Recent Advances

Preface 7
Acknowledgements 7
Part 1: Recent Advances in Early Vertebrates
Richard CLOUTIER:
Hans-Peter SCHULTZE’s contribution to our understanding of lower vertebrate evolution 11
Philippe JANVIER:
Early specializations in the branchial apparatus of jawless vertebrates: a consideration of gill number and size 29
Valentina KARATAJUTE-TALIMAA & Moya MEREDITH SMITH:
Tesakoviaspis concentrica: microskeletal remains of a new order of vertebrate from the Upper Ordovician and Lower Silurian of Siberia 53
Susan TURNER:
Early vertebrates: analysis from microfossil evidence 65
Mark V. H. WILSON & Tiiu MÄRSS:
Toward a phylogeny of the thelodonts 95
Daniel GOUJET & Gavin C. YOUNG:
Placoderm anatomy and phylogeny: new insights 109
Robert K. CARR:
Recognizing paraphyletic stem groups: A case study in the analysis of eubrachythoracid arthrodires (Placodermi) 127
John G. MAISEY:
Endocranial morphology in fossil and recent chondrichthyans 139
Richard LUND & Eileen D. GROGAN:
Two tenaculum-bearing Holocephalimorpha (Chondrichthyes) from the Bear Gulch Limestone (Chesterian, Serpukhovian) of Montana, USA 171
Gavin F. HANKE & Mark V. H. WILSON:
New teleostome fishes and acanthodian systematics 189
Richard CLOUTIER & Gloria ARRATIA:
Early diversification of actinopterygians 217
Min ZHU & Xiaobo YU:
Lower jaw character transitions among major sarcopterygian groups – a survey based on new materials from Yunnan, China 271
Zerina JOHANSON:
Late Devonian sarcopterygian fishes from eastern Gondwana (Australia and Antarctica) and their importance in phylogeny and biogeography 287
Jennifer A. CLACK & Per E. AHLBERG:
A new stem tetrapod from the Early Carboniferous of Northern Ireland 309
Robert L. CARROLL, Catherine BOISVERT, John BOLT, David M. GREEN, Noah PHILIP, Campbell ROLIAN, Rainer SCHOCH & Anna TARENKO:
Changing patterns of ontogeny from osteolepiform fish through Permian tetrapods as a guide to the early evolution of land vertebrates 321
Rainer R. SCHOCH & Andrew R. MILNER:
Structure and implications of theories on the origin of lissamphibians 345
Johannes MÜLLER:
The relationships among diapsid reptiles and the influence of taxon selection 379
Richard A. KISSEL & Robert R. REISZ:
Synapsid fauna of the Upper Pennsylvanian Rock Lake Shale near Garnett, Kansas and the diversity pattern of early amniotes 409

Part 2: Recent Advances in Fishes
Elga MARK-KURIK:
Buchanosteids (Placodermi, Arthrodira) from Central Asia 431
Marius ARSENAULT, Sylvain DESBIENS, Philippe JANVIER & Johanne KERR:
New data on the soft tissues and external morphology of the antiarch Bothriolepis canadensis (WHITEAVES, 1880), from the Upper Devonian of Miguasha, Quebec 439
David K. ELLIOTT, Randal C. REED & Elizabeth J. LOEFFLER:
A new species of Allocryptaspis (Heterostraci) from the Early Devonian, with comments on the structure of the oral area in cyathaspidids 455
Michal GINTER:
Devonian sharks and the origin of Xenacanthiformes 473
Oliver HAMPE, Z. Sarah ABOUSSALAM & R. Thomas BECKER:
Omalodus teeth (Elasmobranchii: Omalodontida) from the northern Gondwana margin (middle Givetian: ansatus conodont Zone, Morocco) 487
Richard LUND & Eileen D. GROGAN:
Five new euchondrocephalan Chondrichthyes from the Bear Gulch Limestone (Serpukhovian, Namurian E2b) of Montana, USA 505
Rodrigo SOLER-GIJÓN:
Development and growth in xenacanth sharks: new data from Upper Carboniferous of Bohemia 533
Colin D. LITTLE & William E. BEMIS:
Observations on the skeleton of the heterocercal tail of sharks (Chondrichthyes: Elasmobranchii) 563
Paulo M. BRITO & Pascal P. DEYNAT:
Freshwater stingrays from the Miocene of South America with comments on the rise of potamotrygonids (Batoidea, Myliobatiformes) 575
Gloria ARRATIA & Richard CLOUTIER:
A new cheirolepidid fish from the Middle-Upper Devonian of Red Hill, Nevada, USA 583
Eric J. HILTON:
The caudal skeleton of Acipenseriformes (Actinopterygii: Chondrostei): recent advances and new observations 599
Christopher FIELITZ:
The phylogenetic relationships of the †Enchodontidae (Teleostei: Aulopiformes) 619
Gloria ARRATIA, Adriana LÓPEZ-ARBARELLO, Guntupalli V. R. PRASAD, Varun PARMAR & Jürgen KRIWET:
Late Cretaceous-Paleocene percomorphs (Teleostei) from India – Early radiation of Perciformes 635
CHANG Mee-mann:
Synapomorphies and scenarios – more characters of Youngolepis betraying its affinity to the Dipnoi 665
Michael D. GOTTFRIED, Raymond R. ROGERS & Kristina CURRY ROGERS:
First record of Late Cretaceous coelacanths from Madagascar 687
Index of taxa 693

Philippe JANVIER:
Early specializations in the branchial apparatus of jawless vertebrates: a consideration of gill number and size
[pp. 29-52, 8 figs., 1 tab.]

The number of gill units and gill compartments in fossil and living agnathans shows an important interspecific and individual variation in living hagfish, but is quite stable in living lampreys and gnathostomes. In Palaeozoic agnathan taxa, interspecific variations seem to occur essentially in galeaspids, anaspids and two anaspid-like taxa, but galeaspids can display up to forty-five pairs of gill compartments, a number unequalled among vertebrates. Current vertebrate phylogenies suggest that the occurrence of "polybranchy" (i.e. a large number of gills, exceeding ten) in agnathans is homoplastic. At any rate, it is certainly a derived character within galeaspids, and it is probable that the generalised condition in most "ostracoderm" taxa was about seven to ten gill compartments. Osteostracans and possibly pituriaspids show that the posterior expansion of the branchial apparatus was constrained by the paired fins and girdles, and the former taxon solved this problem by expanding its branchial apparatus forward, far beyond the level of the optic capsules. Current data on the developmental genetics of the branchial apparatus offer no explanation to the process that leads to polybranchy, except for the fact that the anterior portion of the endoderm plays a major role in patterning the pharynx and that the production of pharyngeal pouches is theoretically unlimited within this region, as long as the individual pouches can become reduced in size without loosing their respiratory function. It is assumed that the common ancestor to all fossil and living vertebrates probably possessed from five to possibly ten gill units, and that polybranchy, though vaguely recalling the numerous pharyngeal slits of cephalochordates, is not a primitive character for the vertebrates.

