Early tetrapod relationships revisited

Biol. Rev. (2003), 78, pp. 251–345. f Cambridge Philosophical Society DOI : 10.1017/S1464793102006103 Printed in the United Kingdom 251 Early tetrapod relationships revisited MARCELLO RUTA1*, MICHAEL I. COATES1 and DONALD L. J. QUICKE2 1 The Department of Organismal Biology and Anatomy, The University of Chicago, 1027 East 57th Street, Chicago, IL 60637-1508, USA ( mruta@midway.uchicago.edu ; mcoates@midway.uchicago.edu ) 2 Department of Biology, Imperial College at Silwood Park, Ascot, Berkshire SL5 7PY, UK and Department of Entomology, The Natural History Museum, Cromwell Road, London SW 7 5BD, UK ( d.quicke@ic.ac.uk ) (Received 29 November 2001; revised 28 August 2002; accepted 2 September 2002) ABSTRACT In an attempt to investigate differences between the most widely discussed hypotheses of early tetrapod relationships, we assembled a new data matrix including 90 taxa coded for 319 cranial and postcranial characters. We have incorporated, where possible, original observations of numerous taxa spread throughout the major tetrapod clades. A stem-based ( total-group ) definition of Tetrapoda is preferred over apomorphy- and node-based ( crown-group ) definitions. This definition is operational, since it is based on a formal character analysis. A PAUP* search using a recently implemented version of the parsimony ratchet method yields 64 shortest trees. Differences between these trees concern : ( 1) the internal relationships of aı̈stopods, the three selected species of which form a trichotomy ; ( 2) the internal relationships of embolomeres, with Archeria crassidisca and Pholiderpeton scutigerum collapsed in a trichotomy with a clade formed by Anthracosaurus russelli and Pholiderpeton attheyi ; ( 3 ) the internal relationships of derived dissorophoids, with four amphibamid species forming an unresolved node with a clade consisting of micromelerpetontids and branchiosaurids and a clade consisting of albanerpetontids plus basal crown-group lissamphibians ; ( 4 ) the position of albenerpetontids and Eocaecilia micropoda, which form an unresolved node with a trichotomy subtending Karaurus sharovi, Valdotriton gracilis and Triadobatrachus massinoti ; (5 ) the branching pattern of derived diplocaulid nectrideans, with Batrachiderpeton reticulatum and Diceratosaurus brevirostris collapsed in a trichotomy with a clade formed by Diplocaulus magnicornis and Diploceraspis burkei. The results of the original parsimony run – as well as those retrieved from several other treatments of the data set (e.g. exclusion of postcranial and lower jaw data ; character reweighting ; reverse weighting ) – indicate a deep split of early tetrapods between lissamphibian- and amniote-related taxa. Colosteids, Crassigyrinus, Whatcheeria and baphetids are progressively more crownward stem-tetrapods. Caerorhachis, embolomeres, gephyrostegids, Solenodonsaurus and seymouriamorphs are progressively more crownward stem-amniotes. Eucritta is basal to temnospondyls, with crown-lissamphibians nested within dissorophoids. Westlothiana is basal to Lepospondyli, but evidence for the monophyletic status of the latter is weak. Westlothiana and Lepospondyli form the sister group to diadectomorphs and crown-group amniotes. Tuditanomorph and microbrachomorph microsaurs are successively more closely related to a clade including proximodistally : ( 1 ) lysorophids ; ( 2 ) Acherontiscus as sister taxon to adelospondyls ; ( 3 ) scincosaurids plus diplocaulids ; ( 4) urocordylids plus aı̈stopods. A data set employing cranial characters only places microsaurs on the amniote stem, but forces remaining lepospondyls to appear as sister group to colosteids on the tetrapod stem in several trees. This arrangement is not significantly worse than the tree topology obtained from the analysis of the complete data set. The pattern of sister group relationships in the crownward part of the temnospondyl-lissamphibian tree re-emphasizes the important role of dissorophoids in the lissamphibian origin debate. However, no specific dissorophoid can be identified as the immediate sister taxon to crown-group lissamphibians. The branching sequence of various stem-group amniotes reveals a coherent set of internested character-state changes related to the acquisition of progressively more terrestrial habits in several Permo-Carboniferous forms. Key words : amniotes, characters, congruence, lissamphibians, parsimony ratchet, taxon exemplar, tetrapods, total-group. * Author for correspondence. Marcello Ruta, Michael I. Coates and Donald L. J. Quicke 252 CONTENTS I. II. III. IV. V. Introduction ................................................................................................................................................ Conflicting cladogram topologies ............................................................................................................. Historical background ................................................................................................................................ Taxonomic definitions ............................................................................................................................... Taxon exemplars ........................................................................................................................................ ( 1 ) Devonian and Lower Carboniferous taxa ..................................................................................... ( a ) Taxonomic sample ..................................................................................................................... ( b ) Remarks ....................................................................................................................................... ( 2 ) Acherontiscidae ................................................................................................................................. ( a ) Taxonomic sample ..................................................................................................................... ( b ) Remarks ....................................................................................................................................... ( 3 ) Adelospondyli .................................................................................................................................... ( a ) Taxonomic sample ..................................................................................................................... ( b ) Remarks ....................................................................................................................................... ( 4 ) Aı̈stopoda ............................................................................................................................................ ( a ) Taxonomic sample ..................................................................................................................... ( b ) Remarks ....................................................................................................................................... ( 5 ) Baphetidae .......................................................................................................................................... ( a ) Taxonomic sample ..................................................................................................................... ( b ) Remarks ....................................................................................................................................... ( 6 ) Colosteidae ......................................................................................................................................... ( a ) Taxonomic sample ..................................................................................................................... ( b ) Remarks ....................................................................................................................................... ( 7 ) Diadectomorpha ................................................................................................................................ ( a ) Taxonomic sample ..................................................................................................................... ( b ) Remarks ....................................................................................................................................... ( 8 ) Embolomeri and Eoherpetontidae ................................................................................................. ( a ) Taxonomic sample ..................................................................................................................... ( b ) Remarks ....................................................................................................................................... ( 9 ) Gephyrostegidae ................................................................................................................................ ( a ) Taxonomic sample ..................................................................................................................... ( b ) Remarks ....................................................................................................................................... ( 10 ) Lysorophia .......................................................................................................................................... ( a ) Taxonomic sample ..................................................................................................................... ( b ) Remarks ....................................................................................................................................... ( 11 ) Microsauria ........................................................................................................................................ ( a ) Taxonomic sample ..................................................................................................................... ( b ) Remarks ....................................................................................................................................... ( 12 ) Nectridea ............................................................................................................................................ ( a ) Taxonomic sample ..................................................................................................................... ( b ) Remarks ....................................................................................................................................... ( 13 ) Seymouriamorpha ............................................................................................................................. ( a ) Taxonomic sample ..................................................................................................................... ( b ) Remarks ....................................................................................................................................... ( 14 ) Temnospondyli .................................................................................................................................. ( a ) Taxonomic sample ..................................................................................................................... ( b ) Remarks ....................................................................................................................................... ( 15 ) Crown-group Lissamphibia ............................................................................................................. ( 16 ) Crown-group Amniota ..................................................................................................................... ( 17 ) Outgroups .......................................................................................................................................... VI. Characters ................................................................................................................................................... VII. Analysis ........................................................................................................................................................ ( 1 ) Character coding ................................................................................................................................. ( 2 ) The parsimony ‘ratchet ’ .................................................................................................................... 253 255 257 259 261 261 261 262 262 262 262 262 262 262 263 263 263 263 263 263 263 263 263 264 264 264 264 264 264 265 265 265 265 265 265 265 265 266 267 267 267 268 268 268 268 268 268 269 270 270 270 270 270 271 Early tetrapod relationships revisited VIII. IX. X. XI. XII. XIII. XIV. 253 ( 3) Results ................................................................................................................................................... ( a) The tetrapod stem-group ............................................................................................................ ( b) The lissamphibian stem-group ................................................................................................... ( c) The amniote stem-group ............................................................................................................. ( d) The affinities of Westlothiana ........................................................................................................ ( e) The status of microsaurs .............................................................................................................. ( f ) The aı̈stopod-nectridean clade ................................................................................................... ( g) Acherontiscus is an adelospondyl .................................................................................................... ( h) The position of lysorophids ......................................................................................................... ( 4) Reweighted analysis ............................................................................................................................ ( 5) Cranial data ......................................................................................................................................... ( 6) Deletion of lower jaw characters ....................................................................................................... ( 7) Reverse weighting ............................................................................................................................... ( 8) Chronology .......................................................................................................................................... Future research ........................................................................................................................................... Conclusions ................................................................................................................................................. Acknowledgements ..................................................................................................................................... References ................................................................................................................................................... Appendix 1. List of specimens examined ................................................................................................ Appendix 2. Character list ........................................................................................................................ Appendix 3. Data matrix .......................................................................................................................... I. INTRODUCTION This paper presents a new, detailed cladistic analysis of early tetrapods as part of an ongoing project aiming to discover sources of conflict between the most widely discussed, published tetrapod phylogenies and to search for correlated character transformations in early tetrapod evolution. A long-term goal of our investigation is to generate and test new hypotheses of relationships using, as far as possible, exhaustive taxon and character combinations not considered in previous studies. Primitive tetrapod interrelationships are a topic of considerable palaeontological and biological interest. Our knowledge of this subject has improved considerably during the last two decades (see Sections II and III). Significant fossil discoveries have cast new light on the pattern of anatomical transformations that occurred at the vertebrate transition from water to land (e.g. Jarvik, 1980, 1996; Clack, 1989, 1994a, b, 1998b; Coates & Clack, 1990, 1991, 1995; Ahlberg, 1991, 1995, 1998; Ahlberg, Luksevics & Lebedev, 1994; Daeschler et al., 1994; Lebedev, 1984; Clack & Coates, 1995; Lebedev & Coates, 1995; Coates, 1996; Ahlberg & Clack, 1998; Ahlberg, Luksevics & Mark-Kurik, 2000; Daeschler, 2000; Shubin & Daeschler, 2001). Revised interpretations of palaeontological and comparative anatomical data have clarified the intrinsic and extrinsic relationships of numerous extinct groups (e.g. Panchen, 1985; Smithson, 1985; Panchen & Smithson, 1987, 1988, 1990; Clack, 1996), and descriptions of 271 275 277 280 282 283 283 284 284 285 287 288 289 290 290 292 293 294 304 304 331 several new taxa are beginning to bridge the morphological and/or stratigraphical gap between Devonian and Carboniferous faunas (e.g. Clack, 1998 a, c, 2001, 2002; Lombard & Bolt, 1995; Thulborn et al., 1996; Clack & Finney, 1997; Paton, Smithson & Clack, 1999; see also Bolt & Lombard, 2000; Clack & Carroll, 2000, and Ruta, Milner & Coates, 2001, for a review). As in the case of other areas of palaeobiology, early tetrapod studies have benefitted from interactions between classical morphology and modern embryology at the interface between evolution and development (e.g. Shubin & Alberch, 1986; Coates, 1991, 1995, 1996; Thorogood, 1991; Duboule, 1994; Shubin, 1995; Sordino & Duboule, 1995; Sordino, van der Hoeven & Duboule, 1995; Tickle, 1995; Cohn et al., 1997; Shubin, Tabin & Carroll, 1997; Coates & Cohn, 1998; Jeffery, 2001). Recently, much interest has centered on comparisons between morphological and molecular analyses, on fossil-based calibrations of molecular clocks, and on the timing of such key events as the phylogenetic split between lissamphibians and amniotes (e.g. Feller & Hedges, 1998; Kumar & Hedges, 1998; Hedges, 2001; Van Tuinen, Porder & Hadly, 2001; Ruta & Coates, in press). Reconstructing the branching sequence of early tetrapods is a necessary prerequisite to inform a wide range of questions, such as: (1) understanding the anatomical, physiological and ecological modifications that accompanied the transition from fish ancestors to four-legged vertebrates; (2) establishing the sequence of character acquisitions that 254 Marcello Ruta, Michael I. Coates and Donald L. J. Quicke shaped the early evolutionary history of lissamphibians and amniotes, from their phylogenetic separation to the diversification of their respective crown-groups; (3) reconstructing the changes in developmental pathways that resulted in new morphologies; (4) formulating and testing hypotheses concerning tempo and mode of evolutionary processes through analysis of character change and degree of character correlation; (5) assessing taxon diversity and morphological disparity through time, as inferred from the shape of cladograms; and (6) providing a framework for comparing and contrasting morphology-based and molecular chronologies of key events in vertebrate evolution. In addition, early tetrapod studies contribute an invaluable source of data for elucidating palaeobiogeographical and palaeoecological patterns in primitive terrestrial biota. The earliest diversification of lissamphibians and amniotes is the main focus of this review. In addition, conflicting hypotheses of tetrapod interrelationships are compared, with emphasis on the lissamphibian origin debate, on the pattern of character acquisition in the amniote stem-group, and on the status and affinities of various groups of lepospondyls. There is no wellsupported, let alone stable, hypothesis of early tetrapod phylogeny. The skepticism expressed by Coates, Ruta & Milner (2000), although questioned by Laurin (2002), remains justifiable. Throughout, we make extensive use of the concepts of stem-group, crown-group and total-group, first introduced by Hennig (1965, 1966, 1969) and elaborated upon by various subsequent workers (e.g. Jefferies, 1979; Craske & Jefferies, 1989; Patterson, 1981, 1993 a, b, 1994). The voluminous literature dealing with the relationships of several living and extinct groups, as well as the large number of papers discussing the theory and practice of cladistic analysis, made us assume that it would be unnecessary to clarify the use of basic cladistic jargon. However, a random browse through the palaeontological literature revealed that the stem-group, crown-group and totalgroup concepts are still misinterpreted and, often, used incorrectly. We can add little to the exhaustive discussion of cladistic terminology in the existing literature (e.g. Jefferies, 1979; Craske & Jefferies, 1989; Smith, 1994; Budd, 2001). Therefore, only succinct definitions are appropriate here. Thus, let A and L be two monophyletic groups of extant organisms (e.g. Amniota and Lissamphibia), and let A and L be sister groups. The group including the latest common ancestor of all extant members of A (or L) plus all of its descendants, both living and extinct, is the crowngroup of A (or L). All those extinct organisms which are more closely related to the extant members of A (or L) than to those of L (or A), but which do not belong in the crown group of A (or L), are part of the stem-group of A (or L). The point of latest common ancestry of A and L marks the separation between the total-group of A and the total-group of L. Any fossil organism that belongs in the total-group of A (or L) shares a more recent common ancestor with some or all of the extant members of A (or L) than it does with any extant member of L (or A). If the organism in question is more closely related, in equal measure, to all extant members of A (or L) than to those of L (or A), then it is a member of the stem-group of A (or L), i.e. it branches from the total-group of A (or L) before the basal node marking the beginning of the crown-group radiation. If it is more closely related to some (but not all) extant members of A (or L) than to others, then it is a member of the crowngroup of A (or L). The stem-group is an extinct and paraphyletic assemblage by definition. Crown-group definitions (sometimes referred to as node-based definitions) represent a particular case of nodal or apomorphy-based definitions (in this case, the node subtends the crown-group; Forey, 2001). Apomorphy-based definitions refer to the crown-group and to the portion of the stem-group subtended by the relevant apomorphy. Such a broader monophyletic assemblage corresponds to Craske & Jefferies’s (1989) scion. Finally, stem-based definitions are simply totalgroup definitions (they encompass both the entire stemgroup and the crown-group). As clearly stated by Donoghue, Forey & Aldridge (2000: p. 237), ‘… Although crown and total groups can be given separate names (deQueiroz & Gauthier, 1992), this approach results in an unnecessarily expanded classification scheme and in one of the two groups (stem-group) being paraphyletic (unless that group is represented by one species only, in which case the need for a higher group name is unnecessary). Generally, only the total group is recognized by formal Linnean rank (Patterson, 1993 b; Smith, 1994). Thus, conodonts belong to the Gnathostomata; they are gnathostomes, albeit without jaws’. Following Donoghue et al.’s (2000) example, Eusthenopteron is a tetrapod (more precisely, a stem-group tetrapod), albeit without limbs. Stem-groups are divided into plesions, which ‘… could be inserted anywhere within the classification without altering the Linnean rank of the crown group’ (Smith, 1994: p. 96). Our use of the plesion concept [a totally extinct monophyletic group; see also Smith (1994)] conforms exclusively to that of Patterson & Rosen (1977), but differs from that of Craske & Jefferies (1989), for whom plesions are in principle paraphyletic assemblages. While Patterson & Rosen’s (1977) concept is based upon the pattern of character acquisition along Early tetrapod relationships revisited the stem-group, Craske & Jefferies’s (1989) concept is formulated within the framework of a more idealistic interpretation of the shape of cladograms, whereby segments of the stem-lineage are also incorporated into the plesions. The distinction between the two plesion concepts revolves around the interpretation of cladograms as (almost) strict representations of phylogenetic trees in Craske & Jefferies (1989) and as formulations of hypotheses of character distributions in Patterson & Rosen (1977). II. CONFLICTING CLADOGRAM TOPOLOGIES Published cladistic analyses of early tetrapods show a congruent phylogenetic signal for some groups, such as colosteids and most Devonian forms (Carroll, 1995; Coates, 1996; Laurin & Reisz, 1997, 1999; Ahlberg & Clack, 1998; Clack, 1998 a, c, 2001; Laurin 1998a–c; Paton et al., 1999; Anderson, 2001). However, little consensus has emerged for the relationships and affinities of many other groups [e.g. lepospondyls; see Carroll (1995), Carroll & Chorn (1995), Laurin & Reisz (1997, 1999), Laurin (1998a–c), Paton et al. (1999) and Anderson (2001)], despite the discoveries of new data and the introduction of more powerful analytical techniques. This conflict triggered the present work. Visual inspection of current phylogenies reveals two distinct sets of tree topologies (see also Ruta et al., 2001). The first set consists of trees which place most fossil tetrapods either in the stem-lissamphibians (e.g. temnospondyls) or in the stem-amniotes (e.g. embolomeres) (Panchen & Smithson, 1987, 1988; Panchen, 1991; Ahlberg & Milner, 1994; Coates, 1996; Clack, 1998a, c, 2001; Paton et al., 1999). In these shortstemmed trees, operational taxonomic units (OTUs) are arranged mostly dichotomously (Fig. 1 A, B). The second set of trees reflects the hypothesis that a greater number of fossil tetrapods traditionally allied to lissamphibians or amniotes show no special relationships to either group (e.g. Laurin & Reisz, 1997, 1999; Ahlberg & Clack, 1998; Laurin, 1998 a–c; Anderson, 2001). In these long-stemmed trees (Fig. 1C, D), OTUs form a largely pectinate pattern (unbalanced trees; Smith, 1994). Twelve major groups of early tetrapods are usually recognized (see also below): adelospondyls; aı̈stopods; baphetids; colosteids; diadectomorphs; embolomeres; gephyrostegids; lysorophids; microsaurs; nectrideans; seymouriamorphs; temnospondyls. Fig. 2 shows the percentage distributions of such groups in the tetrapod, lissamphibian and amniote stemgroups, according to various recent studies. For simplicity, these distributions do not take into account 255 several Devonian and Carboniferous genus- and species-level OTUs. The percentage distributions highlight remarkable differences in the number of groups assigned to the lissamphibian and amniote stemgroups. This is especially evident in a comparison of Carroll’s (1995) and our own analyses, in which eight/ nine different groups are placed within stem-amniotes, with the Laurin & Reisz (1999) and Anderson’s (2001) analyses, in which most groups are almost equally distributed between the tetrapod stem-group and the lissamphibian stem-group. Evaluation of conflicting results of published phylogenies is complicated by the use of very different taxon and/or character samples. In several studies, only a small number of OTUs has been considered. These are sometimes represented by supraspecific terminals (e.g. Carroll, 1995), and/or by few genera or species for each major group (e.g. Coates, 1996; Clack, 1998a, c, 2001; Paton et al., 1999). The size of a matrix also depends upon the focus of a particular phylogenetic analysis (e.g. Laurin & Reisz, 1997, 1999; Clack, 1998 a, c, 2001; 1999; Ahlberg & Clack, 1998; Laurin, 1998a–c; Anderson, 2001). Several theoretical considerations suggest that taxon exemplars should be as diverse as possible (e.g. Nixon & Davis, 1991; deBraga & Rieppel, 1997; Anderson, 2001; Prendini, 2001; see also discussion below). Importantly, a recent study based on simulations of true phylogenies (Salisbury & Kim, 2001) indicates that dense and random taxon sampling increases the probability of retrieving correctly the plesiomorphic condition of characters as well as the ancestral state near the tree root. Furthermore, Salisbury & Kim’s (2001) simulations show that in the analysis of small clades, estimates of ancestral states are strongly affected by cladogram topology and by the number of descendent branches in progressively more distal internal nodes. Increasing the number of taxa (as well as the number of characters) poses additional problems, e.g. (1) poor resolution caused by the amount and distribution of missing entries (Wilkinson, 1995; Kearney, 1998, 2002); (2) computation time required by large and complex data sets (Farris et al., 1996; Goloboff, 1999; Nixon, 1999; Quicke, Taylor & Purvis, 2001, and references therein); (3) accuracy in the search for optimal trees; and (4) high levels of homoplasy. The number of characters is obviously a function of the number of taxa and of the degree of morphological variation both within and between examined groups. In addition, the extent to which observed morphologies are ‘atomised’ even for the same taxon sample varies considerably from author to author, as does the perceived importance, or ‘weight’, assigned to particular structures. Marcello Ruta, Michael I. Coates and Donald L. J. Quicke 256 Eusthenopteron Panderichthys Elginerpeton Ventastega Metaxygnathus Acanthostega Ichthyostega Hynerpeton Greererpeton Eryops Balanerpeton Saxonerpeton Tulerpeton Whatcheeria Crassigyrinus Archeria Proterogyrinus Westlothiana Acanthostega Greererpeton Eucritta Baphetes Whatcheeria Balanerpeton Dendrerpeton TEM. MIC. ANT. (A) COL. BAP. Caerorhachis Pholiderpeton Proterogyrinus Eoherpeton Gephyrostegus Cochleosaurus Phonerpeton Microbrachis Balanerpeton Platyrhinops Discosauriscus Diploceraspis Sauropleura Eocaptorhinus Ophiacodon BAP. TEM. Caerorhachis COL. Caerorhachis Panderichthys Elginerpeton Obruchevichthys Ichthyostega Ventastega Metaxygnathus Acanthostega Whatcheeria Tulerpeton Greererpeton Crassigyrinus Megalocephalus COL. Gephyrostegus Bruktererpeton Crassigyrinus Eoherpeton Proterogyrinus Pholiderpeton Seymouria Microbrachis Saxonerpeton Casineria Petrolacosaurus Westlothiana Paleothyris Captorhinus ANT. SEY. MIC. AMN. Acanthostega Ichthyostega Whatcheeria Crassigyrinus Colosteidae Baphetidae Dendrerpeton Eryops Ecolsonia Tersomius Apateon Amphibamus Doleserpeton (B) COL. BAP. TEM. Caerorhachis ANT. TEM. MIC. TEM. SEY. NEC. AMN. (C) Gephyrostegidae Proterogyrinus Archeria Kotlassia Seymouria Ariekanerpeton Westlothiana Limnoscelis Diadectes Synapsida Captorhinidae Procolophonidae Aistopoda Adelogyrinidae Nectridea Pantylus Rhynchonkos Brachystelechida Lysorophia Triadobatrachus Discoglossidae Pipidae Karaurus Hynobiidae Proteidae Sirenidae Eocaecilia Rhinatrematidae Ichthyophiidae ANT. SEY. DIA. AMN. AIS. ADE. NEC. MIC. LYS. LIS. (D) Fig. 1. Phylogenies of early tetrapods, redrawn and modified from Ruta et al. ( 2001 ), after inclusion of Caerorhachis bairdi. ( A ) Coates’ ( 1996) analysis ; ( B) Paton et al.’s ( 1999) analysis ; ( C ) Ahlberg & Clack’s ( 1998) analysis ; ( D) Laurin’s (1998 b ) analysis. Abbreviations as follows : ADE., Adelospondyli ; AIS., Aı̈stopoda ; AMN., crown-group Amniota ; ANT., Anthracosauria (including Embolomeri and Gephyrostegidae) ; BAP., Baphetidae ; COL., Colosteidae ; DIA., Diadectomorpha ; LIS., crowngroup Lissamphibia ; LYS., Lysorophia; MIC., Microsauria ; NEC., Nectridea ; SEY., Seymouriamorpha ; TEM., Temnospondyli. Early tetrapod relationships revisited 257 12 11 10 9 8 66.64 7 6 66.64 41.65 5 4 8.33 3 2 33.32 16.66 1 Panchen and Smithson (1988) Carroll (1995) 8.33 25 16.66 25 25 Carroll (1995) Coates (1996) 8.33 Clack (1998c) Clack (1998a) 12 8.33 11 10 8.33 9 41.65 8 75 7 41.65 6 33.32 5 4 33.32 16.66 3 2 16.66 16.66 Clack (1998a) Paton et al. (1999) 8.33 8.33 1 Paton et al. (1999) Laurin and Reisz (1999) Anderson (2001) Present analysis Fig. 2. Distribution