Distribution patterns of eye-lenses are compared from forty-two preserved visual surfaces of the new phacopine trilobite Phacops (Phacops) tafilaltensis n. sp. from the Late Devonian of Tafilalt, Morocco. In this species, variation in the lens pattern is most common on the ventral and posterior parts of the visual surface. Post-larval specimens typically have seventeen or eighteen files of lenses, but the number present is not correlated with specimen size. Several individuals display seventeen files in one eye and eighteen in the other. The smaller number of files in one eye may reflect incomplete development of that visual surface. The length of the active section of the generative zone and its pattern of growth control the number of dorso-ventral files of lenses and their relative height. Cases of asymmetry in lens distribution are observed in six individuals. The capacity of these organisms to correct both environmental or genetic stresses appears to be weak for this character. The existence of individuals with eyes exhibiting a different number of files confirms that file number is the result of a genetically controlled developmental program. In Phacops (Phacops) tafilaltensis n. sp., the lenses are normally arranged in a regular pattern of hexagonal close packing, their size varies within the visual surface of individual specimens and their number increases during ontogeny. Variation in the timing of lens emplacement accounts for the observed irregularities in lens distribution, i.e. in ventral and posterior parts of the visual surface. Characteristic patterns of the lens arrangement on the visual surface relate to systematic categories and probably reflect evolutionary relationships.

The development of an atypical Hirnantia Fauna in the late Ordovician of Gondwana was coeval with a slow eustatic fall induced by the abstraction of water into a growing ice sheet. This event is dated as early Hirnantian in age and occurred in tandem with the start of a major mass extinction. A tectonic episode in the Caradoc-Ashgill of North Africa differentiated the continental shelf into highs and lows and may have formed the land required for the accumulation of a permanent snow cover. Depositional lows were filled by regressive shallow-marine deposits in the early Hirnantian. During the mid-Hirnantian, advance and retreat of an ice sheet on the continental shelf resulted in the deposition of glaciomarine sediments above these regressive deposits. The demise of an atypical Hirnantia Fauna is attributed to deglaciation and the associated flooding of the continental shelf by a stratified anoxic water column. This glacioeustatic sea-level rise occurred in the late Hirnantian.

U-Pb zircon geochronology of Mesoproterozoic (Subjotnian) rapakivi complexes in central Sweden yields: 1526 ± 3 Ma (Mullnäset), 1524 ± 3 Ma (Mårdsjö), 1520 ± 3 Ma (Nordsjö) and 1497 ± 6 Ma (Rödön). Together with complexes further S in Sweden, they constitute the westernmost, youngest (1·53−1·47 Ga) belt of rapakivi magmatism in the Fennoscandian shield.

The low initial εNd values (−8·9 to −4·8) of all studied Subjotnian basic, intermediate and silicic rocks, require an input from an old (Archaean) low-radiogenic source component, as evidence for Palaeoproterozoic protoliths in the age range 2·5−2·1 Ga is lacking in this region. Crustal, early Svecofennian + Archaean (roughly 30−40%) sources are suggested for the Subjotnian A-type granites and syenites, where the granites derive from undepleted, granodioritic, and the syenites from monzodioritic (±depleted crustal) protoliths. The basic rocks originate from a depleted mantle acquiring the enriched Nd isotopic signatures during interaction with an Archaean lower crust (20−40%), largely depleted after rapakivi melt extraction. Pb isotope data from feldspars (207Pb/204Pb to 15·018−15·542) support the presence of Archaean components in the magmas.

The results indicate that an Archaean basement is underlying relatively wide areas of Svecofennian formations in central Sweden. This old basement section was most likely rifted off the Archaean craton in the NE in Palaeoproterozoic times.

The tetrapod Caerorhachis bairdi, probably from the Pendleian Limestone Coal Group in the Scottish Midland Valley, is rediagnosed and redescribed, and its affinities are discussed. Caerorachis was originally interpreted as a temnospondyl amphibian, based on characters that are now regarded as primitive for tetrapods, or of uncertain polarity. Several features of Caerorhachis (e.g. gastrocentrous vertebrae, curved trunk ribs, reduced dorsal iliac blade, L-shaped tarsal intermedium) are observed in certain primitive amniotes. In particular, Caerorhachis resembles ‘anthracosaurs’, generally considered to be among the most primitive of stem-group amniotes.