Valentina KARATAJUTE-TALIMAA & Moya MEREDITH SMITH:
Tesakoviaspis concentrica:microskeletal remains of a new order of vertebrate from the Upper Ordovician and Lower Silurian of Siberia
[pp. 53-64, 6 figs.]

Tesakoviaspis concentrica n. gen. et sp. is an Early Silurian vertebrate with a distinctive and unique histology. This represents a new order of a tesselate agnathan. The crown growth is cyclomorial with simple odontodes of smooth pear-shaped, morphology. These microremains, collected from the southern part of the Siberian platform, are attributed to the Late Ordovician and Early Silurian. Some are complex scales, or tesserae, from a microsquamous exoskeleton, which shows variability interpreted as a growth series, as well as species variability. Zones of areal growth can be recognized of many small, round, oval, or pear-shaped tubercles surrounding a single large, round or pear-shaped central primordial tubercle. Each tubercle (odontode) grows centripetally around a separate pulp cavity, later the acellular bony base is added below the odontodes in a series of growth layers. These almost close the pulp cavities except for narrow, central openings, continuous with vertical, vascular canals opening onto the inner surface. The same type of concentric areal growth is found in both tesakoviaspids and astraspids.

The coronal tissue is a new type of atubular dentine termed conical lamellin, with a unique histology and without an enamel or enameloid covering. Parallel growth zones are well demarcated, the first coronal layers appear as numerous, steep sided cones, but become gradually flattened in the circumpulpal layers, these are indicative of a new type of vertebrate dermal production. The acellular basal bone has sequential lines of growth where each layer is coincident with the increase in area of the scale. Vertical lines penetrate all layers of bone from the basal surface, where spaces for Sharpey's fibre bundles existed. The most closely related histology is that of the Mongolepidida. There is no equivalent tissue in other known Ordovician, or Silurian agnathan vertebrates, such as astraspids, eriptychiids, or thelodontids.

Susan TURNER:
Early vertebrates: analysis from microfossil evidence
[pp. 67-94, 7 figs., 2 tabs.]

Vertebrate microfossils provide data for potential origination, range and distribution of taxa, for phylogeny of organisms and for assessing response of taxa to events. Isolated vertebrate microremains (microvertebrates, ichthyoliths), such as thelodont and "shark" scales and "shark" teeth can and do provide evidence about the functional morphology and relationships of higher taxa such as the Thelodonti and Chondrichthyes. Vertebrate microfossils can also provide evidence for at least the earliest occurrences of various morphological features characteristic of taxa. Specialized sensory scales related to free neuromasts are identified from a loganellidid thelodont from the Lower Silurian of Devon Island, Nunavut, Canada. New material of early "shark" remains, scales and teeth from the Devonian of Australia and Canada, including the Early Devonian (Emsian) taxon, Emsolepis hanspeteri n. gen. et sp., from the Receptaculites and Jesse limestones of NSW and Doliodus problematicus (WOODWARD), from the Campbellton Formation of northern New Brunswick, add insights into the earliest stages of chondrichthyans. Doliodus, most recently regarded as a member of the Omalodontida, exhibits the earliest fossil record of the chondrichthyan teeth in situ.

Mark V. H. WILSON & Tiiu MÄRSS:
Toward a phylogeny of the thelodonts
[pp. 95-108, 6 figs., 2 apps.]

We investigated the phylogenetic relationships of thelodonts by compiling data on 25 species in 15 genera, comprising most of the known thelodonts represented by articulated specimens (squamations), together with four outgroup taxa. Character states for 53 characters (83 apomorphic states) were compiled and analyzed using Maximum Parsimony methods to yield 150 shortest trees of 177 steps each. The consensus summary of these trees was judged to be unreliable owing to the uncertain placement of two "wild-card" taxa (one of which was an outgroup). After removal of these two taxa a second analysis yielded 31 shortest trees of 153 steps. The consensus tree suggests a monophyletic Thelodonti in which a basal split separates Phlebolepisplus Erepsilepis and their sister Archipelepis from the remaining taxa. Within the remaining species, a group consisting of Turinia plus Loganellia and their possible sister Phillipsilepis is collectively sister to Shielia plus Lanarkia and their sister the Furcacaudiformes. The latter group is monophyletic and comprises the basal Pezopallichthys and a monophyletic Furcacaudidae. These results must be considered tentative pending addition of new characters and inclusion of data from other species of thelodonts.

Daniel GOUJET & Gavin C. YOUNG:
Placoderm anatomy and phylogeny: new insights
[pp. 109-126, 8 figs.]

New specimens of acanthothoracid placoderms and related forms (Romundina, Murrindalaspis, Brindabellaspis) are described, illustrating aspects of placoderm anatomy in the assumed basal branch according to a working hypothesis of placoderm interrelationships. It is proposed that two median dorsal plates may be primitive for placoderms, this being the defining difference to the osteichthyan shoulder girdle in creating a rigid ring of bones for the placoderm dermal neckjoint. One pectoral fin articulation, an anterior insertion for the internal rectus extra ocular muscle, and two abducens innervated eye muscles, may be primitive for placoderms, and for the Gnathostomata.

Robert K. CARR:
Recognizing paraphyletic stem groups: A case study in the analysis of eubrachythoracid arthrodires (Placodermi)
[pp. 127-138, 9 figs., 1 tab.]

Studies on the origin and diversity of vertebrates are dependent, in part, upon a clear understanding of the phylogenetic relationships of the organisms under study. However, the use of stem and other paraphyletic groupings is common in the literature. Following the advent of phylogenetic systematics, a clearer picture of phylogenetic relationships emerged. Despite what appeared to be a panacea for phylogenetic debates, a number of questions remain. Among eubrachythoracid arthrodires (Placodermi), the sister groups Coccosteomorphi and Pachyosteomorphi are well established; however, the relationship of basal taxa remains unresolved despite methodological advances. This instability is explained by three hypotheses: (1) a historical artifact in the definition of characters; (2) missing data; and (3) a byproduct of the evolutionary patterns in the groups under study.

Eubrachythoracids are evaluated as a case study to test these hypotheses. Pre- and post-cladistic analyses are reviewed. The use of historically biased characters represents a tautology, a preformed group is used to define a phenetic character and then the character is used to validate the group. Missing taxa (including author's choice) affect character polarity. A study group of taxa should be bounded by outgroups and higher nested taxa otherwise phylogenetic conclusions may be misleading. Missing characters obviously affect a character-based analysis. Evolutionary effects include tempo and the nature of morphological change. Sister taxa with different rates of evolution show different sampling biases. An asymmetrical loss of plesiomorphic characters within one clade results in an increased sensitivity to missing taxa resulting in paraphyly. This scenario is modeled.

The model and a review of eubrachythoracid basal taxa support all three hypotheses, in that each hypothesis explains individual cases. Historical biases and missing data are recognized sources of error; however, evolutionary pattern has been under estimated as an influencing factor. It is predicted that these sources of instability are ubiquitous among phylogenetic analyses. Recognition of the precautions and increased fieldwork portend a fruitful future for our understanding of phylogenetic relationships among arthrodires and other craniate taxa.