of primitive tetrapod groups in recently published cladistic analyses, based on their assignment to stemtetrapods (dark grey ), stem-lissamphibians ( white ) or stem-amniotes ( black ). The vertical axis represents the total number of groups ( embolomeres and gephyrostegids are treated as separate groups ). Numbers inside bar diagrams indicate their percentage distributions. Light grey areas refer to groups that have not been examined. If an analysis yields different topologies, group distributions are plotted separately for each topology. The diagrams do not consider several Devonian and Carboniferous genera/species. Clearly, characters should be targeted at the diversity displayed by the memberships of very large exemplars. Finally, inclusion/exclusion of taxa and/or characters may affect the outcome of an analysis in unpredictable ways (e.g. Clack, 1998a, c, 2001; Paton et al., 1999; Ruta et al., 2001). However, various theoretical approaches to character inclusion/exclusion often have been misguided by the lack of an adequate conceptual framework [see Grandcolas et al.’s (2001) discussion of the ‘precise primary homologies’ approach]. Also, taxon removal because of incomplete preservation and missing character scores may be undesirable, because such taxa may have a positive effect on cladogram resolution (Novacek, 1992; Wilkinson, 1995; Kearney, 1998, 2002; Anderson, 2001). III. HISTORICAL BACKGROUND Lack of space prevents an exhaustive treatment of the history behind phylogenetic studies of early terrestrial vertebrates. Therefore, only a brief summary is given 258 Marcello Ruta, Michael I. Coates and Donald L. J. Quicke in this section [see also Laurin (1998b), Carroll (2001) and Ruta et al. (2001)]. Among the first attempts at producing synapomorphy schemes used to reconstruct the broad pattern of relationships between major Palaeozoic tetrapod groups are those by Heaton (1979, 1980), Gardiner (1982, 1983), Holmes (1984), Smithson (1985), Gauthier, Kluge & Rowe (1988 a, b), Panchen & Smithson (1987, 1988) and Godfrey (1989). Panchen & Smithson (1988) proposed a deep phylogenetic split between ‘reptiliomorphs’ (amniote-related groups) and ‘batrachomorphs’ (lissamphibian-related groups). In Panchen & Smithson’s (1988) scheme, ichthyostegids, nectrideans, colosteids and microsaurs are successively more closely related to a temnospondyllissamphibian clade, whereas baphetids (loxommatids), anthracosauroids, seymouriamorphs (the latter two groups sometimes thought to be each other’s closest relatives) and diadectomorphs are successive plesions on the amniote stem-group. The problematic Crassigyrinus was considered to be either a plesion between baphetids and a clade of anthracosauroids plus seymouriamorphs, or the sister taxon to anthracosauroids. Lepospondyl monophyly is explicitly rejected in Panchen & Smithson’s (1988) cladogram. Subsequent studies (e.g. Lombard & Sumida, 1992; Lee & Spencer, 1997) have corroborated to a large extent the scheme of ‘reptiliomorph’ relationships proposed by Panchen & Smithson (1988) and Gauthier et al. (1988b). A series of new studies on Devonian and primitive Carboniferous forms during the mid and late 1990s (Ahlberg, 1991, 1995, 1998; Coates, 1991, 1995, 1996; Ahlberg et al., 1994; Daeschler et al., 1994; Clack & Coates, 1995; Coates & Clack, 1995; Lebedev & Coates, 1995; Lombard & Bolt, 1995; Jarvik, 1996; Clack, 1998 b; Milner & Lindsay, 1998), and the publication of the first, large-scale, computer-assisted cladistic analysis of Palaeozoic tetrapods by Carroll (1995), gave new impetus to early tetrapod research. In Carroll’s (1995) study, the hypothesis of a chronologically deep phylogenetic event leading to the separation between lissamphibians and amniotes is implicit, although no members of the lissamphibian crown-group were included. In addition, lepospondyls form a clade on the amniote stem-group. Following the description of the postcranium of Acanthostega, Coates’ (1996) analysis [an elaborated version of Lebedev & Coates’s (1995) work] followed previous authors’ suggestions that the lissamphibian/amniote split was a deep phylogenetic event. It also corroborated Lebedev & Coates’s (1995) conclusion that such a split can be traced back to the late Devonian, based on Lebedev & Coates’s (1995) and Coates’ (1996) interpretation of Tulerpeton as a primitive ‘reptiliomorph’. Other studies supporting a fundamental dichotomy between Palaeozoic tetrapods, based on different subsets of anatomical characters, are those by Sumida & Lombard (1991), Berman, Sumida & Lombard (1992), Sumida, Lombard & Berman (1992), Lee & Spencer (1997), Sumida (1997) and Berman (2000). Clack (1998 a, c, 2001) and Paton et al. (1999) published detailed character analyses for several Palaeozoic groups with an aim to assess the relationships of such problematic forms as Crassigyrinus, Whatcheeria, Eucritta and Casineria. These works paved the way to further scrutiny of problematic regions of the tetrapod tree. Examples include the lissamphibian-amniote split, the pattern of character acquisition in the crownward part of the tetrapod stem-group and in the basal portions of the lissamphibian and amniote stem-groups, and the placement of ‘difficult’ taxa, such as baphetids. Significantly, the results of Laurin & Reisz’s (1997, 1999) and Laurin’s (1998 a–c) analyses departed radically from those of previous studies. In these works, the tetrapod stem-group became much longer, because a series of groups (e.g. embolomeres, temnospondyls) were removed from amniote or lissamphibian ancestry. As a result, the established pattern of character-state changes along the amniote and lissamphibian stemgroups collapsed. Some of the characters generally considered to be diagnostic of basal ‘reptiliomorphs’ and ‘batrachomorphs’ now informed the order of cladogenetic events preceding the crown-tetrapod radiation. Lepospondyls now formed a paraphyletic array of stem-group lissamphibians, whereas diadectomorphs [as well as Solenodonsaurus in Laurin & Reisz’s (1999) analysis] became the only plausible stem-group amniotes. Some of the conclusions reached by Laurin & Reisz (1997, 1999) and Laurin (1998a–c) were corroborated by Ahlberg & Clack’s (1998) analysis of lower jaw characters, especially with regards to the stemtetrapod affinities of Crassigyrinus, Tulerpeton, Whatcheeria, colosteids and baphetids. Ahlberg & Clack’s (1998) analysis incorporated isolated material into a wider taxon set, and detected patterns of jaw character transformation across the fish–tetrapod transition and the crownward part of the stem-group. However, they also found that lower jaw data are apparently insufficient to retrieve a single origin for several long-accepted Palaeozoic groups, which appear, instead, as para- or polyphyletic assemblages [but see Ruta & Coates (in press)]. The most recent cladistic analyses of early tetrapods are those by Anderson (2001) and Clack (2002). Although few taxa were considered outside lepospondyls, Anderson’s (2001) work generally agrees with Laurin & Reisz’s (1997, 1999) and Laurin’s (1998a–c) Early tetrapod relationships revisited results by placing lepospondyls on the lissamphibian stem, and seymouriamorphs, embolomeres and temnospondyls on the tetrapod stem. Clack’s (2002) analysis encompasses a diverse range of early tetrapod groups, and offers a rather unconventional branching pattern. Whatcheeriids, Crassigyrinus, Eoherpeton, embolomeres and gephyrostegids are successively more closely related to a diverse group including, on the one hand, Westlothiana as sister taxon to lepospondyls, and on the other, seymouriamorphs and temnospondyls as successive sister groups to a clade of colosteids plus Caerorhachis paired with Eucritta plus baphetids. Evaluation of the results of Anderson’s (2001) and Clack’s (2002) studies is beyond the scope of the present review. Persistent conflict indicates that the resolution of several phylogenetic problems must await comprehensive treatment of the expanding tetrapod data base, as well as input from smaller-scale studies targeted at the specific relationships within various groups. As noted by Carroll (2001: p. 1212), ‘We have a great deal of knowledge of the anatomy of a vast array of Paleozoic tetrapods (Heatwole & Carroll, 2000), but the specific interrelationships of the major taxa and their affinities with the modern orders remain impossible to establish with assurance without much more knowledge of fossils from the Lower Carboniferous and from the period between the Lower Permian and the Jurassic’. Carroll’s (2001) statement identifies the problem of discovering unambiguous phylogenetic signal behind the broad spectrum of primitive tetrapod morphologies. This review is intended to resolve some of the current problems, not only by presenting a new hypothesis of relationships, but also by identifying the limits and difficulties of the ongoing debate. Unlike Carroll (2001), we argue that lack of critical fossils from crucial periods of tetrapod history may be less significant than a detailed scrutiny of the evidence available, at least in some regions of the tetrapod tree. IV. TAXONOMIC DEFINITIONS Two issues of taxonomic nomenclature are addressed here. The first relates to the definition and taxonomic content of Tetrapoda, and the second concerns the introduction of ‘… new [taxonomic] names and altered meanings for old names’ (Greene, 2001: p. 738), and their use in phylogenetic systematics. Several definitions of Tetrapoda have been proposed. Laurin & Reisz (1997, 1999), Laurin (1998a–c) and Laurin, Girondot & deRicqlès (2000 a, b) adopt a node-based definition, referring the Tetrapoda exclusively to the crown-group. In several important 259 respects, this resembles Gaffney’s (1979: p.103) explicitly nodal definition of the Neotetrapoda (contra Laurin, 2002). All of these definitions have been established with clear reference to taxon naming within a phylogenetic framework (e.g. deQueiroz & Gauthier, 1990, 1992, 1994; Cantino et al., 1999; Bryant & Cantino, 2002, and references therein). According to Laurin et al. (2000b), the ‘… [phylogenetic nomenclatural] system clarifies the taxonomy … because there is only one valid phylogenetic definition (the first published one) for each taxon name’. We note that this definition prunes the content of the Tetrapoda, relative to previous uses of this term (Coates, 1996; Benton, 2000; Coates et al., 2000; Forey, 2001; see also below). However, as pointed out by Coates et al. (2000: p. 327), not only the biological community, but ‘… the world at large has a say about what is, or is not, a tetrapod’. Tetrapods have long been identified on the basis of limbs with digits, i.e. synonymous with dactyly, but it is now clear that digit presence extends beyond the crown clade. Within Laurin’s (1998b) and Laurin & Reisz’s (1999) preferred tree topologies, several Upper Devonian to Upper Permian dactylous groups, crownward of Panderichthys, are excluded from the Tetrapoda. Instead, these clades now rank among a heterogeneous stem assemblage of ‘non-tetrapod stegocephalians’. The Tetrapoda, sensu Laurin (1998b), is poorly informative in evolutionary as well as general biological discussions of dactylous vertebrates as a whole, although this has been one of the most easily recognized of all vertebrate groups (Goodrich, 1930; Romer, 1966; Gaffney, 1979; Panchen & Smithson, 1987, 1988; Benton, 1988, 2001; Carroll, 1988; Schultze & Trueb, 1991). As an alternative, we advocate the use of a total-group (stem-based) definition of the Tetrapoda. Objections to a stem-based definition have been raised by Ahlberg (1998) and Ahlberg & Clack (1998) (see also Clack, 1998c, 2001). Their arguments can be summarized as follows: (1) there is as yet no consensus on the identity of the Recent sister group of living tetrapods (Forey, 1998; Zardoya & Meyer, 2001); (2) the taxonomic content of the ‘fish-like’ portion of the tetrapod stem-group is not agreed upon [see Zhu & Schultze (2001) and Johanson & Ahlberg (2001) for summary hypotheses]; and (3) digits cannot be used to characterize the basal ‘fish-like’ part of the tetrapod stem-group. Furthermore, although limb bone fragments have been attributed to various Devonian stem-tetrapods (e.g. Ahlberg, 1991, 1998; Ahlberg et al., 1994), the occurrence of digits in such forms is unknown. For these reasons, Ahlberg (1998) and 260 Marcello Ruta, Michael I. Coates and Donald L. J. Quicke Ahlberg & Clack (1998) propose an apomorphy-based definition, but acknowledge that this represents only a temporary solution (see also Clack, 1998 c). Specifically, dactyly is chosen as the key derived character for Tetrapoda – again, in agreement with traditional definitions (see comments in Anderson, 2001). This clade encompasses all taxa crownward of Elpistostegalia (the group including Panderichthys and allied forms). However, once again, the exclusion of various basal groups from Tetrapoda (e.g. the paraphyletic array of ‘osteolepiform fishes’ ; Ahlberg & Johanson, 1998) limits the information content discovered in cladistic analysis, although less so than Laurin’s and Laurin & Reisz’s node-based definition. Such basal groups are named, rather clumsily, ‘non-tetrapod tetrapodomorphs’ (e.g. Cloutier & Ahlberg, 1996; Ahlberg & Johanson, 1998). Operationally, the adoption of a total-group definition (e.g. Jefferies, 1979; Craske & Jefferies, 1989; Budd, 2001; Jeffery, 2001) solves some of the abovementioned nomenclatural problems (see also discussion in Coates, 1996). Thus, the tetrapod stem-group includes any fossil taxon that can be shown, based on a formal character analysis, to be more closely related to lissamphibians and amniotes than to any other living group. This holds true regardless of the presence/ absence of key apomorphic features (such as digits), and the identity of the tetrapods’ extant sister taxon (e.g. Jeffery, 2001). If a fossil is more closely related to either extant lissamphibians or extant amniotes, it is a member of the crown-group by definition (see above). Uncertainty in the placement of extinct forms in the basal part of the crown-group or in the crownward part of the stem-group is irrelevant, because the membership of a total-group only concerns closeness of relationship of any fossil to one particular extant clade (Patterson & Rosen, 1977; Jefferies, 1979; Craske & Jefferies, 1989; Forey, 1993; Patterson, 1993 a, b, 1994; Smith, 1994; Nixon & Carpenter, 2000; Budd, 2001; but see also Brochu & Sumrall, 2001). The approach advocated here is not new. For example, the ‘stemmodified node-based definition’ employed by Meng et al. (1994) and Wyss & Meng (1996) is much the same in its formulation [see comments in Sereno (1998, 1999) and Bryant & Cantino (2002)]. Importantly, a total-group definition can accommodate a node-based phylogenetic definition without sacrificing cladistic information, whereas the converse is not true. From a purely operational perspective, stability in taxonomic content and degree of corroboration of clades (Lee & Spencer, 1997) may become secondary issues. Also, Lee & Spencer (1997) and Sereno (1998, 1999) showed that adoption of crown-clade definitions does not necessarily lead to increased taxonomic stability. Furthermore, we leave aside the issues of naming cladogram nodes within extinct monophyla, and the widespread misuse of plesions [sensu Patterson & Rosen, 1977; see Craske & Jefferies’s (1989) approach to plesion subdivision]. Several examples from the literature show that the often heated debates on assignments of taxonomic names to specific cladogram nodes (e.g. Aves, Mammalia, Tetrapoda) may be ill-conceived. These debates could be avoided by adopting total-group definitions, which are only marginally affected by reshuffling of extinct taxa, or by changes in the fossil membership of crown-groups. Thus, referring to Eusthenopteron as a stem-tetrapod should not be more puzzling or counterintuitive than regarding ceratopian dinosaurs (e.g. Triceratops) as stem-birds, sail-back pelycosaurs (e.g. Dimetrodon) as stem-mammals, or Australopithecus as a stem-human. In all cases, an explicit hypothesis of relationship with a living monophylum is provided, and ‘… the name and rank given to the clade formed by the modern phena is extended to include all stem group members of that taxon’ (Smith, 1994: p. 97). Sereno (1998, 1999) notes that the interpretation of fossils always relies upon identification of one or more key features shared with a Recent taxon, even if such fossils fall outside the crown-group. Thus, using an appropriate name modifier might be suitable in dealing with crown-clades (e.g. living or Recent Mammalia; living or Recent Aves) instead of restricting a wellknown name to the crown-clade exclusively. Apomorphy-based names are problematic in at least two respects. First, a taxon for which a character cannot be coded (e.g. because of unknown or inapplicable information) may fall inside or outside a group defined upon the possession of the character in question. Its inclusion or exclusion from the group depends upon alternative character state optimizations (e.g. accelerated or delayed transformations). Likewise, if a taxon does not show a certain character, but its position in a cladogram is nested between groups that display that character, then optimization implies either parallel acquisitions (delayed transformation), or a single origin followed by secondary loss in the taxon in question (accelerated transformation). Second, key apomorphy definitions can be problematic. Various conditions of particular structures may occur at different stem-group nodes (e.g. integumentary structures preceding true feathers in several theropods; e.g. Xing, Zhong-He & Prum, 2001). Anderson’s (2001) definition of Tetrapoda uses Elpistotegalia and crown-tetrapods as ‘anchor’ taxa, and is argued to be consistent with traditional usage as Early tetrapod relationships revisited well as phylogenetic nomenclature. We acknowledge the rationale behind Anderson’s (2001) usage, but the exclusion of taxa less crownward than Elpistostegalia is somewhat arbitrary. Moreover, the monophyly of Elpistostegalia is questionable, and there is always the potential that incompletely known fossils, such as Elpistostege (Schultze & Arsenault, 1985), could turn out to be more closely related to alternative groups, the consequences of which would depart radically from Anderson’s stated intention. A further nomenclatural issue concerns the application of historically laden names to novel phylogenetic definitions (see also Anderson, 2001). Laurin’s (1998a–c) Anthracosauria is a prime example, because it includes none of the taxa traditionally placed within ‘anthracosaurs’, such as embolomeres, gephyrostegids and, more questionably, seymouriamorphs (Heaton, 1980; Smithson, 1985; Panchen & Smithson, 1987, 1988; Gauthier et al., 1988b; Forey, 2001). Instead, the new definition refers to a clade encompassing Solenodonsaurus, diadectomorphs and crown-amniotes. Consequently, Anthracosaurus russelli Huxley, 1863 is neither an anthracosaur nor a tetrapod, whereas T. H. Huxley himself would be classified as a cotylosaurian anthracosaur. Such long-abandoned terms as Cotylosauria and Stegocephali, traditionally referring to archaic grade-groups, are now re-introduced with a novel content. Thus, Cotylosauria includes Anthracosauria minus Solenodonsaurus (i.e. diadectomorphs plus crown-amniotes), whereas Stegocephali includes all taxa with digits, i.e. tetrapods in the traditional sense (Goodrich, 1930; Gaffney, 1979; Coates, 1996; Ahlberg, 1998; Ahlberg & Clack, 1998; Anderson, 2001). This resurrectionist approach is currently the subject of intense debate [for conflicting views, see Benton (2000), Nixon & Carpenter (2000), Cantino (2000), Coates et al. (2000), Forey (2001), Brochu & Sumrall (2001), Bryant & Cantino (2002), and references therein]. 261 2001; Ruta et al., 2001). Ninety species are included in the present work (see Appendix 1 for a list of the specimens examined). OTUs are chosen according to three criteria: (1) sample of maximally diverse taxon exemplars (Nixon & Davis, 1991; Anderson, 2001; Prendini, 2001); (2) inclusion of the majority of taxa considered in previous studies; (3) use of species as terminals (e.g. Bininda-Emonds, Gittleman & Purvis, 1999; Anderson, 2001). Justification for the exclusion of some species is provided in the relevant taxonomic sections below. The plesiomorphic conditions of various tetrapod groups remain untested in several analyses. Therefore, large exemplars are used when hypotheses of relationships within a particular group are unavailable, or are based on a limited character/taxon sample. Diverse exemplars may also prevent spurious pairing of taxa resulting either from long branch attraction or from a host of convergent characters. For instance, a cluster of ‘absence’ features may discriminate against sister group relationships based on a smaller number of ‘good’ apomorphies. Finally, if members of a group display conflicting character distributions, exemplars should encompass such distributions. Few basal crown-lissamphibians and crownamniotes are considered here. Recent supraspecific OTUs are omitted, since large-scale interrelationships of primitive tetrapods are our main focus. Laurin & Reisz’s (1997, 1999) and Laurin’s (1998 a–c) analyses include several families from each of the three lissamphibian orders. However, comparisons between Palaeozoic and Recent faunas demand a proper evaluation of the primitive condition for several extant groups, and may be impractical, given the aims of this study. As an alternative approach, primitive members of various modern clades could be included to document crowntetrapod diversity in the Mesozoic and Caenozoic. The utility of such a comprehensive data set is nontheless dubious, since convergent features are likely to be widespread. Also, the size of the resulting matrix and the abundant missing entries may introduce severe computation problems. V. TAXON EXEMPLARS The limits and content of several early tetrapod groups are widely agreed upon and, with few exceptions (e.g. anthracosaurs, microsaurs, temnospondyls), their monophyly has not been disputed (Säve-Söderbergh, 1934; Carroll, 1970; Heaton, 1980; Smithson, 1985, 1986, 1994, 2000; Gauthier et al., 1988b; Clack, 1994c, 1998a, c; Smithson et al., 1994; Lombard & Bolt, 1995; Coates, 1996; Laurin & Reisz, 1997, 1999; Laurin, 1998a–c; Ahlberg & Clack, 1998; Paton et al., 1999; Berman, 2000; Bolt & Lombard, 2000; Anderson, (1) Devonian and Lower Carboniferous taxa (a) Taxonomic sample Acanthostega gunnari Jarvik, 1952. Ichthyostega stensioei Säve-Söderbergh, 1932. Tulerpeton curtum Lebedev, 1984. Ventastega curonica Ahlberg, Luksevics & Lebedev, 1994. Caerorhachis bairdi Holmes & Carroll, 1977. Crassigyrinus scoticus Watson, 1929. Eucritta melanolimnetes Clack, 1998 a. 262 Marcello Ruta, Michael I. Coates and Donald L. J. Quicke Westlothiana lizziae Smithson & Rolfe, 1990. Whatcheeria deltae Lombard & Bolt, 1995. (b) Remarks The four Devonian species considered here have been the subjects of intense recent scrutiny. Work on the first discovered Devonian tetrapod, Ichthyostega ( Jarvik, 1980, 1996; see notes and references therein for comments on species status), has been followed by detailed investigations of exceptionally well preserved and abundant material of Acanthostega, including descriptions of the snout and palate (Clack, 1994 a), braincase (Clack, 1994 a, 1998b), stapes (Clack, 1989, 1994 b), skeletal gill apparatus (Coates & Clack, 1991), limbs (Coates, 1991, 1995, 1996; Clack & Coates, 1995; Coates & Clack, 1990, 1995) and postcranium (Coates, 1996). The branching sequence of the majority of Devonian tetrapods is generally agreed upon [but see Lebedev & Coates (1995), Coates (1996), Ahlberg & Clack (1998), Laurin (1998a–c) and Laurin & Reisz (1997, 1999)]. The following taxa, known mainly from lower jaw rami and/or incomplete postcranial remains, are omitted: Metaxygnathus denticulus Campbell & Bell, 1977, Obruchevichthys gracilis Vorobyeva, 1977, Elginerpeton pancheni Ahlberg, 1995 [see also Ahlberg (1991, 1998)], Hynerpeton bassetti Daeschler, Shubin, Thomson & Amaral, 1994, Livoniana multidentata Ahlberg, Lukseviks & Mark-Kurik, 2000 and Densignathus rowei Daeschler, 2000. Crassigyrinus scoticus and Whatcheeria deltae are among the most problematic of early Carboniferous tetrapods [reviews in Clack & Carroll (2000) and Bolt & Lombard (2000)]. Traditionally, they have been regarded as having ‘reptiliomorph’ affinities (e.g. Panchen, 1973, 1985; Panchen & Smithson, 1987, 1988; Lebedev & Coates, 1995; Lombard & Bolt, 1995; Coates, 1996; Clack, 1998a, c, 2001; Paton et al., 1999; Bolt & Lombard, 2000), but evidence in support of their placement among basal stem-amniotes has been challenged repeatedly (e.g. Laurin & Reisz, 1997, 1999; Ahlberg & Clack, 1998; Laurin, 1998a–c; Clack & Carroll, 2000; Clack, 2002). Westlothiana from the uppermost Viséan of East Kirkton is usually considered to be one of the most primitive stem-amniotes (Smithson & Rolfe, 1990; Smithson et al., 1994). However, Laurin & Reisz (1999) placed Westlothiana as the closest outgroup to the tetrapod crown-clade. Caerorhachis, probably from the lowermost Serpukhovian of Scotland (Holmes & Carroll, 1977), was originally described as a basal temnospondyl [see Milner & Sequeira (1994) and Coates (1996) for an alternative view], but has been reinterpreted as a primitive ‘reptiliomorph’ by Ruta et al. (2001). More recently, Clack (2002) has placed this tetrapod as sister group to colosteids. Finally, Eucritta from East Kirkton displays a unique array of baphetid, temnospondyl and ‘anthracosaur’ features (Clack, 1998a, 2001) that account for the instability of the basal part of the tetrapod crown-group (but see comments in Thorley & Wilkinson, 1999). (2) Acherontiscidae (a) Taxonomic sample Acherontiscidae: Acherontiscus caledoniae Carroll, 1969 b. (b) Remarks Re-examination of the single known specimen of Acherontiscus (lowermost Serpukhovian of Scotland) indicates that it is probably an immature or pedomorphic adelospondyl (M. Ruta, personal observations). Vertebral construction is light; ribs are weakly ossified; skull roof, cheek bones, arrangement of circumnarial bones and mandibular shape are consistent with this interpretation. Discussion of the anatomy and relationships of this poorly known form will be presented in a future publication. (3) Adelospondyli (a) Taxonomic sample Adelogyrinidae: Adelospondylus watsoni Carroll, 1967; Adelogyrinus simorhynchus Watson, 1929; Dolichopareias disjectus Watson, 1929. (b) Remarks Adelospondyls, ranging from the upper Viséan to the lowermost Serpukhovian, have been reviewed by Andrews & Carroll (1991), Carroll et al. (1998) and Carroll (2000). Details of the skull roof are known in most species and, in the case of Adelospondylus, a partially preserved palate is also observed. Lower jaws and partial postcranial remains are associated with Adelogyrinus and Adelospondylus. A fourth species, Palaeomolgophis scoticus Brough & Brough, 1967, known from a postcranium and associated partial skull roof, incomplete palate and lower jaws, will be discussed in conjunction with the planned revision of Acherontiscus caledoniae. Adelospondyls display a highly specialized skull roof (e.g. reduction and/or loss of several bones; presence of a squamosotabular element), heavily ossified gill arches, and apparent absence of endochondral Early tetrapod relationships revisited shoulder girdle and limbs. Conversely, the dermal portion of the shoulder girdle is robust. Limb absence may well be a preservational artifact, especially because of the very few specimens known. No pelvic girdle has been observed. Some details of the snout and elongate skull roof and cheek bones resemble those in colosteids (Smithson, 1982; Schultze & Bolt, 1996; Panchen & Smithson, 1987). Similarities with colosteids are also evident in the general morphology and flange-like processes of the ribs. By contrast, the vertebrae are gastrocentrous, like those of microsaurs and lysorophids (see below). (4) Aı̈stopoda (a) Taxonomic sample Lethiscidae: Lethiscus stocki Wellstead, 1982. Ophiderpetontidae: Oestocephalus amphiuminum Cope, 1868. Phlegethontiidae: Phlegethontia linearis Cope, 1871. (b) Remarks Aı̈stopods, ranging from the mid Viséan to the upper part of the Lower Permian, are limbless, snake-like tetrapods characterized by a broad postorbital emargination of the cheek, covered by a sheet of integument with embedded ossicles. The suspensorial configuration led Lund (1978) to suggest the occurrence of a snake-like skull kinetism in Phlegethontia, but Anderson’s (in press) review of phlegethontiid crania indicates that this is incorrect, although limited kinesis near the snout tip may have occurred. The highly specialized nature of aı̈stopods poses problems for a correct assessment of their affinities [Carroll, 1998; see Anderson (2001) for ongoing anatomical and systematic revision of this clade]. Lethiscus is usually regarded as the most basal known aı̈stopod, based on its skull roof pattern (see also Milner, 1994). According to Anderson, Carroll & Rowe (2001), Lethiscus shows similarities with ophiderpetontids, which are paraphyletic relative to other aı̈stopods. Oestocephalus and Phlegethontia are the best known genera within ophiderpetontid and phlegethontiid aı̈stopods, respectively (review in Carroll et al., 1998). The cranial anatomy of Oestocephalus has been recently redescribed by Carroll (1998). McGinnis’ (1967) and Lund’s (1978) classical papers on Phlegethontia are now superseded by Anderson’s (in press) revision of this genus. Several cranial and postcranial characters of aı̈stopods (especially the morphology of the vertebrae in some taxa) indicate possible affinities with nectrideans, in agreement with 263 Thomson & Bossy’s (1970) Holospondyli (=aı̈stopods plus nectrideans) hypothesis. (5) Baphetidae (a) Taxonomic sample Baphetidae: Baphetes kirkbyi Watson, 1929; Megalocephalus pachycephalus (Barkas, 1873). (b) Remarks The interrelationships of baphetids, an uppermost Viséan to uppermost Moscovian group of tetrapods with keyhole-shaped orbits and a closed palate, remain unclear. Unequivocal association of cranial and postcranial material can be established only for Baphetes cf. kirkbyi (Milner & Lindsay, 1998). The lower jaw mesial surface is known in detail only in Megalocephalus (Beaumont, 1977; Ahlberg & Clack, 1998). The aberrant Spathicephalus Watson, 1929 (Beaumont & Smithson, 1998) is morphologically very divergent from remaining baphetids, although it is likely to be the sister taxon to these (Beaumont & Smithson, 1998). Loxomma Huxley, 1862 is in several respects intermediate morphologically between Baphetes and Megalocephalus, but its exclusion from the data set has no impact on the outcome of the analysis. Baphetids have been variously regarded as derived stem-tetrapods, basal stem-lissamphibians, or even basal ‘reptiliomorphs’. Evidence in support of each of these hypotheses is problematic (Beaumont, 1977; Panchen, 1980; Panchen & Smithson, 1987, 1988; Ahlberg & Milner, 1994; Carroll, 1995; Laurin & Reisz, 1997, 1999; Beaumont & Smithson, 1998; Clack, 1998a, c, 2001; Laurin, 1998 a–c; Milner & Lindsay, 1998). Various cranial and postcranial features (e.g. supratemporal–postparietal contact; fang pairs on palatal bones; shape of the humerus) indicate the primitive nature of this group, and contrast with such autapomorphic features as keyhole-shaped orbits, drop-shaped choanae and small temporal notches bordered anteriorly by the supratemporal (e.g. Clack, 1998 a, 2001). (6) Colosteidae (a) Taxonomic sample Colosteidae: Colosteus scutellatus (Newberry, 1856); Greererpeton burkemorani Romer, 1969. (b) Remarks Colosteids range from the upper Viséan to the uppermost Moscovian. Colosteus and Greererpeton are the best known members of the group (Smithson, 1982; Hook, 264 Marcello Ruta, Michael I. Coates and Donald L. J. Quicke 1983; Godfrey, 1989), but only the latter has been considered in most recent analyses of early tetrapod interrelationships. New data on the skull and lower jaw of Greererpeton (Bolt, 1996; Schultze & Bolt, 1996; Ahlberg & Clack, 1998; Bolt & Lombard, 2001) have confirmed previous hypotheses about the primitive status of the group [but see Hook (1983), Holmes (2000) and Carroll (2001) for alternative conclusions]. Plesiomorphic characters (e.g. trunk elongation; rhachitomous vertebrae; L-shaped humerus; anteriorly expanded clavicles) are associated with a unique pattern of circumnarial bones and a notch in the anterior extremity of the