The phylogenetic position of Caerorhachis is considered in the light of recently published cladistic analyses of Palaeozoic tetrapods. Most analyses place Caerorhachis at the base of, or within, ‘anthracosaurs’. When multiple, equally parsimonious solutions are found, its ‘anthracosaur’ affinities are shown in at least some trees, and are supported by several informative and, generally, highly consistent characters. Alternative phylogenetic placements (e.g. sister taxon to temnospondyls) are usually less well corroborated.

If the fundamental evolutionary split of most early tetrapods into stem-group lissamphibians (e.g. temnospondyls) and stem-group amniotes (e.g. ‘anthracosaurs’) is accepted, then the revised interpretation of Caerprhachis sheds light on near-ancestral conditions for Amniota.

Little is known regarding the internal anatomy of thelodonts, with most of the available information coming from a single specimen, the holotype of Turinia pagei (Powrie). Previous descriptions have led to many partial, or conflicting interpretations. Herein, we describe fully the anatomy of T. pagei based on the holotype and additional material. T. pagei possessed a branchial system composed of eight pairs of gills, a buccal/nasohypophyseal region lined with minute denticles, comparable to buccopharyngeal denticles of sharks, and possessed a stomach which is preserved by sediment infill in the holotype specimen. Contrast between the petrological character of the gut infill and the sediment in which the animal is preserved suggests that the gut was infilled in vivo and that T. pagei was probably a deposit feeder. Phylogenetic analysis resolves T. pagei and the Galeaspida as sister-taxa, comprising a sister-group to the Osteostraci plus jawed vertebrates. In contrast to the view that has prevailed hitherto, thelodonts with a dorsoventrally compressed cross-sectional profile comprise a monophyletic group, of which T. pagei is the least derived member. The furcacaudiforms are resolved as an unnatural group, one taxon being the sister taxon to the ‘conventional’ thelodont clade, and the other, the sister taxon to this clade plus galeaspids, osteostracans and jawed vertebrates. The analysis agrees with the earlier view that thelodonts lack distinct synapomorphies, but we argue that distinct synapomorphies are not a requisite of monophyly.

A study of a sample of hipparion teeth from Pliocene marine fossil deposits in Suffolk revealed two species, belonging to different genera of these three-toed horses. Hipparion sp. (of the H. crassum Group) and Proboscidipparion sp. have different stratigraphic ranges on the European continent. Their stratigraphic origin in the Suffolk deposits is discussed, based on a review of the occurrence of similar forms on the European continent.

The sculpture of scales and plates of articulated anaspids from the order Birkeniida is described and used to clarify the position of scale taxa previously left in open nomenclature. The dermal skeleton of a well-preserved squamation of Birkenia elegans Traquair, 1898 from the Silurian of Scotland shows a characteristic finely tuberculated sculpture over the whole body. Rhyncholepis parvula Kiær, 1911, Pterygolepis nitida (Kiær, 1911) and Pharyngolepis oblonga Kiær, 1911, from the Silurian of Norway show three other sculpture types. Northern Hemisphere disarticulated scales and plates are described here, supporting a new anaspid taxonomy that includes both articulated and disarticulated material. The diversity, distribution, evolutionary trends and biostratigraphy of anaspids are described in the context of this new taxonomy, which includes six families (two are new) subdivided into 16 genera (10 are new) and 22 species (15 are new).

New taxa among Birkeniidae Traquair, 1898 are Birkenia robusta sp. nov. and Hoburgilepis papillata gen. et sp. nov.. Rhyncholepididae Kiær, 1924 includes Rhyncholepis butriangula sp. nov., Silmalepis erinacea gen. et sp. nov., Vesikulepis funiforma gen. et sp. nov., Maurylepis lacrimans gen. et sp. nov., and the previously described Schidiosteus mustelensis Pander, 1856 and Rytidolepis quenstedtii Pander, 1856. Tahulalepididae fam. nov. is represented by Tahulalepis elongituberculata gen. et sp. nov. and the revised T. kingi (Woodward, 1947). Septentrioniidae fam. nov. contains Septentrionia lancifera gen. et sp. nov., S. mucronata gen. et sp. nov., S. dissimilis gen. et sp. nov., S. seducta gen. et sp. nov., Liivilepis curvata gen. et sp. nov., Spokoinolepis alternans gen. et sp. nov. and Manbrookia asperella gen. et sp. nov. The family level position of Ruhnulepis longicostata gen. et sp. nov. is uncertain. Pterygolepididae Obruchev, 1964 and Pharyngolepididae Kiær, 1924 remain monogeneric.