John G. MAISEY:
Endocranial morphology in fossil and recent chondrichthyans
[pp. 139-170, 11 figs.]

Craniate endocranial morphology has considerable phylogenetic potential, irrespective of how closely it corresponds to the anatomy of the brain and other cranial soft structures. Endocranial morphology may not be reflected by external features and can only be determined by direct observation; non-invasive CT-scanning is currently the preferred procedure. In this work, features of the principal endocranial and labyrinth cavities are described for the first time in several fossil elasmobranchs, based on high-resolution CT-scanning. Taxa investigated include Tribodus, Egertonodus, Cladodoides, Cobelodus and the early Devonian stem chondrichthyan Pucapampella. The comparison also includes CT-scan data from modern Squalus and Notorynchus, plus non-scan data from chimaeroids and the Paleozoic elasmobranchs Orthacanthus and Tamiobatis. Morphological variation in the telencephalic, hypophyseal and otic regions is described. Hybodonts and modern elasmobranchs share features in their skeletal labyrinth that are associated with semi-directional low-frequency phonoreception; these features are absent in chimaeroids, many Paleozoic elasmobranchs and Pucapampella. The Pennsylvanian shark Cobelodus represents the first known example of a chondrichthyan with a tropibasic braincase, and it shares many specialized cranial features with other tropibasic gnathostomes such as actinopterygians and certain placoderms.

Richard LUND & Eileen D. GROGAN:
Two tenaculum-bearing Holocephalimorpha (Chondrichthyes) from the Bear Gulch Limestone (Chesterian, Serpukhovian) of Montana, USA
[pp. 171-187, 6 figs., 2 apps.]

The isolated Visean (Scotland) and Chesterian (USA) spines named Harpacanthus fimbriatus can now be attributed to the anterior tenaculae of an autodiastylic holocephalimorph chondrichthyan on the basis of specimens from the Upper Chesterian (Serpukhovian) of Montana, USA. The Montana specimens exhibit these organs as three pairs of rostrally located anterior tenaculae. An isolated braincase of the holocephalan Traquairius nudus with a long, antorbitally articulated median anterior tenaculum is also described. All previously know anterior tenaculae have been demonstrated to be male secondary sexual characters, and thus we conclude the same for the tenaculae of these species. The sporadic occurrences and possible developmental origins of anterior tenaculae among the Holocephalimorpha are discussed.

Gavin F. HANKE & Mark. V. H. WILSON:
New teleostome fishes and acanthodian systematics
[pp. 189-216, 12 figs., 2 apps.]

Specimens of two new fish species were collected from the Lower Devonian ichthyofauna of the Mackenzie Mountains, Northwest Territories, Canada. These two species are interesting in that they have monodontode scales, lack teeth, and have an unossified axial, visceral, and appendicular endoskeleton. These characteristics have been suggested to be primitive for jawed fishes. However, the new taxa have combinations of median and paired fin spines which are similar to those of acanthodian fishes. The new taxa show no obvious characteristics to suggest relationship to any particular group of acanthodians, and for the moment, we will not try to determine their relationships, but to use them as outgroups in an analyses of relationships within the class Acanthodii. Our cladistic analysis results suggest that climatiiform fishes are basal relative to acanthodiform and ischnacanthiform taxa. However, in contrast to previously published analyses, the order Climatiiformes appears paraphyletic relative to the other two acanthodian orders. Lupopsyrus pygmaeus is placed as the basal-most acanthodian species, Brochoadmones milesi, Euthacanthus macnicoli, and diplacanthids are relatively derived "climatiiform" fishes, and the heavily armored condition in Climatius reticulatus and Brachyacanthus scutiger appears as a uniquely derived state and not primitive for all acanthodians. In addition, Cassidiceps vermiculatus and Paucicanthus vanelsti seem to be related to acanthodiform fishes based on fin spine structures. Cassidiceps vermiculatus originally was placed with climatiiform fishes in the original description. Given our character coding, we identified several primitive characteristics which were retained in relatively derived acanthodian taxa.

Richard CLOUTIER & Gloria ARRATIA:
Early diversification of actinopterygians
[pp. 217-270, 15 figs., 1 tab., 2 apps.]

During the past decade, new Devonian (e.g., Cheirolepis n. sp., Dialipina salgueiroensis, Limnomis) and Carboniferous actinopterygians (e.g., Aesopichthys, Cyranorhis, Discoserra, Guildayichthys, Kalops, Melanecta, Mesopoma, Proceramala, Wendyichthys, Woodichthys) have been discovered around the world. In addition, new basal osteichthyan taxa, such as Psarolepis, have been found showing puzzling combinations of actinopterygian and sarcopterygian characters. Although the monophyly of the Actinopterygii has never been questioned, the synapomorphies corroborating the monophyly as well as the interrelationships among basal actinopterygians vary among the studies. The Late Silurian Lophosteus, the Early Devonian Dialipina, the Middle-Late Devonian Cheirolepis, the Frasnian Howqualepis, and the recent Polypterus have each been identified as the basal taxon of the actinopterygians. A combined data matrix is compiled in order to include 142 cranial and 43 postcranial characters (new and previously published characters) of 33 basal Paleozoic actinopterygians, five advanced actinopterygians (including the Jurassic Leptolepis, and the living Polypterus, Acipenser, Amia, and Lepisosteus) and four basal osteichthyans. Phylogenetic characters are redefined. Owing to the morphological diversity of fossil and living taxa and the morphological incompleteness of certain species, the large proportion of missing data generates conflictual phylogenetic information. We evaluated the influence of outgroup selection, the inclusion/exclusion of Jurassic and living taxa, and the inclusion/deletion of poorly known characters on the resolution of lower actinopterygians. Although some interrelationships among Devonian and Carboniferous taxa are well corroborated, phylogenetic uncertainties remain for the position of certain taxa.

Min ZHU & Xiaobo YU:
Lower jaw character transitions among major sarcopterygian groups – a survey based on new materials from Yunnan, China
[pp. 271-286, 8 figs.]

The lower jaw materials of stem-group sarcopterygian Achoania and an indeterminable form with certain onychodont features are described for the first time from the Lower Devonian of Yunnan, China. In addition, new observations on lower jaw features in Psarolepis and Styloichthys from Yunnan are reported to expand or amend previous descriptions. The lower jaw features of Achoania corroborate the phylogenetic position of Achoania originally derived from features of the anterior cranial portion. Comparison with available onychodont lower jaws suggests that the indeterminable form from Yunnan and Langdenia from northern Vietnam may represent lower jaw materials of basal onychodonts in the Lockhovian fauna of South China. The lower jaw features of basal sarcopterygian groups, when mapped onto our previous cladogram (ZHU & YU 2002), suggest two major changes in lower jaw character transitions leading from the condition in basal actinopterygians to the condition in dipnomorphs and tetrapodomorphs. The first major change is represented by the occurrence in Psarolepis of a parasymphysial dental plate bearing a tooth whorl with fangs, five coronoids with precoronoid and intercoronoid fossae, and a prearticular extending forward and covering the area mesial to the coronoid series. The second major change is the occurrence of the 3-coronoid condition in Styloichthys, dipnomorphs and tetrapodomorphs. When the precoronoid and intercoronoid fossae are used as landmarks for detecting possible homology among the coronoids, coronoids 2-4 in Psarolepis (and Achoania) probably correspond to the anterior, middle and posterior coronoids found in basal dipnomorphs/tetrapodomorphs while coronoids 1 and 5 may have been lost in the transition from the 5-coronoid condition to the 3-coronoid condition. Functionally, this transition probably involved a shift of emphasis in the bite, with the major biting force moving from the anteriorly positioned parasymphysial whorls (as in Psarolepis, Achoania and onychodonts) to the more posteriorly positioned fangs carried by coronoids and opposing palatal bones (as in Styloichthys and basal dipnomorphs/tetrapodomorphs).

