
of validation remains an issue. As a preliminary step, we have compared data obtained from a FE model of
a cat skull (Felis sylvestris catus) with that from an earlier in vitro analysis of the cat skull using strain
gauges.

The FE model was created with data provided by the High Resolution Computed Tomography
Laboratory, University of Texas, Austin. It was then loaded and constrained to replicate the experimental
set-up used by Buckland-Wright in his classic work on the cat. He recorded five major strain foci � in the
anterior, middle and posterior parts of the zygomatic arch; along the canine root; and on the palatal surface
adjacent to the tooth row. The same five zones were observed on the FE models with comparable strain
values. Peak levels of strain were recorded in the zygomatic arch and at the anterior root of the zygomatic
arch, and these findings match those of more recent in vivo studies of other mammals.

On the basis of molecular-biological evidence from extant animals, the basic cladogram for
deuterostomes reads: (((hemichordates + echinoderms) Xenoturbella) chordates). Moreover there is now
molecular evidence that, if the probable effects of long-distance attraction are discounted, the basic
cladogram for chordates should read: (acraniates (tunicates + vertebrates)). Here, we accept these results
and attempt to place the most primitive deuterostome fossils in the stem groups implied by the cladograms.
In doing so, we find that a stereomic calcitic skeleton of echinoderm type, with each plate a single crystal
of calcite, is not a hallmark of echinoderms as commonly believed. Rather, in the latest common ancestor
of all extant deuterostomes it would have existed, but has been lost at least six times among the
descendants of that animal. In this light, a calcitic skeleton in a carpoid is a symplesiomorphy - it does not
indicate an echinoderm but merely a deuterostome. We further conclude that the carpoids known as
ctenocystoids are stem-group hemichordates, mainly because one recently discovered ctenocystoid must on
functional-morphological grounds have been a burrower comparable to an extant enteropneust, and in
particular to the newly described deep-sea enteropneust Torquarator. The carpoids called Cincta, which
probably gave rise to the ctenocystoids, are also likely members of the hemichordate stem group. The
echinoderm stem group is probably represented by the basically triradial fossils known as helicoplacoids
and Tribrachidium. Their triradiality presumably evolved, in more crownward parts of the echinoderm
stem group, into the standard quinqueradiality of extant echinoderms. The most primitive (most
"rootward") part of the chordate stem group was probably included in the solute carpoids, while more
crownward members of the chordate stem group were the cornutes. The most primitive members of the
chordate crown group were the mitrates. The most primitive known solute was attached by the end of its
tail, whereas more crownward stem-group chordates, and the mitrates also, locomoted rearwards tail first.

New vertebrate records from the Late Cretaceous of Kazakhstan
For the last three summers, we have undertaken exploratory fieldwork in the Aral Sea region of
northeastern Kazakhstan searching for dinosaurs and other fossil vertebrates. This region of Central Asia
remains largely unexplored, although limited details of the geology and palaeontology of the lower Syr-
Dar�ya Uplift are available.

In addition to skeletal remains, we have mapped and recovered abundant vertebrate microfossils
from a series of horizons within the Late Cretaceous (Turonian�Campanian) Bostobynskaya Formation
(Bostobynskaya Svita). Microfossil remains display taphonomic characteristics consistent with their
deposition within floodplain-hosted assemblages and allow us to make initial estimates of the
Bostobynskaya faunal composition. Teeth collected from these horizons confirm the presence of
theropods, hadrosaurs and sauropods in the formation, consistent with previous suggestions of the dinosaur
fauna. Compositional analysis of microfossil collections show that the material is characterised by low
weathering and abrasion states, a high diversity of small fossils that represent aquatic, semi-aquatic and
terrestrial taxa, and an abundance of resistant bioclasts, such as teeth. The sedimentology of the
Bostobynskaya Formation is dominated by crevasse-splay and floodplain facies. New records from these
sites document an important Late Cretaceous vertebrate fauna in an equally important and much
understudied part of Central Asia.

A new ungulate mammal family Olseniidae is proposed. The definition of this family is based on
fossil material, which was excavated during two field campaigns of the Martin-Luther-University to the
Eocene Toru Ajgyr locality, Kyrgyzstan in 1997 and 1998. A nearly complete foot was discovered, which
corresponds to an astragalus of Olsenia mira from the Eocene Shara Murun locality, northern China. The
new Kyrgyz form cf. Olsenia sp. is an early ungulate that combines characteristics of mesonychids,
perissodactyls and artiodactyls: tetradactyl and paraxonic foot; terminal phalanges claw-like, but not
fissured; astragalus with shallow proximal caput and without distal trochlea. This unique character set fills
a gap in the fossil record and gives insights into ungulate phylogeny, which is still not completely
understood.

In the Middle Jurassic, the UK consisted of a series of islands at the western edge of the European
archipelago. The vertebrate faunas of these islands include some of the earliest frogs, salamanders,
albanerpetontids, turtles, lizards, choristoderes and maniraptoran dinosaurs, as well as a diversity of
mammals. However, most localities yield only dissociated bones. The Kilmaluag Formation of Skye also
produces rare associations. Recent fieldwork has extended the known assemblage and added a nearly
complete turtle and partial salamanders. The Skye assemblage contains representatives of 12 major
vertebrate clades and >20 distinct taxa including hybodont sharks; semionotiform fish; primitive caudates;
a turtle; lepidosauromorphs (Marmoretta, lizards); choristoderans (Cteniogenys), crocodiles (goniopholids,
atoposaurs); pterosaurs; sauropod dinosaurs; tritylodont synapsids (Stereognathus), and mammals
(Borealestes, indet. docodont). This assemblage is similar to those from English localities, but appears to
differ in the rarity of crocodiles, and the absence of frogs and albanerpetontids. Salamander and turtle
remains are the most abundant at the site, whereas terrestrial animals (pterosaurs, dinosaurs, mammals) are
relatively rare. This could imply an environmental bias in the record (probably the most likely option)
and/or the operation of a taphonomic filter related to body size (the latter in the case of dinosaurs and
crocodiles).

Asia boasts the greatest known diversity of Late Cretaceous lizards, but tracing the history of this
radiation has been complicated by a paucity of earlier data. New work on Barremian-Albian localities in
China, Mongolia and Central Asia is filling the gaps, but many questions remain. The Early Cretaceous
(Valanginian-Hauterivian) Tetori Group of central Japan has yielded a diverse assemblage of small
vertebrates including fish, amphibians, reptiles, birds and mammals. The most prolific locality is that of
Kaseki-Kabe, Hakusan City, Ishikawa Prefecture. It has produced at least six lizard taxa, with several
associated specimens. The largest and most common is a long-bodied swimmer with anguimorph affinities.
The second is a specialised herbivore with teeth resembling those of the living Iguana. New material of
Sakurasaurus, originally described from Gifu Prefecture, adds important new data. A new phylogenetic
analysis suggests a relationship with the Chinese (Yixian Formation) lizard Yabeinosaurus. The remaining
Kuwajima lizards include two new scincomorphs and a very fragmentary possible gekkotan. Rectangular
osteoderms closely similar to those of the Jurassic/ Cretaceous scincoid paramacellodids are common in the
Kuwajima rocks and probably pertain to one of the scincomorphs, but they have yet to be found in
association.

The flying beams � what biomechanics tells us about pterosaur evolution
Investigation of the biomechanics of organismic structures yields important information for
understanding evolutionary processes. In the present study, pterosaur skulls were analysed using a
combined approach via finite element analysis, static investigations and application of classical beam
theory. Comparison of the biomechanical behaviour of the different skull constructions present in
pterosaurs results in major evolutionary transformations: elongation of rostra; inclination of the occipital
region; variation in tooth morphology; reduction of the dentition and replacement of teeth by a keratinous
hook or rhamphotheca; fusion of nares and antorbital fenestra; and the development of bony and soft-tissue
crests. These transformational processes are discussed for their biomechanical effects biting. They are
interpreted in the light of greater feeding efficiency and reduction of bony mass combined with an
increased stability against bending, shear and torsional loads. As a consequence, certain feeding options are
assigned to the different skull constructions, and previous hypotheses (e.g., skimming) can be verified.
Using the principle of economisation, these processes help to define evolutionary pathways that
can be used for comparison and verification of recent hypothesis of pterosaur phylogenetic systematics.