dentary. Coates (1996) demonstrated the presence of a five-digited manus in Greererpeton. The occurrence of a four-digited manus in Colosteus is based on data from Hook (1983), although the limb material of this genus is less well preserved than that of Greererpeton. Superficially, colosteids resemble certain temnospondyls [e.g. eobrachyopids (=saurerpetontids)], but similarities between the two groups are usually assumed to be convergent. Various skull roof features (e.g. elongate frontals and parietals) resemble those of certain lepospondyls (e.g. adelospondyls, nectrideans) and may indicate a close relationship (e.g. Panchen & Smithson, 1987; Milner, 1993; see also discussion of cranial data analysis below). detailed redescription of Tseajaia campi Vaughn, 1964, this taxon is excluded from the data matrix (Moss, 1972; Walliser, 1998, 1999). The arrangement of bones in the posterior and posterolateral portions of the skull table of diadectomorphs is reminiscent of those in several primitive crown-amniotes (e.g. Berman et al., 1998; Berman, 2000). This is especially evident in the enlargement of the parietals and in the expansion of the supraoccipital. A series of recent papers (Sumida & Lombard, 1991; Berman et al., 1992, 1998; Sumida et al., 1992; Berman, 2000) have clarified several poorly understood aspects of diadectomorph osteology, but the origin and diversification of this group remain problematic. The Pennsylvanian Solenodonsaurus janenschi Broili, 1924 has been generally allied to ‘reptiliomorphs’ (Pearson, 1924; Brough & Brough, 1967; Carroll, 1970; Gauthier et al., 1988b; Laurin & Reisz, 1999), based on its skull table morphology, gastrocentrous vertebrae and curved ribs. It is included in the present work because of its combination of features found in different ‘reptiliomorphs’, such as gephyrostegids, seymouriamorphs, diadectomorphs and basal crownamniotes. According to Laurin & Reisz (1999), Solenodonsaurus is the sister taxon to a clade encompassing diadectomorphs and crown-amniotes (but see Lee & Spencer, 1997). (7) Diadectomorpha (a) Taxonomic sample Diadectidae: Diadectes absitus Berman, Sumida & Martens, 1998. Limnoscelidae: Limnoscelis paludis Williston, 1911. (b) Remarks Diadectomorphs range from the upper Bashkirian to the upper part of the Lower Permian, and are usually regarded as the closest relatives of crown-amniotes (e.g. Lee & Spencer, 1997), based on cranial and postcranial characters (Laurin & Reisz, 1997, 1999; Laurin, 1998a–c), morphology of the occiput (Berman, 2000), and atlas–axis complex (Sumida & Lombard, 1991; Berman et al., 1992; Sumida et al., 1992). According to Berman et al. (1992) and Berman (2000), the highly autapomorphic nature of Diadectes makes this taxon unsuitable for polarising characters at the base of the amniote tree. For this reason, and in agreement with previous studies (e.g. Laurin & Reisz, 1997, 1999; Laurin, 1998a–c), a second diadectomorph – Limnoscelis paludis – is included in this work (Williston, 1912; Romer, 1946; Fracasso, 1987; Berman & Sumida, 1990; Berman et al., 1992; Berman, 2000). Pending a (8) Embolomeri and Eoherpetontidae (a) Taxonomic sample Anthracosauridae: Anthracosaurus russelli Huxley, 1863. Archeriidae: Archeria crassidisca (Cope, 1884). Eogyrinidae: Pholiderpeton attheyi (Watson, 1926); Pholiderpeton scutigerum Huxley, 1869. Eoherpetontidae: Eoherpeton watsoni Panchen, 1975. Proterogyrinidae: Proterogyrinus scheelei Romer, 1970. (b) Remarks Embolomeres and eoherpetontids [uppermost Viséan to lowermost Upper Permian; review in Panchen (1980)] include some of the best known Coal Measures ‘anthracosaurs’ (sensu Smithson, 1985, 1986, and Panchen & Smithson, 1987, 1988; see also Clack, 1994 c). Several authors interpret ‘anthracosaurs’ as a basal radiation of aquatic or semiaquatic, long-bodied and amniote-like taxa (e.g. Panchen & Smithson, 1988; Coates, 1996; Lee & Spencer, 1997; Clack, 1998a, c, 2001; Paton et al., 1999). However, their phylogenetic position relative to amniotes has been questioned (Laurin & Reisz, 1997, 1999; Laurin, 1998 a–c), and the possibility that they fall outside the tetrapod Early tetrapod relationships revisited crown-group cannot be ruled out (Dr J. A. Clack, personal communication to M. Ruta, 2001; see also Clack, 2002). Although embolomere anatomy is known in great detail (e.g. Romer, 1957; Panchen 1964, 1970, 1972, 1973, 1977, 1980; Holmes, 1980, 1984, 1989; Smithson, 1985; Clack, 1987 a, b; Clack & Holmes, 1988), surveys of character distribution have not resulted in a consensus over their intrinsic relationships (e.g. Holmes, 1984, 1989; Smithson, 1985; Clack, 1987a). Silvanerpeton miripedes Clack, 1994c and Eldeceeon rolfei Smithson, 1994, both from the uppermost Viséan site of east Kirkton, are two of the earliest known ‘anthracosauroids’. Several postcranial features (e.g. U-shaped intercentra and pleurocentra; low neural spines; small tabular horns) suggest that they are less derived than embolomeres. (9) Gephyrostegidae (a) Taxonomic sample Gephyrostegidae: Bruktererpeton fiebigi Boy & Bandel, 1973; Gephyostegus bohemicus Jaekel, 1902. (b) Remarks The monophyly of gephyrostegids (lower Bashkirian to uppermost Moscovian) is supported by some recent studies (Paton et al., 1999), and implicitly assumed in others (e.g. Laurin & Reisz, 1997, 1999; Laurin, 1998a–c). Gephyrostegids have long been considered to share characteristics with ‘anthracosaurs’ and higher ‘reptiliomorphs’ (e.g. Carroll, 1970, 1986, 1991b; Heaton, 1980; Smithson, 1985). Several cranial features resemble conditions in embolomeres, seymouriamorphs and various basal crown-amniotes (e.g. protorothyridids and captorhinids). Similarities with primitive amniotes are also evident in palatal bone proportions and in the morphology of the parasphenoid (e.g. Lee & Spencer, 1997). However, they lack a toothed transverse pterygoid flange, long regarded as a key amniote apomorphy (discussions in Carroll, 1970, 1991 b). Likewise, their lower jaws (Ahlberg & Clack, 1998) include a mixture of features otherwise found in embolomeres and basal amniotes. Finally, the postcranium combines primitive features (e.g. U-shaped intercentra and pleurocentra) with several derived ones (e.g. reduced dorsal iliac blade; scapulocoracoid extending posteroventrally with respect to the posterior glenoid margin; L-shaped tarsal intermedium). Although morphological evidence is not strong (e.g. Carroll, 1991 b), gephyrostegids may lie closer to early amniotes than embolomeres (see also Lee & Spencer, 1997). 265 (10) Lysorophia (a) Taxonomic sample Cocytinidae: Brachydectes elongatus Wellstead, 1991; Brachydectes newberryi Cope, 1868. (b) Remarks Lysorophids (upper Bashkirian to upper part of Lower Permian) are among the most enigmatic of all lepospondyls. The most recent account of the group is by Wellstead (1991). Their highly specialized and elongate skulls are characterized by a bar-like, anteroventrally sloping suspensorium and by a large fenestration in the cheek region that becomes confluent with the orbit anteriorly (Bolt & Wassersug, 1975). Other noteworthy features are the extreme reduction and poor ossification of limbs and girdles, the presence of vertebral keels, the occurrence of sutures between neural arches and vertebral bodies and the extreme elongation of the trunk region. As in the case of microsaurs, the occiput of lysorophids is strap-shaped. In addition, lysorophids share various cranial and mandibular characters with one or more microsaur families. For example, the configuration of the mandible and the shape and proportions of the premaxillae are reminiscent of those of brachystelechids (see also Wellstead, 1991). Laurin & Reisz (1997, 1999) and Laurin (1998a–c) place lysorophids as the nearest relatives of crown-lissamphibians. However, most of the characters supporting this position appear to be secondary losses (e.g. those related to certain cranial and palatal bones; see also Carroll & Bolt, 2001). Furthermore, some supposed synapomorphies are dubious. An example is represented by the occurrence of a cheek emargination, which is only superficially similar to that of certain primitive crown-lissamphibians and dissorophoids (Carroll, 2001). In these groups, the maxillary arcade is often incomplete posteriorly and the palatal bones are reduced to slender rods or struts. Conversely, the maxillary arcade of lysorophids is extensively sutured medially with broad palatal bones. Anderson (2001) considers lysorophids to be allied to aı̈stopods – both groups being nested within nectrideans (Fig. 3). Brachydectes is here treated as a composite genus, with anatomical information based on both B. elongatus and B. newberryi (see also Anderson, 2001). (11) Microsauria (a) Taxonomic sample Brachystelechidae: Batropetes Deichmüller, 1882). fritschia (Geinitz & 266 Marcello Ruta, Michael I. Coates and Donald L. J. Quicke Acanthostega gunnari Greererpeton burkemorani Balanerpeton woodi Dendrerpeton acadianum Proterogyrinus scheelei Seymouria baylorensis Limnoscelis paludis Utaherpeton franklini Hyloplesion longicostatum Odonterpeton triangulare Saxonerpeton geinitzi Hapsidopareion lepton Llistrofus pricei Asaphestera intermedia Tuditanus punctulatus Micraroter erythrogeios Pelodosotis elongatum Pantylus cordatus Stegotretus agyrus Sparodus validens Cardiocephalus sternbergi Cardiocephalus peabodyi Euryodus dalyae Euryodus primus Rhynchonkos stovalli Eocaecilia micropoda Batropetes fritschia Carrolla craddocki Quasicaecilia texana Mazon Creek microsaur Microbrachis pelikani Adelogyrinus simorhynchus Scincosaurus crassus Sauropleura scalaris Sauropleura bairdi Sauropleura pectinata Urocordylus wandesfordii Ctenerpeton alveolatum Ptyonius marshii Batrachiderpeton reticulatum Keraterpeton galvani Diceratosaurus brevirostris Diploceraspis burkei Diplocaulus magnicornis Diplocaulus primus Brachydectes elongatus / newberryi Oestocephalus granulosum Phlegethontia longissima COL. Tuditanidae: Asaphestera intermedia (Dawson, 1894); Tuditanus punctulatus Cope, 1874. TEM. ANT. SEY. DIA. MIC. LIS. MIC. ADE. NEC. LYS. AIS. Fig. 3. Anderson’s (2001) cladogram of lepospondyls. Abbreviations as in Fig. 1. Hapsidopareiontidae: Hapsidopareion lepton Daly, 1973; Saxonerpeton geinitzi (Credner, 1890). Goniorhynchidae: Rhynchonkos stovalli (Olson, 1970). Gymnarthridae: Cardiocephalus sternbergi Broili, 1904 a; Euryodus primus Olson, 1939. Hyloplesiontidae: Hyloplesion longicostatum Fritsch, 1876. Microbrachidae: Microbrachis pelikani Fritsch, 1876. Odonterpetontidae: Odonterpeton triangulare Moodie, 1909. Ostodolepidae: specimen BPI 3839 [possibly belonging to Micraroter erythrogeios Daly, 1973; see Carroll & Gaskill (1978)]; Pelodosotis elongatum Carroll & Gaskill, 1978. Pantylidae: Pantylus cordatus Cope, 1871; Stegotretus agyrus Berman, Eberth & Brinkman, 1988. (b) Remarks Microsaurs range from the uppermost Serpukhovian/ lowermost Bashkirian to the upper part of the Lower Permian and are the most diverse of all lepospondyl groups. Carroll & Gaskill (1978, and references therein) monographed the entire group. Gregory, Peabody & Price (1956) is a standard reference for gymnarthrids. Fifteen of the 21 genera examined by Anderson (2001; Fig. 3) are included in the present work. The following taxa, however, are omitted: (1) the brachystelechids Carrolla Langston & Olson, 1986 and Quasicaecilia Carroll, 1990; (2) one species from each of the two genera Cardiocephalus and Euryodus (Gymnarthridae); (3) the hapsidopareiontid Llistrofus Carroll & Gaskill, 1978; (4) the pantylid (fide Anderson, 2001) Sparodus Fritsch, 1876; (5) Utaherpeton Carroll, Bybee & Tidwell, 1991; and (6) an unnamed microsaur from Mazon Creek. A recently described, unnamed microsaur from Goreville, Illinois (Lombard & Bolt, 1999) is also excluded because of incomplete preservation. Differences in the arrangement of skull roof bones are the most distinctive features of microbrachomorph and tuditanomorph microsaurs (Carroll & Gaskill, 1978). Few characters are shared by these two groups, the most important of which is the strap-shaped configuration of the exoccipitals and basioccipital, which is observed also in lysorophids (see above). Microbrachomorphs and tuditanomorphs possess a single bone in the position usually occupied by the intertemporal, supratemporal and tabular in several early tetrapods. In agreement with previous works (e.g. Carroll & Gaskill, 1978), the bone in question is interpreted as a tabular, based upon topological similarity. We note that this conjectural homology assessment affects further character scores that depend upon the morphology and spatial relationships of the tabular. The extrinsic relationships of microsaurs are debated [see Milner (1993) for a summary]. Panchen & Smithson (1988) and Milner (1993) regard the presence of a ‘waisted’, propellor-blade like humerus as a synapomorphy of microsaurs and temnospondyls. However, humerus shape varies considerably both among temnospondyls and, to a lesser extent, among microsaurs. The hypothesis that caecilians evolved from longbodied, presumably burrowing lepospondyls (Carroll & Currie, 1975; Carroll, 2000) has received support in certain recent analyses (e.g. Laurin & Reisz, 1997, 1999; Laurin, 1998a–c; Anderson, 2001). In particular, Early tetrapod relationships revisited Anderson’s (2001) study points to brachystelechids as the nearest relatives of caecilians. Few phylogenetic analyses of microsaurs are available in the literature. Gymnarthrid interrelationships and a family-level analysis of tuditanomorphs are detailed by Schultze & Foreman (1981; see also Milner, 1993), whereas pantylids are discussed by Berman et al. (1988). Laurin & Reisz (1997, 1999) and Laurin (1998a–c) dispute the monophyletic status of microsaurs. In their analyses, Pantylus, Rhynchonkos and Brachystelechidae are progressively more closely related to lysorophids plus crown-lissamphibians, implying ipso facto the paraphyletic status of tuditanomorphs. However, their taxon sample does not represent adequately microsaur diversity, and is biased towards inclusion of taxa with presumed gymnophionan (Rhynchonkos) or generalized lissamphibian similarities (Brachystelechidae) (Carroll & Currie, 1975; Carroll, 2000; Anderson, 2001). In Anderson’s (2001) cladogram, tuditanomorph microsaurs include, in proximo-distal sequence, hapsidopareiontids (monophyletic), tuditanids (monophyletic), ostodolepids (monophyletic), and a clade of pantylids plus gymnarthrids placed as sister group to a clade of rhynchonkids, brachystelechids and Eocaecilia. In the same study, microbrachomorphs are distributed as follows: (1) Utaherpeton and Hyloplesion are basal to all remaining lepospondyls; (2) Odonterpeton is basal to hapsidopareiontids and more derived tuditanomorphs; (3) Microbrachis and the unnamed microsaur from Mazon Creek form the monophyletic sister group to all other (non-microsaur) lepospondyls. (12) Nectridea (a) Taxonomic sample Diplocaulidae: Batrachiderpeton reticulatum (Hancock & Atthey, 1869); Diceratosaurus brevirostris (Cope, 1875); Diplocaulus magnicornis Cope, 1882; Diploceraspis burkei Romer, 1952; Keraterpeton galvani Wright & Huxley, 1866. Scincosauridae: Scincosaurus crassus Fritsch, 1876. Urocordylidae: Ptyonius marshii Cope, 1875; Sauropleura Cope, 1868 [treated as a composite genus with data from two species, S. pectinata Cope, 1868 and S. scalaris (Fritsch, 1883)]; Urocordylus wandesfordii Wright & Huxley, 1866. (b) Remarks Nectrideans are known from the upper Bashkirian to the lowermost Upper Permian, and are usually divided 267 into urocordylids, scincosaurids and diplocaulids. The present work encompasses 75% of the taxa examined by Anderson (2001; Fig. 3). The genus Ctenerpeton Cope, 1897 and some species of Diplocaulus and Sauropleura are excluded. Diagnostic characters of nectrideans are observed almost exclusively in the postcranial skeleton (A. C. Milner, 1980) and relate to vertebral morphology (e.g. configuration of neural and haemal arches; extra-articulations above zygapophyses). Some of these characters are shared with at least some aı̈stopods (Bossy & Milner, 1998; Anderson, in press). A striking aspect of the anatomy of all nectrideans is the extreme elongation of the tail. Hardly any character of the skull roof and palate can be identified as a shared derived feature of the three families. However, there is agreement on the derived status of diplocaulids relative to other nectrideans. Various analyses (A. C. Milner, 1980; Milner, 1993; Bossy & Milner, 1998) place scincosaurids as the sister taxon to diplocaulids, based upon such unique features as the quadratebracing internal shelf of the squamosal. According to Panchen & Smithson (1988), nectrideans are just crownward of ichthyostegids on the lissamphibian stem, based largely on the presence of a four-digited manus. Milner’s (1993) scheme of relationships agrees mostly with that of Panchen & Smithson (1988), except that nectrideans and colosteids appear as sister taxa. Three characters of the skull table are used by Milner (1993) to unite nectrideans with colosteids. Two of these – skull table elongation; broad postorbital-parietal contact – are also present (each one separately or both together) in other lepospondyl taxa (e.g. Acherontiscus; adelospondyls), and their conditions reverse within derived nectrideans. The third character (prefrontal bordering external naris and excluding nasal from naris margin) is problematic. Although the prefrontal enters the nostril in many nectrideans, the nasal does contribute to the nostril in several genera. Nectrideans are similar in several respects to aı̈stopods (e.g. Thomson & Bossy, 1970; Anderson, in press), but no support for a nectridean-aı̈stopod clade has been found in recent studies (Carroll, 1995; Laurin & Reisz, 1997, 1999; Laurin, 1998 a–c; Anderson, 2001). In Anderson’s (2001) cladogram, aı̈stopods and lysorophids form the sister group to diplocaulids, with urocordylids and scincosaurids as progressively more outlying clades. These results are in agreement with the observation that hardly any cranial feature of nectrideans is uniquely shared by all members of this group (Beerbower, 1963; A. C. Milner, 1980; Milner, 1993; Bossy & Milner, 1998), and that similarities with representatives of other clades are widespread (Anderson, in press). 268 Marcello Ruta, Michael I. Coates and Donald L. J. Quicke (13) Seymouriamorpha (a) Taxonomic sample Discosauriscidae: Discosauriscus austriacus (Makowsky, 1876). Kotlassiidae: Kotlassia prima Amalitsky, 1921. Seymouriidae: Seymouria baylorensis Broili, 1904b; S. sanjuanensis Vaughn, 1966. (b) Remarks The phylogenetic position of seymouriamorphs (Pennsylvanian/Permian boundary to late Upper Permian) has been debated for almost a century. Characters of the lower jaw (e.g. rearward extension of splenial; anterior mandibular foramen), skull roof (e.g. suture between parietal and tabular), palate (e.g. transverse pterygoid flange bearing no teeth) and vertebral column (e.g. gastrocentrous vertebrae with cylindrical pleurocentra; swollen neural arches) indicate possible amniote affinities (Gauthier et al., 1988 a, b; Sumida & Lombard, 1991; Sumida et al., 1992; Klembara, 1997; Lee & Spencer, 1997; Sumida, 1997; Ahlberg & Clack, 1998; Berman et al., 2000; Klembara & Bartı́k, 2000, and references therein). However, the construction of the palate and occiput reveal several primitive traits (White, 1939; Laurin, 1995, 1996 b; Berman, 2000). Soft tissue and osteological markers indicate the presence of gill filaments and electroreceptors in certain forms (Ivakhnenko, 1981; Kuznetsov & Ivakhnenko, 1981; Klembara, 1994, 1995). Various features of the limbs (especially the femur) and pelvic girdle (e.g. shape of the ilium and development of iliac shelf ) resemble closely those of diadectomorphs. Among the characters supporting seymouriamorph monophyly are a broad, transverse lamina ascendens of the pterygoid (but this character may be more widely distributed among early tetrapods), small posttemporal fenestrae, otic tubes and slender stapes [summary in Laurin (1998 b, 2000); see also Klembara (1997) and Klembara & Bartı́k (2000)]. In the present work, Seymouria is treated as a composite genus. Anatomical information is based on the works of White (1939), Berman & Martens (1993), Laurin (1995, 1996 b), Berman et al. (2000) and Klembara, Martens & Bartı́k (2001). Seymouria baylorensis and S. sanjuanensis differ in details of the skull roof and postcranium, but a comparative study of these species must await an exhaustive redecription of S. sanjuanensis (Dr J. Klembara, personal communication to M. Ruta, 2001; but see also Berman et al., 2000). The cranial and postcranial anatomy of Discosauriscus have been thoroughly restudied by Klembara (1997, and references therein) and Klembara & Bartı́k (2000, and references therein). Kotlassia has been re-examined by Bystrow (1944). Unfortunately, several osteological details of the latter form are very poorly known and need adequate redescription. Pending a reassessment of several eastern European and western Asian seymouriamorphs [see Ivakhnenko (1981), Kuznetsov & Ivakhnenko (1981), Zhang, Li & Wan (1984), Laurin (1996 a, c, 1998a–c) and Laurin & Reisz (1997, 1999)], these are not considered further here, but the reader should refer to Ivakhnenko (1987) and Novikov, Shishkin & Golubev (2000) for reviews. (14) Temnospondyli (a) Taxonomic sample Amphibamidae: Amphibamus grandiceps Cope, 1865; Doleserpeton annectens Bolt, 1969; Eoscopus lockardi Daly, 1994; Platyrhinops lyelli (Wyman, 1858). Branchiosauridae: Apateon pedestris Meyer, 1844; Leptorophus tener (Schönfeld, 1911); Schoenfelderpeton prescheri Boy, 1986. Cochleosauridae: Chenoprosopus lewisi Hook, 1993; Cochleosaurus florensis Rieppel, 1980. Dendrerpetontidae: Dendrerpeton acadianum Owen, 1853. Dissorophidae: Broiliellus brevis Carroll, 1964; Ecolsonia cutlerensis Vaughn, 1969. Edopidae: Edops craigi Romer, 1935. Eobrachyopidae: Isodectes obtusus (Cope, 1868). Eryopidae: Eryops megacephalus Cope, 1877. Micromelerpetontidae: Micromelerpeton credneri Bulman and Whittard, 1926. Trematopidae: Acheloma cumminsi Cope, 1882; Phonerpeton pricei (Olson, 1941). Trimerorhachidae: Neldasaurus wrightae Chase, 1965; Trimerorhachis cfr. insignis Case, 1935. Family incertae sedis: Balanerpeton woodi Milner & Sequeira, 1994. (b) Remarks The nature and status of temnospondyls (uppermost Viséan to Albian), the most abundant and diverse of all groups of early tetrapods, are intensely debated. Temnospondyls have long played a pivotal role in our understanding of lissamphibian origins (e.g. Bolt, 1969, 1977, 1979, 1991; Milner, 1988, 1990, 1993, 2000; Trueb & Cloutier, 1991; Rocek & Rage, 2000 a; Gardner, 2001). However, some recent analyses have questioned their lissamphibian affinities (Laurin & Reisz, 1997, 1999; Laurin, 1998 a–c). The most distinctive character of temnospondyls is the occurrence of interpterygoid vacuities at least half as wide as the Early tetrapod relationships revisited skull and bordered by triradiate pterygoids (Milner, 1988, 1990, 1993; Milner & Sequeira, 1994; Holmes, 2000). Although palatal vacuities are known in several other groups, those present in the vast majority of temnospondyls have a strongly concave perimeter, including the anteriormost extremity (Edops being one notable exception). These features occur in at least one microsaur and in some nectrideans (Carroll & Gaskill, 1978; A. C. Milner, 1980; Milner, 1993; Bossy & Milner, 1998; Ruta et al., 2001). However, Anderson (2001) has demonstrated recently that regressions of estimated areas of the interpterygoid vacuities over skull lengths in several temnospondyls and microsaurs are significantly different, and that the vacuities of microsaurs are absolutely smaller than those of temnospondyls. Cochleosaurid edopoids possess plectrum- or teardrop-shaped vacuities in the posterior half of the palate, somewhat intermediate between those of Edops and higher temnospondyls (Milner & Sequeira, 1998). We follow Milner & Sequeira (1994, 1998) in considering edopoids as the most basal temnospondyl clade. The large-scale interrelationships of postedopoid temnospondyls remain poorly understood, despite much recent progress (e.g. Schoch & Milner, 2000; Yates & Warren, 2000; Damiani, 2001). In Milner’s (1990) phylogenetic scheme, post-edopoid temnospondyls are divided, in order of increasing affinities with crown-lissamphibians, into a trimerorhachoid complex, a stereospondyl complex (a derived and diverse clade within archegosauroids), an eryopoid complex and a dissorophoid complex (Holmes, 2000; Schoch & Milner, 2000). Eryopoids are variously regarded as a paraphyletic group relative to stereospondyls, or to stereospondyls plus dissorophoids (Milner, 1990; Milner & Sequeira, 1998). Yates & Warren (2000) group trimerorhachoids (their Dvinosauria) with stereospondyls, and place this broader clade (termed the Limnarchia) as sister taxon to an eryopoid– dissorophoid clade (their Euskelia). An exhaustive treatment of temnospondyls is beyond the scope of the present work. Certain studies have explored the interrelationships of several temnospondyl subgroups (e.g. Schoch & Milner, 2000; Yates & Warren, 2000; Damiani, 2001; Steyer, 2002), although a large-scale computerized phylogeny of the whole clade has not been attempted (but see Milner, 1990). For this reason, we use mostly those genera that have been included in previous small-scale analyses (e.g. Berman, Reisz & Eberth, 1985; Dilkes, 1990; Trueb & Cloutier, 1991; Daly, 1994; Milner & Sequeira, 1994, 1998; Godfrey & Holmes, 1995; Godfrey, Fiorillo & Carroll, 1987; Holmes, Carroll & Reisz, 1998; Laurin, 1998a–c; Holmes, 2000). 269 Taxon sample is necessarily limited. It does, however, encompass members of most major temnospondyl groups. The content and limit of some of these groups are still poorly understood. Eryops appears to be a generalized eryopoid (Holmes, 2000), and is undoubtedly one of the best known Palaeozoic tetrapods (Romer, 1922, 1947; Miner, 1925; Sawin, 1941; Moulton, 1974). Trimerorhachoids are the subject of ongoing investigation (see Sequeira, 1998). Finally, dissorophoid interrelationships are still in a state of flux, despite the amount of morphological information available for several families (e.g. Watson, 1940; Carroll, 1964; Boy, 1972, 1974, 1978, 1985, 1986, 1987, 1995; Bolt, 1979, 1991; Milner, 1988, 1990, 1993; Dilkes, 1990; Trueb & Cloutier, 1991; Schoch, 1992; Daly, 1994; Boy & Sues, 2000, and references therein). One of the dissorophoids examined by Laurin & Reisz (1997, 1999) and Laurin (1998a–c) – the genus Tersomius Case, 1910 – is excluded from the present study. Specimens attributed to Tersomius consist of assorted skulls some of which probably belong to immature dissorophids and to various amphibamids (Bolt, 1977; Dr A. R. Milner, personal communication to M. Ruta, 2001). (15) Crown-group Lissamphibia Numerous recent discoveries (e.g. Jenkins & Walsh, 1993; Shubin & Jenkins, 1995; Jenkins & Shubin, 1998; Evans & Sigogneau-Russell, 2001; Gao & Shubin, 2001) add to our knowledge of primitive lissamphibian diversity [reviews in Báez & Basso (1996), Carroll (2000), Milner (2000) and Rocek (2000), and references therein]. Small-scale analyses of early salientians, caudates and gymnophionans (e.g. Báez & Basso, 1996; Evans & Sigogneau-Russell, 2001; Gao & Shubin, 2001; Gao & Wang, 2001) provide a framework for the choice of exemplars. Only certain fossil representatives of the three modern lissamphibian clades are considered here: the Early Jurassic stemgymnophionan Eocaecilia micropoda Jenkins & Walsh, 1993 (review in Carroll, 2000); the Early Triassic stemsalientian Triadobatrachus massinoti (Piveteau, 1936) (Watson, 1940; Hecht, 1960, 1962; Kuhn, 1962; Estes & Reig, 1973; Rage & Rocek, 1986, 1989; Rocek & Rage, 2000b); the Late Jurassic stem-caudate Karaurus sharovi Ivakhnenko, 1978 (review in Milner, 2000). The Early Cretaceous caudate Valdotriton gracilis Evans & Milner, 1996, is also included in the analysis. Additional primitive lissamphibians will be considered in an expanded version of our data set. Published schemes of character distribution [summaries in Estes (1981) and Milner (1988)] support the basal position of 270 Marcello Ruta, Michael I. Coates and Donald L. J. Quicke Triadobatrachus, Karaurus and Eocaecilia relative to other fossil salientians, caudates and gymnophionans, respectively. The family Albanerpetontidae, briefly reviewed by Milner (2000), consists of problematic salamander-like forms variously regarded as caudate relatives or as derived stem-lissamphibians (see also McGowan & Evans, 1995). In a recent paper, Gardner (2001, and references therein) placed albanerpetontids on the stem of the caudate-salientian group [but see also Trueb & Cloutier (1991) and Milner (2000)]. Coding for albanerpetontids is based on information from two of the best preserved species, Albanerpeton inexpectatum Estes & Hoffstetter, 1976 (Gardner, 1999) and Celtedens ibericus McGowan & Evans, 1995. Pending a re-assessment of such problematic groups as batrachosauroidids and scapherpetontids (review in Milner, 2000), these are excluded from the present study. (16) Crown-group Amniota Only three stem-diapsid taxa are considered – the captorhinid Captorhinus aguti Cope, 1882 (Fox & Bowman, 1966; Modesto, 1998); the protorothyridid Paleothyris acadiana Carroll, 1969a (Carroll, 1970, 1991b; Clark & Carroll, 1973); the araeoscelidian Petrolacosaurus kansensis Lane, 1945 (Peabody, 1952; Reisz, 1977, 1981). These appear to be generalized basal crown-amniotes, and their anatomy is known in sufficient detail (Carroll & Baird, 1972). Additional species from the basal portions of the synapsid and reptile branches of the amniote crown-group will be considered in an expanded version of our data set (see also Gauthier et al., 1988b). The diversification of primitive amniotes has been the subject of intense revision over the last ten years (e.g. Carroll & Currie, 1991). The fundamental split of amniotes into Synapsida (mammals and their extinct relatives) and Reptilia (turtles, lizards, snakes, birds, crocodiles and their extinct relatives) is widely accepted (Reisz, 1986; Hopson, 1991). However, the branching pattern in the basal part of crown-Reptilia has not reached a satisfactory consensus. The core of the problem revolves around the position of turtles and the placement of several Permo-Carboniferous and Triassic groups (e.g. Ivakhnenko, 1987; Gauthier et al., 1988a, b; Reisz & Laurin, 1991; Lee, 1993, 1995, 1996; Laurin & Reisz, 1995; Rieppel & deBraga, 1996; deBraga & Rieppel, 1997; Rieppel & Reisz, 1999). Paton et al. (1999) interpreted the Scottish upper Viséan tetrapod Casineria kiddi as the earliest known amniote, and placed it in a polytomy with Westlothiana, Captorhinus, Petrolacosaurus and Paleothyris. In Paton et al.’s (1999: p. 512) words, although the results of their phylogenetic analysis ‘… are not very robust, [they] nonetheless appear to place Casineria not only on the amniote stem but also among the true amniotes of the Late Carboniferous … It could thus be an amniote, predating not only the earliest true amniotes from the Westphalian, but also the earliest previously known stem-amniote, Westlothiana, from East Kirkton’. (17) Outgroups Recent comprehensive analyses (Cloutier & Ahlberg, 1996; Ahlberg & Johanson, 1998; Zhu & Schultze, 2001; Johanson & Ahlberg, 2001) have repeatedly and consistently found panderichthyids and tristichopterids to be successively more outlying sister groups to the limbed tetrapods. In agreement with these studies, and contra Rosen et al.’s (1981) hypothesis (for a detailed and comprehensive analytical criticism, see Panchen & Smithson, 1987), the tristichopterid Eusthenopteron foordi Whiteaves, 1881 (Andrews & Westoll, 1970; Jarvik, 1980, and references therein) and the panderichthyid Panderichthys rhombolepis (Gross, 1930) (Vorobyeva, 1977, 1992, 2000; Vorobyeva & Schultze, 1991; Ahlberg, Clack & Luksevics, 1996; Ahlberg & Clack, 1998, and references therein) are used to polarize characters. VI. CHARACTERS We are in the process of compiling a new, expanded matrix for early tetrapods based upon the data set presented here, and including a detailed character discussion. To aid cross-reference between elements of the present and future matrices, each character, as stated in Appendix 2, is preceded by a bold number identifying its position in the current data matrix (see Appendix 3), and by an italicized, abbreviated name and number for the osteological feature to which it refers (this second number will remain in future