Immediately W of the islands of Uist in the Outer Hebrides is a wide, low-gradient submarine shelf on which glacifluvial deposits of Devensian age and calcareous sand accumulated. Following moulding by Devensian ice, the Lateglacial landscape of the littoral zone of the Uists became a series of low-lying bedrock ridges and basins. The analysis of sub-tidal organic deposits has shown that in the early Holocene this landscape supported water bodies, marshes and a vegetation mosaic of Betula-Corylus woodland and Calluna vulgaris-herbaceous taxa open ground. The submergence of this littoral area during the Holocene Marine Transgression, together with wave action typical of the position on the Atlantic margin, led to the transfer onshore of submarine shelf deposits, so creating machair (sand plain) landscapes. This brought about vegetational changes, eventually creating calcareous grasslands. The timing of these events was asynchronous, being location- and site-specific due to variations in the configuration of the littoral zone. Although the date of the initial transfer of sand is unknown, evidence from the sub-tidal deposits indicates that a major incursion of sand, in North Uist, occurred c. 7600 BP (8450–8340 cal BP). The same source also suggests that a further major sand movement took place during the period 5800–4200 cal BP, a period of widespread sand drift in NW Europe. The analyses of the sub-tidal deposits have also reinforced the current theory of machair evolution.

A re-examination of the lectotype of Sus giganteus Falconer & Cautley, 1847, reveals that it is a suine, as originally thought by its discoverers, and not a tetraconodont as recently published. It may belong to the genus Propotamochoerus as proposed by Azzaroli in 1989. This reassessment radically modifies tetraconodont systematics published recently by Van der Made (1999).

Devonian rhizodontid fishes from East Gondwana include Gooloogongia loomesi Johanson & Ahlberg 1998 from near Canowindra, New South Wales (NSW), Australia and Aztecia mahalae gen. nov., sp. nov. from southern Victoria Land, Antarctica. Gooloogongia loomesi is known from well-preserved cranial and postcranial material, and as such represents the most complete rhizodont known. Newly discovered elements (braincase, pelvic girdle) are described for the first time. Aztecia mahalae gen. nov., sp. nov. is based on shoulder girdle material formerly assigned to Notorhizodon Young et al. 1992, which, based on skull and lower jaw morphology, is a member of the Tristichopteridae. The presence of several plesiomorphic characters suggests that Gooloogongia occupies a basal phylogenetic position within the Rhizodontida, while Aztecia n. gen. possesses a more derived shoulder girdle. We argue that the Rhizodontida evolved on the Gondwanan landmass. Notorhizodon occupies a relatively derived position within the Tristichopteridae, but is contemporary with the earliest and phylogenetically most basal Laurussian members of the group. This shows that the tristichopterids achieved a worldwide distribution considerably earlier than previously thought.

A new genus and species of primitive echinozoan, Bromidechinus rimaporus, is described from the Upper Ordovician of Oklahoma, USA. This has a unique plate arrangement. There is a single perradial series of imperforate plates bounded on either side by a column of perforate ambulacral plates. A double column of interambulacral plates separates ambulacral zones. The sparse record of Ordovician echinozoans is reviewed and cladistic analysis suggests that Bromidechinus represents a lineage that diverged prior to the split between bothriocidarids and main-line echinoids. This leads to a revised interpretation of the earliest stages of morphological evolution of echinoids.