Recently, it has been suggested that stem-group gnathostomes and basal taxa of crown-group
gnathostomes may each possess a unique pattern for the establishment of both tooth initiation sites and
further replacement tooth production. Few studies have focussed on the early events that prime these initial
odontogenic sites across the vertebrates. The initial sequential pattern that lays down the order of tooth site
specification is thought to vary across the major vertebrate clades. Subsequently, the mechanism involved
in producing further generations of teeth may also show fundamental differences across the vertebrates.
Within the rainbow trout there are different tooth-bearing locations, each with a different temporospatial
order for the initial tooth positions. Early components of the mechanisms that lay down the pattern for a
spatiotemporally ordered dentition have been identified in fish. In the trout dental system individual teeth
are initiated within a common epithelial odontogenic band for each toothed region, localised by the
expression of Shh and Pitx-2. Gene expression is restricted to the basal epithelium of each odontogenic
band. Within this band tooth buds are located and form in alternate tooth positions for each dentate region,
thus initialising the primary order and that of the replacement dentition. Both gene expression data and
morphological observations will be compared for osteichthyan (Oncorhynchus mykiss) and chondrichthyan
fish (Scyliorhinus canicula), both for initial dental pattern and their replacement mechanisms. This forms
the basis for understanding the evolution of tooth initiation order and further lifetime production in early
vertebrates.

The Gorgonopsia, a group of mammal-like reptiles from the Upper Permian of South Africa, is
revised. This work aimed to resolve major issues in the alpha-level taxonomy of these animals, which
remained unclear in several areas despite the enormous efforts of Sigogneau-Russell (1970, 1989). The 22
genera that were thought to be valid at that time have been further investigated and compared. As a result,
the number of genera has been reduced and the total species number has also been modified. The mounted
skeleton of ?Aelurognathus parringtoni was re-examined and investigated functionally and
morphologically. The questionable relationships of this East African specimen are resolved: it is shown to
have a close affinity with the Russian form Sauroctonus progressus. On this basis, it was possible to
establish a sound, computer-based phylogeny for this group for the first time.

?Aelurognathus parringtoni was compared with the sabertoothed cat Smilodon in functional terms.
As both taxa possess extremely elongated canines, I sought to identify other possible functional
convergences in the rest of the skeleton. It was discovered that this particular gorgonopsid shows some
functional similarities to Smilodon, but this must be considered in the light of the different evolutionary
histories of each of these animals.

Kuehneotherium is one of the oldest of the "symmetrodontans", a grouping of primitive mammals
characterized by the possession of a reversed-triangle molar pattern. In spite of its importance to early
mammalian phylogeny, Kuehneotherium has proved difficult to interpret, due to the fragmentary nature of
the material.

A reconstruction of the dentition is now given and the phylogenetic position of Kuehneotherium is
re-examined. A number of features are also described which suggest that diphyodonty had been only
recently established.

The Early Jurassic fissure fills of South Wales reveal previously unexpected taxonomic diversity in
the kuehneotheriids, including two new species of Kuehneotherium. There are also a number of interesting
teeth with plesiomorphic characters and the relationship of this kuehneotheriid with Kuehneotherium is
discussed.

Well-preserved pliosaurs (Reptilia: Sauropterygia) are rare, and those with high-quality
braincase material are extremely uncommon. A new and undescribed genus and species of
pliosaur from the Lower Cretaceous of Colombia, South America, is one of the very few pliosaur
specimens which preserves the otic capsules and braincase substantially complete, articulated and
largely uncrushed. The 3D cranium of this new specimen has been acid-prepared, revealing the
otic capsules and details of the internal osseous labyrinth. However, the bones surrounding the
osseous labyrinth are both fragile and fused, and understanding the detailed osteology and
visualising the precise morphology of the semicircular canals is challenging. In order to obtain
fuller information from the fossil, the otic capsules were CT scanned and the images manipulated
with the software VG Studio Max in order to produce a virtual 3D �cast� of the semicircular
canals. The generation of this 3D computer model of the otic capsules and the osseous labyrinth of
this pliosaur, displays for the first time, full details of this region of the cranium in a
sauropterygian marine reptile.

Associations among functionally or ontogenetically-related traits may be a major influence on
morphological evolution. The study of morphological integration, via quantitative analyses of trait
correlations, allows for investigation of broad patterns of trait associations, with data that are comparable to
developmental and quantitative genetic studies. I present analyses of cranial integration for 108 species of
extant and fossil mammals, including monotremes, marsupials and placentals. 3D data were gathered for 59
landmarks to assess cranial modularity and the relationship of integration to several influences on cranial
morphology: phylogeny, diet, encephalisation and heterochrony.
All large clades display a strong correlation between similarity of integration and phylogeny, while
diet�s influence on integration is generally restricted to less-inclusive clades. Cranial modularity is
conserved across therian mammals, with six discrete groups of traits. The anterior oral/nasal, molar and
basicranial groups are strongly integrated, while the orbit and zygomatic/pterygoid groups are weakly
integrated, probably reflecting differences in developmental complexity. The integration of the cranial vault
is variable across taxa, but it is not significantly correlated with encephalisation. Monotremes display only
two modules (anterior oral/nasal and basicranium), with little correlation among other landmarks,
demonstrating that cranial modularity has been modified during mammalian evolution.

Progress in the morphology and systematics of rauisuchian archosaurian reptiles
Rauisuchia is a large group of Triassic crocodilian-line archosaurian reptiles. Rauisuchian
systematics is highly confused and contentious, so that it is a matter of debate whether the group is mono-,
para- or polyphyletic. The unsatisfactory state of rauisuchian systematics is rooted in inadequate low-level
taxonomy, and a lack of detailed morphological documentation. Recent work has made progress in
describing new taxa, revising previous material, and cautiously proposing precise phylogenetic hypotheses.
A summary of recent and ongoing morphological and phylogenetic work is presented, with particular
attention paid to Arizonasaurus and Batrachotomus, and new data on crocodilian-line archosaur braincase
anatomy. This recent work indicates that rauisuchians are polyphyletic, and that early crocodilian-line
archosaurs were more diverse than previously recognised.

Reconstructions of the cranial musculature of extinct animals are a means of learning more about
the characteristic feeding adaptations of individual species. Through comparison with modern taxa the
feeding behaviour can be deduced, which largely increases our understanding of their way of life. Muscle
reconstructions of extinct animals like plesiosaurs, who do not have a living relative or analogy, are more
difficult, but still help towards creating more complete reconstructions of these animals and how they were
adapted to their environment. So far, only the cranial musculature of some pliosauroid species have been
reconstructed. In this talk, the first reconstruction of a plesiosauroid, namely the German Seeleyosaurus
guilelmiimperatoris from the Toarcian Posidonian shale is presented. Based on skull reconstructions and
preserved muscle scars the different cranial muscles are reconstructed and the operational mode of the jaw
is explained. The differences and similarities to the pliosauroids and the implications for feeding
adaptations are discussed.

Butselia biveri was first described on isolated molars from the Belgian early Oligocene site of
Hoogbutsel. The monotypic genus was subsequently discovered in similar-aged strata in the Isle of Wight,
UK. Quinet & Misonne (1965) placed Butselia in its own family in Gill�s (1885) insectivoran suborder
Zalambdodonta (tenrecs and golden moles), because its molars are semi-zalambdodont. Butler (1972)
instead allied it with Plesiosorex (extinct lipotyphlan family Plesiosoricidae). Apart from European
Butselia, all other pre-late Oligocene family members are Asian, the earliest, according to McKenna &
Bell�s (1997) classification, being Ernosorex and Pakilestes from the early to middle Eocene. The
occurrences of Butselia then slightly post-dated an important faunal turnover, the �Grande Coupure�, when
many taxa from Asia had dispersed into Europe. Therefore, stratigraphical position was consistent with an
Asian origin for Butselia as a plesiosoricid.

Two new finds impact on the relationships of Butselia. One is a tooth of the genus in Isle of Wight
strata predating the �Grande Coupure�, suggesting European not Asian roots. The other is an isolated astragalus judged to belong to Butselia.
This astragalus is unlike any lipotyphlan and, whereas plesiosoricid
postcranials are unknown, it suggests either that the Plesiosoricidae are not lipotyphlans or that Butselia is
not a plesiosoricid.

The commercial sale of palaeontological material has emerged over recent years as one of the most
controversial and fiercely contested debates in palaeontology. This paper explores a sector of the fossil
market, focusing on live auctions and the emergence of a new market in fossils from the early 1990s
through to 2005. The findings are based primarily on quantitative and qualitative data drawn from the
major auction houses and catalogues.