versions). A key feature of subsequent data sets will be to provide detailed treatments of each aspect of the anatomy of primitive tetrapods. Work in this direction has already begun (e.g. Lombard & Bolt, 1999; Bolt & Chatterjee, 2000; Bolt & Lombard, 2001). VII. ANALYSIS (1) Character coding The theoretical and practical problems associated with different regimes of character coding are intensely Early tetrapod relationships revisited debated topics (e.g. Scotland & Pennington, 2000). In the present work, most characters are binary and refer to the presence or absence of a structure (or condition of a structure). Multistate characters are coded as unordered (non-additive) in all analyses. All characters are equally weighted and optimized using ACCTRAN. A discussion of the results implied by different coding methods (cf. Pleijel, 1995) is outside the aims of this work, and will be detailed elsewhere. The data matrix includes unknown scores for inapplicable characters. In this respect, the coding regime is similar to Forey & Kitching’s (2000) contingent method. Optimization of state changes (available upon request from the authors) often leads to undesirable results, as in the case of inapplicable scores. For example, an unknown score for the condition of a certain bone (e.g. suture pattern between intertemporal and cheek region in a taxon that lacks an intertemporal; e.g. Solenodonsaurus) may be fully optimized on a branch that subtends a taxon in which the bone in question is absent. (2) The parsimony ‘ratchet’ A data matrix consisting of 90 taxa coded for 319 osteological characters (224 cranial and 95 postcranial) was built in MacClade 3.0.5 (Maddison & Maddison, 1992), which was also used to manipulate trees in experiments of taxon pruning and regrafting, in the analysis of suboptimal cladograms, and in comparisons between conflicting positions for various taxa. Cladistic analyses were performed on a PowerMac G4 computer using PAUP* 4.0b10 (Swofford, 1998). Because PAUP* only supports MINSTEPS, tree lengths reported treat all polytomies as soft. The widespread occurrences of missing entries and the moderately large size of the complete data matrix made it likely a priori that finding the optimum tree(s) under parsimony optimality criteria would be difficult. Therefore, we employed a range of tree searching strategies to maximize our chances of finding optimal islands. In order to cover a wide range of tree space in a practical length of time (days) we carried out 40000 random stepwise additions followed by TBR (tree bisection-reconnection) branch-swapping searching, but holding only one tree in memory at any one time (i.e. MAXTREES=1) (Quicke et al., 2001). These searches hit trees of the shortest length recovered over 150 times. Searching on each tree with unlimited MAXTREES recovered the same island of trees. No shorter trees were recovered by employing the iterative re-weighting strategy proposed by Quicke et al. (2001). Searching on subsets of characters (e.g. see below for a discussion of cranial character analysis, removal 271 of lower jaw characters and reverse weighting tree search strategy) was essentially the same from a methodological point of view, except that only 5000 random stepwise additions were used. (3) Results A parsimony analysis with all characters unordered and equally weighted yielded 64 equally parsimonious trees, constituting a single island, with a length of 1375 steps [ensemble consistency index (CI) excluding uninformative characters=0.2392; ensemble retention index (RI)=0.6727; ensemble rescaled consistency index (RC)=0.1654]. A strict consensus (Fig. 4) shows the following unresolved relationships: (1) the node subtending all included species of aı̈stopods; (2) an internal node within the embolomeres, with a trichotomy subtending Archeria crassidisca, Pholiderpeton scutigerum and a clade formed by Anthracosaurus russelli and Pholiderpeton attheyi; (3) the node subtending derived dissorophoids, with four amphibamid species collapsed in a polytomy with a clade including micromelerpetontids and branchiosaurids and a clade encompassing albanerpetontids and basal crown-group lissamphibians; (4) the node leading to albanerpetontids and Eocaecilia micropoda, both forming an unresolved trichotomy with a collapsed clade including Karaurus sharovi, Valdotriton gracilis and Triadobatrachus massinoti; and (5) an internal node within diplocaulid nectrideans, with Batrachiderpeton reticulatum and Diceratosaurus brevirostris collapsed in a polytomy with a clade formed by Diplocaulus magnicornis and Diploceraspis burkei. One of the 64 fundamental trees (Fig. 5, also shown as circular cladogram in Fig. 6) was chosen to discuss character distribution at selected nodes (character-state distribution for all trees is available upon request from the authors). Overall tree topology is fairly balanced, as evidenced by the Colless index, Ic=0.357, which is closer to the value of a fully dichotomous cladogram (0) than to that of a completely pectinate cladogram (1) (Heard, 1992; Colless, 1995). This contrasts with rather higher Ic values obtained from other recent tetrapod phylogenies, such as those of Carroll (1995; Ic=0.83 or 0.75, depending upon tree topology), Coates (1996; Ic=0.7), Laurin & Reisz (1999; Ic=0.49), Paton et al. (1999; Ic=0.46, 0.44 or 0.42, depending upon tree topology) and Anderson (2001; Ic=0.42). For brevity, only ACCTRAN-optimized characters are taken into account in the remainder of the paper (unless otherwise specified). We used two simple methods to evaluate the amount of phylogenetic signal present in the matrix. The first method is based on comparisons between the Marcello Ruta, Michael I. Coates and Donald L. J. Quicke 272 2 2 2 2 1 3 6 1 1 1 2 5 5 3 2 1 1 1 2 4 2 5 3 1 3 1 1 1 Lissamphibian-amniote phylogenetic split 4 2 2 1 2 1 1 1 1 1 2 3 1 3 5 3 1 1 1 1 4 1 1 1 1 1 4 1 1 1 1 1 1 3 1 1 4 1 13 2 3 2 1 2 ALB. Crown-group LISSAMPHIBIA EMB. GEP. SEY. DIA. Crown-group AMNIOTA MIC. LYS. ACH. ADE. NEC. Holospondyli 1 TEM. Lepospondyli 1 1 BAP. Anthracosauria 4 1 COL. Crown-group TETRAPODA 1 Stem-group TETRAPODA 2 Eusthenopteron foordi Panderichthys rhombolepis Ventastega curonica Acanthostega gunnari Ichthyostega stensioei Tulerpeton curtum Colosteus scutellatus Greererpeton burkemorani Crassigyrinus scoticus Whatcheeria deltae Baphetes kirkbyi Megalocephalus pachycephalus Eucritta melanolimnetes Edops craigi Chenoprosopus lewisi Cochleosaurus florensis Isodectes obtusus Neldasaurus wrightae Trimerorhachis insignis Balanerpeton woodi Dendrerpeton acadianum Eryops megacephalus Acheloma cumminsi Phonerpeton pricei Ecolsonia cutlerensis Broiliellus brevis Amphibamus grandiceps Doleserpeton annectens Eoscopus lockardi Platyrhinops lyelli Micromelerpeton credneri Apateon pedestris Leptorophus tener Schoenfelderpeton prescheri ALBANERPETONTIDAE Eocaecilia micropoda Karaurus sharovi Triadobatrachus massinoti Valdotriton gracilis Caerorhachis bairdi Eoherpeton watsoni Proterogyrinus scheelei Archeria crassidisca Pholiderpeton scutigerum Anthracosaurus russelli Pholiderpeton attheyi Bruktererpeton fiebigi Gephyrostegus bohemicus Solenodonsaurus janenschi Kotlassia prima Discosauriscus austriacus Seymouria baylorensis/sanjuanensis Diadectes absitus Limnoscelis paludis Captorhinus aguti Paleothyris acadiana Petrolacosaurus kansensis Westlothiana lizziae Batropetes fritschia Tuditanus punctulatus Pantylus cordatus Stegotretus agyrus Asaphestera intermedia Saxonerpeton geinitzi Hapsidopareion lepton Micraroter erythrogeios Pelodosotis elongatum Rhynchonkos stovalli Cardiocephalus sternbergi Euryodus primus Microbrachis pelikani Hyloplesion longicostatum Odonterpeton triangulare Brachydectes elongatus/newberryi Acherontiscus caledoniae Adelospondylus watsoni Adelogyrinus simorhynchus Dolichopareias disjectus Scincosaurus crassus Keraterpeton galvani Batrachiderpeton reticulatum Diceratosaurus brevirostris Diplocaulus magnicornis Diploceraspis burkei Ptyonius marshii Sauropleura pectinata/scalaris Urocordylus wandesfordii Lethiscus stocki Oestocephalus amphiuminum Phlegethontia linearis AIS. Fig. 4. Strict consensus of 64 equally parsimonious trees deriving from the total data set. Numbers at nodes represent decay index values. Abbreviations as in Fig. 1 with the following additions : ACH., Acherontiscidae ; ALB., Albanerpetontidae ; EMB., Embolomeri ; GEP., Gephyrostegidae. Early tetrapod relationships revisited 273 96 61 85 68 95 54 90 52 83 57 77 69 57 55 68 70 83 77 90 77 66 100 65 Eusthenopteron foordi Panderichthys rhombolepis Ventastega curonica Acanthostega gunnari Ichthyostega stensioei Tulerpeton curtum Colosteus scutellatus Greererpeton burkemorani Crassigyrinus scoticus Whatcheeria deltae Baphetes kirkbyi Megalocephalus pachycephalus Eucritta melanolimnetes Edops craigi Chenoprosopus lewisi Cochleosaurus florensis Isodectes obtusus Neldasaurus wrightae Trimerorhachis insignis Balanerpeton woodi Dendrerpeton acadianum Eryops megacephalus Acheloma cumminsi Phonerpeton pricei Ecolsonia cutlerensis Broiliellus brevis Platyrhinops lyelli Eoscopus lockardi Micromelerpeton credneri Apateon pedestris Leptorophus tener Schoenfelderpeton prescheri Amphibamus grandiceps Doleserpeton annectens ALBANERPETONTIDAE Eocaecilia micropoda Valdotriton gracilis Karaurus sharovi Triadobatrachus massinoti Caerorhachis bairdi Eoherpeton watsoni Proterogyrinus scheelei Anthracosaurus russelli Pholiderpeton attheyi Archeria crassidisca Pholiderpeton scutigerum Bruktererpeton fiebigi Gephyrostegus bohemicus Solenodonsaurus janenschi Kotlassia prima Discosauriscus austriacus Seymouria baylorensis / sanjuanensis Diadectes absitus Limnoscelis paludis Captorhinus aguti Paleothyris acadiana Petrolacosaurus kansensis Westlothiana lizziae Batropetes fritschia Tuditanus punctulatus Pantylus cordatus Stegotretus agyrus Asaphestera intermedia Saxonerpeton geinitzi Hapsidopareion lepton Micraroter erythrogeios Pelodosotis elongatum Rhynchonkos stovalli Cardiocephalus sternbergi Euryodus primus Microbrachis pelikani Hyloplesion longicostatum Odonterpeton triangulare Brachydectes elongatus / newberryi Acherontiscus caledoniae Adelospondylus watsoni Adelogyrinus simorhynchus Dolichopareias disjectus Scincosaurus crassus Keraterpeton galvani Batrachiderpeton reticulatum Diceratosaurus brevirostris Diplocaulus magnicornis Diploceraspis burkei Ptyonius marshii Sauropleura pectinata / scalaris Urocordylus wandesfordii Oestocephalus amphiuminum Lethiscus stocki Phlegethontia linearis Fig. 5. One of the fundamental trees deriving from the original parsimony run. Numbers at nodes refer to bootstrap percentage values for clades with bootstrap support greater than 50 %. Remaining, unlabelled nodes are collapsed in a bootstrap 50% majority-rule consensus tree. 274 Marcello Ruta, Michael I. Coates and Donald L. J. Quicke Fig. 6. Circular dendrogram of tetrapod interrelationships using the branching sequence shown in Fig. 5. Selected taxa illustrated as indicators of morphological diversity ( figures not drawn to the same scale ). CI value associated with the real data set, and the expected CI value for an ideal data set of the same size (Sanderson & Donoghue, 1989). The latter value, referred to as CIexpected, is related to the number of taxa (n) through the following simple equation: CIexpected =0:9x0:022n+0:000213n2 : ð1Þ The above formula presents some theoretical problems. A graph of the above equation, with CIexpected plotted as a function of the number of taxa, shows that the former decreases continuously for n comprised between 0 and approximately 51. A continuous increase in CIexpected values is found for n comprised between approximately 51 and approximately 107 taxa. Thus, the equation represents increasing levels of homoplasy in a data set only within a restricted taxon number interval, with CIexpected reaching its minimum theoretical value when only 51 taxa are considered. Homoplasy levels decrease for n greater than 51. Obviously, the equation is not valid for n greater than 107, because CIexpected would be greater than 1. For 90 taxa, CIexpected=0.645. The fact that the CI obtained from our data set is considerably Early tetrapod relationships revisited 275 lower than the expected CI value may not necessarily imply low levels of phylogenetic information. It does, however, suggest that ‘noise’ is pervasive. The second method used to measure the amount of ‘noise’ in the data is based on the index of Klassen, Mooi & Locke (1991), which compares the CI of a real data set with that of a random matrix of the same size. The index, referred to as CIrandom, is calculated as follows: CIrandom =2:937nx0:9339 : ð2Þ Briefly, CIrandom values that are lower than CI values associated with a real data set imply the presence of phylogenetic signal in the latter. For 90 taxa, CIrandom =0.044. This value is considerably lower than 0.239 retrieved in the original analysis, suggesting that despite the broad range of fossil taxa included, nonrandom matrix structure exceeds the level of background noise. Most cladogram nodes are collapsed in a bootstrap 50% majority-rule consensus tree based on 10000 replicates employing the fast stepwise addition option of PAUP*. Bootstrap percentage values greater than 50% are plotted on the selected tree (Fig. 5). Decay index values (Bremer support values) are indicated on the strict consensus tree in Fig. 4. As expected, there is, usually, a good match between bootstrap percentage and decay index. Bremer supports were estimated by running up to 10 000 random additions holding no more than one tree for TBR swapping at any one time. Searches were terminated either when they reached the MPT (most parsimonious tree) length of +1 (i.e. Bremer support=1) or when the length of the shortest tree found had been hit at least 40 times. (a) The tetrapod stem-group The branching order of post-panderichthyid Devonian taxa (Fig. 5; see also lower left sector of Fig. 6) is in broad agreement with the results of several previous works (but see Ahlberg & Clack, 1998). A sister group relationship between Acanthostega and Ichthyostega (e.g. Laurin & Reisz, 1999) can be obtained at the cost of three additional steps. Several authors (Lombard & Bolt, 1995; Lebedev & Coates, 1995; Coates, 1996; Clack, 1998 a, c; Paton et al., 1999) have considered Tulerpeton, Crassigyrinus, Whatcheeria and baphetids to be amniote relatives. However, as the morphological and taxonomical data base for early tetrapods expands, the systematic affinities of these tetrapods are changing (e.g. Clack, 2002). The stem-group topology recovered by the present analysis agrees with Ahlberg & Clack’s (1998), Laurin & Reisz’s (1997, 1999) and Laurin’s (1998a–c) results, although the position of the abovementioned taxa is very weakly supported (Panchen, 1973, 1985, 1991; Panchen & Smithson, 1987, 1988; Lombard & Bolt, 1995; Clack, 1996, 1998a, c, 2001, 2002; Bolt & Lombard, 2000; Clack & Carroll, 2000). Thus, if baphetids are placed on the amniote stem as sister group to Caerorhachis plus more derived stemamniotes, then tree length increases by five steps only. With four extra steps, baphetids can be grafted to the lissamphibian stem as sister group to Eucritta plus temnospondyls. Only two additional steps are required to place a clade consisting of Eucritta and baphetids on the lissamphibian stem as sister group to temnospondyls, and five to place the same clade on the amniote stem as sister group to Caerorhachis plus more derived stemamniotes. However, tree length increases by as many as 32 steps if we impose Clack’s (2001) tree topology, wherein Whatcheeria and Gephyrostegus are successive sister taxa to Crassigyrinus and embolomeres, and the Eucritta-baphetid clade is placed as sister taxon to all of these groups. Tulerpeton and baphetids are known from incomplete material [but see Lebedev & Coates (1995) and Milner & Lindsay (1998)]. The position of Tulerpeton as a primitive stem-amniote in Lebedev & Coates’s (1995) and Coates’ (1996) cladograms implies that the lissamphibian-amniote phylogenetic split had occurred by the late Devonian (Famennian; see also discussion below). Conversely, the present study favours a stemtetrapod placement for Tulerpeton. Thus far, all studies of Tulerpeton have assumed the coherent nature of the contributory material, ranging from the near-articulated postcranium to isolated palatal and lower jaw fragments (Lebedev, 1984; Lebedev & Clack, 1993; Lebedev & Coates, 1995). We were interested to test the placement and coherence of Tulerpeton as a natural taxon. For this purpose, we built a data matrix in which Tulerpeton was divided into two taxa, Tulerpeton1 consisting of cranial and lower jaw data (with postcranial data coded as unknown), and Tulerpeton2, consisting of postcranial data (with cranial and lower jaw data coded as unknown). The strict consensus of the resulting 256 equally parsimonious trees at 1403 steps (CI=0.2545; RI=0.6727; RC=0.1755) resembles that recovered from the original analysis, but places Tulerpeton1, Tulerpeton2, Crassigyrinus and colosteids in a polytomous node between Ichthyostega and Whatcheeria. Inspection of an Adams consensus and of an agreement subtree [i.e. a taxonomically ‘pruned’ cladogram showing the largest subset of taxa for which all fundamental trees agree upon relationships (Swofford, 1998)] reveals that Tulerpeton1 is a rogue taxon, but that Tulerpeton2 is unequivocally placed 276 Marcello Ruta, Michael I. Coates and Donald L. J. Quicke between Ichthyostega and colosteids. Therefore, despite the unstable position of Tulerpeton1, we conclude that both this taxon and Tulerpeton2 fall in the tetrapod stemgroup. These results provide insufficient evidence to warrant the treatment of the cranial and postcranial data as belonging to separate OTUs, although conclusive evidence can only come from more complete Tulerpeton material. The tetrapod crown-group is supported by the following character-state changes (character consistency index expressed as c.i.), none of which represent unambiguous synapomorphies: 51 (c.i.=0.2; 0p1), presence of postparietal-exoccipital suture; 78 (c.i.=0.125; 0)1), postorbital broadly crescentic, narrowing posteriorly to a point; 93 (c.i.=0.333; 1)0), jugal without deep, V-shaped indentation along its dorsal margin; 115 (c.i.=0.286; 3p4), absence of lateral line system on skull roof; 116 (c.i.=0.286; 3)4), absence of mandibular canal; 208 (c.i.=0.25; 1p0), mid coronoid with denticles; 212 (c.i.=0.2; 1)0), posterior coronoid with denticles; 214 (c.i.=0.25; 0)1), posterior coronoid with posterodorsal process; 238 (c.i.=0.167; 0p1), latissimus dorsi process aligned with ectepicondyle; 240 (c.i.=0.5; 0p1), absence of ventral humeral ridge. Eucritta and Caerorhachis bracket the base of the tetrapod crown-group (Fig. 5; see also upper left sector of Fig. 6), thus corroborating previous interpretations of their primitive nature (Clack, 1998a, 2001; Ruta et al., 2001). The mosaic of characters that both taxa share with such diverse groups as baphetids, ‘anthracosaurs’ and temnospondyls indicates strongly that they may be phylogenetically close to the divergence of lissamphibian and amniote clades (Clack, 1998a, 2001; Ruta et al., 2001). Consistent with this interpretation, only two steps are added to tree length if Eucritta is placed in any of the following positions (locations of other taxa being unchanged): sister taxon to baphetids; crownward of baphetids on the tetrapod stem; sister taxon to Caerorhachis plus more crownward stem-amniotes. With three extra steps, Eucritta can be placed between Caerorhachis and embolomeres. With four extra steps, it can be paired with Caerorhachis. Tree length increases by four steps if Caerorhachis is placed either crownward of baphetids on the tetrapod stem or as sister taxon to Eucritta plus more crownward temnospondyls. Alternative placements for Eucritta and Caerorhachis (further away from the crown-group basal node) involve additional steps. This deep split between lissamphibian- and amnioterelated taxa has interesting implications for the distribution of certain characters long regarded as ‘reptiliomorph’ or ‘anthracosauroid’ apomorphies (Lombard & Bolt, 1995; Clack, 1998a, c; Paton et al., 1999; Bolt & Lombard, 2000). Several of these may in fact represent tetrapod plesiomorphies at a post-colosteid level, that persist in basal crown-group members. Here, we focus on four such characters: orbit shape, skull roof suture patterns, vertebral body construction and the number of manus digits. Orbits with an irregularly shaped outline (i.e. neither round nor elliptical) are widespread among early tetrapods. A plausible functional explanation for these shapes has not been found (Clack, 1987 b), although several alternatives for the marked antorbital vacuities of baphetids have been offered (Beaumont, 1977; Bjerring, 1986; Milner & Lindsay, 1998), including development of broad insertion areas for jaw musculature; presence of salt glands; and presence of electrosensory organs. Irregular orbits are also present in Crassigyrinus, Whatcheeria, Eucritta and, among embolomeres, Anthracosaurus, Carbonoherpeton, Eoherpeton and Pholiderpeton attheyi (Panchen, 1972, 1975, 1977, 1980, 1985; Beaumont, 1977; Klembara, 1985; Clack, 1987 a, b, 1998a, c, 2001; Beaumont & Smithson, 1998; Milner & Lindsay, 1998). According to Clack (1998 a), the antorbital vacuities of baphetids may represent exaggerated (possibly peramorphic) versions of the small anteroventral orbital embayments of Eucritta (see also Clack, 2001). A similar embayment is also observed in the embolomere Palaeoherpeton (Panchen, 1964, 1980). In Crassigyrinus, the orbit outline is near rhomboidal (Panchen, 1985; Clack, 1998 c), and includes a small but distinct anteroventral corner or embayment. A less pronounced version of such rhomboidal orbits is found in Eoherpeton (Smithson, 1985). In baphetids, the dorsal (orbital) margin of the jugal includes a characteristic, deep notch (Beaumont, 1977). However, jugal notches are also present in Anthracosaurus, Carbonoherpeton, Pholiderpeton attheyi and Whatcheeria (Panchen, 1964, 1972, 1977, 1980; Klembara, 1985; Clack, 1987a, b, 1998 b, c; Lombard & Bolt, 1995). According to our analysis, angular orbits are a transitory condition, since they occur in a series of stem-tetrapods, some basal stem-amniotes and one stem-lissamphibian. They are not an unambiguous shared derived feature of discrete monophyletic groups. One of the orbit characters employed by Clack (1998c, 2001) relates to the occurrence of an anteroventral orbit corner (our character 105), observed in Crassigyrinus, Eucritta and Whatcheeria. Under ACCTRAN optimization, this character appears to be transitional in the portion of the stem-group comprised between colosteids and baphetids (i.e. nodes leading to Crassigyrinus and Whatcheeria) and is acquired in parallel by Eucritta. However, if DELTRAN is used, then Crassigyrinus, Eucritta and Whatcheeria are shown to have Early tetrapod relationships revisited acquired an anteroventral orbit corner three times independently. The distribution of characters describing skull table suture patterns corroborates earlier hypotheses that alternative configurations, including mutually exclusive contacts between supratemporal and postparietal, and between parietal and tabular, have diagnostic value for lissamphibian and amniote relatives (Panchen, 1980; Panchen & Smithson, 1988). Leaving aside the question of bone homologies in the skull table of several lepospondyls, we note that a supratemporalpostparietal suture (complement arrangement of bones as expressed in character 39, related to the parietaltabular contact), is conserved as a primitive character (under ACCTRAN and DELTRAN) in the temnospondyl-lissamphibian clade (but see Boy, 1986). On the amniote branch (with the possible exclusion of Caerorhachis; Holmes & Carroll, 1977; Ruta et al., 2001), a parietal-tabular contact is observed in all taxa in which these bones are recognisable as separate ossifications (but see Smithson, 1986), including some lepospondyls (e.g. the urocordylid nectridean Sauropleura). Several models have been proposed to explain the derivation of different vertebral centra from one another or from hypothetical archetypes. Study of primitive tetrapods has clarified the polarity of this character complex (e.g. Coates, 1996). A rhachitomous pattern (or derivations thereof) is ubiquitous in stemtetrapods and among most temnospondyls. The unusual vertebral construction observed in some specimens of Whatcheeria (Lombard & Bolt, 1995) appears to be a simple modification (although not necessarily in a strict phylogenetic sense) of the multipartite centrum of such taxa as colosteids and several temnospondyls (Godfrey, 1989). The morphology of the poorly preserved centra of Tulerpeton (Lebedev & Coates, 1995) and baphetids (Milner & Lindsay, 1998) is also consistent with a rhachitomous model, despite little information on their postcranial skeletons. A gastrocentrous pattern dominates among amniote-like taxa (as well as some derived temnospondyls, e.g. Bolt, 1991; Holmes, 2000; Boy & Sues, 2000). This ranges from the simple construction of Caerorhachis (Holmes & Carroll, 1977; Ruta et al., 2001) and several basal ‘anthracosaurs’ (small intercentra and U-shaped pleurocentra, e.g. Silvanerpeton and Eldeceeon; Clack, 1994b; Smithson, 1994), to the massive, disc-like intercentra and pleurocentra of various embolomeres (e.g. Archeria; Holmes, 1989), and the pleurocentrumdominated vertebrae of several lepospondyls, seymouriamorphs, diadectomorphs and crown-amniotes (White, 1939; Romer, 1956, 1966; Carroll & Gaskill, 277 1978; Sumida, 1997; Carroll, 1988; Carroll et al., 1998; Benton, 2000; Klembara & Bartı́k, 2000). In our original data set, conditions describing the number of digits in the manus were treated as independent characters, thus imposing no linkage between them. However, we also explored the effects of multistate coding for digit number. The highest number of digits (eight in Acanthostega) was given state 0 whereas the lowest (three in microbrachomorphs) was given state 5. The character was treated as unordered, thus allowing free transformations between different conditions. A PAUP* run gave 384 equally parsimonious trees at 1400 steps (CI=0.2558; RI=0.6723; RC=0.1762). A strict consensus is almost identical to that obtained from the original analysis, except for a considerable loss of resolution among microbrachomorph and some tuditanomorph microsaurs. Reconstruction of the character-state changes relative to the number of digits on a selected cladogram shows that character optimization is equivocal (under both ACCTRAN and DELTRAN) in the post-panderichthyid part of the tetrapod stem-group. This is not unexpected given the unavailability of data in several stem-tetrapod taxa (Ichthyostega, Crassigyrinus, Whatcheeria, baphetids), and the conflicting distribution of states among colosteids (Hook, 1983; Coates, 1996). Among crowntetrapods, the presence of a manus with no more than four digits is acquired in parallel in the temnospondyllissamphibian clade and in the Westlothiana-lepospondyl clade. Within lepospondyls, digit number decreases further in microbrachomorphs. Therefore, a five-digited manus does not identify any particular clade. Rather, this condition appears to be transitional among several basal stem-amniotes and primitive crown-amniotes. (b) The lissamphibian stem-group The basal node of the temnospondyl-lissamphibian clade (including Eucritta) is supported by several homoplastic features, some of which relate to optimizations of missing character scores: 144 (c.i.=0.083; 0p1), pterygoid with posterolateral flange; 178 (c.i.= 0.5; 0p1), absence of parasymphysial plate; 188 (c.i.=0.2; 1p0), rearmost extension of mesial lamina of splenial closer to anterior end of lower jaw than to adductor fossa; 215 (c.i.=0.167; 0p1), posterior coronoid exposed in lateral view; 229 (c.i.=0.2; 1p0), posterior margin of interclavicle not drawn out into parasternal process; 230 (c.i.=0.2; 1p0), parasternal process not elongate and parallel-sided; 257 (c.i.=0.154; 2)1), radius approximately as long as ulna; 280 (c.i.=0.143; 1)0), ribs mostly straight in at least part of the trunk; 314 (c.i.=0.25; 0p1), presence 278 Marcello Ruta, Michael I. Coates and Donald L. J. Quicke of no more than four digits in manus. The shortest path leading from the base of the tetrapod crown-group to the node subtending crown-group lissamphibians includes 115 character-state changes over 12 internodes with an average of 9.6 changes for each internode. Comparisons between crownward stem-tetrapods and temnospondyls reveal the conservative morphology of various skeletal characters in the latter (e.g. skull roof suture pattern; vertebral construction). Unfortunately, the Mississippian record of temnospondyls is sparse (worse than that of stem-group amniotes), presumably as a result of palaeoecological factors. The significance of the near-complete material of Balanerpeton is thus emphasised as a unique glimpse of conditions in the earliest members of the group. Crownward of Eucritta, the branching pattern of the basal part of the temnospondyl tree (Fig. 5; see also upper and middle left sectors in Fig. 6) conforms mostly to the results of several alternative recent analyses (Milner, 1990; Milner & Sequeira, 1994, 1998; Holmes et al., 1998; Holmes, 2000). The Permo-Carboniferous edopoids are a basal clade of long-snouted forms known mostly from cranial material. Several derived edopoids (e.g. Cochleosaurus and Chenoprosopus) are characterized by broad dorsomesial extensions of the premaxillae, teardrop-shaped choanae, and elongation of the preand interchoanal regions (Milner & Sequeira, 1998). The anteriorly sutured pterygoids and the absence of premaxillary alary processes represent possible plesiomorphic characters. Various features of the skull and palate of the recently described Adamanterpeton ohioensis indicate the primitive condition from which more derived edopoids might have arisen (Milner & Sequeira, 1998). For example, its rather narrow palatal vacuities are proportioned similarly to those of Eucritta and Caerorhachis (Holmes & Carroll, 1977; Holmes, 2000; Ruta et al., 2001). The phylogenetic position of Balanerpeton and Dendrerpeton has been debated (Milner, 1980; Milner & Sequeira, 1994; Holmes et al., 1998; Holmes, 2000). Only two extra steps are required to pair Dendrerpeton with Balanerpeton, as in Holmes et al.’s (1998) phylogeny, or to place them as successively more closely related taxa to eryopoids plus higher temnospondyls. With three extra steps, Dendrerpeton and Balanerpeton can be placed, in that order, as successively more closely related taxa to trimerorhachoids and higher temnospondyls, as in Milner & Sequeira’s (1994) cladogram. If Balanerpeton is paired with trimerorhachoids or inserted between edopoids and trimerorhachoids, then only one extra step is added to tree length. If these latter rearrangements are applied to Dendrerpeton, then tree length increases by two and three steps, respectively. Dendrerpeton is sister taxon to a clade encompassing eryopoids, dissorophoids and crown-lissamphibians. In the light of recent work on D. acadianum (Holmes et al., 1998), the position of this taxon sheds new light on the early diversification of eryopoids and dissorophoids. In particular, it calls for a reassessment of the distribution of such key dissorophoid/salientian features as the occurrence of a posterodorsal process of the quadrate (Lombard & Bolt, 1979; Bolt & Lombard, 1985; Milner, 1988, 1990; Bolt, 1991; Daly, 1994). The interrelationships of dissorophoids depart significantly from those of previous studies (e.g. Milner, 1990, 1993; Trueb & Cloutier, 1991; Daly, 1994). Dissorophids emerge as paraphyletic and branch from the lissamphibian stem between trematopids (monophyletic) and a poorly resolved clade consisting of amphibamids, micromelerpetontids, branchiosaurids, albanerpetontids and crown-lissamphibians. The position of Ecolsonia – crownward of trematopids – agrees with one of the two alternative hypotheses of relationship of this taxon postulated by Berman et al. (1985), but contrasts with its relatively derived position in Daly’s (1994) analysis. The most surprising results concern the derived portion of the lissamphibian stem. Thus, amphibamids are paraphyletic with respect to a micromelerpetontid-branchiosaurid clade. Together, these taxa are paired with albanerpetontids plus crown-lissamphibians (Fig. 6, mid-lower left sector). Inspection of an agreement subtree reveals that the only unequivocal pattern of sister group relationships among derived dissorophoids consists of Leptorophus and Schoenfelderpeton as sister groups, with Apateon, Micromelerpeton and Eoscopus as progressively more ‘outlying’ taxa. This broader clade joins albanerpetontids as sister taxon to Karaurus plus Triadobatrachus in the agreement subtree. This result may reflect a genuine pattern of relationships, or may be due to lack of additional characters. As pointed out by Milner (1993), conflicting character distributions suggest that several dissorophoid lineages approached the condition of basal crown-lissamphibians independently and to varying degrees. The amount of character convergence in crownward stem-lissamphibians might explain why our analysis fails to retrieve a sister group relationship between one or few specific dissorophoids and crown taxa in some of the most parsimonious solutions. In others, however, including the tree used for character discussion, crownlissamphibians plus albanerpetontids are paired with a clade consisting of Amphibamus plus Doleserpeton (e.g. Bolt, 1969, 1979, 1991; Trueb & Cloutier, 1991; Milner, 1993). The sequence of cladogenetic events in the crownward part of the temnospondyl tree Early tetrapod relationships revisited re-emphasizes the importance of dissorophoids in the lissamphibian origin debate [but see Laurin & Reisz (1997, 1999) and Laurin (1998a–c) for a contrasting opinion]. The node subtending derived dissorophoids plus crown-lissamphibians is supported by the following character-state changes: 29 (c.i.