Zerina JOHANSON:
Late Devonian sarcopterygian fishes from eastern Gondwana (Australia and Antarctica) and their importance in phylogeny and biogeography
[pp. 287-308, 7 figs., 3 apps.]

Canowindridae, Tristichopteridae and Rhizodontida are three important sarcopterygian groups from the Upper Devonian of eastern Gondwana. These taxa are part of the tetrapod stem-group, being more closely related to the Tetrapoda than to either lungfish or coelacanths. Recent descriptions of faunas have provided important new morphological information, clarifying relationships within these groups and relative to other stem-tetrapods. Eastern Gondwanan tristichopterids represent the more derived nodes of that group, whereas the co-occurring rhizodontids represent some of the more basal nodes of the Rhizodontida. These Gondwanan taxa have phylogenetic affinities with Euramerican taxa, supporting the proximity of Euramerican and Gondwanan areas in the Late Devonian, contrary to palaeomagnetic evidence. This may be extended into the Middle Devonian, based on the phylogenetic position of the Euramerican rhizodont Sauripterus among Gondwanan representatives of the group, including the Antarctic taxon Aztecia, as well as the position of the Antarctic Notorhizodon among Euramerican and Gondwanan taxa within the tristichopterid clade.

Jennifer A. CLACK & Per E. AHLBERG:
A new stem tetrapod from the Early Carboniferous of Northern Ireland
[pp. 309-320, 3 figs., 2 tabs.]

We report on the first Carboniferous tetrapod specimen discovered in Northern Ireland, and one of the most primitive tetrapods known from the UK. The specimen consists of a partial left jaw ramus showing sufficient features to diagnose it as a new taxon. The specimen was discovered in 1843 by Portlock, described as Holoptychius and housed with the British Geological Survey. Recent work on tetrapodomorph lower jaws has shown that this specimen belongs to an undoubted tetrapod that bears resemblances to deep-skulled forms such as the whatcheeriids and Crassigyrinus. The precise locality is uncertain, but it probably derives from a site near Maghera, County Londonderry. Palynological evidence is equivocal, ranging from late Tournaisian through early Westphalian. The Altagoan Formation, from which most of Portlock's material seems to have been collected, is usually regarded as late Tournaisian or early Viséan (Courceyan or Chadian). The specimen is almost certainly older than those from Jarrow in Eire (Westphalian A), and is certainly more primitive than any tetrapod described from there. The specimen extends the geographical range of known Early Carboniferous tetrapods, which have now been found much further west in the British Isles than previously reported, and suggests the possibility of further discoveries in this region.

The jaw displays a distinctive character combination, including coronoids with an organised tooth row, lack of postsplenial-prearticular contact, an open mandibular lateral line sulcus, and a stepped rather than straight dentary-angular suture. A preliminary analysis of lower jaw characters places the new taxon in the neighbourhood of "Tulerpeton" (late Famennian jaw material from Andreyevka, Russia) and Whatcheeria, above all other Devonian tetrapods, and below Crassigyrinus, Greererpeton, Megalocephalus and an anthracosaur-temnospondyl clade. It may belong to an early and wide-ranging post-Devonian tetrapod radiation.

Robert L. CARROLL, Catherine BOISVERT, John BOLT, David M. GREEN, Noah PHILIP, Campbell ROLIAN, Rainer SCHOCH & Anna TARENKO:
Changing patterns of ontogeny from osteolepiform fish through Permian tetrapods as a guide to the early evolution of land vertebrates
[pp. 321-343, 12 figs., 4 apps.]

Larval stages of both recent and fossil amphibians provide an added dimension – ontogenetic change – that can be used for phylogenetic analysis. This is especially important in regard to the origin and early radiation of tetrapods. A continuous growth series in Eusthenopteron shows that a close sister taxon of tetrapods had direct development, without a recognizable larval stage. Gilled larvae of labyrinthodonts are known no earlier than the Upper Carboniferous, although they may have evolved by the Viséan. There is little evidence as to whether the larval stages of temnospondyls and seymouriamorphs evolved from a common ancestry, or by convergence. All lepospondyls show direct development, with even the smallest fossils resembling the adults in body proportions and in possessing highly ossified vertebral centra but lacking external gills. The ancestry of the three "lissamphibian" orders can all be traced to separate Paleozoic lineages, based on both adult anatomy and the pattern of larval development. Frogs and primitive salamanders share similarities with temnospondyls in having gilled larvae, and in the rate and sequence of vertebral development. The sequence of ossification of the skull bones of conservative extant salamanders is identical with that of the early larval stages of the branchiosaur Apateon. The adult cranial morphology of extant salamanders could have been achieved by truncation of ossification prior to the appearance of additional skull bones that developed at a later stage in Paleozoic temnospondyls. The ancestors of salamanders may have been facultatively neotenic, with long periods of larval development, but the antecedents of frogs may have gone through a period of direct development before evolving the highly specialized tadpole larvae. Caecilians resemble lepospondyls in the rapidity of vertebral ossification. Most caecilians lose the external gills soon after hatching, which might have been the case in lepospondyls.

Rainer R. SCHOCH & Andrew R. MILNER:
Structure and implications of theories on the origin of lissamphibians
[pp. 345-377, 4 figs., 2 tabs.]

The origin of the Lissamphibia is the subject of continuing debate, and there is no current consensus. Albeit often considered a different problem, the intrarelationships of crown-lissamphibians have a strong influence on the identification of phylogenetic polarity for the lissamphibian stem, as cladistic reanalyses indicate. Regarding crown-group relationships, there is a majority view that salamanders and anurans are sister groups (Batrachia Hypothesis), supported by morphological and, more recently, molecular studies. The only competing alternative so far generated by molecular systematic analyses is the Procera Hypothesis (salamanders + caecilians). Both albanerpetontids and the stem-groups of the three extant clades contribute significantly to the structure of tree topologies by adding information that often reverses phylogenetic polarity, such as the presence of a stegokrotaphic skull and retention of separate intercentra by Eocaecilia.