Auctions play an important role in signalling the presence of a market, and can act as a socially
acceptable process for determining price when there are high levels of uncertainty and ambiguity.
Originally marketed as the �collectibles of the 1990s�, the auction market was built on the popular success
of Michael Crichton�s Jurassic Park and exciting new scientific discoveries. Fossils at auction have since
undergone a transformation, emerging as a more stable and successful d�cor market in the UK. This paper
describes and explores the characteristics of the current UK auction market: price trends, supply, location
of material, and type of material being sold, analysing the emergence and development of a market for
fossils presented as d�cor and objet d�art.

This paper also explores differences between UK auctions and the more controversial US auction
market for fossils, comparing price ranges, types of material sold, and presentation. Finally, a brief mention
and comparison is made to one of the online auction houses, eBay.

Dicynodonts had a derived feeding apparatus that permitted propaliny and allowed effective
grinding of plant material. Changes in skull structure across Dicynodontia were previously correlated with
changes in masticatory function, and hence, diet. Derived cranial features of Lystrosaurus were
hypothesized to increase the efficiency of its masticatory system relative to a generalised, purely Permian
dicynodont such as Oudenodon. These features possibly allowed Lystrosaurus to cross the Permo-Triassic
boundary due to its ability to exploit new, more resistant vegetation.

This study will examine quantitatively the functional significance of cranial form of Oudenodon
and Lystrosaurus using Finite Element Analysis (FEA). By determining the distribution and magnitude of
cranial stress, FEA can establish if morphological differences have biomechanical significance related to
changes in masticatory function. To implement FEA, several aspects of dicynodont skull anatomy must be
first examined and include: i) investigating qualitative differences in skull morphology correlated with
changes in masticatory function; ii) creating three-dimensional skull models; iii) examining the histology
and determining the approximate material properties of dicynodont cranial bone; and iv) establishing the
attachment areas of the adductor muscles. These combined parameters will allow FEA to accurately model
the distribution and magnitude of stresses experienced during mastication for both genera.

Rhynchocephalians were globally distributed during the early Mesozoic, but the living Sphenodon
(the Tuatara) is restricted to New Zealand off-shore islands. Traditionally, Rhynchocephalia were
considered morphologically uniform, but new research has revealed two independent aquatic radiations,
differences in body proportions, novel cranial architecture, and a wide range of tooth morphologies. Basal
taxa undergo a transition from simple pleurodont teeth towards larger acrodont teeth, probably representing
a shift in primary jaw function from prey capture toward prey reduction. Derived taxa possessed more
complex teeth, powerful bites, and sophisticated shearing mechanisms: orthal (e.g., Clevosaurus) or
propalinal (e.g., Sphenodon). This diversity in feeding apparatus prompted the idea that lepidosaurian
skulls exhibit a trend towards accessing larger food items: squamates use skull kinesis whereas
rhynchocephalians reduced food items using harder bites. Rhynchocephalian skull shape was analysed
using geometric morphometrics. This identifies associated shape differences in landmark constellations.
Changes in cranial structure can subsequently be illustrated using thin plate spline deformation grids.
Outgroup comparison demonstrates that rhynchocephalian clades exhibit differences in skull proportions
and jaw joint position. Modifications in both of these features are closely associated with increased biting
performance, indicating that acquisition of tougher food items was of major importance to
rhynchocephalian skull evolution.

The Lower Cretaceous (Aptian-Albian) southern high-latitude deposits of Australia have yielded a
diverse range of marine reptile fossils. Identifiable taxa include the ubiquitous Cretaceous ichthyosaur
Platypterygius, and at least five distinct kinds of plesiosaur � a new Leptocleidus-like pliosauroid, the large
pliosaurid Kronosaurus, indeterminate elasmosaurids, polycotylids, and a new genus and species of
cryptocleidoid. Notably, the polycotylid material is amongst the oldest from anywhere in the world and
might suggest a Southern Hemisphere origin for the group. The fossils are derived primarily from the lower
Aptian to lower Albian Bulldog Shale and Wallumbilla Formation. These units are famous for producing
opal and represent shallow epicontinental marine environments. Isolated plesiosaur (indeterminate
pliosauroid) remains have also been recovered from high-latitude non-marine strata of the Wonthaggi,
Eumeralla and Griman Creek formations. These deposits are early Aptian to middle Albian in age and
comprise fluviatile/estuarine sediments laid down in inland rift valleys and coastal flood plains near the
Cretaceous southern polar circle. Estimates of palaeolatitude place most of southern Australia at around 60
to 85� South during the late Early Cretaceous. Sedimentary structures (including glacial erratics,
glendonites, cryoturbated terrestrial sediments and growth-banded wood), fossils, isotope data and climatic
modelling indicate highly seasonal cool-cold conditions (possibly with winter freezing). This contrasts
markedly with climate regimes typically tolerated by modern aquatic reptiles, but suggests that some
Mesozoic forms might have possessed adaptations (e.g., elevated metabolic levels and/or annual migration)
to cope with low average water temperatures.

There has been a renewed interest in the application of ecomorphological techniques to bovid
postcranial remains as a way to infer past ecological conditions at sites pertaining to hominin evolution.
However, the best habitat discriminators are often complete long bones that are rarely found intact in fossil
assemblages, reducing the efficiency of this method. Furthermore, only a handful of elements have been
analysed in an ecomorphological context and the remainder of the postcranial skeleton has been neglected
despite the relatively high occurrence of the complete recovery of some of the smaller and denser elements
like the carpals and tarsals.

In order to increase the amount of material available for an ecomorphological
reconstruction, the majority of bovid postcranial elements were assessed for their ability to
discriminate between seven habitat types in a series of discriminant function analyses. A global
sample of extant bovids (n = 205), cervids (n = 14) and tragulids (n = 5) comprise this
comparative dataset. The baseline of chance accuracy for the DFAs (i.e., the percentage of correct
predictions that can be expected when habitat assignments are randomised) was determined. This
baseline served as the cut-off point between good and bad habitat predictors. The good predictor
elements from both the Upper Laetoli Beds and the Upper Ndolanya Beds at Laetoli were
analysed. Summaries of the number of specimens predicted to belong to each habitat type indicate
that at the time of deposition of the Laetoli Beds the area had heavy woodland-bushland cover
with some lighter tree and bush cover and grass available. The results also indicate that during
Ndolanya times the environment had become more open and the grassland component of the
environment had increased significantly. Light woodland-bushland and an abundance of grass
cover dominated the landscape, although tracts of land with denser vegetation likely existed,
agreeing with earlier suggestions that the area was a semi-arid bushland.

Research into dietary adaptations of teeth has largely focused on crown structure but little is known
about subocclusal morphology. Here I present a study on tooth root morphology and size in selected
carnivore species with distinct diets and feeding behaviour. I explore the hypothesis that root attachment
area is closely linked to dietary specialisation. The maxillary and mandibular dentitions of six species
(Canis familiaris, Panthera pardus, Crocuta crocuta, Ursus americanus, Ailuropoda melanoleuca and
Phoca vitulina) were scanned using computed tomography and reconstructed using three-dimensional
image processing. The results showed that the magnitude of root surface area in all species strongly
depends on the feeding behaviour and occlusal forces irrespective of feeding category. Species which
engage in the mastication of tough foods such as the herbivorous giant panda Ailuropoda melanoleuca and
the osteophagous spotted hyaena Crocuta crocuta exhibit postcanine roots with the relatively largest
surface area when corrected for craniofacial size. In contrast, those species that feed on a soft diet like the
common seal Phoca vitulina bear small roots with less attachment area. Moreover, the size and the
distribution of root surface area along the tooth row seem to be strongly related to other structures of the
masticatory apparatus such as temporomandibular joint morphology. Further investigation of tooth root
morphology may offer a valuable contribution to the reconstruction of feeding behaviour and diets of other
related extant and extinct taxa.

Previous studies of the earliest Eocene (Ypresian) Fur Formation of Denmark have revealed a
diverse fossil avian fauna. Among them are a number of articulated, three-dimensionally preserved
skeletons, which represent some of the best and earliest fossil remains of anatomically modern birds
(Neornithes). Situated immediately after the Palaeocene-Eocene extinction event, these fossils will provide
important insights on the diversification and evolution of modern birds during the Palaeogene. Also,
incorporation of these well-preserved fossils into phylogenetic analyses will provide an important key to
the interrelationships within and between clades of modern birds. However, until now the fossil specimens
remain largely unpublished.