The first whipspider (Arachnida: Amblypygi) and three new examples of whipscorpions (Arachnida: Uropygi: Thelyphonida) are described from Lower Cretaceous (Aptian) strata of the Crato Formation, Ceará State, Brazil. The whipspider is described as Britopygus weygoldti gen. et sp. nov. It resembles members of the extant family Charinidae, but cannot be unequivocally placed. Two of the whipscorpions are referred to Mesoproctus rowlandi Dunlop, 1998 and one could be a subadult, since, like juveniles of extant species, it has a relatively broad anterior sternum. There are differences between these two specimens in pedipalp and flagellum morphology, but both characters are variable within extant species. The third and largest whipscorpion is incomplete and referred to Mesoproctus sp. Its carapace length is equal to or exceeds that of the largest known extant genus Mastigoproctus Pocock, 1894.

The stem tetrapod Eucritta melanolimnetes is described in full from five specimens. The animal is compared in detail with the contemporary and superficially similar temnospondyl Balanerpeton woodi, and with baphetids (=loxommatids). Growth of Eucritta skulls approaches isometry, and lies on the same trajectory as skulls belonging to the genus Baphetes. The function of the temporal embayment is discussed; it was probably not associated with an air-borne hearing device. The axial skeleton of Eucritta was ossified in larger specimens. Ribs in Eucritta gain expanded distal ends with growth, and the animal probably breathed by buccal pumping. An updated cladistic analysis is presented, incorporating data from a fifth specimen. Eucritta remains as a probable basal baphetid, but the sister group to the baphetids remains uncertain.

The fossiliferous bed in the Upper Comley Sandstone, described by Cobbold & Pocock (in 1934) as the ‘Paradoxides forchhammeri Grit’ and correlated by them with the Andrarum Limestone of the late Middle Cambrian (P. forchhammeri Stage) in Scandinavia, is shown to be a shallow-water development of the older punctuosus Zone (medial Middle Cambrian, P. paradoxissimus Stage). It is here re-named the ‘Rushton Brook Bed’. The trilobite fauna, formerly considered comparable with that of the Andrarum Limestone, is reviewed and found taxonomically distinct, the supposed similarities being attributable to facial and preservational factors.

This paper reconstructs the characteristics of rivers which deposited proglacial fluvial sediments in east-central Scotland during the Late Devensian. Sediment depositional architecture and geometry, together with the relationship between high-stage and falling and low-stage depositional elements, were used to relate the proglacial sediments to the glacial meltwater discharge regime. The proglacial river systems studied were dominated by ‘normal’ ablation controlled discharge, rather than by high magnitude flood events. Consequently there is a great deal of spatial and vertical variability. Deposition occurred during short intervals of rapid aggradation, so that relatively fine-grained falling-stage sediments, as well as coarser, bar-core sediments are well preserved. Models relating the characteristics of the final deposit to the nature of the river are presented. These emphasise the role of stage changes and aggradation rates in controlling sediment architecture in braided fluvial deposits.

A new genus and species of felid, Ginsburgsmilus napakensis, is described from Napak, Uganda. Material from Songhor, Kenya, previously identified as Afrosmilus turkanae by Schmidt-Kittler (1987) is assigned to this species. A new species of Afrosmilus, A. hispanicus, is described from Artesilla, Spain. These new felid species combine sabre-tooth canines of maochairodont type with cheek teeth similar to those of Pseudailurus. Phylogenetically they are close to Afrosmilus turkanae which shows a greater specialisation towards markedly machairodont forms. On the basis of dental features, we propose a close relationship between these two genera (which we group into the new tribe Afrosmilini) and the tribe Barbourofelini (Sansanosmilus and Barbourofelis), the transition between the two being represented in the fossil record by Prosansanosmilus peregrinus Heizmann et al. 1980.

Pegmatites at the East Kemptville Sn-Zn-Cu-Ag deposit occur within a F- and P-rich, 370 Ma topaz-muscovite leucogranite, the most evolved phase of the chemically zoned Davis Lake Pluton. Structural observations and geochronology indicate that the leucogranite intruded into an active shear/fault zone environment. Pegmatites are preferentially located near the roof zone of the intrusion and include: (1) quartz-feldspar pods (≤ 1–2 m) aligned parallel to a foliation in the leucogranite. Such pegmatites have marginal aplites or may be cored by aplites; (2) aplite-pegmatite zones with well-developed crenulate and unidirectional solidification textures; and (3) muscovite-bordered miarolitic cavities lined with quartz-feldspar ± cassiterite. Associated with areas of pegmatites are irregular- to cuspate-shaped zones enriched in muscovite. Mineralogically the pegmatites are simple, consisting of clear to black quartz, albite, microcline perthite and muscovite; topaz may be enriched in aplites coring pegmatites and coarse euhedral cassiterite is rare.