=0.1; 0)1), maxilla entering orbit margin; 73 (c.i.=0.077; 0)1), parietal–parietal width greater than distance between the posterior margin of the skull table and the posterior margin of the orbits, measured along the midline; 104 (c.i.=0.105; 2p1), minimum interorbital distance smaller than maximum orbit diameter; 126 (c.i.=0.2; 0p1), presence of distinct posterolateral process of the vomer bordering more than half of the posterior margin of the choana; 150 (c.i.=0.333; 0p1), quadrate ramus of pterygoid straight, rod-like and gently tapering distally; 249 (c.i.=0.25; 0)1), slender and elongate humerus, the length of which is more than three times the width of its distal end; 252 (c.i.=0.125; 0)1), width of entepicondyle less than half the length of the humerus; 283 (c.i.=1; 0)1), longest trunk ribs poorly ossified, slender rods, the length of which is smaller than the length of three mid-trunk vertebrae. Laurin & Reisz’s (1997, 1999) and Laurin’s (1998a–c; Fig. 1d) analyses deserve further comment. These authors consider only a limited sample of putative shared derived characters linking dissorophoids to lissamphibians. Despite the inclusion of such key taxa as Doleserpeton, their temnospondyl exemplar does not adequately encompass internested sets of lissamphibian apomorphies identified in previous studies (e.g. Milner, 1988, 1990, 1993; Bolt, 1991; Trueb & Cloutier, 1991; Daly, 1994; Gardner, 2001). Laurin’s (1998a–c) and Laurin & Reisz’s (1997, 1999) lepospondyl-lissamphibian clade is mostly supported by ‘absence’ characters (e.g. losses of certain cranial and mandibular bones). ‘Absence’ data pose special problems in phylogenetic reconstructions (Poe & Wiens, 2000). It is difficult to assess, a priori, whether they contain true phylogenetic signal. Clusters of ‘absence’ features may bias the results of an analysis in favour of sister group relationships between taxa that display few, if any, derived characters, by means of swamping any signal derived from alternative character sets. In the case of lissamphibians and lysorophids, Laurin’s (1998a–c) and Laurin & Reisz’s (1997, 1999) analyses place emphasis on the simplification of the skull roofing pattern (e.g. bone reduction and/or loss). In fact, examples of such simplification occur repeatedly throughout osteichthyan clades, and often show distinct phylogenetic trends (e.g. synapsids; Sidor, 2001, and further examples therein). Although putative temnospondyl-lissamphibian synapomorphies (e.g. dental features) are also included 279 in Laurin’s (1998a–c) and Laurin & Reisz’s (1997, 1999) analyses, several other characters used in previous studies (e.g. configuration of various palatal elements) are omitted (e.g. Milner, 1988, 1990, 1993; Bolt, 1991; Trueb & Cloutier, 1991; Daly, 1994). This may be significant. We decided to assess the impact of ‘absence’ characters against a larger set of putative temnospondyllissamphibian synapomorphies than that used by Laurin (1998a–c) and Laurin & Reisz (1997, 1999). Milner (1993) found that in four genera of Amphibamidae, he could discern as few as one (in Platyrhinops) and as many as six (in Doleserpeton) synapomorphies with crown-lissamphibians. When crown-lissamphibians were placed as sister taxon to Doleserpeton, Milner (1993) found that four characters related to dentition, palate and vertebrae originated only once within dissorophoids. In the present analysis, at least some of the fundamental trees (including that in Fig. 4) show that as many as nine characters support a sister group relationship between a clade including Amphibamus plus Doleserpeton, and a clade including albanerpetontids plus crown-lissamphibians: 32 (c.i.=0.5; 0)1), maxillary facial process shaped like a rectangular flange; 127 (c.i.=0.167; 0)1), palatine without fangs; 133 (c.i.=0.333; 0p1), palatine poorly ossified, slender and strut-like; 134 (c.i.=0.167; 0p1), absence of ectopterygoid; 183 (c.i.=0.1; 0)1), dentary without anterior pair of fangs; 218 (c.i.=0.5; 0p1), presence of pedicely on marginal teeth; 258 (c.i.=0.091; 1)0), absence of olecranon process on ulna; 293 (c.i.=0.333; 0p1), trunk pleurocentra fused midventrally; 296 (c.i.=0.125; 0p1), neural spines of trunk vertebrae fused to centra. Placing albanerpetontids plus crown-lissamphibians as sister group to Doleserpeton requires only one extra step, as does Gardner’s (2001) preferred tree topology (albanerpetontids on the common stem-group of salientians and caudates, with caecilians as sister group to remaining lissamphibians). Based on Gardner’s (2001) branching scheme, the following characters unite crown-lissamphibians with Doleserpeton (with or without implied reversals within crown-group lissamphibians): vomer with transverse patch of small teeth posteromesial to choana (125); absence of ectopterygoid (134); pterygoid sutured with maxilla (146); trunk pleurocentra fused midventrally (293); trunk pleurocentra fused middorsally (294); neural spines of trunk vertebrae fused to centra (296). We performed further tests to evaluate the significance of alternative taxon arrangements in the crownward part of the temnospondyl branch. In particular, Carroll & Bolt’s (2001) hypothesis of separate 280 Marcello Ruta, Michael I. Coates and Donald L. J. Quicke origins of caudates and salientians from among dissorophoids postulates that Doleserpeton and Apateon are the most crownward plesions on the salientian and caudate stem-groups, respectively. For simplicity, dissorophoid relationships were left unchanged. We constrained Triadobatrachus to appear as sister taxon to Doleserpeton, and caudates plus caecilians (with albanerpetontids as a more outlying group) as sister taxon to Apateon. This arrangement entails 18 extra steps, but represents a considerably worse fit for the total data than the shortest trees overall (Templeton test: P=0.0027; Kishino–Hasegawa test: P=0.0026; Winning-sites test: P=0.0046). Forcing Eocaecilia to appear as sister taxon to the microsaur Rhynchonkos [see Carroll & Currie (1975), Carroll (2000) and references therein], but leaving the rest of the ingroup topology unchanged, entails 29 extra steps. Again, such a topology represents a considerably worse fit for the total data than the most parsimonious trees (Templeton test: P=0.0001; Kishino–Hasegawa test: P=0.0001; Winning-sites test: P=0.0002). Similar results (Templeton test: P<0.0001; Kishino–Hasegawa test: P< 0.0001; Winning-sites test: P=0.0001) are obtained if a Rhynchonkos-Eocaecilia clade is created with Triadobatrachus as sister taxon to Doleserpeton and caudates plus albanerpetontids as sister goup to Apateon (tree length increases by 42 steps). Tests of Laurin’s (1998a–c) and Laurin & Reisz’s (1997, 1999) hypothesized sister group relationship between lysorophids and crown-group lissamphibians are described below (Section VII.3h). A final remark concerns the position of albanerpetontids and caudates. In the present analysis, albanerpetontids appear to be the most crownward plesion on the lissamphibian stem [but see Trueb & Cloutier (1991) and Gardner (2001)]. However, only one extra step is required to place albanerpetontids as sister taxon to Eocaecilia, or as a stem-group member of the salientian-caudate clade. Pairing albanerpetontids with either Triadobatrachus or caudates increases tree length by four steps. At six, seven and eight extra steps, albanerpetontids can be placed as sister taxon to Valdotriton, Karaurus or Triadobatrachus, respectively. A caecilian-caudate clade requires three extra steps. Although this clade has not been retrieved in several traditional, morphology-based schemes of lissamphibian relationships (e.g. Estes, 1981; Duellmann & Trueb, 1986; Milner, 1988), it is nonetheless found in some morphological studies [discussion in Milner (1988), and references therein], as well as in some recent molecular analyses (e.g. Feller & Hedges, 1998, and references therein). Feller & Hedges (1998) erected the clade Procera for the monophyletic group including salamanders plus caecilians, and listed some osteological and soft anatomical features in support of it. However, evaluation of the osteological evidence must await a redescription of Eocaecilia (but see Carroll, 2000) as well as a reexamination of several early salamander-like taxa (e.g. batrachosauroidids; scapherpetontids; the problematic Ramonellus Nevo & Estes, 1969; review in Milner, 2000). The Procera hypothesis has some interesting implications for the assessment of character distribution among primitive crown-group lissamphibians. For instance, Laurin (1998b) reasoned that the presence of a tympanum, deduced to have existed in at least some temnospondyls, cannot be used as a valid argument to support derivation of lissamphibians from temnospondyls. Because the tympanum is present only in salientians (frogs), the conventional phylogenetic arrangement of salientians as sister taxon to caudates (with caecilians as sister taxon to the remaining two orders) would entail unparsimonious independent losses of a tympanum in caecilians and salamanders (which could nevertheless have happened), or its loss at the base of the lissamphibian crown-group followed by reacquisition in frogs, depending upon character optimization. However, if salamanders and caecilians are indeed sister groups, only a single loss event at the base of the Procera is required under all character optimizations. Furthermore, Milner (1988; quoting Smirnov, 1986) points out the fundamental similarities between the early developmental stages of the salientian and caudate ears and the fact that the adult ear of frogs may represent the likely primitive condition for both groups, the caudate ear being secondarily reduced. (c) The amniote stem-group Eleven character-state changes support the basal node of the amniote stem-group. Once again, none of these synapomorphies is unambiguous. These changes include: 5 (c.i.=0.2; 0p1), premaxillae less than twothirds the width of the skull; 66 (c.i.=0.1; 1p0), supratemoral contact with squamosal smooth; 68 (c.i.=0.2; 0p1), tabulars with subdermal blade-like postero-lateral horns; 107 (c.i.=0.125; 0)2), pineal foramen situated anterior to interparietal suture mid length; 117 (c.i.=0.333; 0p1), ventral, exposed surface of vomers narrow, elongate and strip-like, without extensions anterolateral or posterolateral to choana and two and a half to three times longer than wide; 234 (c.i.=0.125; 0p1), scapulocoracoid extending ventral to posteroventral margin of glenoid; 253 (c.i.=0.125; 0p1), length of humeral shaft portion proximal to entepicondyle greater than the width of humeral head; 261 (c.i.=0.2; 0)1), ilium with Early tetrapod relationships revisited transverse pelvic ridge; 276 (c.i.=0.25; 0)1), tarsus with L-shaped proximal element; 293 (c.i.=0.333; 0)1), pleurocentra fused midventrally; 315 (c.i.= 0.25; 0p1), presence of no more than five digits in manus. The shortest path leading from the base of the tetrapod crown-group to the node subtending crown-group amniotes includes 72 character-state changes over eight internodes with an average of nine changes for each internode. The branching sequence of taxa in the proximal half of the amniote stem (Fig. 5) reflects the conventional view that embolomeres are an early offshoot of (perhaps secondarily) aquatic, longbodied amniotes, and that gephyrostegids are more crownward, and presumably more terrestrial forms (Fig. 6, uppermost sector). See Laurin & Reisz (1997, 1999) and Laurin (1998a–c) for alternative views. Likewise, the pattern of sister group relationships in the crownward part of the amniote tree is in partial agreement with several previous hypotheses (e.g. Gauthier et al., 1988b; Sumida & Lombard, 1991; Berman et al., 1992; Sumida et al., 1992; Laurin & Reisz, 1997, 1999; Laurin, 1998a–c; Berman, 2000). The position of seymouriamorphs reflects traditional theories of primitive amniote relationships (e.g. Heaton, 1980; Gauthier et al., 1988b; Sumida & Lombard, 1991; Berman et al., 1992; Sumida et al., 1992; Lee & Spencer, 1997; Berman, 2000), and emphasizes the key role of this group in understanding the evolution of several amniote characters. Among the internested changes leading to the condition of several crownamniotes are: rearward shift and reduction/loss of posterior bones of skull table; widening of parietals; enlargement of the transverse pterygoid flanges; ‘swollen’ neural arches; consolidation of the atlas-axis complex; reduction and loss of intercentra; rearward position of neural arches relative to the position of pre- and postzygapophyses; development of an iliac shelf; progressive reduction of entepicondylar length and width relative to humeral shaft; modifications of humeral and femoral processes (Sumida & Lombard, 1991; Berman et al., 1992; Sumida et al., 1992; Sumida, 1997; Berman et al., 1998; Berman, 2000). In this context, it is interesting to note that Solenodonsaurus appears in a less crownward position than that retrieved by some recent analyses (notably, Laurin & Reisz, 1999). The present position reflects in part the ‘transitional’ nature of this tetrapod (Carroll, 1970; Gauthier et al., 1988 b; see above). However, it also conflicts, in part, with the distribution of certain cranial (e.g. absence of intertemporal) and trunk features (e.g. long, curved ribs) occurring in the crownward part of the amniote tree (although some of these characters are already found in some embolomeres). 281 Character-state changes at the node subtending Solenodonsaurus and more crownward amniotes are as follows: 40 (c.i.=0.167; 0p1), presence of suture between parietal and postorbital; 60 (c.i.=0.167; 0p1), intertemporal absent as separate ossification; 61 (c.i.=0.167; 0p1), intertemporal interdigitating with cheek (it is noteworthy that the occurrence of this character-state change provides no phylogenetic information for this branch whatsoever, since it derives from optimization of a morphological condition that is linked to a more generalized character; such an optimization exemplifies problems deriving from missing entries to signify inapplicable characters, and introduces a bias in the computation of branch length); 66 (c.i.=0.1; 0p1), interdigitating contact between supratemporal and squamosal; 68 (c.i.=0.2; 1p0), tabulars without subdermal blade-like postero-lateral horns; 89 (c.i.=0.083; 0)1), jugal entering ventral margin of skull roof; 104 (c.i.=0.105; 1)0), interorbital distance greater than maximum orbit diameter; 118 (c.i.=0.125; 0p1), vomer without fang pair; 127 (c.i.=0.167; 0p1), palatine without fang pair; 197 (c.i.=0.25; 0p1), angular reaching posterior end of lower jaw; 216 (c.i.=0.25; 0p1), posterodorsal process of posterior coronoid contributing to highest point of lateral margin of adductor fossa; 231 (c.i.=0.143; 0)1), interclavicle wider than long; 239 (c.i.=0.111; 0p1), humerus with distinct supinator process projecting anteriorly; 247 (c.i.=0.2; 0)1), humerus with expanded proximal and distal ends; 294 (c.i.= 0.25; 0)1), trunk pleurocentra fused middorsally; 296 (c.i.=0.125; 0)1), trunk neural spines fused to centra. Several recently described ‘reptiliomorph’ taxa from the Mississippian deserve additional comment. Eldeceeon (Smithson, 1994) and Silvanerpeton (Clack, 1994 c), both from the uppermost Viséan of Scotland, are similar to stratigraphically younger embolomeres, and may indeed be the latter’s plesiomorphic sister taxa. Further preparation of the Eldeceeon and Silvanerpeton material is likely to illuminate character distribution patterns at the base of the ‘reptiliomorph’ radiation. A third Mississippian tetrapod, Casineria (Paton et al., 1999) from the upper Viséan of Scotland, might represent a more derived amniote than Eldeceeon and Silvanerpeton. Casineria was initially excluded from our data set because the large number of missing entries (86.5% of total number of characters) indicated that it would behave as a ‘rogue’ taxon. This was confirmed by a parsimony run of the original data set after inclusion of Casineria. The resulting 2208 most parsimonious trees at 1407 steps yield a mostly unresolved strict consensus, although the temnospondyl–lissamphibian relationships 282 Marcello Ruta, Michael I. Coates and Donald L. J. Quicke are the same as those in the original analysis (including the branching sequence of temnospondyl taxa and the position of Eucritta at the base of the lissamphibian stemgroup). Conversely, all stem-group tetrapods crownward of Tulerpeton, and several plesions in the amniote stem-group are collapsed into a large polytomy. Inspection of the Adams consensus reveals a far less dramatic pattern of sister group relationships by relocating ‘… taxa in conflicting positions … to the most inclusive node that they have in common among the fundamental cladograms’ (Kitching et al., 1998: p. 199). Such relocations affect, among others, Acherontiscus ( joining colosteids in an unresolved node between Tulerpeton and Crassigyrinus) and, unsurprisingly, Casineria. Loss of resolution also affects, in part, microsaurs. In the Adams consensus, Casineria forms a polytomy with a clade of diadectomorphs plus crown-amniotes and a clade of Westlothiana plus lepospondyls. Random scanning through different samples of trees from the pool of 2208 fundamental topologies shows Casineria in one or the other of four possible positions: (1) sister taxon to Westlothiana plus lepospondyls; (2) sister taxon of lepospondyls (this arrangement is retrieved in 77% of the fundamental trees, as shown by a 50% majority-rule consensus); (3) nested within crown-amniotes (e.g. as sister taxon to Captorhinus in some trees); (4) nested within microsaurs (in several alternative positions within tuditanomorphs and microbrachomorphs). Paton et al.’s (1999) conclusions regarding the affinities of Casineria are partly supported by the present study. Both analyses fail to resolve the position of Casineria relative to such diverse taxa as basal crown-amniotes, Westlothiana and (in the present study) microsaurs. Clearly, a more precise phylogenetic assessment of this tetrapod is not possible in the absence of cranial and more complete postcranial material. Available evidence from limb proportions, shape of the ilium, configuration of the vertebral centra and ribs suggest that Casineria should be regarded as a ‘reptiliomorph’ of uncertain phylogenetic affinities. (d) The affinities of Westlothiana Westlothiana (Fig. 6, lower right sector) is prominent because it combines generalized amniote-like features with lepospondyl characters. The two current interpretations of the phylogenetic position of Westlothiana, regarded either as a primitive amniote (Smithson, 1989; Smithson & Rolfe, 1990; Smithson et al., 1994) or as the most crownward plesion in the tetrapod stemgroup (Laurin & Reisz, 1999), are contrasted with the hypothesis of relationships presented here (Fig. 5; see also Clack, 2002). According to Smithson et al. (1994), Westlothiana branches from the amniote stem between seymouriamorphs and diadectomorphs. Smithson et al.’s (1994) hypothesis is compatible with our results, except for the fact that Westlothiana is basal to a lepospondyl clade. Tree branch manipulation within MacClade shows that relocating Westlothiana between seymouriamorphs and the clade including diadectomorphs plus crown-amniotes requires fewest additional steps (four) compared with alternatives. The position of Westlothiana in the shortest trees is supported by 14 character-state changes. Most of these are, however, homoplastic and/or based on optimization of missing or inapplicable entries: 27 (c.i.=0.2; 0p1), portion of lacrimal lying anteroventral to orbit abbreviated; 49 (c.i.=0.125; 1p0), total width of postparietal smaller than four times its length; 82 (c.i.=0.333; 0)1), presence of kink in anteromedial margin of postorbital; 116 (c.i.=0.286; 4p0), mandibular canal totally enclosed; 141 (c.i.=0.5; 1)0), absence of transverse flange of pterygoid; 163 (c.i.=0.5; 0p1), exoccipitals forming with basioccipital a concave, continuous and strap-shaped articular surface; 170 (c.i.=0.25; 1)0), absence of posterolaterally directed, ridge-like thickenings (ridges ending in basal tubera) on basal plate of parasphenoid; 215 (c.i.=0.167; 1p0), posterior coronoid not exposed in lateral view; 235 (c.i.=0.25; 0p1), absence of glenoid foramen on scapulocoracoid; 255 (c.i.=0.167; 1p0), length of humerus greater than that of two and a half mid trunk vertebrae; 267 (c.i.=0.143; 1)0), absence of a distinct rugose area on the fourth trocanter; 312 (c.i.=0.25; 0)1), height of neural arch in midtrunk vertebrae smaller than the length between pre- and postzygapophyses; 314 (c.i.=0.25; 0p1), presence of no more than four digits in manus; 315 (c.i.=0.333; 1p0), absence of five digits in manus. In Anderson’s (2001) analysis, the microsaur Utaherpeton is identified as the most basal lepospondyl. Aside from considerations of the status of microsaurs (discussed further below), it is noteworthy that Utaherpeton and Westlothiana are similar in several respects. Comparisons between these two taxa are necessarily limited by their poor preservation. However, they resemble each other in the shape and relative proportions of the bones in the preorbital region of the skull, in the morphology of the mandible, in the vertebral construction, and in the shape of the puboischiadic plate (especially with regards to the ischium/pubis length ratio). Some of these features are also found in several microsaur taxa, especially primitive tuditanomorphs. Major differences between Utaherpeton and Westlothiana (e.g. in the morphology and proportions of limb elements) may reflect in part the immature condition of Utaherpeton Early tetrapod relationships revisited (Carroll et al., 1991; Carroll & Chorn, 1995). The amniote affinities of microsaurs [e.g. Olson, 1962; Vaughn, 1962; Brough & Brough, 1967; but see also Romer (1950), Carroll & Baird (1968), Carroll & Gaskill (1978), and references therein] are also supported by Paton et al.’s (1999) analysis, although the latter includes only a limited sample of microsaurs. Taken together, these observations offer an alternative perspective on the significance of Westlothiana for our understanding of the evolutionary history of primitive amniotes. The conjectured amniote or microsaur affinities of Westlothiana are no longer mutually exclusive. In fact, Westlothiana now appears as something of a keystone taxon, contributing to a more inclusive and explanatory hypothesis of early amniote diversity. It seems that, early in amniote history, certain terrestrial forms became elongate (although not necessarily small, contra Carroll, 1996) and displayed skink-like overall body proportions, similar to those of certain later microsaurs (e.g. rhynchonkids, ostodolepids and, possibly, gymnarthrids). Such proportions might be interpreted as adaptations to a burrowing life-style, at least in some of the above-mentioned taxa. (e) The status of microsaurs Our scheme of relationships supports microsaur paraphyly (Fig. 5; see also lower mid-right sector in Fig. 6), but differs from Anderson’s (2001) analysis in the branching order of the tuditanomorph families as well as in the fact that microbrachomorphs other than brachystelechids form a clade. This clade (admittedly a poorly supported one) is the sister taxon to remaining lepospondyls. Brachystelechids are paired with a monophyletic tuditanomorph assemblage. The arrangement of tuditanomorphs mostly agrees with Schultze & Foreman’s (1981) and Milner’s (1993) hypotheses. There are, however, some differences between those studies and the present result. First, we found no evidence for a monophyletic Tuditanidae as defined by Carroll & Gaskill (1978), although this may be due, in part, to limited character choice and poor preservation of Tuditanus (Carroll & Baird, 1968). Tuditanus, Pantylidae and Asaphestera are progressively more closely related to other tuditanomorphs. Second, Hapsidopareiontidae are paraphyletic, with Saxonerpeton and Hapsidopareion as successive sister taxa to a clade including ostodolepids, rhynchonkids and gymnarthrids. Hapsidopareiontids, ostodolepids and gymnarthrids have emarginated cheeks (character 112). The greater or lesser degree of emargination in hapsidopareiontids and ostodolepids, respectively, may represent a trend towards acquisition of a secondarily closed and solid 283 cheek, like that observed in the rhynchonkid-gymnarthrid clade [possibly adaptated to a burrowing life-style (Carroll & Gaskill, 1978; Milner, 1993)]. Among gymnarthrids, Cardiocephalus shows a rather shallow emargination, the presence of which was disputed by Carroll & Gaskill (1978). The configuration of the cheek region of Cardiocephalus is comparable to that of ostodolepids. ACCTRAN optimization shows that the emargination of Cardiocephalus is convergent with that of hapsidopareiontids and ostodolepids. Cheek morphology varies in gymnarthrids. Thus, Euryodus and Sparodus display a conventional cheek with a straight ventral margin, whereas Pariotichus and Cardiocephalus show a shallow cheek embayment (Carroll & Gaskill, 1978; Schultze & Foreman, 1981; Carroll et al., 1998). ( f ) The aı̈stopod-nectridean clade The present analysis retrieves a scincosaurid-diplocaulid clade (Fig. 5; see also upper mid-right sector in Fig. 6) – in agreement with A. C. Milner’s (1980) and Milner’s (1993) hypotheses – and an aı̈stopodurocordylid clade. Only two extra steps are required to reconstruct a monophyletic Nectridea or to place aı̈stopods as sister taxon to diplocaulids. Tree length increases by 11 steps when the nectridean genera are arranged as in Anderson’s (2001) analysis (constraining nectrideans to be monophyletic but without changing the position of lysorophids and aı̈stopods), and by 17 steps when lysorophids and aı̈stopods are sister group to diplocaulids [Anderson’s (2001) topology]. The present work supports in part Thomson & Bossy’s (1970) concept of Holospondyli, but forces us to explore further characters that may re-establish nectridean monophyly. An emended diagnosis of nectrideans that takes into account the position of aı̈stopods does not seem to be warranted. However, the possibility that aı̈stopod ancestry is rooted into basal nectrideans cannot be entirely ruled out. If a tree topology is reconstructed that matches Anderson’s (2001; Fig. 3) arrangement of lepospondyl taxa down to genus level (including the position of caecilians among derived tuditanomorphs), then tree length increases by 55 steps, and the tree is a significantly worse fit for the data than the most parsimonious trees (P<0.0001 for Templeton, Kishino–Hasegawa and Winning-sites tests). Similar statistical test results are obtained if a further constraint is imposed on Anderson’s (2001) topology by maintaining his arrangement of lepospondyls but forcing lissamphibian monophyly at a deeper level (i.e. with caecilians’ ancestry rooted into lepospondyls, and salientians’ and caudates’ 284 Marcello Ruta, Michael I. Coates and Donald L. J. Quicke ancestry rooted into dissorophoid temnospondyls; see also discussion in Milner, 1988, 1993). The new constraint entails 85 extra steps. The aı̈stopod-nectridean clade is supported by the following character-state changes (several of which represent optimized missing entries): 16 (c.i.=0.067; 1p0), prefrontal less than three times as long as wide; 136 (c.i.=0.091; 1p0), ectopterygoid without denticles; 138 (c.i.=0.125; 1p0), ectopterygoid with tooth row; 282 (c.i.=0.333; 1p0), absence of elongate posterodorsal flange in most trunk ribs; 286 (c.i.=0.5; 0)1), presence of extra articulations above zygapophyses in at least some trunk and caudal vertebrae; 287 (c.i.=1; 0)1), neural and haemal spines rectangular to fan-shaped in lateral view; 288 (c.i.=1; 0)1), neural and haemal spines facing each other dorsoventrally; 289 (c.i.=0.5; 0)1), haemal spines fused to caudal centra; 296 (c.i.=0.125; 0)1), neural spines of trunk vertebrae fused to centra; 311 (c.i.=0.5; 1)0), absence of capitular facets on posterior rim of vertebral midtrunk centra. The sister group relationship between aı̈stopods and urocordylids is based on five character-state changes (again, several reversals and optimized missing entries are implied within the clade): 63 (c.i.=0.25; 1)0), presence of supratemporal; 104 (c.i.=0.105; 0)1), interorbital distance smaller than maximum orbit diameter; 137 (c.i.=0.2; 1p0), ectopterygoid longer than palatine; 267 (c.i.=0.143; 0p1), presence of distinct rugose area on fourth trochanter; 290 (c.i.=1; 0p1), presence of extra articulations on haemal spines. (g) Acherontiscus is an adelospondyl As many as 14 characters indicate a sister group relationship between Acherontiscus and adelospondyls: 13 (c.i.=0.143; 0)1), nasal length less than one-third the frontal length; 38 (c.i.=0.143; 0p1), anterior margin of frontal deeply wedged between posterolateral margins of nasals; 67 (c.i.=0.25; 0p1), absence of tabular as separate ossification; 77 (c.i.=0.333; 0)1), postorbital excluded from margin of orbit; 86 (c.i.=1; 0p1), single squamosotabular in the position of squamosal and tabular; 104 (c.i.=0.105; 0p2), interorbital distance subequal to maximum orbit diameter; 107 (c.i.=0.125; 2p0), position of pineal foramen behind interparietal suture mid point; 115 (c.i.=0.286; 4)1), lateral line system on skull roof mostly enclosed, with short sections in grooves; 116 (c.i.=0.286; 4p3), mandibular canal entirely in grooves; 128 (c.i.=0.067; 0p1), palatine with denticles; 151 (c.i.=0.25; 0p1), distinct anterior digitiform process of palatal ramus of pterygoids; 167 (c.i.=0.5; 0p1), basioccipital circular and recessed; 191 (c.i.=0.143; 1p0), presence of postsplenial; 281 (c.i.