The Fur Formation is chiefly a diatomite deposited in a subtropical sea with dysoxic to anoxic
oxygen bottom conditions. This has resulted in other remarkable states of fossil preservation; fossilisation
of soft tissue structures, such as imprints of the scales of the leg or carbonised imprints of feathers are not
uncommon. Among the previously unpublished fossil material from the Fur Formation is an articulated
charadriiform (shorebird), which is the geologically oldest well-preserved fossil representative of the order.
A preliminary cladistic analysis indicates that the specimen is closely related to modern gulls (Laridae:
Larini), and it will be used to shed further light on the interrelationships and evolution of the order
Charadriiformes

There is a wealth of evidence that terrestrial vertebrates were affected by mass extinction events
during the Permo-Triassic (at the end-Permian, end-Carnian and end-Triassic). However, statistically
speaking, these results are sometimes hard to demonstrate. Two datasets, one of 814 genera and the other of
208 families, were compiled from the literature and each taxon assigned to categories of body size, diet,
habitat and geographic range. Traditional metrics of diversity, extinction and origination failed to recover
evidence of mass extinction. However, where stage-crossing taxa were considered in isolation the two
major events (end-Permian and end-Triassic) become more prominent. Chi-square tests compared pre- and
post-extinction faunas with the end-Permian and end-Carnian exhibiting significant ecological change.
Jablonski's model of alternating macroevolutionary regimes is weakly supported, but not contradicted.
Similarly, differences between events seem to support their contingency.

Further analysis of the datasets included Pearson chi-square tests of variable independence that
show that the selected 'characters' are redundant due to rampant and significant association. Comparison
between the datasets suggests neither the familial nor the generic level is the superior species-proxy in
tetrapod evolution. Despite mostly negative ramifications for future research, potentially rewarding
avenues are briefly discussed.

The fossil catfish fauna from Palaeocene/Eocene strata in the Republic of Mali, west Africa are
presented. They are compared with similar faunas from other west African localities. The implications for
the timing and extent of the trans-Sahara seaway are explored.

In this study, inter-specific patterns of differences in lumbar vertebral size and shape between
Gorilla gorilla, Pan troglodytes, Pongo pygmaeus and Homo sapiens are explored and discussed in relation
to inter-specific differences in locomotor repertories, body size and phylogenetic history. Geometric
morphometric methods are employed to assess the high-dimensional complexity of the overall form of
lumbar vertebrae and the lumbar spine as a whole. Results indicate that interspecific differences in
vertebral size between great ape taxa have a relationship with interspecific differences in body size.
However, differences in vertebral size between humans and great apes are related to interspecific
differences in locomotor functions. With regards to hominoids as a whole, humans have the most different
vertebral shape followed by Pongo, whereas the African apes are relatively similar to each other. Interspecific
differences in locomotor functions have a strong relationship with differences in vertebral shape
but not with those in body size. Phylogeny does not have a strong relationship either with differences in
vertebral size or shape between the hominoid taxa in the study. We conclude that in hominoids, lumbar
vertebral size and shape are greatly influenced by the various highly specialized locomotor repertoires and
loading patterns characteristic of each taxon. Understanding the range and patterning of morphological
variation in the extant hominoid taxa is essential in order to make valid functional and phylogenetic
interpretations of fossil hominid skeletal material. Thus, these conclusions add valuable insights relevant to
the interpretation of fossil hominid vertebral material in relation to locomotor functions.

Excellent new specimens of the alligatoroid eusuchian Acynodon iberoccidentalis are described
from late Campanian-early Maastrichtian localities of southern France. A complete skull and a mandible
allow identification of new diagnostic characters. Acynodon is the only known globidontan alligatoroid
from Europe for this period. These specimens add enough new data to allow comparison with previously
described North American basal globidontans such as Brachychampsa, Albertochampsa and
Stangerochampsa. Phylogenetic affinities of Acynodon are examined by addition of the new identified
characters to a previously published data matrix.

The relatively short snout, the heterodont dentition, presenting anterior peg-like and posterior
crushing teeth, the medial expansion of a shelf on the mandible and the robustness of the posterior portion
of the mandible suggest that Acynodon attained some degree of feeding specialization. Previous hypotheses
concerning diet among basal globidontans are explored. North American globidontans showing posterior
crushing and anterior piercing teeth have been viewed as generalists. The anterior peg-like teeth of
Acynodon reinforce the hypothesis of a different diet.

A new species of Monjurosuchus, a choristoderan reptile, is described from the Lower Cretaceous
Kuwajima Formation (Tetori Group) of Japan. This is the first report of the genus from Japan and this is
also the first report of the same tetrapod genus shared between the Jehol Group and Tetori Group. The new
species is characterized by unique postorbital structures, such as reduction of the quadratojugal and a spiky
flange on the squamosal. A phylogenetic analysis including the new animal was conducted, which also
incorporated data on a poorly known long-necked choristodere from China (Hyphalosaurus lingyanensis)
for the first time. A new phylogenetic hypothesis is proposed that differs significantly from previous
studies. Choristoderans are formed mainly by two monophyletic groups. One group, Neochoristodera,
includes Champsosaurus, Tchoiria, Simoedosaurus and Ikechosaurus. It is characterized by snout
elongation and a relatively small orbit. The other, unnamed, group includes Lazarussuchus,
Monjurosuchus, Hyphalosaurus and Shokawa. These species are recognized as a monophyletic group for
the first time. The group is mainly characterized by a dorsally directed large orbit and closed lower
temporal fenestrae. Members of the group remained as small animals less than 1 m in total length, but
exhibited high morphological diversity, for example, giving rise to long-necked forms such as
Hyphalosaurus and Monjurosuchus.

The Purbeckian crocodilians from Cherves-De-Cognac (Berriasian, southwestern France)
The site of Cherves-de-Cognac yields a diversified fauna, which includes all the groups of
vertebrates: Osteichthyes, Chondrichthyes, Amphibia, Reptilia (Lepidosauromorpha, Testudines,
Crocodylia, Dinosauria), Aves and Mammalia, with hundreds of macro-remains and thousands microremains.
Crocodilians are particularly common, with four families represented by at least seven species.
The medium to large-sized Goniopholis is the most abundant crocodilian, with 11 skulls and numerous
post-cranial elements, referable to G. simus and G. crassidens. The longirostral Pholidosaurus is rarer,
represented by two complete skulls and numerous isolated teeth, referable to P. purbeckensis. Bernissartia
fagesii is very common with abundant isolated teeth. A small-toothed mandible can be attributed to a new
Bernissartiidae. The Atoposauridae are common with four sub-complete specimens and numerous isolated
teeth referable to Theriosuchus pusillus, but also with numerous and strange folded teeth, which could
represent a new Theriosuchus species.

This crocodilian assemblage is deposited in a deltaic environment, contemporaneous to the Middle
Purbeck Beds of Dorset. The quantitative study of the vertebrate assemblage shows that these crocodilians
were inland dwellers, except maybe Goniopholis, which could have been a brackish-water dweller. This
crocodilian community exploited diverse trophic resources, with the opportunistic Goniopholis, the
ichthyophagous Pholidosaurus, the crushing Bernissartiidae and the dwarf "micropredator" Atoposauridae.

The spiky-headed temnospondyl amphibian Stegops divaricata from the Middle Pennsylvanian
coal of Linton, Ohio has remained neglected and enigmatic for several decades. It has been argued to be the
ancestor of the Permian Zatrachydidae, also spiky-headed temnospondyls, although there are few
resemblances other than the spikes. An examination of previously undescribed material of Stegops, along
with a re-evaluation of the original specimens, permits a redescription and partial systematic assignment of
it. All specimens have bony spikes on the tabular, quadratojugal and angular, but in apparent dimorphism,
only some have squamosal and supratemporal spikes. A phylogenetic analysis of 52 characters in 15
temnospondyl taxa places Stegops within the dissorophoid clade but leaves its position uncertain within
that clade. The Zatrachydidae, represented by Acanthostomatops, fall outside the Dissorophoidea, and the
zatrachydid affinities of Stegops asserted by previous workers are based on homoplasious similarities in
ornamentation. Internal relationships of the Dissorophoidea remain unresolved and Stegops shares
conflicting similarities with Amphibamidae in some resolutions and with an Ecolsonia + Dissorophidae +
Trematopidae clade in others.