The major hypotheses of lissamphibian origin (temnospondyl, lepospondyl, and polyphyletic) rely on different character-sets, and true total evidence has not been reached yet. At first sight, the Temnospondyl and Lepospondyl hypotheses appear to involve diametrically opposite interpretations of character evolution, especially regarding the middle ear and vertebral centrum, two character-sets often employed in the study of lissamphibians. Here we discuss the most important among the different sets of characters, analyze their problems and strengths, and test their distribution in the three different phylogenetic topologies. According to this test, most characters can be explained to have evolved equally parsimoniously in the lepospondyl and temnospondyl hypotheses, while the Polyphyletic Hypothesis forms a plausible alternative in only a few cases. While there is some quantitative support in the numerous absence characters (loss of bones) for the Lepospondyl Hypothesis, the evolution of the palate and dentition appears much more plausible (parsimonious) under the Temnospondyl Hypothesis. Surprisingly, vertebral characters do not favour the Lepospondyl Hypothesis, and some of them are well explained by all concepts, including polyphyly.

We conclude that (i) the large number of absence characters is problematic for several reasons while most of them are invalidated by incongruent distribution, (ii) vertebral characters are not decisive at all and their proper understanding requires further studies of development, and (iii) some characters of the middle ear and palate, and dentition appear to be highly informative but are challenged by difficulties in identifying the primitive condition for salamanders. Altogether, our analysis of the palaeobiological implications of the current lissamphibian origin hypotheses favour the Temnospondyl Hypothesis, with the Salientia-Procera Hypothesis of internal relationships involving fewer unparsimonious assumptions than the Batrachia-Gymnophiona Hypothesis.

We suggest that future cladistic analyses will have to deal with a thorough reanalysis of salamander ancestry, total evidence including albanerpetontids, the consideration of ontogenetic changes in morphology in the coding of many character-states, and the inclusion of additional character sources, such as ossification sequences among extant and fossil taxa.

Johannes MÜLLER:
The relationships among diapsid reptiles and the influence of taxon selection
[pp. 379-408, 8 figs., 3 apps.]

In recent years, there was significant progress in our understanding of diapsid phylogeny, but also some new and intensively discussed problems arose, of which the most prominent example is the suggested diapsid affinity of turtles. The reliability of several results, however, suffered from the problem that only a few taxa were included in the respective analyses. For that reason, a new analysis of the interrelationships of diapsid reptiles was performed, with the main focus on taxa that were either rarely entered in the latest analyses (e.g. drepanosaurs or thalattosaurs) or where strongly differing opinions towards a phylogenetic placement exist (e.g. turtles, ichthyosaurs, sauropterygians). In total, 33 taxa and 184 informative characters were taken into consideration. The most important results are a positioning of thalattosaurs and ichthyosaurs outside the lepidosauromorph/archosauromorph dichotomy but with a certain signal towards archosauromorphs, a placement of turtles close to lepidosaurs, a grouping of kuehneosaurs with drepanosaurs outside saurians, the paraphyly of prolacertiforms, and a loss of stable relationships at the base of neodiapsids. It is noted that the tree topology depends much more on the choice of taxa than on the number and definition of characters, which means that the removal of some taxa has an effect on the position of other diapsids. Testing the influence of several "problematic" forms shows, however, that this is not strongly correlated with the number of characters for which a taxon can be coded. With regard to these phenomena, it is postulated that in the case of large-scale phylogenetic analyses the stability of the results should not only be tested by using methods like bootstrap or jackknifing, but also by evaluating the effect of a respective taxon on the overall tree topology.

Richard A. KISSEL & Robert R. REISZ:
Synapsid fauna of the Upper Pennsylvanian Rock Lake Shale near Garnett, Kansas and the diversity pattern of early amniotes
[pp. 409-428, 10 figs., 1 tab.]

Stream channel infills near Garnett, Kansas preserve the most abundant and diverse assemblage of amniotes known from the Carboniferous. In addition to the diapsid Petrolacosaurus kansensis and the synapsids Haptodus garnettensis, Xyrospondylus ecordi, and Ianthasaurus hardestiorum, at least three additional synapsid taxa can be recognized, including a new eupelycosaur genus that is represented by a partial skull and an isolated maxilla. Ianthodon schultzei n. gen. et sp. possesses three premaxillary and seventeen or eighteen maxillary teeth, and the individual teeth are generally conical in outline with only a slight recurvature near the tip. On the lingual surface, the teeth possess densely packed longitudinal fluting that extends to the tip of each tooth. No distinct caniniform tooth is present on the maxilla, but there is a modest caniniform region along the anterior one-third of the bone. I. schultzei is also characterized by a tall lacrimal, and the pineal foramen is found at the midpoint of the median parietal suture. Available evidence suggests that Ianthodon is a member of Sphenacodontia and is positioned within the "haptodont" grade of synapsids. Phylogenetic evidence and a review of occurrence data from Garnett and the 21 other reported Pennsylvanian amniote-bearing localities indicate that the early history of Amniota is characterized by: 1) a basal dichotomy into synapsids and sauropsids; and 2) an uneven rate of diversification, with synapsids rapidly diversifying and quickly developing a dominant role in the terrestrial ecosystems of the late Paleozoic.

Elga MARK-KURIK:
Buchanosteids (Placodermi, Arthrodira) from Central Asia
[pp. 431-438, 5 figs., 1 tab.]

Remains of two buchanosteids from the Early Devonian of Central Asia, Kazakhstan and Uzbekistan, are described. The skull-roof of Buchanosteus nuricus n. sp. comes from the Qaraghandy region, Kazakhstan. The Zinzilban Gorge section, Uzbekistan has also yielded a suborbital plate, belonging to Buchanosteus sp. Buchanosteids are very widely distributed Early Devonian brachythoracids. They were initially described from eastern Gondwana (Australia), but are now also known from China, western Gondwana (Iran, Saudi Arabia), Kazakhstan, the southern and western Siberian + Kara-Tajmyr Continents (Uzbekistan, Siberian Arctic) and the eastern part of Euramerica (Spitsbergen, South Urals). The buchanosteids from Central Asia are fairly similar to the type species, Buchanosteus confertituberculatus (CHAPMAN), and indicate some connection between the Emsian paleobasins.

Marius ARSENAULT, Sylvain DESBIENS, Philippe JANVIER & Johanne KERR:
New data on the soft tissues and external morphology of the antiarch Bothriolepis canadensis (Whiteaves, 1880), from the Upper Devonian of Miguasha, Quebec
[pp. 439-454, 8 figs.]

The euantiarch Bothriolepis canadensis from the Late Devonian of Miguasha is revisited on the basis of some new material which yields additional information about its internal anatomy and overall external morphology. New evidence for soft tissue preservation in this species is provided by the discovery of blood vessel imprints beneath the plates of the thoracic armour in several specimens. The pattern of some of these blood vessel imprints, in the anterior part of thoracic armour raises the question of their possible relations to the presumed paired pharyngeal diverticles, or "lungs", previously described in this species. The nature of the more or less metameric grooves on the internal surface of the dorsal plates of the thoracic armour is discussed, and it is suggested that they may have housed lymphatic vessels. A new reconstruction of the armour of B. canadensisis proposed, on the basis of two specimens which have not been subjected to dorsoventral flattening. The overall morphology of B. canadensis appears thus to be in better agreement with Patten's than with Stensiö's reconstruction; i.e., with a rather deep thoracic armour and more tilted skull roof.