=0.5; 0p1), presence of spur-like posterodorsal processes in at least some trunk ribs. Additional characters from the mandible and postcranial skeleton are currently being examined in conjunction with a revision of Acherontiscus. We found a remarkable array of similarities (proportions of several skull roof bones; vertebral and rib morphology) between Acherontiscus and adelospondyls, on one side, and colosteids, on the other. We were, therefore, interested to discover that only five additional steps are required to shift the Acherontiscus-adelospondyl clade retrieved in the original parsimony run to a stem-tetrapod position, as sister taxon to colosteids. This topology does not represent a significantly worse fit of the data than the fundamental trees (Templeton test: P=0.3532; Winning-site test: P=0.4583; Kishino–Hasegawa test: P=0.354; the results are based on comparisons between the first tree obtained from the original analysis and a constrained tree in which the clade Acherontiscus plus adelospondyls is sister group to colosteids; comparisons with additional selected trees [10, 20, 30, 40, 50, 60] also imply values of P40.05). In the light of such findings, we are exploring the significance of additional characters in evaluating the question of lepospondyl monophyly and the affinities of adelospondyls (see also below). If independently corroborated by other characters, such a hypothesis of relationships suggests that tendency towards body elongation and limb reduction/loss occurred early in the evolutionary history of tetrapods, and was acquired convergently in several crown-group lineages (e.g. microsaurs; embolomeres; urocordylids; aı̈stopods). (h) The position of lysorophids The highly specialized lysorophids share characters with one or more representatives of various lepospondyl groups, although they are generally considered to be closely related to microsaurs (Wellstead, 1991; Carroll et al., 1998). Several cranial characters of Brachydectes resemble those of Batropetes (Carroll & Gaskill, 1978; Carroll, 1991a). However, placing Batropetes as sister taxon to Brachydectes requires nine steps. This topology is only slightly worse than the most parsimonious trees, although the level of significance is not high (Templeton test: P=0.0389; Winning-sites test: P=0.0636; Kishino–Hasegawa test: P=0.0388). Nine steps are also required to place lysorophids as sister taxon to aı̈stopods as in Anderson’s (2001) analysis (again, with a low level of significance). Early tetrapod relationships revisited We were interested to compare our tree topology (Fig. 4), with particular reference to the position of lysorophids, with Laurin & Reisz’s (1997, 1999) and Laurin’s (1998a–c) preferred topology (Fig. 1 D), in which lysorophids are the closest relatives of crowngroup lissamphibians. To this purpose, we carried out three separate exercises, in which taxa were rearranged in order to match as closely as possible Laurin & Reisz’s (1997, 1999) and Laurin’s (1998 a–c) cladogram. For simplicity, tests were performed using Laurin & Reisz’s (1999) tree, since their study supersedes those of Laurin & Reisz (1997) and Laurin (1998a–c). In the first exercise, we kept intrinsic relationships within major groups mostly unaltered, but rearranged such groups according to the branching sequence favoured by Laurin & Reisz (1999). In particular, Rhynchonkos, Batropetes and Brachydectes were placed as a series of progressively more crownward plesions on the lissamphibian stem-group. The new arrangement entails 74 additional steps and is a considerably worse fit for the whole character set than the most parsimonious trees (P<0.0001 for Templeton, Winning-sites and Kishino–Hasegawa tests). In the second exercise, our taxon sample was stripped down to resemble Laurin & Reisz’s (1997, 1999) and Laurin’s (1998 a–c) taxon matrix (except for the exclusion of living groups and of the genus Tersomius; see above). All taxa belonging to a given group were kept in the analysis if such a group was represented by a supraspecific OTU in Laurin & Reisz’s (1999) dataset (e.g. Gephyrostegidae; Aı̈stopoda; Nectridea). The reduced matrix yielded 24 equally parsimonious trees at 985 steps (CI=0.3326; RI=0.648; RC=0.2296) supporting a derivation of lissamphibians from dissorophoids. Nine of these trees were compared with Laurin & Reisz’s (1999) topology. In all cases examined, their favoured branching pattern represents a considerably worse fit for the total data than the topology retrieved from the original parsimony run (significance at P<0.0001 for Templeton, Winning-sites and Kishino–Hasegawa tests). The strict consensus of the 24 fundamental trees resembles that of the original analysis (Fig. 4), except in the following features: (1) embolomeres and gephyrostegids are sister groups, as in Laurin & Reisz (1999); (2) seymouriamorphs are paraphyletic, with Seymouria and Kotlassia as successive plesions on the amniote stem-group; and (3) Westlothiana, Batropetes, Rhynchonkos, Pantylus and Brachydectes are successively more closely related to a clade including, proximodistally, Scincosaurus, diplocaulids, urocordylids and a monophyletic adelospondyl-aı̈stopod group. In the third exercise, we constrained crown-lissamphibians (without albanerpetontids) to appear as sister 285 group to lysorophids without changing the relationships of the other taxa. When we compared the resulting tree (at 22 extra steps) with the fundamental cladograms, we found significant differences (Templeton test: P=0.0008; Winning-sites test: P=0.0013; Kishino– Hasegawa test: P=0.0007). However, it is important to note that if albanerpetontids are grouped with crown-lissamphibians, and this clade is placed as sister group to lysorophids, then tree length increases by only 10 steps, and the new topology is not fundamentally different from the shortest cladograms overall (Templeton test: P=0.1736; Winning-sites test: P=0.2207; Kishino–Hasegawa test: P=0.174). As explained below (see Section VII.7), this result depends upon the unstable ‘balance’ beween different character sets that support alternative, conflicting, positions for crown-lissamphibians. Specifically, the character signal supporting the lissamphibian-dissorophoid relationship is diluted by the pervasive noise associated with a host of reversals and ‘absence’ features (especially cranial features). In fact, albanerpetontid crania, like lysorophid examples, include unusually few bones, and, despite gross morphological differences, we argue that it is the apparent shared pattern of simplification that forces these taxa together. (4) Reweighted analysis Reweighting characters by their consistency index values (best fit) yields one tree (CI=0.4068; RI=0.7666; RC=0.3279; Fig. 7) which differs from the fundamental trees (consensus in Fig. 4) in two main respects. First, Crassigyrinus and Whatcheeria are sister taxa and branch from the tetrapod stem between Tulerpeton and colosteids. We strongly suspect that this reweighting procedure reveals a likely new clade of stem-tetrapods. Crassigyrinus has long been considered as a ‘peculiar aberrant form’ (Milner et al., 1986: p. 4), but at least some of these peculiarities are now emerging as possible synapomorphies for a discrete, Whatcheerialike assemblage of archaic Mississippian forms with a plausible Late Devonian origin (see also Clack, 2002). Second, baphetids and Eucritta are sister taxa (cf. Clack, 2001) and form the most crownward stemtetrapod plesion. With Eucritta snapped to the baphetids, edopoids assume the most basal position on the lissamphibian stem, thereby emphasizing the extreme patchiness of the early lissamphibian record (Milner & Sequeira, 1998). It is also noteworthy that the reweighted analysis resolves the branching pattern of derived temnospondyls in favour of a sister group relationship between a clade encompassing Amphibamus plus Doleserpeton and a clade of branchiosaurids with 286 Marcello Ruta, Michael I. Coates and Donald L. J. Quicke Eusthenopteron foordi Panderichthys rhombolepis Ventastega curonica Acanthostega gunnari Ichthyostega stensioei Tulerpeton curtum Crassigyrinus scoticus Whatcheeria deltae Colosteus scutellatus Greererpeton burkemorani Eucritta melanolimnetes Baphetes kirkbyi Megalocephalus pachycephalus Edops craigi Chenoprosopus lewisi Cochleosaurus florensis Isodectes obtusus Neldasaurus wrightae Trimerorhachis insignis Balanerpeton woodi Dendrerpeton acadianum Eryops megacephalus Acheloma cumminsi Phonerpeton pricei Ecolsonia cutlerensis Broiliellus brevis Platyrhinops lyelli Eoscopus lockardi Micromelerpeton credneri Leptorophus tener Apateon pedestris Schoenfelderpeton prescheri Amphibamus grandiceps Doleserpeton annectens ALBANERPETONTIDAE Eocaecilia micropoda Triadobatrachus massinoti Karaurus sharovi Valdotriton gracilis Caerorhachis bairdi Eoherpeton watsoni Proterogyrinus scheelei Anthracosaurus russelli Pholiderpeton attheyi Archeria crassidisca Pholiderpeton scutigerum Bruktererpeton fiebigi Gephyrostegus bohemicus Solenodonsaurus janenschi Kotlassia prima Discosauriscus austriacus Seymouria baylorensis / sanjuanensis Diadectes absitus Limnoscelis paludis Captorhinus aguti Paleothyris acadiana Petrolacosaurus kansensis Westlothiana lizziae Batropetes fritschia Tuditanus punctulatus Pantylus cordatus Stegotretus agyrus Asaphestera intermedia Saxonerpeton geinitzi Hapsidopareion lepton Micraroter erythrogeios Pelodosotis elongatum Rhynchonkos stovalli Cardiocephalus sternbergi Euryodus primus Microbrachis pelikani Hyloplesion longicostatum Odonterpeton triangulare Brachydectes elongatus / newberryi Acherontiscus caledoniae Adelospondylus watsoni Adelogyrinus simorhynchus Dolichopareias disjectus Scincosaurus crassus Keraterpeton galvani Batrachiderpeton reticulatum Diceratosaurus brevirostris Diplocaulus magnicornis Diploceraspis burkei Ptyonius marshii Sauropleura pectinata / scalaris Urocordylus wandesfordii Lethiscus stocki Oestocephalus amphiuminum Phlegethontia linearis Fig. 7. Single tree deriving from reweighting characters by their consistency index ( c.i. ) values (best fit ). Early tetrapod relationships revisited Leptorophus as sister taxon to Apateon plus Schoenfelderpeton. This broader group is paired with albanerpetontids plus crown-lissamphibians. Progressively less crownward taxa include Micromelerpeton, Eoscopus and Platyrhinops. (5) Cranial data The results of a PAUP* analysis applied to cranial and mandibular characters were examined in order to assess the influence of different character partitions on tree topology. However, the postcranial character set could not be processed successfully, due to time- and memoryconsuming computer requirements. The all-cranial version of the data set produced 1188 fundamental trees at 1022 steps (CI=0.2485; RI=0.6787; RC=0.1707). Despite the extremely poor resolution of a strict consensus, the monophyletic status of several groups is corroborated, although the arrangement of several taxa departs significantly from that of the original analysis. An Adams consensus (Fig. 8) reveals the instability of taxa including Acherontiscus, Caerorhachis, Whatcheeria, adelospondyls, aı̈stopods, lysorophids and urocordylids. Also noteworthy is the unstable position of various ‘reptiliomorphs’ (e.g. Gephyrostegus, Seymouria, Solenodonsaurus, Kotlassia and Limnoscelis). On the lissamphibian stem, the relationships of the most crownward temnospondyls differ from those of the original parsimony run in that Broiliellus is sister taxon to crown-lissamphibians. Progressively less crownward taxa include: (1) a clade formed by Amphibamus and Doleserpeton; (2) Platyrhinops; (3) a clade formed by Eoscopus and Micromelerpeton as successive sister taxa to branchiosaurids. As in the reweighted analysis (using the complete character set; Fig. 7), the cranial data support the polyphyly of amphibamids. The most striking result of the cranial analysis is the identification of two distinct monophyletic groups of lepospondyls in several trees. The first group consists of microsaurs, placed as sister taxon to Westlothiana on the amniote stem, in agreement with the original parsimony run. The second group consists of a heterogeneous assemblage of aı̈stopods, lysorophids, adelospondyls and nectrideans, forming the sister group to colosteids in at least some trees. In the first group, pantylids are paired with gymnarthrids, as in Anderson’s (2001) cladogram, whereas Odonterpeton and Batropetes form a clade between Saxonerpeton and Hapsidopareion. The pantylid–gymnarthrid sister group relationship is supported by their similar tooth morphology and by their general skull proportions. The match between the cladogenetic event sequence and the stratigraphical appearance of microsaur families is better than that obtained when using the 287 Eusthenopteron foordi Panderichthys rhombolepis Ventastega curonica Ichthyostega stensioei Acanthostega gunnari Tulerpeton curtum Colosteus scutellatus Greererpeton burkemorani Acherontiscus caledoniae Brachydectes elongatus / newberryi Adelospondylus watsoni Adelogyrinus simorhynchus Dolichopareias disjectus Ptyonius marshii Sauropleura pectinata / scalaris Urocordylus wandesfordii Lethiscus stocki Oestocephalus amphiuminum Phlegethontia linearis Scincosaurus crassus Keraterpeton galvani Diceratosaurus brevirostris Batrachiderpeton reticulatum Diplocaulus magnicornis Diploceraspis burkei Crassigyrinus scoticus Whatcheeria deltae Baphetes kirkbyi Megalocephalus pachycephalus Eucritta melanolimnetes Caerorhachis bairdi Edops craigi Chenoprosopus lewisi Cochleosaurus florensis Isodectes obtusus Neldasaurus wrightae Trimerorhachis insignis Balanerpeton woodi Dendrerpeton acadianum Eryops megacephalus Acheloma cumminsi Phonerpeton pricei Ecolsonia cutlerensis Eoscopus lockardi Micromelerpeton credneri Apateon pedestris Leptorophus tener Schoenfelderpeton prescheri Platyrhinops lyelli Amphibamus grandiceps Doleserpeton annectens Broiliellus brevis ALBANERPETONTIDAE Eocaecilia micropoda Karaurus sharovi Triadobatrachus massinoti Valdotriton gracilis Eoherpeton watsoni Pholiderpeton scutigerum Archeria crassidisca Proterogyrinus scheelei Anthracosaurus russelli Pholiderpeton attheyi Bruktererpeton fiebigi Gephyrostegus bohemicus Discosauriscus austriacus Seymouria baylorensis / sanjuanensis Kotlassia prima Solenodonsaurus janenschi Diadectes absitus Limnoscelis paludis Captorhinus aguti Paleothyris acadiana Petrolacosaurus kansensis Westlothiana lizziae Hyloplesion longicostatum Microbrachis pelikani Tuditanus punctulatus Asaphestera intermedia Saxonerpeton geinitzi Batropetes fritschia Odonterpeton triangulare Hapsidopareion lepton Micraroter erythrogeios Pelodosotis elongatum Rhynchonkos stovalli Cardiocephalus sternbergi Euryodus primus Pantylus cordatus Stegotretus agyrus Fig. 8. Adams consensus of 1188 equally parsimonious trees obtained after removal of postcranial characters. total data set. The pairing of Odonterpeton and Batropetes is rather unexpected. However, characters in common to both genera are the robust aspect of the mandible and the sloping of the posterior cheek margin. The material of Odonterpeton is imperfectly known and requires thorough redescription. A stem-tetrapod position for various lepospondyls, such as those within the second group, has been proposed by Milner (1993), who speculated that nectrideans might be progenetically dwarf relatives of 288 Marcello Ruta, Michael I. Coates and Donald L. J. Quicke colosteids. The large size of certain long-horned diplocaulids would, we assume, result from subsequent peramorphosis. Body elongation, increase in vertebral count and morphological uniformity, and limb reduction/loss are often associated with swimming and/ or burrowing lifestyles. Aı̈stopods may represent the very nadir of these morphological trends. We explored further the effects of placing lepospondyls other than microsaurs in a stem-tetrapod position (see also above for a discussion of adelospondyls). If the adelospondyls-nectrideans-aı̈stopods clade deriving from the original parsimony run (i.e. using the complete set of characters) is grafted to colosteids in order to match the results of the all-cranial analysis, the resulting tree topology is found to be only a slightly worse fit for the data than the most parsimonious cladograms overall (Templeton test: P=0.0197; Winning-sites test: P=0.01; Kishino–Hasegawa test: P=0.0224; the results are based on comparisons between all trees obtained from the original analysis and a constrained tree forcing lepospondyls other than microsaurs into a stem-tetrapod position). The hypothesis that some lepospondyl lineages diversified early in tetrapod history, before the lissamphibian-amniote split, cannot be entirely ruled out. However, this hypothesis requires independent testing using several new characters, and will be dealt with elsewhere. (6) Deletion of lower jaw characters To assess the impact of lower jaw morphology on cladogram topology, we ran a cladistic analysis excluding mandibular characters. The strict consensus of 2160 trees at 1220 steps (CI=0.2322; RI=0.6737; RC=0.1607; Fig. 9) is only slightly less resolved than that based on the total data set. It differs from the latter in the following respects: (1) Ventastega and Whatcheeria are sister taxa and form the most crownward plesion on the tetrapod stem-group; (2) Eucritta plus baphetids form the most basal clade in the lissamphibian stem-group; (3) Edops and cochleosaurids form an unresolved node with higher temnospondyls; (4) Balanerpeton and trimerorhachoids likewise form an unresolved node with higher temnospondyls; (5) crown-lissamphibians are more deeply nested into the derived portion of the temnospondyl tree, and form the sister group to a fully resolved clade in which amphibamids are a paraphyetic assemblage relative to a micromelerpetontid-branchiosaurid clade; (6) within crown-lissamphibians, relationships are resolved in favour of a salientian-caudate clade, with Karaurus and Valdotriton as sister taxa; however, albenerpetontids Eusthenopteron foordi Panderichthys rhombolepis Acanthostega gunnari Ichthyostega stensioei Tulerpeton curtum Colosteus scutellatus Greererpeton burkemorani Crassigyrinus scoticus Ventastega curonica Whatcheeria deltae Eucritta melanolimnetes Baphetes kirkbyi Megalocephalus pachycephalus Edops craigi Chenoprosopus lewisi Cochleosaurus florensis Isodectes obtusus Neldasaurus wrightae Trimerorhachis insignis Balanerpeton woodi Dendrerpeton acadianum Eryops megacephalus Acheloma cumminsi Phonerpeton pricei Ecolsonia cutlerensis Broiliellus brevis Platyrhinops lyelli Eoscopus lockardi Amphibamus grandiceps Doleserpeton annectens Micromelerpeton credneri Apateon pedestris Leptorophus tener Schoenfelderpeton prescheri ALBANERPETONTIDAE Eocaecilia micropoda Triadobatrachus massinoti Karaurus sharovi Valdotriton gracilis Caerorhachis bairdi Eoherpeton watsoni Archeria crassidisca Pholiderpeton scutigerum Proterogyrinus scheelei Anthracosaurus russelli Pholiderpeton attheyi Bruktererpeton fiebigi Gephyrostegus bohemicus Solenodonsaurus janenschi Kotlassia prima Discosauriscus austriacus Seymouria baylorensis/sanjuanensis Diadectes absitus Limnoscelis paludis Captorhinus aguti Paleothyris acadiana Petrolacosaurus kansensis Westlothiana lizziae Asaphestera intermedia Batropetes fritschia Cardiocephalus sternbergi Euryodus primus Hapsidopareion lepton Rhynchonkos stovalli Saxonerpeton geinitzi Tuditanus punctulatus Micraroter erythrogeios Pelodosotis elongatum Pantylus cordatus Stegotretus agyrus Microbrachis pelikani Hyloplesion longicostatum Odonterpeton triangulare Brachydectes elongatus/newberryi Acherontiscus caledoniae Adelospondylus watsoni Adelogyrinus simorhynchus Dolichopareias disjectus Scincosaurus crassus Keraterpeton galvani Batrachiderpeton reticulatum Diceratosaurus brevirostris Diplocaulus magnicornis Diploceraspis burkei Ptyonius marshii Sauropleura pectinata/scalaris Urocordylus wandesfordii Lethiscus stocki Oestocephalus amphiuminum Phlegethontia linearis Fig. 9. Strict consensus of 2160 equally parsimonious trees obtained after removal of lower jaw characters. and Eocaecilia are collapsed in a polytomy with the remaining lissamphibians; and (7) most tuditanomorphs are collapsed in a large polytomy, except for ostodolepids and pantylids; loss of resolution among tuditanomorphs is due to the unstable positions of Batropetes, Hapsidopareion, Rhynchonkos and Tuditanus. Ahlberg & Clack (1998) have recently discussed the results of a cladistic analysis of early tetrapods based on lower jaw features (see also Daeschler, 2000). The relationships of several taxa in their work differ Early tetrapod relationships revisited substantially from those of previous analyses. One of the most unexpected results is the fact that such traditional groups as temnospondyls and ‘anthracosaurs’ appear as polyphyletic arrays of taxa. The conflict between the phylogenetic signal provided by the lower jaw and the cladogenetic pattern based on other data may be rooted into the paucity of mandibular characters (see also comments in Ruta et al., 2001). Although it is possible to identify, as Ahlberg & Clack (1998) did, a series of morphological trends affecting the evolution of lower jaws in passing from stem-tetrapods to basal crown-tetrapods, the degree of character resolution may be insufficient to yield hypotheses of relationships (both between and within groups) that match those based on other skeletal features. Thus, while a host of cranial and postcranial features support a single origin for temnospondyls, their lower jaws change little within various lineages in this group. Differences between such lineages are most prominent in the relative size and position of Meckelian foramina, and in the proportions of infradentaries and coronoids [see Schoch & Milner (2000), Yates & Warren (2000) and Damiani (2001)]. The observation that certain primitive characters (e.g. parasymphysial plate) are retained in a variety of otherwise very distinctive groups (e.g. baphetids, colosteids, some embolomeres) suggests that the lower jaw underwent modifications at a slower rate than other parts of the skeleton. Interestingly, such modifications appear to be largely decoupled from changes that affected the skull roof, palate and postcranium both in taxa spanning the fish–tetrapod transition and in some basal members of the crowngroup. At higher levels of the tetrapod hierarchy, and especially among amniote-related taxa, important modifications are clustered consistently around specific cladogram nodes (e.g. number and proportions of coronoids; extensions of mesial laminae of infradentary bones; enlargement and reduction in number of Meckelian foramina; decrease in the curvature of the posterior two-thirds of the lower margin). As Ahlberg & Clack (1998: p. 42) pointed out, ‘… it is curious to observe that the trends towards reduction of the coronoids and the endoskeletal components of the jaw continue in the synapsids, but not the other amniote lineages. A modern lizard or crocodile jaw is still in most respects comparable to that of the Early Permian Eocaptorhinus’. On the lissamphibian stem, on the other hand, modifications were less drastic than in the amniote stem – hence the overall similarities between the lower jaws of various basal and derived temnospondyls (regardless of absolute size). Furthermore, although the lower jaws of several crown-lissamphibians are highly derived relative to those of their Palaeozoic 289 counterparts (Ahlberg & Clack, 1998), Schoch (1998) has shown that identification of homologous features is possible when embryological data are combined with sequences of growth stages recovered from the fossil record. Intriguingly, Schoch’s (1992) analysis of morphological changes in the development of two different species of Apateon reveals striking similarities between this dissorophoid and certain primitive salamanders, e.g. at the level of the parasphenoid, quadrate ramus of pterygoid and vomer. (7) Reverse weighting Trueman’s (1998) reverse weighting procedure assists searches for conflicting signals within data sets. The method is based on successive removals of unambiguous synapomorphies (i.e. characters with c.i.=1) following parsimony runs. Synapomorphy stripping should reveal residual or masked phylogenetic signal within the remaining character set. We note that PAUP* calculates c.i. values on the basis of each individual tree, and not on performance of characters across the entire tree set (providing that several trees are found). It is thus possible that characters with a c.i. of 1 may support nodes that do not occur in a strict consensus. Therefore, we propose that such characters should be retained for subsequent rounds of reverse weighting. While these characters perform as if signalconsistent within a particular tree, they exhibit homoplastic distribution across the entire tree set. Consequently, they remain a source of alternative phylogenetic patterns, and any list of characters with a c.i. of 1 should be plotted on a strict consensus of the entire tree set (including the particular tree from which the c.i.’s were obtained). This seems to be a more discriminatory procedure than that applied by other workers (see Rieppel, 2000). Fifty-five characters identified as unambiguous synapomorphies at the end of the original parsimony run were excluded (1, 4, 8, 19, 22, 26, 31, 52, 54, 55, 65, 72, 76, 86, 87, 92, 99, 108, 111, 152, 157, 161, 162, 164, 166, 185, 199, 211, 217, 223, 225, 226, 227, 236, 241, 260, 263, 264, 271, 274, 275, 277, 278, 283, 287, 288, 290, 291, 305, 307, 308, 309, 310, 316, 318). The new, reduced data set yielded 1608 trees at 1341 steps (CI=0.2266; RI=0.6543; RC=0.1512), a strict consensus of which is very poorly resolved. The most important feature of these trees is the fact that crownlissamphibians and albanerpetontids now appear to be nested within lepospondyls. Remaining portions of the tree, however, remain largely unchanged. A selected tree shows that the position of lissamphibians is accounted for by reversals as well as by optimizations of 290 Marcello Ruta, Michael I. Coates and Donald L. J. Quicke missing characters. An agreement subtree shows that albanerpetontids plus crown-lissamphibians maintain the same relationships with adelospondyls in all trees, with urocordylids, diplocaulids, Scincosaurus, pantylids and a heterogeneous microsaur clade (with microbrachomorphs nested within remaining tuditanomorphs) as progressively more distantly related taxa. Also evident in the agreement subtree is the paraphyletic status of gephyrostegids, seymouriamorphs and diadectomorphs, and the fact that edopoids branch from the temnospondyl tree between Balanerpeton and eryopoids. A further five characters (167, 210, 301, 302, 303) were identified as unambiguous synapomorphies in a selected tree obtained from the reduced data set, but no further analysis was carried out due to computation time (the further five characters are unambiguous synapomorphies in some distal portions of the cladogram, and their removal is likely to produce further decrease in the resolution of a strict consensus). Deep nodes in the strict consensus appear to be entirely supported by homoplastic changes. Character reversals and optimized missing entries have a profound effect on the analysis. They introduce diffused ‘noise’ which accumulates when sets of ‘presence’ characters, such as certain lissamphibian-temnospondyl synapomorphies, are removed. These results show that evidence for a lissamphibian-dissorophoid clade, based on the whole data set, is consistently stronger than the alternative hypothesis of a sister group relationship between lissamphibians and lepospondyls. (8) Chronology A minimum hypothesis of the timing of the principal events in tetrapod phylogeny, as implied by the cladogram in Figs 5 and 6, is shown in Fig. 10. The date of the crown-group node is effectively pegged by the occurrence of the earliest known aı̈stopod Lethiscus (Wellstead, 1982). This material is slightly older than the late Viséan East Kirkton locality (Rolfe, Clarkson & Panchen, 1994), and thus in excess of 338 million years before present. A quick inspection of Fig. 10 shows a proliferation of taxa during the Viséan, but this pattern results from, on the one hand, the problematic nature of the few known scattered remains of Tournaisian tetrapods (Clack & Carroll, 2000), and on the other, the disproportionate influence of key faunas, such as East Kirkton (Rolfe et al., 1994). It is noteworthy that all recent phylogenetic analyses identify basal crown taxa within the diversity of mid-Viséan tetrapods, and it therefore appears that the proposed date for the crown-node (on the basis of morphological data) is unusually robust. Subsequent changes appear more likely to result from new fossil discoveries than from new or alternative phylogenies (cf. Laurin, 1998 b; Paton et al., 1999). The inferred timing of the lissamphibian-amniote split falls some 20 million years short of the estimate derived from molecular data (Hedges, 2001); the significance or otherwise of this is dicussed in greater detail elsewhere (Ruta & Coates, in press). VIII. FUTURE RESEARCH Despite the increase in early tetrapod data, numerous aspects of their evolutionary history remain obscure. The earliest phases of diversification of several groups, including those with an extensive fossil record (e.g. temnospondyls), are poorly understood. Furthermore, the earliest known members of various taxa display almost complete arrays of apomorphies found in later representatives (e.g. microsaurs; nectrideans; Carroll & Gaskill, 1978; A. C. Milner, 1980; Milner, 1993; Carroll et al., 1998), thus providing little information on their ancestry and sister taxon relationships. Some tetrapods show puzzling combinations of characters shared with members of two or more different groups (e.g. Caeorhachis; Crassigyrinus; Eucritta; Whatcheeria; Lombard & Bolt, 1995; Clack, 1996, 1998a, c, 2001; Bolt & Lombard, 2000; Ruta et al., 2001). Still others are extremely specialized from their first appearance in the fossil record (e.g. adelospondyls; aı̈stopods; lysorophids; Wellstead, 1982, 1991; Carroll, 1998; Carroll et al., 1998). All of these observations, coupled with the vagaries of fossil preservation, suggest no imminent, unequivocal, solution to the problem of evaluating the phylogenetic position of several crucial taxa. Recent research is beginning to shed new light on the anatomy and relationships of rare and problematic forms, such as lepospondyls (Carroll, 1998; Carroll et al., 1998; Anderson, 2001, in press; Anderson et al., 2001). Several issues related to lepospondyl interrelationships are likely to undergo extensive revision in the near future. Published analyses of lepospondyls reveal a disconcerting lack of agreement, to the point that almost any pattern of relationships has been proposed (Smithson, 1985; Panchen & Smithson, 1987; Milner, 1993; Carroll, 1995; Carroll & Chorn, 1995; Laurin & Reisz, 1997, 1999; Laurin, 1998a–c; Anderson, 2001). The most challenging task posed by lepospondyl studies consists of identifying good symapomorphies with other major groups. Comparisons between the results generated by different character set partitions (see above) show that much work is still needed to unravel the confounding signal produced by convergence. However, these comparisons already suggest that at least some of Gelasian Piacenzian Zanclean Messinian Tortonian Serravallian Langhian Burdigalian Aquitanian Chattian Rupelian Priabonian Barthonian Lutetian Ypresian Thanetian Danian Maastrichtian Coniacian Santonian Campanian Albian Cenomanian Turonian Hauterivian Aptian Barremian Valanginian Berriasian Tithonian Kimmeridgian Oxfordian Callovian Bathonian Bajocian Aalenian Toarcian Pliensbachian Sinemurian Hettangian Rhetian Norian Carnian Ladinian Anisian Olenekian Induan Changhsingian Wuchiapigian Capitanian Wordian 1.75 3.4 5.3 7.3 11 14.3 15.8 20.3 23.5 28 33.7 37 40 46 53 65 72 83 87 88 92 96 108 113 117 123 131 135 141 146 154 160 164 170 175 184 191 200 203 220 230 233 240 250 295 Colosteus scutellatus Baphetes kirkbyi Megalocephalus pachycephalus Edops craigi Cochleosaurus florensis Chenoprosopus lewisi Isodectes obtusus Neldasaurus wrightae Trimerorhachis insignis Balanerpeton woodi Dendrerpeton acadianum Eryops megacephalus Acheloma cumminsi Phonerpeton pricei Ecolsonia cutlerensis Broiliellus brevis Platyrhinops lyelli Eoscopus lockardi Micromelerpeton credneri Apateon pedestris Leptorophus tener Schoenfelderpeton prescheri Amphibamus grandiceps Caerorhachis bairdi Doleserpeton annectens Eoherpeton watsoni Eucritta melanolimnetes Proterogyrinus scheelei Anthracosaurus russelli Pholiderpeton attheyi ALBANERPETONTIDAE Valdotriton gracilis Karaurus sharovi 291 Triadobatrachus massinoti Archeria crassidisca Pholiderpeton scutigerum Bruktererpeton fiebigi Gephyrostegus bohemicus Solenodonsaurus janenschi Kotlassia prima Discosauriscus austriacus Seymouria baylorensis / sanjuanensis Diadectes absitus Limnoscelis paludis Captorhinus aguti Paleothyris acadiana Petrolacosaurus kansensis Westlothiana lizziae Batropetes fritschia Tuditanus punctulatus Pantylus chordatus Stegotretus agyrus Asaphestera intermedia Saxonerpeton geinitzi Hapsidopareion lepton Micraroter erythrogeios Pelodosotis elongatum Rhynchonkos stovalli Cardiocephalus sternbergi Euryodus primus Microbrachis pelikani Hyloplesion longicostatum Odonterpeton triangulare Brachydectes elongatus / newberryi Acherontiscus caledoniae Adelospondylus watsoni Adelogyrinus simorhynchus Scincosaurus crassus Dolichopareias disjectus Keraterpeton galvani Batrachiderpeton reticulatum Diceratosaurus brevirostris Diplocaulus magnicornis Diploceraspis burkei Ptyonius marshii Sauropleura pectinata / scalaris Urocordylus wandesfordii Oestocephalus amphiuminum Phlegethontia linearis Lethiscus stocki Eocaecilia micropoda Early tetrapod relationships revisited PLIOCENE MIOCENE OLIGOCENE EOCENE PALEOCENE LATE CRETACEOUS EARLY CRETACEOUS LATE JURASSIC MIDDLE JURASSIC EARLY JURASSIC LATE TRIASSIC MIDDLE TRIASSIC EARLY TRIASSIC LATE PERMIAN Roadian Kungurian Artinskian Sakmarian 325 345 355 370 375 380 