Eotyrannus lengi Hutt et al., 2001 is a coelurosaurian theropod from the Wessex Formation
(Hauterivian-Barremian: Lower Cretaceous) of the Isle of Wight. New preparation and investigation have
allowed the preliminary description to be substantially updated and E. lengi has proved more complete than
initially reported. It seems that E. lengi was more aberrant than suspected, and in many details is highly
apomorphic. However, cladistic analysis, incorporating information from new basal tyrannosauroids
including Dilong and Appalachiosaurus, supports inclusion of E. lengi within Tyrannosauroidea.
Unfortunately, E. lengi�s skull is fragmentary, but newly recognised elements include a surangular
and palatine. The premaxillary teeth are U-shaped; unfused interdental plates were present; and both the
dentaries and nasals reveal newly recognised - and surprising - apomorphies. Cervical centra are
pleurocoelous, unlike the caudal dorsal centra, but the probable first sacral vertebra was also pleurocoelous.
Several caudal vertebrae are preserved. New information on the scapulae shows that their dorsal end was
expanded. Both humeri are present, as are an incomplete radius and ulna and multiple manual elements. At
least some of the manual digits can be reconstructed, and a partial metacarpus and well-preserved
trochleated distal carpal also provide new information on forelimb structure and function.

The effects of captivity on the morphology of mammals is not well understood, yet it has important
implications for any study of the anatomy of exotic creatures that are only readily available as zoo-derived
specimens in museum collections. This paper concentrates on skeletal remains rather than soft tissues and
will briefly discuss a number of relevant examples in the literature before presenting in detail two studies of
big cat morphology. The first is the effects of captivity on the crania of lions (Panthera leo) and leopards
(Panthera pardus), which has shown that there are morphometric differences between wild and captive
animals. The second study is on cheetah (Acinonyx jubatus) postcrania, where radiography has been used to
examine cortical bone thickness in wild, modern captive-bred and historical wild caught and subsequently
captive samples. The implications of these studies will be discussed, both in terms of the light they shed on
captivity related morphological changes, but also the wider implications for the study of comparative
anatomy.

Tyrannosaurids, such as Tyrannosaurus rex, have fused, robust nasals in comparison to other
theropod dinosaurs. Our interest is in asking why are the nasals fused. From the results of two-dimensional
Finite Element Analyses (FEA) of the skull of T. rex in lateral view, we interpret the fused nasals as an
adaptation against high stress during feeding. We carried out two-dimensional FEA on the dorsal view of
the T. rex skull for further examination of the mechanical function of the nasals. Two models of the fused
and unfused nasals were compared when experiencing shear load along the tooth rows. The results include:
i) shearing strain concentration in the nasals is significantly less and distributed over a larger area,
including the frontals and parietals, if the nasals are fused; and ii) shearing stress is high at the nasals in
both cases. Our analyses confirm that the fused nasals are mechanically suitable for coping with shearing
strain in tearing. The shearing strain in the nasals is improved by fused nasals. The fused nasals are
observed in basal tyrannosaurids such as Dilong paradoxus. It is suggested that morphological change of
the nasals that occurred in the tyrannosaurid lineage was related to the stress distributions created by
shearing strain during tearing.

The thalattosuchian crocodilian Pelagosaurus typus Bronn, 1841 from the Upper Lias of England
is properly documented and described for the first time. The material under study is part of a historical
collection founded by Charles Moore (1814-1881) at Strawberry Bank (Ilminster, England) in the late
1800�s. The identification of pelagosaurs in England extends the range of this species across the English
Channel. Pelagosaurus was a very small, exceedingly long-snouted, gracile crocodile whose diet probably
consisted of small fish, crustaceans and possibly insects. Laterally placed eyes suggest that this species
actively pursued its prey rather than sitting and waiting at the water surface. Predation on the small-bodied
fish Leptolepis is evident due to stomach contents of a small juvenile pelagosaur that confirm this
interpretation. The exact phylogenetic position of Pelagosaurus has been debated: however, evidence from
this current study reveals that this species is most closely allied with the family Teleosauridae. Pelagosaurs
possess the majority of teleosaurid apomorphies, including: four premaxillary teeth; small prefrontal;
lachrymal visible in dorsal view; presence of the mandibular fenestrae; dermal armour; and a straight tail.
Accordingly, Pelagosaurus can be seen as a primitive small-bodied teleosaurid.

The order Carnivora occupies many habitats, ranging from arboreal olingos to aquatic seals.
Despite their diverse locomotor styles, carnivorans are often divided into terrestrial fissipeds and aquatic
pinnipeds. Individual tarsals contain information about locomotor function. Interlocking facets indicate
flexibility; lengths of processes reveal mechanical advantages. Extensive data document the association of
form and function, but tarsal analysis has been anecdotal because these blocky bones are difficult to
quantify. Consequently, a method for 3D surface analysis was devised. A flexible tube of points � similar
to a fishnet stocking � was applied to the scanned 3D surface and analyzed using standard Procrustes
methods. Unlike ordinary geometric morphometrics, the results represent the complete bone, not an
abstract representation.

The terrestrial locomotion of fissipeds was found to constrain their tarsal morphology. Fissipeds
occupy a broad, distinct zone of morphospace from which pinnipeds departed. Mapping of
calcaneal morphology onto independent phylogenetic trees shows that phylogeny has retraced
paths across the zone many times, with only the pinniped branch escaping. But, the rate of
evolution was no higher in pinnipeds, indicating that their derived morphology is not due to rapid
evolution, but to evolution along a different trajectory. Abandoning terrestrial locomotion allowed
transformations selectively prohibited in fissipeds.

Since its discovery and initial description, Heterodontosaurus tucki has remained a
problematic animal in terms of its cranial biomechanics during feeding. This small herbivorous
dinosaur from the Early Jurassic upper Elliot Formation of South Africa features a closely packed
magazine of cheek teeth with oblique occlusal surfaces that suggest H. tucki processed plant
matter using a transverse power stroke. Because diapsid jaw musculature does not allow a side-toside
motion of the lower jaw against the upper jaw, several fundamentally different theories have
been put forward to explain how H. tucki chewed. These include rotation or scissoring of the
dentaries against the predentary, propaliny and lateral rotation of the cheek region. I am utilizing
FEA (Finite Element Analysis) to approach the problem of cranial biomechanics and feeding in H.
tucki in a way that will allow the various proposed hypotheses to be objectively tested. CT scans
from the cranium of H. tucki have been used in the construction of a three-dimensional finite
element model. Forces simulating those encountered during each of the proposed feeding
mechanisms will be applied to the model and stress and strain patterns generated by different
feeding strategies will then be analysed to determine which of the proposed modes of mastication
best corresponds to skull morphology.

In the quarry of Cherves-de-Cognac (Charente, Southwest France) forty-four marly and calcareous
levels are exposed. Dated from the Berriasian (Early Cretaceous), they are partly equivalent to the English
Purbeck Limestone Group and reveal an evaporitic coastal environment evolving from lagoonal to deltaic
conditions of deposition. Most of the levels yield an abundant and diversified microvertebrate fauna,
reaching 40,000 teeth per ton in the richest stratum. All vertebrate groups are present: osteichthyans,
chondrichthyans, amphibians, reptiles (turtles, crocodilians and dinosaurs), birds and mammals. Mammals
are exclusively present in one bed at a rate of 40 teeth per ton. They represent 0.15% of the dental microremains,
0.37% of the amniotes and 38.5% of the strictly terrestrial ones. Furthermore, they are diversified,
since from the 24 vertebrate families registered, four are mammalians: Triconodontidae, Spalacotheriidae,
Dryolestidae and one family of multituberculates. The quantitative study of the biodiversity of Cherves-de-
Cognac shows that these mammals constitute an important part of the ecological diversity.