David K. ELLIOTT, Randal C. REED & Elizabeth J. LOEFFLER:
A new species of Allocryptaspis(Heterostraci) from the Early Devonian, with comments on the structure of the oral area in cyathaspidids
[pp. 455-472, 7 figs., 1 tab.]

Allocryptaspis sandbergi n. sp. is described from the Lower Devonian Water Canyon Formation of Utah and a review of the genus shows that the previously described A. flabelliformis is not a valid species. The presence of plates from the oral cover in the material of A. sandbergi n. sp. enables a new reconstruction of the oral area to be made. This indicates that the previous reconstruction for this genus, and for cyathaspidids as a whole, is incorrect in that they did not possess finger-like oral plates but had an oral cover consisting of a small number of large plates. An assessment of cyathaspidid mouthparts as known shows that these animals must have been initially benthonic suspension feeders that radiated into deposit feeding and epilithic grazing.

Michal GINTER:
Devonian sharks and the origin of Xenacanthiformes
[pp. 473-486, 3 figs.]

Interrelationships between major groups of Devonian chondrichthyans and the origin of Late Palaeozoic xenacanthiform sharks are considered on the grounds of tooth morphology and biostratigraphy. Suggestion that cladodont-toothed sharks appeared later in the evolution than the diplodont ones is combined with the idea of a phoebodontiform ancestry of xenacanthiforms. Such a phylogenetical model requires the shortest unexplained breaks in stratigraphic record and the lowest morphological barriers between closely related taxa, as far as dentition characters are concerned.

Oliver HAMPE, Z. Sarah ABOUSSALAM & R. Thomas BECKER:
Omalodus teeth (Elasmobranchii: Omalodontida) from the northern Gondwana margin (middle Givetian: ansatus conodont Zone, Morocco)
[pp. 487-504, 7 figs., 1 tab.]

Teeth of Omalodus schultzei n. sp. were discovered at the Moroccan locality of El Atrous in the southern part of the Tafilalt area. The elasmobranch genus Omalodus, known so far from North America, Poland, Siberia and Mauritania, is a typical representative of the Middle varcus conodont Zone. Sedimentological and biofacies data imply that Omalodus schultzei n. sp. avoided open pelagic settings and preferred to live near reefal margins, as known for a wide range of recent sharks. Cladistic analyses suggest a close relationship with the Givetian forms Portalodus, Aztecodus and Anareodus from Antarctica.

Richard LUND & Eileen D. GROGAN:
Five new euchondrocephalan Chondrichthyes from the Bear Gulch Limestone (Serpukhovian, Namurian E2b) of Montana, USA
[pp. 505-531, 12 figs., 2 tabs.]

The Class Chondrichthyes has historically been classified into two distinct crown groups, Elasmobranchii and Holocephali. Cladistic analyses result in the Holocephali being placed as the crown group of the Euchondrocephali. Several other clades have been associated as stem groups of the subclass Euchondrocephali, but insufficient data has rendered the relationships of these clades, and the basal phylogenetic structure of the euchondrocephalans, obscure. Five new euchondrocephalan chondrichthyans from the Bear Gulch Limestone member of the Heath Formation (Big Snowy Group, Serpukhovian, Upper Chesterian, Namurian E2b) that share similar dentitions and also share autodiastylic suspensoria are described below. The new family Gregoriidae is erected, encompassing the species Gregorius rexi, Bealbonn rogaire, Srianta srianta, S. iarlais and S. dawsoni. The Gregoriidae plesiomorphously and individually share teeth conforming to the several nominal species of the putative genus "Desmiodus" ST. JOHN & WORTHEN 1875. This renders "Desmiodus" a nomen vanum. The Gregoriidae vary in neurocranial proportions, symphysial elements, suspensorial details and postcranial morphologies. Cladistic analyses show that the Gregoriidae form a monophyletic clade that has a basal position relative to the Orodus + Helodus-Petalodontomorpha clade of the Euchondrocephali.

Rodrigo SOLER-GIJÓN:
Development and growth in xenacanth sharks: new data from Upper Carboniferous of Bohemia
[pp. 533-562, 13 figs., 1 tab.]

Articulated, nearly complete xenacanthids from the Westphalian D of Bohemia (Czech Republic) representing pre-adult and adult stages provide new information about the early development of xenacanths. Most of the specimens are juveniles of Orthacanthus bohemicus and adults of Xenacanthus parallelus. Study of the new material and comparison with other specimens previously described lead to a revision of several characters related to the dorsal spine and unpaired fins that are often used in phylogenetic analyses.

The new material indicates that the dorsal spine of members of Xenacanthidae, hitherto regarded as the cranial (occipital) spine, was initially formed in front of the dorsal fin, behind the pectoral girdle, in a fashion similar to the spine of the more primitive xenacanths Lebachacanthidae and Diplodoselachidae. The subsequent occipital location of the spine in sub-adult and adult of Xenacanthidae was achieved through differential growth of the proximal part of the spine early in ontogeny. In contrast, the dorsal spine of Diplodoselachidae and Lebachacanthidae is always located behind the pectoral girdle.

The new results support the presence of two dorsal and anal fins (anterior and posterior) as a synapomorphy of the clade formed by Lebachacanthidae and Xenacanthidae. The long anterior dorsal fin has traditionally been considered the only dorsal fin in xenacanths. This is the state in Diplodoselachidae only. The posterior dorsal fin has been considered the dorsal lobe of a diphycercal caudal fin or the dorsal part of the superior lobe of the heterocercal caudal fin. On the other hand, the posterior anal fin described here has previously been identified as the inferior lobe of a heterocercal caudal fin. There is no true caudal fin in Lebachacanthidae and Xenacanthidae, in contrast to the primitive Diplodoselachidae where a heterocercal caudal fin is well defined. The development of the posterior dorsal and anal fins is a novelty for Lebachacanthidae and Xenacanthidae and correlates with the caudal elongation of the straight vertebral column. A careful analysis of ontogenetic changes in the postcranial skeleton is needed to more clearly establish character states relating to the position and structure of anatomical elements.

Colin D. LITTLE & William E. BEMIS:
Observations on the skeleton of the heterocercal tail of sharks (Chondrichthyes: Elasmobranchii)
[pp. 563-573, 5 figs., 1 tabs.]

This paper presents new illustrations and descriptions of the anatomy of the heterocercal tail of twelve species of sharks represented by multiple adult specimens. Nine species from the family Carcharhinidae were examined (bull shark, Carcharhinus leucas; blacktip shark, C. limbatus; dusky shark, C. obscurus; sandbar shark, C. plumbeus; tiger shark, Galeocerdo cuvieri; Atlantic sharpnose shark, Rhizoprionodon terraenovae; bonnethead, Sphyrna tiburo; great hammerhead, S. mokarran; and scalloped hammerhead, S. lewini). We also studied one species of Alopiidae (common thresher shark, Alopias vulpinus), as well as one species of Ginglymostomatidae (nurse shark, Ginglymostoma cirratum) and one species of Triakidae (smooth dogfish, Mustelus canis). Our most interesting observations concern anatomical relationships of the hemal arches and spines that support the ventral fin-web of the hypochordal lobe of the caudal fin and the modified neural arches and spines that support the epichordal portion of the caudal fin. Not surprisingly, the patterns of these skeletal elements differ in many details from the patterns described previously for the heterocercal caudal fin of actinopterygians such as paddlefishes, and these differences offer insight into general aspects of the anatomy of heterocercal caudal fins. Additionally, variation in caudal anatomy documented here suggests its potential as a source of phylogenetic characters at the species, generic, family and ordinal levels. This paper is a contribution to a symposium honoring Hans-Peter Schultze and his many contributions to the anatomy and systematics of fossil and living fishes.