390 400 410 Greererpeton burkemorani Crassigyrinus scoticus Whatcheeria deltae MIDDLE PERMIAN EARLY PERMIAN Gzhelian Asselian Kazimovian Bashkirian Serpukhovian Visean Tournaisian Famennian Frasnian Givetian Eifelian Emsian Pragian Lochkovian PENNSYLVANIAN Moscovian MISSISSIPPIAN LATE DEVONIAN MIDDLE DEVONIAN EARLY DEVONIAN (Clack & Finney, 1997; Clack, 2002). Future work on the earliest known Carboniferous faunas (Thulborn et al., 1996; Clack & Carroll, 2000) may provide further insight into the pattern of character distribution near the base of the crown-group. Fig. 10. The selected tree shown in Fig. 5 plotted on a stratigraphic timescale, modified from Ruta & Coates (in press ). the hypothesized sister group relationships may be correct (e.g. Thomson & Bossy, 1970; Wellstead, 1991). Commonly held assumptions about the polarity of several characters in the most crownward part of the tetrapod stem-group are challenged by new discoveries Eusthenopteron foordi Panderichthys rhombolepis Ventastega curonica Acanthostega gunnari Ichthyostega stensioei Tulerpeton curtum 292 Marcello Ruta, Michael I. Coates and Donald L. J. Quicke Theories of lissamphibian origin, or origins, continue to be the source of considerable conflict. As Carroll & Bolt (2001: p.38A) have noted, the problem with certain recently published phylogenetic hypotheses (e.g. Laurin & Reisz, 1997, 1999; Laurin, 1998a–c) is that they ‘… support a monophyletic origin of the modern [lissamphibian] orders, but imply a large number of biologically improbable character reversals and an immediate common ancestor with very few characters shared with any of the derived orders’ (see also Carroll, 2001). Investigations of patterns of skeletal ossification in several derived temnospondyls (notably, dissorophoids), and comparisons with developmental data from modern as well as extinct lissamphibians (e.g. Schoch, 1992, 1995, 1998; Carroll, Kunst & Albright, 1999; Boy & Sues, 2000; Carroll, 2001; Carroll & Bolt, 2001; Chipman & Tchernov, 2002) provide a large amount of new morphological information that is likely to be incorporated in comprehensive, morphologybased cladistic analyses. Particularly challenging topics include the single or multiple origins of lissamphibians from among dissorophoids (Bolt, 1969, 1977, 1979, 1991; Lombard & Bolt, 1979; Bolt & Lombard, 1985; Milner, 1988, 1993; Trueb & Cloutier, 1991; Carroll, 2001; Carroll & Bolt, 2001; Dr A. R. Milner, personal communication to M. Ruta, 2001), and the ancestry of caecilians (review in Carroll, 2000). Interrelationships of various putative stem-amniote groups remain controversial. The impact on phylogenetic reconstruction of several Permian and late Carboniferous forms (e.g. chroniosuchids; kotlassiids; nycteroleterids) cannot be properly evaluated without a detailed revision of their osteology. The morphological gap between diadectomorphs and primitive crown-amniotes is puzzling, despite the fact that several characters of the former foreshadow the condition of basal crown-amniotes (Berman & Sumida, 1990; Sumida & Lombard, 1991; Berman et al., 1992, 1998; Sumida et al., 1992; Sumida, 1997; Berman, 2000). Future research targeted at the most primitive ‘anthracosauroids’ (e.g. Clack, 1994 c; Smithson, 1994) may yield new data on the early radiation of ‘reptiliomorphs’, with particular emphasis on the skeletal modifications that accompanied the transition from semiaquatic/aquatic (e.g. embolomeres) to more fully terrestrial forms (e.g. gephyrostegids; Solenodonsaurus). The position of seymouriamorphs is interesting in this context. Although some of their characters appear to be more primitive than those observed in certain early terrestrial ‘reptiliomorphs’, this may simply reflect the paedomorphic or juvenile condition of some forms (e.g. discosauriscids). On balance, similarites between diadectomorphs and seymouriamorphs are striking, and point to a more derived placement of the latter on the amniote stem relative to embolomeres and gephyrostegids. IX. CONCLUSIONS (1) In an attempt to investigate conflicts between the most widely discussed hypotheses of early tetrapod relationships, we assembled a large data matrix encompassing character sets from each published study. This has not been a literature-based exercise. We have incorporated, where possible, original observations of numerous taxa spread throughout the major clades, as well as data from the redescriptions of Tulerpeton (Lebedev & Coates, 1995), Acanthostega (Coates, 1996) and Caerorhachis (Ruta et al., 2001). The results of our new analysis indicate a deep phylogenetic split between lissamphibian- and amniote-related groups. A series of Lower Carboniferous early tetrapods branch from the tetrapod stem. These include colosteids, Crassigyrinus, Whatcheeria and baphetids, in order of increasing proximity to the crown-group. Some of these taxa (notably Crassigyrinus, Whatcheeria and baphetids) have been allied to amniotes in certain analyses, but their ‘reptiliomorph’ characters are now emerging as generalized tetrapod features (see also Clack, 2002). The tetrapod crown-group is bracketed at its base by Eucritta and Caerorhachis, notorious for their debated affinities (Holmes & Carroll, 1977; Milner & Sequeira, 1994; Coates, 1996; Clack, 1998 a; Holmes, 2000) and for their curious mixture of features otherwise regarded as unique to mutually exclusive groups, such as baphetids, embolomeres and temnospondyls (Clack, 2001; Ruta et al., 2001). (2) Despite the large number of ‘absence’ features seemingly shared by crown-lissamphibians and certain lepospondyls (notably, lysorophids), a sister group relationship between lissamphibians and dissorophoid temnospondyls best accounts for the distribution of putative synapomorphies in these two groups. Several alternative hypotheses of lissamphibian ancestry imply a worse fit of the total data. However, no specific dissorophoid can be identified as the nearest relative of crown-lissamphibians. Rather, these are paired with a heterogeneous clade including amphibamids, micromelerpetontids and branchiosaurids. Such an arrangement probably results from the fact that different lineages of dissorophoids approached the condition of basal lissamphibians independently (Milner, 1993). The pattern of sister group relationships in the crownward part of the temnospondyl branch re-emphasizes Early tetrapod relationships revisited the importance of dissorophoids in the debate about lissamphibian origin. The evolutionary implications of these results have yet to be explored in depth. Initially recognized patterns include the morphological conservatism of stem-lissamphibians relative to the diversity of stem-amniotes. Comparisons between dissorophoids and various living and extinct caudates show that the latter appear as generalized in their postcranial and cranial features as their supposed ancestors from among derived temnospondyls (Bolt, 1969, 1977, 1979, 1991; Milner, 1988, 1990, 1993, 2000; Trueb & Cloutier, 1991; Schoch, 1992, 1995, 1998; Carroll, 2001). Most importantly, we have tried to show that comprehensive treatments of the available evidence from the fossil record, rather than the use of just some key characters, can overturn hypotheses of relationships based on clusters of ‘absence’ features. Alternative patterns of relationships, based on a large proportion of such features (e.g. Laurin & Reisz, 1997, 1999; Laurin, 1998 a–c), appear to be less informative with regards to the ancestry of some or all of the lissamphibian orders. As explained by Carroll (2001, p. 1207), Laurin & Reisz’s (1997, 1999) and Laurin’s (1998a–c) hypothesized sister group relationship between lysorophids and crown-lissamphibians relies upon some characters that lysorophids share ‘… with each of the three groups [of lissamphibians]: greatly elongate body with much reduced limbs in common with the earliest known caecilian, a fenestrate skull, vaguely comparable with those of frogs and salamanders, and loss of many similar skull bones, but the total configuration is that of a chimaera that has no unique derived characters in common with any of the individual orders’. (3) The branching sequence of stem-group amniotes reveals a coherent series of internested character-state changes leading up to the condition of basal crownamniotes. In particular, changes in body proportions account for a progressive tendency towards the acquisition of terrestrial habits (e.g. gephyrostegids; some seymouriamorphs; diadectomorphs). The interrelationships of primitive amniotes are largely in agreement with the conventional view that ‘anthracosaurs’ (i.e. embolomeres and gephyrostegids), seymouriamorphs and diadectomorphs are successively more closely related to crown-amniotes. However, the analysis also shows Westlothiana and lepospondyls to be amniote relatives, although support for lepospondyl monophyly is weak. (4) Cranial data are in conflict with total data with regard to the position of lepospondyls other than microsaurs. The placement of microsaurs on the amniote stem persists even when postcranial data are omitted. 293 The relationships of remaining lepospondyls change significantly under these conditions, since they are relocated on the tetrapod stem, as sister group to colosteids. Such an arrangement is not significantly worse than the topology based on the whole character suite. Importantly, it emphasizes similarities between colosteids and various lepospondyls (notably adelospondyls). Additional characters may identify some of the lepospondyls as stem-tetrapod offshoots. X. ACKNOWLEDGEMENTS For access to collections, exchange of photographic, artistic and written documentation, use of facilities and many invaluable discussions, we thank : Drs Andrew Milner and Sandra Sequeira ( School of Biological and Chemical Sciences, Birkbeck College, University of London, UK ) ; Dr Jason Anderson (Department of Zoology, University of Toronto, Mississauga, Canada ) ; Dr David Berman ( Section of Vertebrate Paleontology, Carnegie Museum of Natural History, Pittsburgh, USA) ; Dr Robert Carroll ( Redpath Museum, McGill University, Montreal, Quebec, Canada ); Dr Helen Chatterjee ( Grant Museum of Zoology, University College London, UK ); Dr Jenny Clack ( University Museum of Zoology, University of Cambridge, UK ) ; Dr Ross Damiani ( Bernard Price Institute for Palaeontological Research, University of the Witwatersrand, South Africa ); Dr James Gardner (Royal Tyrrell Museum of Palaeontology, Drumheller, Alberta, Canada ) ; Dr Jonathan Jeffery ( Institute of Evolutionary and Ecological Sciences, University of Leiden, The Netherlands ) ; Dr Jozef Klembara ( Department of Ecology, Comenius University, Bratislava, Slovakian Republic) ; Dr Angela Milner and Ms Sandra Chapman ( Palaeontology Department, The Natural History Museum, London, UK ) ; Dr Sean Modesto ( Section of Vertebrate Paleontology, Carnegie Museum of Natural History, Pittsburgh, USA ) ; Dr Roberta Paton ( National Museums of Scotland, Edinburgh, UK) ; Prof. Stuart Sumida ( Department of Biology, California State University at San Bernardino, USA ) ; Dr Anne Warren ( Department of Zoology, La Trobe University, Melbourne, Australia ) ; Dr Adam Yates ( Department of Earth Sciences, University of Bristol, UK ). We are particularly grateful to Mr Jack Conrad ( Department of Organismal Biology and Anatomy, University of Chicago, USA ), Dr Andrew Milner and Prof. Stuart Sumida for their constructive criticism of earlier drafts of this work. We are indebted to Drs Jason Anderson and Jenny Clack for their numerous insightful comments and editorial remarks and for saving us from some major howlers. Dr Anderson kindly supplied a pre-publication version of his work in press on phlegethontiid aı̈stopods and shared with the senior author invaluable information. This work was funded by BBSRC Advanced Research Fellowship no. 31/AF/13042 awarded to Michael I. Coates. 294 Marcello Ruta, Michael I. Coates and Donald L. J. Quicke XI. REFERENCES AHLBERG, P. E. ( 1991 ). Tetrapod or near-tetrapod fossils from the Upper Devonian of Scotland. Nature 354, 298–301. AHLBERG, P. E. ( 1995 ). Elginerpeton pancheni and the earliest tetrapod clade. Nature 373, 420– 425. AHLBERG, P. E. ( 1998 ). Postcranial stem tetrapod remains from the Devonian of Scat Craig, Morayshire, Scotland. Zoological Journal of the Linnean Society 122, 99–141. AHLBERG, P. E. & CLACK, J. A. ( 1998 ). Lower jaws, lower tetrapods – a review based on the Devonian genus Acanthostega. Transactions of the Royal Society of Edinburgh : Earth Sciences 89, 11– 46. AHLBERG, P. E., CLACK, J. A. & LUKSEVICS, E. ( 1996 ). Rapid braincase evolution between Panderichthys and the earliest tetrapods. Nature 381, 61–64. AHLBERG, P. E. & JOHANSON, Z. ( 1998 ). Osteolepiforms and the ancestry of tetrapods. Nature 395, 792–794. AHLBERG, P. E., LUKSEVICS, E. & LEBEDEV, O. ( 1994 ). The first tetrapod finds from the Devonian ( Upper Famennian ) of Latvia. Philosophical Transactions of the Royal Society of London, Series B 343, 303–328. AHLBERG, P. E., LUKSEVICS, E. & MARK-KURIK, E. ( 2000 ). A neartetrapod from the Baltic Middle Devonian. Palaeontology 43, 533–548. AHLBERG, P. E. & MILNER, A. R. ( 1994 ). The origin and early diversification of tetrapods. Nature 368, 507–514. AMALITSKY, V. P. ( 1921 ). Seymouridae. North Dvina Excavations of Prof. V. P. Amalitsky 2, 1–14. ANDERSON, J. S. ( 2001 ). The phylogenetic trunk : maximal inclusion of taxa with missing data in an analysis of the Lepospondyli ( Vertebrata, Tetrapoda ). Systematic Biology 50, 170–193. ANDERSON, J. S. ( in press ). Revision of the aı̈stopod genus Phlegethontia ( Tetrapoda : Lepospondyli ). Journal of Paleontology. ANDERSON, J. S., CARROLL, R. L. & ROWE, T. ( 2001 ). [Abstract] New information on Lethiscus stocki ( Tetrapoda, Lepospondyli, Aı̈stopoda ) from high resolution computerized tomography and a phylogenetic analysis of Aı̈stopoda. Journal of Vertebrate Paleontology 21 ( Suppl. 3 ), 28A. ANDREWS, S. M. & CARROLL, R. L. ( 1991 ). The order Adelospondyli : Carboniferous lepospondyl amphibians. Transactions of the Royal Society of Edinburgh : Earth Sciences 82, 239–275. ANDREWS, S. M. & WESTOLL, T. S. ( 1970 ). The postcranial skeleton of Eusthenopteron foordi Whiteaves. Transactions of the Royal Society of Edinburgh 68, 207–329. BÁEZ, A. M. & BASSO, N. G. ( 1996 ). The earliest known frogs of the Jurassic of South America : review and cladistic appraisal of their relationships. Münchner Geowissenschaftliche Abhandlungen, Reihe A ( Geologie und Paläontologie ) 30, 131–158. BARKAS, T. P. ( 1873 ). Illustrated Guide to the Fish, Amphibian, Reptilian and Supposed Mammalian Remains of the Northumberland Carboniferous Strata. Hutchins, London. BEAUMONT, E. H. ( 1977 ). Cranial morphology of the Loxommatidae ( Amphibia : Labyrinthodontia ). Philosophical Transactions of the Royal Society of London, Series B 280, 29–101. BEAUMONT, E. H. & SMITHSON, T. R. ( 1998 ). The cranial morphology and relationships of the aberrant Carboniferous amphibian Spathicephalus mirus Watson. Zoological Journal of the Linnean Society 122, 187–209. BEERBOWER, J. R. ( 1963 ). Morphology, paleo-ecology and phylogeny of the Permo-Pennsylvanian amphibian Diploceraspis. Bulletin of the Museum of Comparative Zoology, Harvard 130, 31–108. BENTON, M. J. ( ed. ) ( 1988 ). The Phylogeny and Classification of the Tetrapods. Clarendon Press, Oxford. BENTON, M. J. ( 2000 ). Stems, nodes, crown clades, and rank-free lists : is Linnaeus dead ? Biological Reviews 75, 633–648. BENTON, M. J. ( 2001 ). Vertebrate palaeontology. Chapman & Hall, London. BERMAN, D. S. ( 2000 ). Origin and early evolution of the amniote occiput. Journal of Paleontology 74, 938–956. BERMAN, D. S., EBERTH, D. A. & BRINKMAN, D. B. ( 1988 ). Stegotretus agyrus a new genus and species of microsaur ( amphibian ) from the Permo-Pennsylvanian of New Mexico. Annals of the Carnegie Museum 57, 293–323. BERMAN, D. S., HENRICI, A. C., SUMIDA, S. S. & MARTENS, T. ( 2000 ). Redescription of Seymouria sanjuanensis ( Seymouriamorpha ) from the Lower Permian of Germany based on complete, mature specimens with a discussion of paleoecology of the Bromacker locality assemblage. Journal of Vertebrate Paleontology 20, 253–268. BERMAN, D. S. & MARTENS, T. ( 1993 ). First occurrence of Seymouria ( Amphibia : Batrachosauria ) in the Lower Permian Rotliegend of central Germany. Annals of the Carnegie Museum 62, 63–79. BERMAN, D. S., REISZ, R. R. & EBERTH, D. A. ( 1985 ). Ecolsonia cutlerensis, an early Permian dissorophid amphibian from the Cutler Formation of north-central New Mexico. Circular of the New Mexico Bureau of Mines and Mineral Resources 191, 1–31. BERMAN, D. S. & SUMIDA, S. S. ( 1990 ). A new species of Limnoscelis ( Amphibia, Diadectomorpha ) from the late Pennsylvanian Sangre de Cristo Formation of central Colorado. Annals of the Carnegie Museum 59, 303–341. BERMAN, D. S., SUMIDA, S. S. & LOMBARD, R. E. ( 1992 ). Reinterpretation of the temporal and occipital regions in Diadectes and the relationships of diadectomorphs. Journal of Paleontology 66, 481– 499. BERMAN, D. S., SUMIDA, S. S. & MARTENS, T. ( 1998 ). Diadectes ( Diadectomorpha : Diadectidae ) from the early Permian of central Germany, with description of a new species. Annals of the Carnegie Museum 67, 53–93. BININDA-EMONDS, O. R. P., GITTLEMAN, J. L. & PURVIS, A. ( 1999 ). Building large trees by combining phylogenetic information : a complete phylogeny of the extant Carnivora ( Mammalia ). Biological Reviews of the Cambridge Philosophical Society 74, 143–175. BJERRING, H. C. ( 1986 ). Electric tetrapods ? In Studies in Herpetology ( ed. Z. Rocek ), pp. 29–36. Charles University, Prague. BOLT, J. R. ( 1969 ). Lissamphibian origins : possible protolissamphibian from the Lower Permian of Oklahoma. Science 166, 888–891. BOLT, J. R. ( 1977 ). Dissorophoid relationships and ontogeny, and the origin of the Lissamphibia. Journal of Paleontology 51, 235–249. BOLT, J. R. ( 1979 ). Amphibamus grandiceps as a juvenile dissorophid : evidence and implications. In Mazon Creek Fossils ( ed. M. H. Nitecki ), pp. 529–563. Academic Press, New York. BOLT, J. R. ( 1991 ). Lissamphibian origins. In Origins of the Higher Groups of Tetrapods : Controversy and Consensus ( eds. H.-P. Schultze and L. Trueb ), pp. 194–222. Cornell University Press, Ithaca. BOLT, J. R. ( 1996 ). [Abstract] Cranial kinesis in Mississippian colosteid amphibians. Journal of Vertebrate Paleontology 16 ( Suppl. 3 ), 23A. BOLT, J. R. & CHATTERJEE, S. ( 2000 ). A new temnospondyl amphibian from the late Triassic of Texas. Journal of Paleontology 74, 670–683. BOLT, J. R. & LOMBARD, R. E. ( 1985 ). Evolution of the tympanic ear and the origin of frogs. Biological Journal of the Linnean Society 24, 83–99. Early tetrapod relationships revisited BOLT, J. R. & LOMBARD, R. E. ( 2000 ). Palaeobiology of Whatcheeria deltae, a primitive Mississippian tetrapod. In Amphibian Biology, 4 : Palaeontology( eds. H. Heatwole and R. L. Carroll ), pp. 1044–1052. Surrey Beatty & Sons, Chipping Norton. BOLT, J. R. & LOMBARD, R. E. ( 2001 ). The mandible of the primitive tetrapod Greererpeton, and the early evolution of the tetrapod lower jaw. Journal of Paleontology 75, 1016–1042. BOLT, J. R. & WASSERSUG, R. J. ( 1975 ). Functional morphology of the skull in Lysorophus, a serpentiform Paleozoic amphibian ( Lepospondyli ). Paleobiology 1, 320–332. BOSSY, A. K. & MILNER, A. C. ( 1998 ). Order Nectridea Miall, 1875. In Handbuch der Paläoherpetologie, Teil 1 : Lepospondyli ( ed. P. Wellnhofer ), pp. 73–212. Pfeil, Munich. BOY, J. A. ( 1972 ). Die Branchiosaurier ( Amphibia ) des saarpfälzischen Rotliegenden ( Perm, SW-Deutschland ). Abhandlungen des Hessischen Landesamtes für Bodenforschung 65, 1–137. BOY, J. A. ( 1974 ). Die Larven der rhachitomen Amphibien ( Amphibia : Temnospondyli ; Karbon-Trias ). Paläontologische Zeitschrift 48, 236–268. BOY, J. A. ( 1978 ). Die Tetrapodenfauna ( Amphibia, Reptilia ) des saarpfälzischen Rotliegenden ( Unter-Perm, SW-Deutschland ). 1 Branchiosaurus. Mainzer Geowissenschaftliche Mitteilungen 7, 27–76. BOY, J. A. ( 1985 ). Über Micropholis, den letzten Überlebenden der Dissorophoidea ( Amphibia, Temnospondyli ; Unter-Trias ). Neues Jahrbuch für Geologie und Paläontologie, Monatshefte 1985, 29– 45. BOY, J. A. ( 1986 ). Studien über die Branchiosauridae ( Amphibia : Temnospondyli ). 1. Neue und wenig bekannte Arten aus dem mitteleuropäischen Rotliegenden ( ?oberstes Karbon bis unteres Perm ). Paläontologische Zeitschrift 60, 131–166. BOY, J. A. ( 1987 ). Studien über die Branchiosauridae ( Amphibia : Temnospondyli ; Ober-Karbon – Unter-Perm ). 2. Systematische Übersicht. Neues Jahrbuch für Geologie und Paläontologie, Abhandlungen 174, 75–104. BOY, J. A. ( 1995 ). Über die Micromelerpetontidae ( Amphibia : Temnospondyli ). 1. Morphologie und Paläoökologie des Micromelerpeton credneri ( Unter-Perm ; SW-Deutschland ). Paläontologische Zeitschrift 69, 429– 457. BOY, J. A. & BANDEL, K. ( 1973 ). Bruktererpeton fiebigi n. gen. n. sp. ( Amphibia : Gephyrostegida ) der erste Tetrapode aus dem Rheinisch-Westfalischen Karbon ( Namur B ; W.-Deutschland ). Palaeontographica, Abteilung A 145, 29–77. BOY, J. A. & SUES, H.-D. ( 2000 ). Branchiosaurs : larvae, metamorphosis and heterochrony in temnospondyls and seymouriamorphs. In Amphibian Biology, 4 : Palaeontology ( eds. H. Heatwole and R. L. Carroll ), pp. 1150–1197. Surrey Beatty & Sons, Chipping Norton. BROCHU, C. A. & SUMRALL, C. D. ( 2001 ). Phylogenetic nomenclature and paleontology. Journal of Paleontology 75, 754 –775. BROILI, F. ( 1904a ). Permische Stegocephalen und Reptilien aus Texas. Palaeontographica 51, 1–120. BROILI, F. ( 1904b ). Stammreptilien. Anatomischer Anzeiger 25, 577–587. BROILI, F. ( 1924 ). Ein Cotylosaurier aus der oberkarbonischen Gaskohle von Nürschan in Böhmen. Sitzungsberichte Bayerische Akademie der Wissenschaften 1924, 3–11. BROUGH, M. C. & BROUGH, J. ( 1967 ). Studies on early tetrapods. I. The Lower Carboniferous microsaurs. II. Microbrachis, the type microsaur. III. The genus Gephyrostegus. Philosophical Transactions of the Royal Society of London, Series B 252, 107–165. BRYANT, H. N. & CANTINO, P. D. ( 2002 ). A review of criticisms of phylogenetic nomenclature : is taxonomic freedom the fundamental issue ? Biological Reviews 77, 39–55. BUDD, G. ( 2001 ). Climbing life’s tree. Nature 412, 487. 295 BULMAN, O. M. B. & WHITTARD, W. F. ( 1926 ). On Branchiosaurus and allied genera ( Amphibia ). Proceedings of the Zoological Society of London 1926, 533–579. BYSTROW, A. P. ( 1944 ). On Kotlassia prima ( Amalitsky ). Bulletin of the Geological Society of America 55, 379– 416. CAMPBELL, K. S. W. & BELL, M. W. ( 1977 ). A primitive amphibian from the Late Devonian of New South Wales. Alcheringa 1, 369–381. CANTINO, P. D. ( 2000 ). Phylogenetic nomenclature : addressing some concerns. Taxon 49, 85–93. CANTINO, P. D., BRYANT, H. N., DEQUEIROZ, K., DONOGHUE, M. J., ERIKSSON, T., HILLIS, D. M. & LEE, M. S. Y. ( 1999 ). Species names in phylogenetic nomenclature. Systematic Biology 48, 790–807. CARROLL, R. L. ( 1964 ). Early evolution of the dissorophid amphibians. Bulletin of the Museum of Comparative Zoology, Harvard University 131, 161–250. CARROLL, R. L. ( 1967 ). An adelogyrinid lepospondyl amphibian from the Upper Carboniferous. Canadian Journal of Zoology 45, 1–16. CARROLL, R. L. ( 1969a ). A middle Pennsylvanian captorhinomorph, and the interrelationships of primitive reptiles. Journal of Paleontology 43, 151–170. CARROLL, R. L. ( 1969b ). A new family of Carboniferous amphibians. Palaeontology 12, 537–548. CARROLL, R. L. ( 1970 ). The ancestry of reptiles. Philosophical Transactions of the Royal Society of London, Series B 257, 267–308. CARROLL, R. L. ( 1986 ). The skeletal anatomy and some aspects of the physiology of primitive reptiles. In The Ecology and Biology of Mammal-like Reptiles ( eds. N. Hotton, III, P. D. Maclean, J. J. Roth and E. C. Roth ), pp. 25–45. Smithsonian Institution Press, Washington. CARROLL, R. L. ( 1988 ). Vertebrate Paleontology and Evolution. Freeman, New York. CARROLL, R. L. ( 1990 ). A tiny microsaur from the Lower Permian of Texas : size constraints in Paleozoic tetrapods. Palaeontology 33, 893–909. CARROLL, R. L. ( 1991a ). Batropetes from the Lower Permian of Europe – a microsaur, not a reptile. Journal of Vertebrate Paleontology 11, 229–242. CARROLL, R. L. ( 1991b ). The origin of reptiles. In Origins of the Higher Groups of Tetrapods : Controversy and Consensus ( eds. H.-P. Schultze and L. Trueb ), pp. 331–353. Cornell University Press, Ithaca. CARROLL, R. L. ( 1995 ). Problems of the phylogenetic analysis of Paleozoic choanates. In Studies on Early Vertebrates : ( VIIth International Symposium, Parc de Miguasha, Quebec ) ( eds. M. Arsenault, H. Lelièvre and P. Janvier ), pp. 389–445. Bulletin du Muséum National d ’Histoire Naturelle, Paris ( Séries 4 ) 17 ( C ). CARROLL, R. L. ( 1996 ). Elongate early tetrapods : why ‘ Lizzie ’ had little limbs. In Studies on Carboniferous and Permian Vertebrates ( ed. A. R. Milner ), pp. 139–148. Special Papers in Palaeontology 52. CARROLL, R. L. ( 1998 ). Cranial anatomy of ophiderpetontid aı̈stopods : Palaeozoic limbless amphibians. Zoological Journal of the Linnean Society 122, 143–166. CARROLL, R. L. ( 1999 ). Homology among divergent Paleozoic tetrapod clades. In Homology ( eds. G. R. Bock and G. Cardew ), pp. 47–64. Wiley, Chichester. CARROLL, R. L. ( 2000 ). Eocaecilia and the origin of caecilians. In Amphibian Biology, 4 : palaeontology ( eds. H. Heatwole and R. L. Carroll ), pp. 1402–1011. Surrey Beatty & Sons, Chipping Norton. CARROLL, R. L. ( 2001 ). The origin and early radiation of terrestrial vertebrates. Journal of Paleontology 75, 1202–1213. 296 Marcello Ruta, Michael I. Coates and Donald L. J. Quicke CARROLL, R. L. & BAIRD, D. ( 1968 ). The Carboniferous amphibian Tuditanus [Eosauravus] and the distinction between microsaurs and reptiles. American Museum Novitates 2337, 1–50. CARROLL, R. L. & BAIRD, D. ( 1972 ). Carboniferous stem-reptiles of the Family Romeriidae. Bulletin of the Museum of Comparative Zoology, Harvard University 143, 321–364. CARROLL, R. L. & BOLT, J. R. ( 2001 ). [Abstract] The Paleozoic divergence of frogs and salamanders. Journal of Vertebrate Paleontology 21 ( Suppl. 3 ), 38A. CARROLL, R. L., BOSSY, K. A., MILNER, A. C., ANDREWS, S. M. & WELLSTEAD, C. F. ( 1998 ). Handbuch der Paläoherpetologie : Teil 1, Lepospondyli. Pfeil, Munich. CARROLL, R. L., BYBEE, P. & TIDWELL, W. D. ( 1991 ). The oldest microsaur ( Amphibia ). Journal of Paleontology 65, 314–322. CARROLL, R. L. & CHORN, J. ( 1995 ). Vertebral development in the oldest microsaur and the problem of ‘ lepospondyl ’ relationships. Journal of Vertebrate Paleontology 15, 37–56. CARROLL, R. L. & CURRIE, P. J. ( 1975 ). Microsaurs as possible apodan ancestors. Zoological Journal of the Linnean Society 57, 229–247. CARROLL, R. L. & CURRIE, P. J. ( 1991 ). The early radiation of diapsid reptiles. In Origins of the Higher Groups of Tetrapods : Controversy and Consensus ( eds. H.-P. Schultze and L. Trueb ), pp. 354 – 424. Cornell University Press, Ithaca. CARROLL, R. L. & GASKILL, P. ( 1978 ). The Order Microsauria. Memoirs of the American Philosophical Society 126, 1–211. CARROLL, R. L., KUNST, A. & ALBRIGHT, K. ( 1999 ). Vertebral development and amphibian evolution. Evolution & Development 1, 1–13. CASE, E. C. ( 1910 ). New or little known reptiles and amphibians from the Permian ( ? ) of Texas. Bulletin of the American Museum of Natural History 28, 163–181. CASE, E. C. ( 1935 ). Description of a collection of associated skeletons of Trimerorhachis. Contributions of the Museum of Paleontology, University of Michigan 4, 227–274. CHASE, J. N. ( 1965 ). Neldasaurus wrightae, a new rhachitomous labyrinthodont from the Texas Lower Permian. Bulletin of the Museum of Comparative Zoology, Harvard University 133, 153–225. CHIPMAN, A. D. & TCHERNOV, E. ( 2002 ). Ancient ontogenies : larval development of the Lower Cretaceous anuran Shomronella jordanica ( Amphibia : Pipoidea ). Evolution & Development 4, 86–95. CLACK, J. A. ( 1987a ). Pholiderpeton scutigerum Huxley, an amphibian from the Yorkshire Coal Measures. Philosophical Transactions of the Royal Society of London, Series B 318, 1–107. CLACK, J. A. ( 1987b ). Two new specimens of Anthracosaurus ( Amphibia : Anthracosauria ) from the Northumberland Coal Measures. Palaeontology 30, 15–26. CLACK, J. A. ( 1989 ). Discovery of the earliest-known tetrapod stapes. Nature 342, 425– 427. CLACK, J. A. ( 1994a ). Acanthostega gunnari, a Devonian tetrapod from Greenland ; the snout, palate and ventral parts of the braincase, with a discussion of their significance. Meddelelser om Grønland : Geoscience 31, 1–24. CLACK, J. A. ( 1994b ). Earliest known tetrapod braincase and the evolution of the stapes and fenestra ovalis. Nature 369, 392–394. CLACK, J. A. ( 1994c ). Silvanerpeton miripedes, a new anthracosauroid from the Viséan of East Kirkton, West Lothian, Scotland. Transactions of the Royal Society of Edinburgh : Earth Sciences 84, 369–376. CLACK, J. A. ( 1996 ). The palate of Crassigyrinus scoticus, a primitive tetrapod from the Lower Carboniferous of Scotland. In Studies on Carboniferous and Permian Vertebrates ( ed. A. R. Milner ). Special Papers in Palaeontology 52, 55–64. CLACK, J. A. ( 1998a ). A new Early Carboniferous tetrapod with a mélange of crown-group characters. Nature 394, 66–69. CLACK, J. A. ( 1998b ). The neurocranium of Acanthostega gunnari and the evolution of the otic region in tetrapods. Zoological Journal of the Linnean Society 122, 61–97. CLACK, J. A. ( 1998c ). The Scottish Carboniferous tetrapod Crassigyrinus scoticus ( Lydekker ) – cranial anatomy and relationships. Transactions of the Royal Society of Edinburgh : Earth Sciences 88, 127–142. CLACK, J. A. ( 2001 ). Eucritta melanolimnetes from the Early Carboniferous of Scotland, a stem tetrapod showing a mosaic of characteristics. Transactions of the Royal Society of Edinburgh : Earth Sciences 92, 75–95. CLACK, J. A. ( 2002 ). An early tetrapod from ‘ Romer’s Gap ’. Nature 418, 72–76. CLACK, J. A. & CARROLL, R. L. ( 2000 ). Early Carboniferous tetrapods. In Amphibian Biology, 4 : Palaeontology ( eds. H. Heatwole and R. L. Carroll ), pp. 1030–1043. Surrey Beatty & Sons, Chipping Norton. CLACK, J. A. & COATES, M. I. ( 1995 ). Acanthostega – a primitive aquatic tetrapod ? In Studies on Early Vertebrates : ( VIIth International Symposium, Parc de Miguasha, Quebec ) ( eds. M. Arsenault, H. Lelièvre and P. Janvier ), pp. 359–373. Bulletin du Muséum National d ’ Histoire Naturelle, Paris ( Series 4 ) 17 ( C ). CLACK, J. A. & FINNEY, S. M. ( 1997 ). [Abstract] An articulated tetrapod specimen from the Tournaisian of western Scotland. Journal of Vertebrate Paleontology 17 ( Suppl. 3 ), 38A. CLACK, J. A. & HOLMES, R. ( 1988 ). The braincase of the anthracosaur Archeria crassidisca with comments on the interrelationships of primitive tetrapods. Palaeontology 31, 85–107. CLARK, J. & CARROLL, R. L. ( 1973 ). Romeriid reptiles from the Lower Permian. Bulletin of the Museum of Comparative Zoology, Harvard University 144, 353– 407. CLOUTIER, R. & AHLBERG, P. E. ( 1996 ). Morphology, characters, and the interrelationships of basal sarcopterygians. In Interrelationships of Fishes ( eds. M. L. J. Stiassny, L. R. Parenti and G. D. Johnson ), pp. 325–337. Academic Press, London. COATES, M. I. ( 1991 ). New palaeontological contributions to limb ontogeny and phylogeny. In Developmental Patterns of the Vertebrate Limb ( eds. J. R. Hinchliffe, J. M. Hurle and D. Summerbell ), pp. 325–337. Plenum, New York. COATES, M. I. ( 1995 ). Fish fins or tetrapod limbs – a simple twist of fate ? Current Biology 5, 844–848. COATES, M. I. ( 1996 ). The Devonian tetrapod Acanthostega gunnari Jarvik : postcranial anatomy, basal tetrapod interrelationships and patterns of skeletal evolution. Transactions of the Royal Society of Edinburgh : Earth Sciences 87, 363– 421. COATES, M. I. & CLACK, J. A. ( 1990 ). Polydactyly in the earliest known tetrapod limbs. Nature 347, 66–69. COATES, M. I. & CLACK, J. A. ( 1991 ). Fish-like gills and breathing in the earliest known tetrapod. Nature 352, 234–236. COATES, M. I. & CLACK, J. A. ( 1995 ). Romer’s Gap – tetrapod origins and terrestriality. In Studies on early vertebrates : ( VIIth International Symposium, Parc de Miguasha, Quebec ) ( eds. M. Arsenault, H. Lelièvre and P. Janvier ), pp. 373–388. Bulletin du Muséum National d’Histoire Naturelle, Parisation ( Séries 4 ) 17 ( C ). COATES, M. I. & COHN, M. J. ( 1998 ). Fins, limbs and tails : outgrowth and patterning in vertebrate evolution. Bioessays 20, 371–381. COATES, M. I., RUTA, M. & MILNER, A. R. ( 2000 ). Early tetrapod evolution. Trends in Ecology and Evolution 15, 327–328. COHN, M. J., PATEL, K., KRUMLAUF, R., WILKINSON, D. G., CLARKE, J. D. W. & TICKLE, C. ( 1997 ). Hox9 genes and vertebrate limb specification. Nature 387, 97–101. Early tetrapod relationships revisited COLLESS, D. H. ( 1995 ). Relative symmetry of cladograms and phenograms : an experimental study. Systematic Biology 44, 102–108. COPE, E. D. ( 1865 ). On Amphibamus grandiceps, a new batrachian from the Coal-measures. Proceedings of the Academy of Natural Sciences, Philadelphia 1865, 134–137. COPE, E. D. ( 1868 ). Synopsis of the extinct Batrachia of North America. Proceedings of the Academy of Natural Sciences, Philadelphia 1868, 208–221. COPE, E. D. ( 1871 ). Observations on the extinct batrachian fauna of the Carboniferous of Linton, Ohio. Proceedings of the American Philosophical Society 12, 177. COPE, E. D. ( 1874 ). Supplement to the extinct Batrachia, Reptilia and Aves of North America. 