Spinosaurid theropods such as Baryonyx walkeri possess a cranium unlike any other archosaur,
combining an elongated crocodilian-like snout and bony palate with a narrow yet tall domed skull complete
with antorbital fenestration. Here we present a study of the functional morphology of the spinosaur
cranium, with specific reference to the function of broad and tubular snouted extant crocodilians.
Simple Finite Element (FE) models of a broad, flat �platyrostral� snout and a tall, domed
�oreinirostral� snout were used to test the prediction that a secondary palate strengthens the skull whilst an
antorbital fenestra weakens it. Generally this is the case during bilateral and unilateral biting, and
oreinirostral snouts were almost always stronger than the platyrostral condition. To see if these predictions
held true for more complex morphology, a Finite Element (FE) model of the snout of Baryonyx walkeri
was created from CT scan information using �Simpleware� FE-model generating software. Further FEmodels
of the snout of Gavialis gangeticus and Alligator mississipiensis were created as representatives of
extant crocodilian taxa bearing a terminal rosette of teeth or a typical platyrostral snout respectively.
We assess whether spinosaurids were indeed �crocodile-mimics�, in the sense they were adapted
for torsional feeding behaviour with structural and functional parallels to extant crocodilian taxa. By
digitally manipulating FE-models we test i) if the spinosaurid bony palate confers structural strength to the
skull, and during what kind of feeding behaviour, and ii) if the antorbital fenestra weakens the snout,
suggesting non-structural constraints influence archosaur cranial evolution.

A recently found specimen of the protorosaurian diapsid Tanystropheus shows fossilised skin and
wide patches of black phosphatic material at the base of the tail. The latter are filled with small spherules of
calcium carbonate, which may precipitate within a corpse in an alkaline environment supersaturated with
carbonate, as may occur in stagnant water by decomposition of a consistent amount of proteins. This mass
of flesh at the base of the tail, surely added considerable weight to the posterior part of the body, shifting
the centre of mass posteriorly. Combining these new data with re-analysis of the body architecture of
Tanystropheus and comparisons with the recently discovered Chinese protorosaurian Dinocephalosaurus,
leads to the suggestions that: i) Tanystropheus was able to keep the neck raised off of the ground without
tilting forward and ii) that it lacked any evident adaptation for effective aquatic locomotion. Consequently,
it is feasible that Tanystropheus did not live permanently in water as previously suggested, but more
probably moved along the shoreline, using the long neck to survey water from above to avoid being spotted
by potential prey. This interpretation is also consistent with the stratigraphy and taphonomy of
Tanystropheus remains from the Grenzbitumenzone.

It has generally been presumed that the stapes of early tetrapods supported the braincase against the
palatoquadrate, with the notable exception of the Late Devonian Ichthyostega. This functional
interpretation is based on the stapedial morphology of the Early Carboniferous colosteid Greererpeton and
similar morphology in the Late Carboniferous embolomere Pholiderpeton.

The Late Carboniferous baphetid Kyrinion martilli, from the northeast of England, clearly does not
have a supporting stapes. CT scanning and 3D computer reconstruction of the internal cranial anatomy of
Kyrinion has revealed a dorsolaterally directed, somewhat laminar, stapes preserved in its original position.
The stapes articulated with the parasphenoid at its proximal end and also sat in the fenestra ovalis. Its distal
end did not contact the palatoquadrate complex. It is proposed that the stapes was in intimate contact with a
space between the braincase and the pterygoid, which presumably contained a spiracle. This arrangement
may have allowed the stapes to pass some airborne vibrations to the inner ear.

Kyrinion�s middle ear provides evidence for a functional transition to the tympanum-coupled
stapes believed to be present in temnospondyls; the first tetrapod group to evolve an apparatus adapted to
the perception of airborne sound.

A major taxonomic split occurred within the Squamata early in its history, dividing the clade into
Scleroglossa and Iguania. One fundamental difference between them concerns locomotor ability; many
iguanians show an ability to run bipedally, whereas, with few exceptions, scleroglossans generally do not.
Conversely, limb reduction or loss has occurred repeatedly in scleroglossans, a characteristic seen in no
other lepidosaurs.

The pelvis is fundamental to tetrapod locomotion; it is a point of attachment for important hindlimb
and axial musculature. Its shape is therefore likely to be influenced by locomotor habits. There is, however,
also the question of descent: one might reasonably expect two groups sharing a recent common ancestor to
share morphological and physiological similarities.

These issues have been explored through geometric morphometric analysis of the shape of the
ilium. Ilia from a wide variety of lizard species were photographed, digitised, and then subjected to
Principal Components Analysis. Consideration of the complete data set showed a distinct phylogenetic
clustering at the family level (i.e., iliac morphology is generally conservative within a family), whilst
functional effects were also observed in groups employing specialised locomotion (e.g., chameleons, and in
arboreal versus terrestrial varanids). Clear shape changes are also revealed which mirror the trends towards
limb reduction and bipedality.

Feeding function is, perhaps, one of the most interesting aspects of the cranial skeleton and many
studies have been conducted for both extant and extinct animals. The general skull morphology of theropod
dinosaurs is widely consistent, but there is also a significant amount of diversity. Differences in jaw margin
morphology indicate a variety of feeding applications, strategies, and behaviours. However, despite
speculation, the mechanical significance of this diversity has rarely been analysed and compared.
1The hypothesis that jaw margin morphology affects bite force distribution along the jaw margin has been
tested through basic mechanics with the aid of computer simulations. Jaw margins were approximated as
simple mathematical equations and bite forces were calculated along their entirety, which in theory reflects
the gradual changes and distributions of bite forces. Calculations showed that bite forces are transmitted
most efficiently in accordance to certain jaw morphologies. There is also a general trend for the highest
efficiency to be in the anterior regions of the jaws, where bite forces are naturally and unavoidably lower.
This analysis showed that jaw margin morphology does affect bite force distribution and in some cases
compensates for anatomical loss of bite force.

A parsimony analysis was performed after sampling 17 recent talpid genera for 157 morphological
characters, using shrews and hedgehogs as outgroups. Independent of which alternative coding for some
contested dental homologies were used, the main analysis resulted in one most parsimonious tree of 473
steps. A series of features support talpid monophyly, for which Uropsilus is the most basal taxon. A
Japanese shrew mole clade (Urotrichus, Dymecodon) is placed after Uropsilus, and excludes Neurotrichus.
The monophyly of the desmans is well supported. Scaptonyx is sister group to a monophyletic Eurasian
fossorial mole clade, while Scaptochirus is sister group to Mogera. The monophyly of Scalopini was
supported, with Scalopus, Scapanus as sister groups. Based on our results and those of previous authors, it
is clear that several migrations must have occurred between Eurasia and North America during the
Cenozoic, and that strict fossoriality evolved at least twice independently in talpids. Evolutionary plasticity
and propensity to migrate across continents is further demonstrated by preliminary examination of the
talpid fossil record in a phylogenetic context.

The independent evolution of a bony tube enclosing the stapedial artery is likely to be correlated
with adaptations to low frequency hearing. Another rare feature in the middle ear of several species is a
bullate stapes. Different degrees of hypsodonty characterize different taxa. A lot of morphological variation
in ankle characters was found, which can be used in isolation to distinguish most clades. The enlargement
of prehallux and prepollex are likely epigenetically coupled.

The Osteostraci (cephalaspids) are a specialised and intriguing group of agnathan fishes that are
currently thought to be the sister taxon of the jawed vertebrates. Although lacking jaws, the majority of
Osteostraci possess a number of gnathostome apomorphies, including paired appendages, which some
authors argue to be homologous to the paired appendages of gnathostomes.

Phylogenetic analysis of the intra-relationships of this group can help us answer questions about
this proposed homology. Whilst previous authors have investigated such intra-relationships, in-depth
systematic studies utilising modern cladogram construction techniques such as PAUP have yet to be
conducted on the group. Such studies will also shed light on the debate as to whether paired appendages
evolved once and were subject to reversions or evolved many times within and outside of the group.
The Silurian thyestidiens are a key monophyletic group containing both forms with paired
appendages and without. They have long courted controversy regarding their position as either the most
derived or most basal osteostracans. Here, a new phylogeny of the thyestidiens is presented based on a
study of British, Estonian, Russian and Scandinavian material. This phylogeny provides the basis for an
analysis of the interrelationships of this group and to trace the evolution of key vertebrate innovations,
crucially including paired appendages.