Paulo M. BRITO & Pascal P. DEYNAT:
Freshwater Stingrays from the Miocene of South America with comments on the rise of potamotrygonids (Batoidea, Myliobatiformes)
[pp. 575-582, 3 figs.]

Remains of potamotrygonids are found in both late Miocene Solimões (Acre Basin, Brazil, western Amazon region) and Paraná Formations (Paraná Basin, Argentina). The fact that typical potamotrygonid remains occur in these basins, implies that for a time the Paraná drainage system was linked with the paleo-Amazonas-Orinoco system, thus providing tentatively a minimum age of about 11.8-10 Ma. As these river systems were isolated by the Early Miocene, we predict a pre-Miocene origination and subsequent dispersion of potamotrygonids. This hypothesis points to an Late Cretaceous or Early Tertiary (Paleocene to Oligocene) origin for potamotrygonids during one of several marine transgressions from the Caribbean region.

Gloria ARRATIA & Richard CLOUTIER:
A new cheirolepidid fish from the Middle-Upper Devonian of Red Hill, Nevada, USA
[pp. 583-598, 10 figs.]

A new species of Cheirolepis is described from the Middle-Upper Devonian of Nevada, USA. Comparisons with Cheirolepis trailli and C. canadensis show that the new species is characterized by a unique combination of characters, e.g., a small orbit, a short spiracular slit, a very elongate and narrow dermosphenotic, a long sutural connection between dermosphenotic and supratemporal, and cranial ornamentation including numerous ganoine ridges and few, small tubercles. A pineal foramen is absent, as are the intertemporal and quadratojugal bones. A series of small pores confirm the presence of a jugal pit line; however, a continuation of the jugal pit line and the infraorbital canal has not been observed. The new material enriches the available information on the morphology of Cheirolepis, particularly on the palatoquadrate region and the pectoral girdle, and reveals that there is variation in the cranial morphology of the species of Cheirolepis that cannot be ignored, e.g., the relationship of the dermosphenotic with other bones, and the presence of notch(es) in the nasal bone and preorbital bone. The structure of the rostral region of Cheirolepis still remains unclear due to incomplete preservation.

Eric J. HILTON:
The caudal skeleton of Acipenseriformes (Actinopterygii: Chondrostei): recent advances and new observations
[pp. 599-617, 7 figs.]

The caudal morphology of Acipenseriformes (sturgeons, paddlefishes, and their fossil relatives) is often cited in broad scale comparisons as representative of the plesiomorphic condition for Actinopterygii, yet there is significant skeletal variation within the group that has not been fully appreciated. In this paper, the caudal skeletal morphology of Acipenseriformes is reviewed, and characters used in previous systematic analyses, as well as newly discovered and newly defined characters, are discussed in the context of current phylogenetic hypotheses. Characters identified and their suggested phylogenetic distribution include: 1) dorsal and ventral lobes of caudal fin being nearly equilobate (†Protopsephurus and Polyodon + †Crossopholis, or Polyodontidae with a reversal in Psephurus; independently derived in †Chondrosteus); 2) elongation and flattening of the caudal peduncle (Scaphirhynchus + †Protoscaphirhynchus); 3) caudal peduncle completely armored (Scaphirhynchus + †Protoscaphirhynchus); 4) presence of a cercus or caudal fin filament (Scaphirhynchini); 5) absence of epaxial caudal fin rays (independently derived in †Yanosteus and Scaphirhynchini); 6) reduction in size and number of dorsal caudal fulcra (reduced in size in most †Peipiaosteidae, only a single dorsal caudal fulcrum in †Peipiaosteus pani); 7) forking of anterior dorsal caudal fulcra rounded (Polyodontidae, unknown or inapplicable in stem-group Acipenseriformes); 8) median ridge on the ventral surface of the first dorsal caudal fulcrum (Polyodontinae + †Paleopsephurus); 9) single ventral caudal fulcrum (Acipenseroidei); 10) lateral line sensory canal following the base of the caudal fin rays posteriorly (Acipenseroidei or Acipenseroidei + †Chondrosteus; also found in certain other basal actinopterygians); 11) lateral line sensory canal always enclosed in simple ossified tubes the entire length of caudal fin (Polyodontidae); 12) reduction or loss of rhombic caudal scales (†Peipiaosteidae); and 13) foraminae or notches for caudal vessels in cartilaginous portion of parhypural (independently derived in Polyodontinae or Polyodontinae + †Paleopsephurus and †Peipiaosteidae).

Christopher FIELITZ:
The phylogenetic relationships of the †Enchodontidae (Teleostei: Aulopiformes)
[pp. 619-634, 8 figs., 2 tabs., 1 app.]

A phylogenetic hypothesis of relationships among certain fossil aulopiform fishes and their relationship to extant aulopiform fishes was inferred by using morphology. A consensus of three equally parsimonious trees found that fossil aulopiform fishes consisting of †Cimolichthys, †Enchodus, †Eurypholis, †Palaeolycus, †Parenchodus, †Rharbichthys, and †Saurorhamphus are the sister group to the Alepisauridae (Alepisaurusand Omosudis). †Rharbichthys was basal to a clade consisting of †Palaeolycus + ((†Eurypholis + †Saurorhamphus) + (†Enchodus + †Parenchodus)). The position of †Parenchodus suggests that it is a highly derived species of †Enchodus. The position of †E. marchesettii as the basal species suggests that the genus arose in the mid-Tethys Ocean in what is now the Middle East. The clade consisting of (†E. venator + †E. shumardi) + (†E. gracilis + (†E. petrosus + (†E. gladiolus + †E. dirus))) suggests that these species originated from North America with ancestors of †E. venator and †E. gracilis migrating to the western edge of the Tethys Sea.

Gloria ARRATIA, Adriana LÓPEZ-ARBARELLO, Guntupalli V. R. PRASAD, Varun PARMAR & Jürgen KRIWET:
Late Cretaceous-Paleocene percomorphs (Teleostei) from
India – Early radiation of Perciformes
[pp. 635-663, 18 figs.]