Chemically the leucogranites marginal to pegmatites are similar to the host leucogranite, but some aphanitic felsic dykes indicate extreme differentiation, as reflected by REE depletion and P-enrichment. Enrichment of pegmatitic feldspar in P2O5 (to 1 wt. %) indicates that melts contained 2–3 wt. % P2O5, whereas muscovite chemistry (to 5 wt. % F) reflects the F-rich nature of the melt. Trace and REE contents of pegmatitic feldspars are consistent with local segregation of volatile-rich melts to form pegmatites rather than extreme crystal-chemical fractionation.

Fluid inclusion studies of pegmatitic quartz and cassiterite indicate the presence of a highly saline brine (40 wt. % eq. NaCl; Na > K > Fe > Mn > Ca > Sr) of magmatic origin. Isochores modelled for a 40 wt. % eq. NaCl fluid constrain pegmatite formation at 550–600°C, thus the depressed solidus is consistent with the volatile-rich (F, P, H2O) nature of the melt. Stable isotopes demonstrate that a magmatic fluid (δ18OH2O = +5·5 to +10‰, δDH2O = −33 to −41‰ for 450–500°C) equilibrated with the pegmatites and the system cooled abruptly at c. 425°C.

The low volume of pegmatite at East Kemptville suggests that the melt was not near volatile saturation, instead pegmatite generation is interpreted to have resulted from rapid decompression related to the active shear zone setting of the granite. Although feldspar chemistry reflects a local segregation model, the chemistry of aphanitic dyke rocks indicates that the leucogranite did evolve into a more fractionated melt. The local presence of cassiterite in pegmatites and miarolitic cavities indicates that locally, saturation of Sn occurred, but not throughout the EKL where elevated Sn contents are attributed to infiltration of mineralising fluids (i.e. deuteric alteration).

This paper evaluates the taxonomy, biostratigraphy, and palaeogeographical significance of the Cambrian bradoriid arthropods of China, the majority of which occur in the lower Cambrian of SW China. Of bradoriid faunas world-wide, Chinese occurrences yield the greatest number of specimens and a comparatively high diversity at all taxonomic levels. Nevertheless, taxonomic diversity is much less than previously supposed. Some 80 bradoriid genera and nearly 300 species have been proposed on the basis of Chinese material. By contrast, in our study, which encompasses all of the important Chinese bradoriid faunas, we recognise only 16 genera and 21 species, including those treated under open nomenclature. Interpretation of deformed specimens as discrete species and lack of application of the rules of the International Code of Zoological Nomenclature resulted in taxonomic splitting and a proliferation of names. There are an additional 12 poorly known monotypic genera of uncertain systematic status that are listed but not treated further herein. One phosphatocopid species, a group originally thought closely related to the Bradoriida, is also described.

Most Chinese bradoriid material is known from Yunnan Province; the group also occurs in Guizhou, Henan, Hubei, Hunan, Liaoning, Qinghai, Shaanxi, Sichuan, Xinjiang and Zhejiang provinces. The first bradoriids occur just below the Abadiella trilobite Biozone. They are most prolific and diverse in the Qiongzhusian Stage, constituting the most abundant animal group; the succeeding Canglangpuian Stage contains fewer individuals and species. A previously proposed bradoriid biozonal scheme lacks rigour and is of little practical value: of the five supposed biozones, two correspond to trilobite zones and three are based on taxa that herein are considered to belong to a single species.

Palaeogeographically the bradoriids occur in the Middle and especially the Western subprovinces of the Cambrian of the SW China (Yangtze) Platform. Almost all of the bradoriid genera and species are endemic to that region. The palaeogeographical links with other bradoriid faunas are mostly within the Redlichiid trilobite Realm, with areas such as N China, Australia and parts of central Asia.