Sauropods show a variety of different neck types, ranging from moderately long to extremely long
necks. Soft-part reconstructions, biomechanical investigations and calculations of neck flexibility are used
to work out constructional differences in these neck types and the biomechanical role of pneumatic spaces.
Sauropod neck constructions mainly differ from each other by the proportions of cervical vertebrae and
proportions and morphology of cervical ribs, which influence the mobility of the necks.
In the neck of sauropods, a dorsal, lateral and ventral muscle mass can be reconstructed of which
the ventral one was the smallest. Along with the ventral cervical muscle mass and a cranially oriented
scapulocostal muscle, ventral flexion of the neck was conducted mostly by gravity. The dorsal ligament
system elastically suspended the neck and guaranteed elastic recoil after lateral and ventral movements.
The cervical vertebrae were surrounded and penetrated by pneumatic diverticula, which also interdigitated
with neck muscles. The reconstructed arrangement of pneumatic diverticula around the cervical vertebrae,
in particular hose-like pneumatic diverticula inside the gap of bifurcated neural spines, suggests that the
cervical pneumatic system in sauropods could have contributed to neck bracing as pneumatically
stabilizing structures.

The time and place of origin of the crown strepsirrhine or �toothcombed� prosimian clade
containing the Malagasy lemurs, African galagos, and Afro-Asian lorises has long been a matter of great
interest to palaeoprimatologists. There is general agreement among specialists that the extinct Laurasian
adapiform primates (notharctids, adapids and sivaladapids) are stem strepsirrhines, but it is not clear
whether these taxa are paraphyletic with respect to crown Strepsirrhini or if they represent that clade�s
monophyletic sister group. Here, I present the results of a phylogenetic analysis of over 100 living and
extinct primates based on 359 craniodental, postcranial and soft tissue characters, which tests various
hypotheses bearing on the nature of the proposed crown strepsirrhine-adapiform relationship.
Interrelationships of Laurasian adapiforms (particularly notharctids) are found to be unstable, but there is
no support for the hypothesis that these taxa are paraphyletic with respect to crown Strepsirrhini. Instead,
crown strepsirrhines are found to be nested within a paraphyletic group of Afro-Arabian stem strepsirrhines
(e.g., Djebelemur, �Anchomomys� milleri and Plesiopithecus), some of which have evidently been
misidentified as members of otherwise Eurasian radiations. These results support the hypotheses that crown
Strepsirrhini arose in Afro-Arabia and that the lemuriform presence in Madagascar is due to dispersal.

Among the ichthyofauna of Pietraroja, Ichthyodectiformes have been reported only from a post53rd
cranial partial skeleton attributed to the species Chirocentrites cf. coroninii, at the beginning of the 20th
Century. Recently, the finding of a complete and articulated skull and pectoral girdle has been attributed to
the family Ichthyodectidae and confirmed the fossil record of this family in the Lower Cretaceous of
southern Italy. Preliminary researches on the cranial material assign it to the genus Cladocyclus.
The preservation of the specimen is exceptional, as it is usual with fossils coming from the
Pietraroja area, as it preserves not only a fully articulated skull and pectoral girdle, but also soft tissue
preservation of muscles around the ribs. Most possibly this specimen is the most complete ichthyodectid in
the southern Italian fossil record. Geochemical studies seem to indicate that the specimen has been
phosphatized, which may be a common process of fossilization in the Pietraroja area, although no other
study in this sense has been carried out in the area. Furthermore, even the other material attributed to
Chirocentrites coroninii is under examination, and preliminary results exclude this genus and point out that
the old material may actually be attributed to the genus Cladocyclus as well.

Theropod forelimb claws were often the main weapon during predation, therefore the
understanding of their structure and use seems to be of great relevance for the study of dinosaur
palaeobiology. We have examined the forelimb claws of 30 theropods from a geometric point of view and
have found that the basic descriptor for all of them is represented by a 3rd degree polynome. In order to
create control groups we have applied the same kind of analysis to forelimb claws of different living
predatory taxa (felids and ursids), but for these samples we have found a number of different descriptors in
the same family, quite different from the uniform results obtained for theropod dinosaurs. Similar results
have been found for hindlimb claws of birds, which also show several different descriptors. Therefore, all
the theropod forelimb claws can be described within a single family of curves, all generated from a 3rd
degree polynome, which may have either functional morphological and/or ontogenetic importance.

The study of pterosaur bone histology dates back to the middle of the 19th Century, but systematic
studies only began recently. Pterosaur bones are predominantly composed of highly vascular fibrolamellar
bone, indicating that pterosaurs grew rapidly. However, cranial and pedal bones contain LAGs (lines of
arrested growth) that record pauses in bone deposition. Pterosaurs had determinate growth and deposited an
endosteal lamella and a periosteal EFS (external fundamental system) at maturity. Pterosaurian epiphyseal
growth plates contain endosteal bone and calcified cartilage. Although endosteal reworking is extensive,
secondary osteons are rare. Some smaller elements contain an orthogonal plywood-like tissue, composed of
alternating lamellae, which may have biomechanical significance. Not all bones are supported internally by
trabeculae; in some cases endosteal ridges may provide reinforcement. Pterosaurian reproductive mineral
dynamics probably did not involve the deposition of a specialised bony tissue within the lumen of the long
bones. Comparative bone histology can distinguish between bones of small theropod dinosaurs and
pterosaurs, but cannot reliably separate birds from pterosaurs or distinguish between pterosaur taxa.

In 1930, an expedition led by Frederick Migeod excavated a large sauropod dinosaur at Tendaguru,
Tanzania. This specimen, BMNH R5937, has attracted surprisingly little study: it has been the subject of
only one significant publication, and none of the material has been figured. Migeod's (1931) account
described the specimen as �of the Brachiosaurus type�, but also mentioned non-brachiosaurid features
including bifurcated neural spines and �wings� in the anterior dorsals. If his interpretations are reliable,
other proportions are also non-brachiosaurid, e.g., humeri only two thirds as long as the scapulocoracoid.
The prepared material includes two dorsal and four cervical vertebrae, two further dorsal centra and an
unidentified long-bone fragment. The cervicals are proportionally longer than in B. brancai. Unopened
jackets contain additional material including at least three more cervical vertebrae, a �wing� from a dorsal,
sacrals, a scapula and an ilium. Of the prepared material, the two dorsals are the best preserved, and
probably represent D8 and D9. They have several laminae not usually associated with Brachiosaurus,
including long, parallel spinoprezygapophyseal laminae and a configuration of the divided
spinopostzygapophyseal laminae in which the median branches merge into a postspinal lamina and the
lateral branches join the spinodiapophyseal laminae in a compound lateral lamina. However, lamination
varies greatly along the dorsal column of the B. brancai type specimen, so comparing BMNH R5937 with
it is difficult. The more complete dorsal column of Cetiosaurus oxoniensis also shows significant variation
between adjacent dorsals, casting doubt on the widely assumed taxonomic informativeness of vertebral
laminae.

In recent years Lower Cretaceous sediments of the Jehol Group in northeast China have yielded
many spectacular vertebrate fossils, among them more than 100 specimens of pterosaurs. Important finds
include eggs with embryos and several examples of exceptionally well preserved soft tissues, but the most
striking aspect of this pterosaur assemblage is its unparalleled diversity. Representatives of two clades of
basal pterosaurs (Anurognathidae and Scaphognathinae) and almost all major pterodactyloid clades
(Ctenochasmatinae, Gnathosaurinae, Lonchodectidae, Tapejaridae, Azhdarchidae, Istiodactylidae,
Ornithocheiridae and ?Pteranodontia) have been found in these deposits and form the most diverse
pterosaur assemblage yet known. This assemblage demonstrates that pterosaurs were an important
component of continental vertebrate biotas, while the widely varying size and morphologies of the 12+
species described so far hints at considerable ecological complexity within Early Cretaceous pterosaur
communities. The Jehol Biota also shows that these pterosaurs lived at the same time and in the same
environments as a remarkably diverse assemblage of birds. This contradicts the view that the basal
radiation of birds led to the displacement of pterosaurs from many habitats and suggests that, in terms of
their evolutionary history, there was little direct interaction between these two major groups of flying
vertebrates.

Chinshakiangosaurus chunghoensis is an enigmatic and poorly understood sauropodomorph from
the Lower Jurassic Fengjiahe Formation of Yunnan, China. Examination of the left dentary of
Chinshakiangosaurus reveals that it possesses an unusual combination of �prosauropod� and �sauropod�
character states. Cladistic analysis places Chinshakiangosaurus as one of the most basal sauropods known
currently. Mapping of dentary and dental characters onto the most parsimonious topologies yields insights
into the sequence of acquisition of a number of feeding-related characters. For example, it seems that basal
sauropodomorphs (traditional prosauropod taxa) possessed a fleshy cheek that attached to the mandible
along a marked ridge, and that the same structure was present in the most basal sauropods. The early
sauropod skull developed a lateral plate that reinforced the bases of the tooth crowns labially, and
possessed teeth with �wrinkled� enamel and a concavity on the mesial portion of the lingual part of each
crown, while retaining a fleshy cheek and a relatively weak symphysis. More advanced sauropods
(eusauropods) lost the cheek, perhaps in order to increase the gape of the jaws in response to a change in
feeding style that involved collection of larger quantities of poor quality foliage.