A new taxon of Percoidei, Indiaichthys bamanborensis n. gen. et sp. and a Percomorpha indet. from the intertrappean beds at Bamanbor, India (Late Cretaceous to early Paleocene age) are described and their systematic positions are discussed. The Indian material is represented by articulated specimens that are some of the most complete Late Cretaceous–Paleocene perciforms recovered until now. Due to its generalized morphological pattern, with some resemblances to certain centropomids (e.g., development of the ascendent and articular processes of the premaxilla, two epurals, and two uroneurals) and its unique combination of features (e.g., partial fusion of distal radials in the most anterior dorsal pterygiophores, dentigerous arm of premaxilla very deep posteriorly, co-occurrence of two epurals and two uroneurals), Indiaichthys cannot be assigned to any of the fossil or extant percoid families and it is left in a family indeterminate. In contrast to Indiaichthys, Percormorpha indet. is known by one incomplete but nevertheless rather informative specimen. According to the available information, it shares with some centropomids the peculiar morphology of the posterior region of the lower jaw. However, due to its incompleteness it cannot be assigned to any percomorph group.

Articulated Late Cretaceous-Paleocene perciforms are represented by few specimens that have been recovered in different continents. For instance, the extinct genera Eoserranus and Indiaichthys n. gen. (India), Nardoichthys (Italy), and Saldenioichthys (Argentina) and the extant genus Percichthys (Bolivia). Due to their combination of characters Nardoichthys and Saldenioichthys are interpreted as Perciformes incertae sedis, and Eoserranus and Indiaichthys as Percoidei incertae sedis. The oldest articulated perciform, Nardoichthys, was recovered in sediments of about 71 Ma in age; in contrast, the oldest perciform otoliths are about 85 Ma old and were recovered in Austria. Because there are only a few Late Cretaceous records of otoliths and skeletal remnants of extinct perciforms, it is apparently that many of the extant families arose during the Eocene-Miocene. However, the fossil record is still poor in comparison to the enormous diversification of perciforms today. Even considering the incompleteness of the fossil record, the available information indicates that perciforms underwent a tremendous radiation during the last five millions years.

CHANG Mee-mann:
Synapomorphies and scenarios – more characters of Youngolepis betraying its affinity to the Dipnoi
[pp. 665-686, 13 figs., 2 apps.]

The close relationship between the Porolepiformes and Dipnoi, based on the unusual combination of characters in Youngolepis and Diabolepis, was suggested nearly 20 years ago, and has since been well corroborated by many workers. The remaining question, however, is whether Youngolepis is closer to the Dipnoi or to the Porolepiformes. Although new morphological features in the related forms tend to further support the relationship, the problem is yet to be completely settled. There are strongly competing alternative views. Some hold that dipnoans (with or without Diabolepis), Youngolepis and Powichthys occupy either the basal rhipidistian or basal sarcopterygian position. Others, though accepting the dipnoan affinity of the Porolepiformes (including Powichthys, Youngolepis and Diabolepis) with some reservations, reject any particular close relationship between Youngolepis, Diabolepis and the dipnoans.

During the reexamination of the grinding series and the specimens of Youngolepis praecursor, several additional morphological features were identified. Most of them were described in 1982. Yet these have somehow eluded most workers on the subject. These features have not only confirmed but also strengthened the point of view that Youngolepis and the Porolepiformes are stem dipnoans.

Michael D. GOTTFRIED, Raymond R. ROGERS & Kristina CURRY ROGERS:
First record of Late Cretaceous coelacanths from Madagascar
[pp. 687-691, 3 figs.]

Coelacanth material is present in the Upper Cretaceous (?Santonian/Coniacian) Ankazomihaboka sandstones of northwestern Madagascar. The specimens include a median extrascapular assigned here to Axelrodichthys, and lower jaw elements with a distinctive dermal ornament that closely resembles that of both Axelrodichthysand Mawsonia. Both genera are also known from Aptian/Albian Cretaceous deposits in Africa and South America, and from the Cenomanian of Africa. The Ankazomihaboka specimens represent the youngest fossil record for coelacanths from any Gondwanan site, and, because the Ankazomihaboka beds are freshwater/terrestrial in origin, this occurrence indicates that at least some coelacanths still inhabited freshwater habitats in the Late Cretaceous, in contrast to the deep-water marine refugia in which Recent Latimeria survive. The Madagascar coelacanth material also provides further evidence for an emerging pattern of vertebrate faunal and biogeographic relationship between Madagascar and South America during the latter stages of the Cretaceous.

The first discoveries of Early Paleozoic fishes took place in Scotland and in the Baltic area (Estonia, Latvia, and Lithuania) at the beginning of the 19th century. The first early vertebrate remains recorded from Scotland were of Carboniferous age and are now referred to the sarcopterygian Rhizodus and Megalichthys. Later, discoveries of additional Scottish and Baltic Devonian localities made these regions (and also European workers) the main source of information on early vertebrates for a long time. This situation reached its most important development with the contributions of Erik STENSIÖ and other Swedish (e.g., Gustav WÄNGSJÖ, Erik JARVIK) and Danish colleagues (e.g., Eigil NIELSEN), who organized important collecting expeditions (e.g., Podolia, Spitsbergen and Greenland). New material from these localities and others (such as the Devonian localities of eastern Canada) allowed STENSIÖ and his followers (the so-called Swedish School) to produce some fascinating morphological work and to build up some hypotheses about the origin of early tetrapods that still today are a source of discussion.

New scientific findings have the potential to produce considerable changes in previous interpretations. Vertebrates are not an exception. Based on information gathered over almost two centuries it has long been believed that the origin of vertebrates occurred "sometime" during the earliest Paleozoic, "somewhere" in the northern Hemisphere. However, discoveries of early vertebrates in the Southern Hemisphere, in Australia and in Bolivia, led to a new understanding of the early history of the group. These new discoveries have been remarkable in stimulating new collecting. Recent progress has included the discovery of the "earliest" forms in the Lower Cambrian of southern China and by new and controversial interpretations of the conodonts.

The most recent decade has led to new findings that concern not only the earliest vertebrates, but also most fish groups as well as lower tetrapods. They shed new light on the origin and diversification of basal vertebrates (e.g., heterostracans, anaspids) and gnathostomes (e.g., placoderms, chondrichthyans, acanthodians, actinopterygians, sarcopterygians and earlier tetrapods). Critical fossils have been discovered in different parts of the world, such as China, North America including the Canadian Arctic, Morocco and Australia. This new material has significant impacts on previous character interpretation and distribution, as well as on previous phylogenetic hypotheses.

We thought that it would be refreshing for many of us to gather information on these recent advances concerning our understanding of the origin and early radiation of vertebrates. For this purpose we organized the Symposium "Recent Advances in the Origin and Early Radiation of Vertebrates, A Symposium Honoring Hans-Peter Schultze" dealing with the subject at the 62nd Annual Meeting of the Society of Vertebrate Paleontology, that was held in Norman, Oklahoma, on October 9, 2002. The symposium included both oral contributions and related posters most of which are included in this book.

This book brings together many of these recent discoveries and new interpretations to commemorate the retirement of Hans-Peter SCHULTZE from the Museum für Naturkunde in Berlin. H.-P. SCHULTZE has worked on most groups of lower vertebrates ranging from conodonts to early tetrapods. He has collected in most of the crucial sites around the world. He is one of the most productive researchers in paleoichthyology and is considered by many to be the leading figure in this field.