Micromammalian fossils belonging to 15 taxa have been collected from the Osborne Member, Late
Eocene, northwest Isle of Wight (UK), and the taphonomic history of the two most abundant species,
Isoptychus sp. and Thalerimys fordi, has been studied. The two theridomyid species show similar patterns
of mortality, element representation, breakage, etching and weathering and, therefore, share similar
taphonomic history. The majority of the individuals belong to the vulnerable age groups (young and old)
and died of biological causes. The most likely process for the accumulation of the remains is predation by a
mammalian carnivore. A premolar of a mammalian carnivore, Cynodictis cf. lacustris, was found in the
horizon and this animal is the most probable predator of the theridomyids. Extended fragmentation of fresh
bone and puncture marks are attributed to chewing by the predator. Some of the material was exposed on
the surface for some time before burial and during that period some of the specimens were chewed by the
glirid Glamys priscus, also present in the fauna. The remains may have been subjected to trampling, as
indicated by breakage of dry bones and absence of complete large bones. The theridomyid specimens have
not been transported for long distances by water and, thus, the two rodents belong to the local
micromammalian community.

Diversity, locomotor adaptations and evolution of the marsupial postcranial skeleton are rarely
analysed on a broad scale. Nevertheless, evolutionary, functional and palaeontological inferences resting on
a representative sample of any clade are valuable. This study investigated intrinsic hand proportions in the
diverse and speciose marsupial order of Diprotodontia. Fifty-two measurements of 90 specimens
representing 50 species were taken, size-adjusted, and evaluated by Principal Components Analysis (PCA).
Confidence for variable loadings was established by determining their standard error using a bootstrap
method. The first principal component (PC1) distinguished between arboreal and terrestrial species, and
was mainly correlated with outer finger length and median width. The second principal component
separated several major diprotodontian higher-level clades. Comparison of phylogenetically independent
contrasts created using a composite phylogeny showed significant differences in PC1 (arboreal vs.
terrestrial) values between possums, koalas and the remaining Diprotodontia. This suggests that long and
slim phalangeal elements and ectaxony are adaptations to arboreality, as suggested by of numerous nonphylogenetic
functional studies on mammals. The best lifestyle predictors, useful even for single
phalangeal elements, are ratios between most positive (phalangeal length) and most negative (phalangeal
width) PC1-correlated measurements. In contrast, the commonly used ratios between proximal and
intermediate phalanges did not distinguish between arboreal and terrestrial taxa. Several other potential
adaptations, e.g., for burrowing or gliding, could not be tested in a phylogenetic framework due to the small
sample size, but are congruent with findings across tetrapods. This study shows that in the diprotodontian
hand, similarities in lifestyle result in anatomical similarities, suggesting that convergence is common.

The relationship between attractiveness and markers of masculinity in the male face is not clearly
understood, both in the context of facial growth and development and with regard to hypotheses of sexual
selection. There is considerable variability reported for female preferences for male faces, but new
evidence from study of facial growth suggests that �masculinity� could have been misrepresented. A
comparison of male and female ontogenetic trajectories, calculated from dimensions recorded from the
facial and basicranial skeleton of a native Southern African population of Homo sapiens, tests whether
growth, uniformly across the face, complies with a model of ontogenetic scaling in men and women. This
study identifies one region of the upper face (height of the nasomaxillary complex) that does not
correspond to predictions based on the ontogenetic scaling of facial measures. This shortening of the male
nasomaxillary region, a change associated with puberty, indicates that sex steroids can affect regions of the
face differently, but not in relation to previously identified �hormone markers�. Furthermore, this facial
dimorphism is clearly evident in Homo erectus, and fossil crania can be sexed accordingly. The findings
imply that the adolescent skeletal growth spurt, uniquely derived in humans relative to other primates, had
evolved in Homo erectus. Male facial attractiveness and female preferences seem to have played a more
important role in shaping human evolution than has been previously recognised.

The potential for gaining a deeper understanding of pterosaur locomotion - both aerial and
terrestrial - has greatly increased in recent years, thanks to the discovery of several near-complete, threedimensionally
preserved fossil skeletons in the Santana Formation of Brazil, most of them belonging to the
predominantly Early Cretaceous family the Ornithocheiridae. Given the usual crushed state of pterosaur
fossils, these 3D specimens offer an unprecedented opportunity to study joint function. Using ten such
specimens and a novel functional morphological technique, a representative ornithocheirid (Anhanguera
santanae) was accurately reconstructed in three dimensions, and the range of movement at the various
joints was determined. The analysis has revealed a number of hitherto unknown and unique aspects of
pterosaur flight control, in particular the role of the pteroid bone - a slender element that articulated at the
wrist - and the use of shoulder rotation. In birds and bats the forelimb is typically rotated back to increase
the angle of attack of the wing for low-speed flight, and rotated forwards for high-speed flight. The
opposite may have been the case in pterosaurs.

In summer 2004, digging activity in a quarry near the M�nchehagen tracksite, close to Hannover,
revealed a large bedding surface with abundant tridactyl dinosaur tracks. To date, approximately 80 m2 of
the new tracksite has been uncovered, yielding five trackways belonging to the ichnogenus Iguanodontipus
and two �allosaurid� theropod trackways. The new tracksite is highly significant because of i) its excellent
preservation of theropod and ornithopod tracks, ii) the general scarcity of longer trackways in the Early
Cretaceous of northern Germany, and iii) the unusual walking directions of the track-producers.
The trackway layer is a fine-grained mudstone with superbly preserved tracks as well as ripple
marks. The iguanodontid tracks (n = 37; longest trackway: n = 18) measure 24-44!cm in length and width,
thus corresponding to sub-adult animals with a body length of about 5!m. Interesting features of the
trackways are �gliding� structures and tracks, where the deep mud was possibly squeezed around and
between the toes and hoofs, respectively, during movement. These structures together with the short pace
(49-77!cm; mean: 69!cm) and stride length (104-156!cm; mean = 134!cm) might indicate that the
iguanodontids walked carefully in the unstable sediment. All Iguanodontipus trackways run in different
directions: three of them cross each other; none of them represent a straight line of walking.
One of the theropod trackways consists of five tracks, the other of two tracks. The width of the
tracks is 23-27!cm (mean = 24!cm); the length is 28-40!cm (mean = 35!cm). The pace of the longer
trackway is 102-113!cm (mean = 107!cm) and the stride is 210-220!cm (mean = 216!cm), indicating a
relatively fast running animal. Ongoing research will clarify if these small theropod trackways represent a
new ichnogenus.

The diverse turtle fauna from the Urumaco Formation (late Miocene) includes the world�s largest
extinct turtle Stupendemys geographicus, the first reliable reported occurrence of a trionychid in South
America, a fossil matamata species, and at least two species of pelomedusoids, including Bairdemys
venezuelensis. New fossils of Bairdemys show a great deal of cranial variation in features previously
thought to be diagnostic for particular species. The lower jaw of Bairdemys has a large triturating surface,
which together with the great development of a secondary palate is probably correlated with a diet
consisting of hard molluscs. A large nesting site demonstrates that B. venezuelensis was a colonial nester,
which laid its eggs in beaches and lived in a near-shore marine environment, providing the first direct
evidence of the palaeoecology of this species. Several new carapaces and postcranial elements of
Stupendemys provide new information about the palaeobiology of this species. A histological study of a
new carapace (CL 3.3mts) reveals that apparently this species did not follow any outstanding way of
growth to reach its gigantic size. Sharpey`s fibres are present only in the external cortex of the neural
hinting at some kind of anchoring of the overlying soft tissue (probably the upper strata of the dermis) that
resides just below the keratinous shields. Comparisons of postcranial anatomy with fossils attributed to
Stupendemys from Brazil, show that at least two species of this genus inhabited in the Miocene of the
neotropics. This work was supported by the National Geographic Society.