1. Catalogue of the air-breathing Vertebrata from the Coal Measures of Linton, Ohio. Transactions of the American Philosophical Society 15, 261–278. COPE, E. D. ( 1875 ). The extinct Batrachia of Ohio. Proceedings of the Academy of Natural Sciences, Philadelphia 1875, 16. COPE, E. D. ( 1877 ). Description of extinct Vertebrata from the Permian and Triassic Formations of the United States. Proceedings of the American Philosophical Society 17, 182–193. COPE, E. D. ( 1882 ). Third contribution to the history of the Vertebrata of the Permian formation of Texas. Proceedings of the American Philosophical Society 20, 405– 461. COPE, E. D. ( 1884 ). Fifth contribution to the knowledge of the fauna of the Permian Formation of Texas and the Indian Territory. Proceedings of the American Philosophical Society 22, 28– 47. COPE, E. D. ( 1897 ). On new Paleozoic Vertebrata from Illinois, Ohio, and Pennsylvania. Proceedings of the American Philosophical Society 36, 71–90. CRASKE, A. J. & JEFFERIES, R. P. S. ( 1989 ). A new mitrate from the Upper Ordovician of Norway, and a new approach to subdividing a plesion. Palaeontology 32, 69–99. CREDNER, H. ( 1890 ). Die Stegocephalen und Saurier aus dem Rothliegenden des Plauen’schen Grundes bei Dresden. IX Theil. Hylonomus geinitzi Cred. Petrobates truncatus Cred. Discosaurus Permianus Cred. Zeitschrift Deutsche Geologische Gesellschaft 42, 240–277. DAESCHLER, E. B. ( 2000 ). Early tetrapod jaws from the late Devonian of Pennsylvania, USA. Journal of Paleontology 74, 301–308. DAESCHLER, E. B., SHUBIN, N., THOMSON, K. S. & AMARAL, W. W. ( 1994 ). A Devonian tetrapod from North America. Science 265, 639–642. DALY, E. ( 1973 ). A lower Permian vertebrate fauna from Southern Oklahoma. Journal of Paleontology 47, 562–589. DALY, E. ( 1994 ). The Amphibamidae ( Amphibia : Temnospondyli ), with a description of a new genus from the Upper Pennsylvanian of Kansas. Miscellaneous Publications of the Museum of Natural History, University of Kansas 85, 1–59. DAMIANI, R. J. ( 2001 ). A systematic revision and phylogenetic analysis of Triassic mastodonsauroids ( Temnospondyli : Stereospondyli ). Zoological Journal of the Linnean Society of London 133, 379– 482. DAWSON, J. W. ( 1894 ). Preliminary note on recent discoveries of batrachian and other air-breathers in the Coal Formation of Nova Scotia. The Canadian Record of Science 6, 1–7. DEBRAGA, M. & RIEPPEL, O. ( 1997 ). Reptile phylogeny and the interrelationships of turtles. Zoological Journal of the Linnean Society of London 120, 281–354. DEQUEIROZ, K. & GAUTHIER, J. ( 1990 ). Phylogeny as a central principle in taxonomy : phylogenetic definitions of taxon names. Systematic Zoology 39, 307–322. 297 DEQUEIROZ, K. & GAUTHIER, J. ( 1992 ). Phylogenetic taxonomy. Annual Review of Ecology and Systematics 23, 449– 480. DEQUEIROZ, K. & GAUTHIER, J. ( 1994 ). Towards a phylogenetic system of biological nomenclature. Trends in Ecology and Evolution 9, 27–31. DILKES, D. W. ( 1990 ). A new trematopid amphibian ( Temnospondyli : Dissorophoidea ) from the Lower Permian of Texas. Journal of Vertebrate Paleontology 10, 222–243. DONOGHUE, P. C. J., FOREY, P. L. & ALDRIDGE, R. J. ( 2000 ). Conodont affinity and chordate phylogeny. Biological Reviews 75, 191–251. DUBOULE, D. ( 1994 ). How to make a limb ? Science 226, 575–576. DUELLMAN, W. E. & TRUEB, L. ( 1986 ). Biology of Amphibians. McGraw-Hill, New York. ESTES, R. ( 1981 ). Handbuch der Paläoherpetologie : Teil 2, Gymnophiona, Caudata. Gustav Fischer Verlag, Stuttgart. ESTES, R. & HOFFSTETTER, R. ( 1976 ). Les urodéles du Miocène de la Grive-St. Alban ( Izère, France ). Bulletin du Muséum National d’Histoire Naturelle, Paris ( Sciences de la Terre ) 57, 297–343. ESTES, R. & REIG, O. ( 1973 ). The early fossil record of frogs : a review of the evidence. In Evolutionary biology of the anurans ( ed. J. L. Vial ), pp. 1–64. University of Missouri Press, Columbia. EVANS, S. E. & MILNER, A. R. ( 1996 ). A metamorphosed salamander from the Lower Cretaceous of Spain. Philosophical Transactions of the Royal Society of London, Series B 351, 627–646. EVANS, S. E. & SIGOGNEAU-RUSSELL, D. ( 2001 ). A stem-group caecilian ( Lissamphibia : Gymnophiona ) from the Lower Cretaceous of North Africa. Palaeontology 44, 259–273. FARRIS, J. S., ALBERT, V. A., KÄLLERSJÖ, M., LIPSCOMB, D. & KLUGE, A. G. ( 1996 ). Parsimony jackknifing outperforms neighborjoining. Cladistics 12, 99–124. FELLER, A. E. & HEDGES, S. B. ( 1998 ). Molecular evidence for the early history of living amphibians. Molecular Phylogeny and Evolution 9, 509–516. FOREY, P. L. ( 1993 ). Therapsids and transformation series. Nature 361, 596–597. FOREY, P. L. ( 1998 ). History of the Coelacanth Fishes. Chapman & Hall, London. FOREY, P. L. ( 2001 ). The PhyloCode : description and commentary. Bulletin of Zoological Nomenclature 58, 81–96. FOREY, P. L. & KITCHING, I. J. ( 2000 ). Experiments in coding multistate characters. In Homology and Systematics : Coding Characters for Phylogenetic Analysis ( eds. R. Scotland and R. T. Pennington ), pp. 54–80. Taylor & Francis, London. FOX, R. C. & BOWMAN, M. C. ( 1966 ). Osteology and relationships of Captorhinus aguti ( Cope ) ( Reptilia : Captorhinomorpha ). University of Kansas Paleontological Contributions 15, 1–80. FRACASSO, M. ( 1987 ). Braincase of Limnoscelis paludis Williston. Postilla 201, 1–22. FRITSCH, A. ( 1876 ). Über die Fauna der Gaskohle des Pilsner und Rakonitzer Beckens. Sitzungsberichte der Königlichen Böhmischen Gesellschaft der Wissenschaften Prague 1875, 70–79. FRITSCH, A. ( 1883 ). Fauna der Gaskohle und der Kalksteine der Permformation Böhmens. Band 1 ( 1879–1883 ). Rivnác, Prague. GAFFNEY, E. S. ( 1979 ). Tetrapod monophyly : a phylogenetic analysis. Bulletin of the Carnegie Museum, Natural History 13, 92–105. GAO, K.-Q. & SHUBIN, N. H. ( 2001 ). Late Jurassic salamanders from northern China. Nature 410, 574–577. GAO, K.-Q. & WANG, Y. ( 2001 ). Mesozoic anurans from Liaoning Province, China, and phylogenetic relationships of archaeobatrachian anuran clades. Journal of Vertebrate Paleontology 21, 460–476. 298 Marcello Ruta, Michael I. Coates and Donald L. J. Quicke GARDINER, B. G. ( 1982 ). Tetrapod classification. Zoological Journal of the Linnean Society 74, 207–232. GARDINER, B. G. ( 1983 ). Gnathostome vertebrae and the classification of the Amphibia. Zoological Journal of the Linnean Society 79, 1–59. GARDNER, J. D. ( 1999 ). Redescription of the geologically youngest albanerpetontid ( ?Lissamphibia ) : Albanerpeton inexpectatum Estes and Hoffstetter, 1976, from the Miocene of France. Annales de Paléontologie 85, 57–84. GARDNER, J. D. ( 2001 ). Monophyly and affinities of albanerpetontid amphibians ( Temnospondyli ; Lissamphibia ). Zoological Journal of the Linnean Society 131, 309–352. GAUTHIER, J. A., KLUGE, A. G. & ROWE, T. ( 1988a ). Amniote phylogeny and the importance of fossils. Cladistics 4, 105–209. GAUTHIER, J. A., KLUGE, A. G. & ROWE, T. ( 1988b ). The early evolution of the Amniota. In The Phylogeny and Classification of the Tetrapods, 1 : Amphibians, Reptiles, Birds ( ed. M. J. Benton ), pp. 103–155. Clarendon Press, Oxford. GEINITZ, H. B. & DEICHMÜ LLER, J. V. ( 1882 ). Die fossilen Saurier in dem Kalke des Rothliegenden von Niederhässlich im Plauen ’ schen Grunde beiDresden. Mineral-geologische Praehistorische Museum Dresden. GODFREY, S. J. ( 1989 ). The postcranial skeletal anatomy of the Carboniferous tetrapod Greererpeton burkemorani. Philosophical Transactions of the Royal Society, Series B 323, 75–133. GODFREY, S. J. & HOLMES, R. ( 1995 ). The Pennsylvanian temnospondyl Cochleosaurus florensis Rieppel, from the lycopod stump fauna of Florence, Nova Scotia. Breviora 500, 1–25. GODFREY, S. J., FIORILLO, A. R. & CARROLL, R. L. ( 1987 ). A newly discovered skull of the temnospondyl amphibian Dendrerpeton acadianum Owen. Canadian Journal of Earth Sciences 24, 796–805. GOLOBOFF, P. A. ( 1999 ). Analyzing large data sets in reasonable times : solutions for composite optima. Cladistics 15, 415– 428. GOODRICH, E. S. ( 1930 ). Studies on the Structure and Development of Vertebrates. Macmillan & Co, London. GRANDCOLAS, P., DELEPORTE, P., DESUTTER-GRANDCOLAS, L. & DAUGERON, C. ( 2001 ). Phylogenetics and ecology : as many characters as possible should be included in the cladistic analysis. Cladistics 17, 104–110. GREENE, H. W. ( 2001 ). Improving taxonomy for us and the other fishes. Nature 411, 738. GREGORY, J. T., PEABODY, F. E. & PRICE, L. I. ( 1956 ). Revision of the Gymnarthridae, American Permian microsaurs. Bulletin of the Peabody Museum of Natural History 10, 1–77. GROSS, W. ( 1930 ). Die Fische des Mitteleren Old Red SüdLivlands. Geologische Paläontologische Abhandlungen 18, 121–156. GUBIN, Y. M. ( 1983 ). The first eryopids from the Permian of the East European Platform. Paleontological Journal 17, 105–110. HANCOCK, A. & ATTHEY, T. ( 1869 ). On a new labyrinthodont from the Northumberland coal field, and on the occurrence in the same locality of Anthracosaurus russelli. Annals and Magazine of Natural History 4, 182–189. HEARD, S. B. ( 1992 ). Patterns in tree balance among cladistic, phenetic and randomly generated phylogenetic trees. Evolution 46, 1818–1826. HEATON, M. J. ( 1979 ). Cranial anatomy of primitive captorhinid reptiles from the Late Pennsylvanian and Early Permian, Oklahoma and Texas. Bulletin of the Oklahoma Geological Survey 127, 1–80. HEATON, M. J. ( 1980 ). The Cotylosauria : a reconsideration of a group of archaic tetrapods. In The Terrestrial Environment and the Origin of Land Vertebrates ( ed. A. L. Panchen ), pp. 497–551. Academic Press, London. HEATWOLE, H. & CARROLL, R. L. ( eds ) ( 2000 ). Amphibian Biology, 4 : Palaeontology. Surrey Beatty & Sons, Chipping Norton. HECHT, M. K. ( 1960 ). The history of the frogs. Anatomical Record 138, 365. HECHT, M. K. ( 1962 ). A reevaluation of the early history of the frogs. Part I. Systematic Zoology 11, 39– 44. HEDGES, S. B. ( 2001 ). Molecular evidence for the early history of living vertebrates. In Major Events in Early Vertebrate Evolution : Palaeontology, Phylogeny, Genetics and Development ( ed. P. E. Ahlberg ), pp. 119–134. Taylor & Francis, London. HENNIG, W. ( 1965 ). Phylogenetic systematics. Annual Review of Entomology 10, 97–116. HENNIG, W. ( 1966 ). Phylogenetic Systematics. University of Illinois Press, Urbana. HENNIG, W. ( 1969 ). Die Stammesgeschichte der Insekten. Kramer, Frankfurt. HOLMES, R. B. ( 1980 ). Proterogyrinus scheelei and the early evolution of the labyrinthodont pectoral limb. In The Terrestrial Environment and the Origin of Land Vertebrates ( ed. A. L. Panchen ), pp. 351–376. Academic Press, London. HOLMES, R. B. ( 1984 ). The Carboniferous amphibian Proterogyrinus scheelei Romer, and the early evolution of tetrapods. Philosophical Transactions of the Royal Society of London, Series B 306, 431–527. HOLMES, R. B. ( 1989 ). The skull and axial skeleton of the Lower Permian anthracosauroid amphibian Archeria crassidisca Cope. Palaeontographica Abteilung A 207, 161–206. HOLMES, R. B. ( 2000 ). Palaeozoic temnospondyls. In Amphibian Biology, 4 : Palaeontology ( eds. H. Heatwole and R. L. Carroll ), pp. 1081–1120. Surrey Beatty & Sons, Chipping Norton. HOLMES, R. B. & CARROLL, R. L. ( 1977 ). A temnospondyl amphibian from the Mississippian of Scotland. Bulletin of the Museum of Comparative Zoology, Harvard University 147, 489–511. HOLMES, R. B., CARROLL, R. L. & REISZ, R. R. ( 1998 ). The first articulated skeleton of Dendrerpeton acadianum ( Temnospondyli, Dendrerpetontidae ) from the lower Pennsylvanian locality of Joggins, Nova Scotia, and a review of its relationships. Journal of Vertebrate Paleontology 18, 64–79. HOOK, R. W. ( 1983 ). Colosteus scutellatus ( Newberry ), a primitive temnospondyl amphibian from the Middle Pennsylvanian of Linton, Ohio. American Museum Novitates 2770, 1– 41. HOOK, R. W. ( 1993 ). Chenoprosopus lewisi, a new cochleosaurid amphibian ( Amphibia : Temnospondyli ) from the Permo-Carboniferous of north-central Texas. Annals of the Carnegie Museum 62, 273–291. HOPSON, J. A. ( 1991 ). Systematics of the nonmammalian Synapsida and implications for patterns of evolution in synapsids. In Origins of the Higher Groups of Tetrapods : Controversy and Consensus ( eds. H.-P. Schultze and L. Trueb ), pp. 635–693. Cornell University Press, Ithaca. HUXLEY, T. H. ( 1862 ). On New Labyrinthodonts from the Edinburgh Coal-Field. Quarterly Journal of the Geological Society 18, 291–296. HUXLEY, T. H. ( 1863 ). Description of Anthracosaurus russelli, a new labyrinthodont from the Lanarkshire coal field. Quarterly Journal of the Geological Society 19, 56–68. HUXLEY, T. H. ( 1869 ). On a new labyrinthodont from Bradford. Quarterly Journal of the Geological Society 25, 309–311. IVAKHNENKO, M. F. ( 1978 ). Urodelans from the Triassic and Jurassic of Soviet Central Asia. Paleontological Journal 1978, 84 –89. IVAKHNENKO, M. F. ( 1981 ). Discosauriscidae from the Permian of Tadzhikistan. Paleontological Journal 1981, 90–102. IVAKHNENKO, M. F. ( 1987 ). Permian parareptiles of USSR. Trudy Paleontologicheskogo Instituta Akademii Nauk 223, 1–160. [In Russian] Early tetrapod relationships revisited JAEKEL, O. ( 1902 ). Ueber einen neuen paläozoischen Tetrapodentypus : Gephyrostegus bohemicus n. g. n. sp. Zeitschrift der Deutschen Geologischen Gesellscahft 54, 127–132. JARVIK, E. ( 1952 ). On the fish-like tail in the ichthyostegid stegocephalians with descriptions of a new stegocephalian and a new crossopterygian from the Upper Devonian of East Greenland. Meddelelser om Grønland 114, 1–90. JARVIK, E. ( 1980 ). Basic Structure and Evolution of Vertebrates. Academic Press, London and New York. JARVIK, E. ( 1996 ). The Devonian tetrapod Ichthyostega. Fossils & Strata 40, 1–206. JEFFERIES, R. P. S. ( 1979 ). The origin of chordates – a methodological essay. In The Origin of Major Invertebrate Groups ( ed. M. R. House ), pp. 443–477. Systematics Association Special Volume, Series 12. JEFFERY, J. E. ( 2001 ). Pectoral fins of rhizodontids and the evolution of pectoral appendages in the tetrapod stem-group. Biological Journal of the Linnean Society 74, 217–236. JENKINS, F. A., JR. & SHUBIN, N. H. ( 1998 ). Prosalirus bitis ( Shubin and Jenkins 1995 ) and the anuran caudopelvic mechanism. Journal of Vertebrate Paleontology 18, 495–510. JENKINS, F. A., JR. & WALSH, D. ( 1993 ). An Early Jurassic caecilian with limbs. Nature 365, 246–250. JOHANSON, Z. and AHLBERG, P. E. ( 2001 ) ‘ Devonian rhizodontids and tristichopterids ( Sarcopterygii ; Tetrapodomorpha ) from East Gondwana ’, Transactions of the Royal Society of Edinburgh : Earth Sciences 92 : 43–74. KEARNEY, M. ( 1998 ). [Abstract] Systematics of the amphisbaenian family Rhineuridae ; missing data and resolution. Journal of Vertebrate Paleontology 18 ( Suppl. 3 ), 55A. KEARNEY, M. ( 2002 ). Fragmentary taxa, missing data, and ambiguity : mistaken assumptions and conclusions. Systematic Biology 51, 369–381. KITCHING, I. J., FOREY, P. L., HUMPHRIES, C. J. & WILLIAMS, D. M. ( 1998 ). Cladistics : the Theory and Practice of Parsimony Analysis. Oxford University Press, New York. KLASSEN, G. J., MOOI, R. D. & LOCKE, A. ( 1991 ). Consistency indices and random data. Systematic Zoology 40, 446– 457. KLEMBARA, J. ( 1985 ). A new embolomerous amphibian ( Anthracosauria ) from the Upper Carboniferous of Florence, Nova Scotia. Journal of Vertebrate Paleontology 5, 293–302. KLEMBARA, J. ( 1994 ). Electroreceptors in the Lower Permian tetrapod Discosauriscus austriacus. Palaeontology 37, 609–626. KLEMBARA, J. ( 1995 ). The external gills and ornamentation of skull roof bones of the Lower Permian tetrapod Discosauriscus ( Kuhn 1933 ) with remarks to its ontogeny. Paläontologische Zeitschrift 69, 265–281. KLEMBARA, J. ( 1997 ). The cranial anatomy of Discosauriscus Kuhn, a seymouriamorph tetrapod from the Lower Permian of the Boskovice Furrow ( Czech Republic ). Philosophical Transactions of the Royal Society of London, Series B 352, 257–302. KLEMBARA, J. & BARTÍ K, I. ( 2000 ). The postcranial skeleton of Discosauriscus Kuhn, a seymouriamorph tetrapod from the Lower Permian of the Boskovice Furrow ( Czech Republic ). Transactions of the Royal Society of Edinburgh : Earth Sciences 90, 287–316. KLEMBARA, J., MARTENS, T. & BARTÍ K, I. ( 2001 ). The postcranial remains of a juvenile seymouriamorph tetrapod from the Lower Permian Rotliegend of the Tambach Formation of central Germany. Journal of Vertebrate Paleontology 21, 521–527. KUHN, O. ( 1962 ). Die vorzeitlichen Frösche und Salamander, ihre Gattungen und Familien. Jahresheft des Vereins für Vaterländische Naturkunde in Württemberg 117, 327–372. 299 KUMAR, S. & HEDGES, S. B. ( 1998 ). A molecular timescale for vertebrate evolution. Nature 392, 917–920. KUZNETSOV, V. V. & IVAKHNENKO, M. F. ( 1981 ). Discosauriscids from the Upper Paleozoic in Southern Kazakhstan. Paleontological Journal 1981, 101–108. LANE, H. H. ( 1945 ). New mid-Pennsylvanian reptiles from Kansas. Transactions of the Kansas Academy of Sciences 47, 381–390. LANGSTON, W., JR. & OLSON, E. C. ( 1986 ). Carrolla craddocki a new genus and species of microsaur from the Lower Permian of Texas. Texas Memorial Museum, Pearce-Sellard Series 43, 1–20. LAURIN, M. ( 1995 ). Comparative cranial anatomy of Seymouria sanjuanensis ( Tetrapoda : Batrachosauria ) from the Lower Permian of Utah and New Mexico. PaleoBios 16, 1–8. LAURIN, M. ( 1996a ). A reappraisal of Utegenia, a Permo-Carboniferous seymouriamorph ( Tetrapoda : Batrachosauria ) from Kazakhstan. Journal of Vertebrate Paleontology 16, 374–383. LAURIN, M. ( 1996b ). A redescription of the cranial anatomy of Seymouria baylorensis, the best known seymouriamorph ( Vertebrata : Seymouriamorpha ). PaleoBios 17, 1–16. LAURIN, M. ( 1996c ). A reevaluation of Ariekanerpeton, a Lower Permian seymouriamorph ( Tetrapoda : Seymouriamorpha ) from Tadzhikistan. Journal of Vertebrate Paleontology 16, 653–665. LAURIN, M. ( 1998a ). A reevaluation of the origin of pentadactyly. Evolution 52, 1476–1482. LAURIN, M. ( 1998b ). The importance of global parsimony and historical bias in understanding tetrapod evolution. Part I. Systematics, middle ear evolution, and jaw suspension. Annals des Sciences Naturelles, Zoologie 19, 1–42. LAURIN, M. ( 1998c ). The importance of global parsimony and historical bias in understanding tetrapod evolution. Part II. Vertebral centrum, costal ventilation, and paedomorphosis. Annals des Sciences Naturelles, Zoologie 19, 99–114. LAURIN, M. ( 2000 ). Seymouriamorphs. In Amphibian Biology, 4 : Palaeontology ( eds. H. Heatwole and R. L. Carroll ), pp. 1064–1080. Surrey Beatty & Sons, Chipping Norton. LAURIN, M. ( 2002 ). Tetrapod phylogeny, amphibian origins, and the definition of the name Tetrapoda. Systematic Biology 51, 364–369. LAURIN, M., GIRONDOT, M. & de RICQLÈ S, A. ( 2000a ). Early tetrapod evolution. Trends in Ecology and Evolution 15, 118–123. LAURIN, M., GIRONDOT, M. & de RICQLÈ S, A. ( 2000b ). Reply. Trends in Ecology and Evolution 15, 328. LAURIN, M. & REISZ, R. R. ( 1995 ). A reevaluation of early amniote phylogeny. Zoological Journal of the Linnean Society 113, 165–223. LAURIN, M. & REISZ, R. R. ( 1997 ). A new perspective on tetrapod phylogeny. In Amniote Origins : Completing the Transition to Land ( eds. S. S. Sumida and K. L. M. Martin ), pp. 9–59. Academic Press, London. LAURIN, M. & REISZ, R. R. ( 1999 ). A new study of Solenodonsaurus janenschi, and a reconsideration of amniote origins and stegocephalian evolution. Canadian Journal of Earth Sciences 36, 1239–1255. LEBEDEV, O. A. ( 1984 ). First discovery of a Devonian tetrapod vertebrate in USSR. Doklady Akademii Nauk SSSR 278, 1470–1473. LEBEDEV, O. A. & CLACK, J. A. ( 1993 ). Upper Devonian tetrapods from Andreyevka, Tula region, Russia. Palaeontology 36, 721–734. LEBEDEV, O. A. & COATES, M. I. ( 1995 ). The postcranial skeleton of the Devonian tetrapod Tulerpeton curtum Lebedev. Zoological Journal of the Linnean Society 114, 307–348. LEE, M. S. Y. ( 1993 ). The origin of the turtle body plan : bridging a famous morphological gap. Science 261, 1716–1720. 300 Marcello Ruta, Michael I. Coates and Donald L. J. Quicke LEE, M. S. Y. ( 1995 ). Historical burden in systematics and the interrelationships of ‘ Parareptiles ’. Biological Reviews of the Cambridge Philosophical Society 70, 459–547. LEE, M. S. Y. ( 1996 ). Correlated progression and the origin of turtles. Nature 379, 812–815. LEE, M. S. Y. & SPENCER, P. S. ( 1997 ). Crown-clades, key characters and taxonomic stability : when is an amniote not an amniote ? In Amniote Origins : Completing the Transition to Land ( eds. S. S. Sumida and K. L. M. Martin ), pp. 61–84. Academic Press, London. LOMBARD, R. E. & BOLT, J. R. ( 1979 ). Evolution of the tetrapod ear : an analysis and reinterpretation. Biological Journal of the Linnean Society 11, 19–76. LOMBARD, R. E. & BOLT, J. R. ( 1995 ). A new primitive tetrapod Whatcheeria deltae from the Lower Carboniferous of Iowa. Palaeontology 3, 471–494. LOMBARD, R. E. & BOLT, J. R. ( 1999 ). A microsaur from the Mississippian of Illinois and a standard format for morphological characters. Journal of Paleontology 73, 908–923. LOMBARD, R. E. & SUMIDA, S. S. ( 1992 ). Recent progress in understanding early tetrapods. American Zoologist 32, 609–622. LUND, R. ( 1978 ). Anatomy and relationships of the family Phlegethontiidae ( Amphibia, Aı̈stopoda ). Annals of the Carnegie Museum 47, 53–79. MADDISON, W. P. & MADDISON, D. R. ( 1992 ). MacClade : Analysis of Phylogeny and Character Evolution. Version 3.0.1. Sinauer Associates, Sunderland, MA. MAKOWSKY, A. ( 1876 ). Über einen neuen Labyrinthodonten ‘ Archegosaurus austriacus nov. spec. ’ Sitzungsberichte der Kaiserichen Akademie der Wissenschaften 73, 155–166. MCGINNIS, H. J. ( 1967 ). The osteology of Phlegethontia, a Carboniferous and Permian aı̈stopod amphibian. Publications of Geological Sciences of the University of California 71, 1–46. MCGOWAN, G. & EVANS, S. E. ( 1995 ). Albanerpetontid amphibians from the Early Cretaceous of Spain. Nature 373, 143–145. MENG, J., WYSS, A. R., DAWSON, M. R. & ZHAL, R.-J. ( 1994 ). Primitive fossil rodent from Inner Mongolia and its implications for mammalian phylogeny. Nature 370, 134–136. MEYER, H. V. ( 1844 ). Über die fossilen Knochen aus dem Tertiar-Gebilde des Cerro de San Isidro bei Madrid. Neues Jahrbuch für Mineralogie 1844, 289–310. MILNER, A. C. ( 1980 ). A review of the Nectridea ( Amphibia ). In The Terrestrial Environment and the Origin of Land Vertebrates ( ed. A. L. Panchen ), pp. 377–405. Academic Press, London. MILNER, A. C. ( 1994 ). The aı̈stopod amphibian from the Viséan of East Kirkton, West Lothian, Scotland. Transactions of the Royal Society of Edinburgh : Earth Sciences 84, 363–368. MILNER, A. C. & LINDSAY, W. ( 1998 ). Postcranial remains of Baphetes and their bearing on the relationships of the Baphetidae ( =Loxommatidae ). Zoological Journal of the Linnean Society 122, 211–235. MILNER, A. R. ( 1980 ). The temnospondyl amphibian Dendrerpeton from the Upper Carboniferous of Ireland. Palaeontology 23, 125–141. MILNER, A. R. ( 1988 ). The relationships and origin of living amphibians. In The Phylogeny and Classification of the Tetrapods, 1 : Amphibians, Reptiles, Birds ( ed. M. J. Benton ), pp. 59–102. Clarendon Press, Oxford. MILNER, A. R. ( 1990 ). The radiations of temnospondyl amphibians. In Major Evolutionary Radiations ( eds. P. D. Taylor and G. P. Larwood ), pp. 321–349. Clarendon Press, Oxford. MILNER, A. R. ( 1993 ). The Paleozoic relatives of lissamphibians. Herpetological Monographs 7, 8–27. MILNER, A. R. ( 2000 ). Mesozoic and Tertiary Caudata and Albanerpetontidae. In Amphibian Biology, 4 : Palaeontology ( eds. H. Heatwole and R. L. Carroll ), pp. 1412–1444. Surrey Beatty & Sons, Chipping Norton. MILNER, A. R. & SEQUEIRA, S. E. K. ( 1994 ). The temnospondyl amphibians from the Viséan of East Kirkton, West Lothian, Scotland. Transactions of the Royal Society of Edinburgh : Earth Sciences 84, 331–361. MILNER, A. R. & SEQUEIRA, S. E. K. ( 1998 ). A cochleosaurid temnospondyl amphibian from the Middle Pennsylvanian of Linton, Ohio, U.S.A. Zoological Journal of the Linnean Society 122, 261–290. MILNER, A. R., SMITHSON, T. R., MILNER, A. C., COATES, M. I. & ROLFE, W. D. I. ( 1986 ). The search for early tetrapods. Modern Geology 10, 1–28. MINER, R. W. ( 1925 ). The pectoral limb of Eryops and other primitive tetrapods. Bulletin of the American Museum of Natural History 51, 145–312. MODESTO, S. P. ( 1998 ). New information on the skull of the Early Permian reptile Captorhinus aguti. PaleoBios 18, 21–35. MOODIE, R. L. ( 1909 ). Carboniferous air-breathing vertebrates of the United States National Museum. Proceedings of the United States National Museum 37, 11–28. MOSS, J. L. ( 1972 ). The morphology and phylogenetic relationships of the lower Permian tetrapod Tseajaia campi Vaughn ( Amphibia : Seymouriamorpha ). Publications of Geological Sciences of the University of California 98, 1–68. MOULTON, J. M. ( 1974 ). A description of the vertebral column of Eryops based on the notes and drawings of A. S. Romer. Breviora 428, 1–44. NEVO, O. & ESTES, R. ( 1969 ). Ramonellus longispinus, an Early Cretaceous salamander from Israel. Copeia 1969, 540–547. NEWBERRY, J. S. ( 1856 ). Description of several new genera and species of fossil fishes from the Carboniferous strata of Ohio. Proceedings of the Academy of Natural Sciences, Philadelphia 8, 96–100. NIXON, K. C. ( 1999 ). The Parsimony Ratchet, a new method for rapid parsimony analysis. Cladistics 15, 407– 414. NIXON, K. C. & CARPENTER, J. M. ( 2000 ). On the other ‘ Phylogenetic Systematics ’. Cladistics 16, 298–318. NIXON, K. C. & DAVIS, J. I. ( 1991 ). Polymorphic taxa, missing values and cladistic analysis. Cladistics 7, 233–241. NOVACEK, M. J. ( 1992 ). Fossils, topologies, missing data, and the higher level phylogeny of eutherian mammals. Systematic Biology 41, 58–73. NOVIKOV, I. V., SHISHKIN, M. A. & GOLUBEV, V. K. ( 2000 ). Permian and Triassic anthracosaurs from Eastern Europe. In The Age of Dinosaurs in Russia and Mongolia ( eds. M. J. Benton, M. A. Shishkin, D. M. Unwin and E. N. Kurochkin ), pp. 60–70. Cambridge University Press, Cambridge. OLSON, E. C. ( 1939 ). The fauna of the Lysorophus pockets in the Clear Fork Permian, Baylor County, Texas. Journal of Geology 47, 389–397. OLSON, E. C. ( 1941 ). The family Trematopsidae. Journal of Geology 49, 149–176. OLSON, E. C. ( 1962 ). Les problèmes de l’origine des reptiles. Colloquium International du C.N.R.S. Problèmes de Paléontologie et Evolution des Vertébrés 104, 157–174. OLSON, E. C. ( 1970 ). New and little known genera and species of vertebrates from the Lower Permian of Oklahoma. Fieldiana : Geology 18, 359– 426. OWEN, R. ( 1853 ). Notes on the above-described fossil remains. Quarterly Journal of the Geological Society 9, 66–67. Early tetrapod relationships revisited PANCHEN, A. L. ( 1964 ). The cranial anatomy of two Coal Measure anthracosaurs. Philosophical Transactions of the Royal Society of London, Series B 247, 593–637. PANCHEN, A. L. ( 1970 ). Handbuch der Paläoherpetologie : Teil 5a, Anthracosauria. Gustav Fischer Verlag, Stuttgart. PANCHEN, A. L. ( 1972 ). The skull and skeleton of Eogyrinus attheyi Watson ( Amphibia : Labyrinthodontia ). Philosophical Transactions of the Royal Society of London, Series B 263, 279–326. PANCHEN, A. L. ( 1973 ). On Crassigyrinus scoticus Watson, a primitive amphibian from the Lower Carboniferous of Scotland. Palaeontology 16, 179–193. PANCHEN, A. L. ( 1975 ). A new genus and species of anthracosaur amphibian from the Lower Carboniferous of Scotland and the status of Pholidogaster pisciformis Huxley. Philosophical Transactions of the Royal Society of London, Series B 269, 581–640. PANCHEN, A. L. ( 1977 ). On Anthracosaurus russelli Huxley ( Amphibia : Labyrinthodontia ) and the family Anthracosauridae. Philosophical Transactions of the Royal Society of London, Series B 279, 447–512. PANCHEN, A. L. ( 1980 ). The origin and relationships of the anthracosaur Amphibia from the late Palaeozoic. In The Terrestrial Environment and the Origin of Land Vertebrates ( ed. A. L. Panchen ), pp. 319–350. Academic Press, London. PANCHEN, A. L. ( 1985 ). On the amphibian Crassigyrinus scoticus Watson from the Carboniferous of Scotland. Philosophical Transactions of the Royal Society of London, Series B 309, 505–568. PANCHEN, A. L. ( 1991 ). The early tetrapods : classification and the shapes of cladograms. In Origins of the Higher Groups of Tetrapods : Controversy and Consensus ( eds. H.-P. Schultze and L. Trueb ), pp. 110–144. Cornell University Press, Ithaca. PANCHEN, A. L. & SMITHSON, T. R. ( 1987 ). Character diagnosis, fossils, and the origin of tetrapods. Biological Reviews of the Cambridge Philosophical Society 62, 341–438. PANCHEN, A. L. & SMITHSON, T. R. ( 1988 ). The relationships of early tetrapods. In The Phylogeny and Classification of the Tetrapods 1 : Amphibians, Reptiles, Birds ( ed. M. J. Benton ), pp. 1–32. Clarendon Press, Oxford. PANCHEN, A. L. & SMITHSON, T. R. ( 1990 ). The pelvic girdle and hind limb of Crassigyrinus scoticus ( Lydekker ) from the Scottish Carboniferous and the origin of the tetrapod pelvic skeleton. Transactions of the Royal Society of Edinburgh : Earth Sciences 81, 31– 44. PATON, R. L., SMITHSON, T. R. & CLACK, J. A. ( 1999 ). An amniotelike skeleton from the Early Carboniferous of Scotland. Nature 398, 508–513. PATTERSON, C. ( 1981 ). Significance of fossils in determining evolutionary relationships. Annual Review of Ecology and Systematics 12, 195–223. PATTERSON, C. ( 1993a ). Bird or dinosaur ? Nature 365, 21–22. PATTERSON, C. ( 1993b ). Naming names. Nature 366, 518. PATTERSON, C. ( 1994 ). Bony fishes. In Major Features of Vertebrate Evolution ( eds. D. R. Prothero and R. M. Schoch ), pp. 321–349. Short Courses Paleont. 7. University of Tennessee, Knoxville. PATTERSON, C. & ROSEN, D. E. ( 1977 ). Review of ichthyodectiform and other Mesozoic teleost fishes and the theory and practice of classifying fossils. Bulletin of the American Museum of Natural History 158, 81–172. PEABODY, F. E. ( 1952 ). Petrolacosaurus kansensis Lane, a Pennsylvanian reptile from Kansas. University of Kansas Paleontological Contributions 1, 1–41. PEARSON, H. S. ( 1924 ). Solenodonsaurus Broili, a seymouriamorph reptile. Annals and Magazine of Natural History 14, 338–343. 301 PIVETEAU, J. ( 1936 ). Une forme ancestrale des Amphibiens Anoures dans le Trias inférieur de Madagascar. Comptes Rendus de l’Académie des Sciences 102, 1607–1608. PLEIJEL, F. ( 1995 ). On character coding for phylogeny reconstruction. Cladistics 11, 309–315. POE, S. & WIENS, J. J. ( 2000 ). Character selection and the methodology of morphological phylogenetics. In Phylogenetic Analysis of Morphological Data ( ed. J. J. Wiens ), pp. 20–36. Smithsonian Institution Press, London. PRENDINI, L. ( 2001 ). Species or supraspecific taxa as terminals in cladistic analysis ? Groundplans versus exemplars revisited. Systematic Biology 50, 290–300. QUICKE, D. L. J., TAYLOR, J. & PURVIS, A. ( 2001 ). Changing the landscape : a new strategy for estimating large phylogenies. Systematic Biology 50, 60–66. RAGE, J.-C. & ROCEK, Z. ( 1986 ). Triadobatrachus revisited. In Studies in Herpetology ( ed. Z. Rocek ), pp. 255–258. Charles University, Prague. RAGE, J.-C. & ROCEK, Z. ( 1989 ). Redescription of Triadobatrachus massinoti ( Piveteau, 1936 ) an anuran amphibian from the early Triassic. Palaeontographica Abteilung A 206, 1–16. REISZ, R. R. ( 1977 ). Petrolacosaurus, the oldest known diapsid reptile. Science 196, 1091–1093. REISZ, R. R. ( 1981 ). A diapsid reptile from the Pennsylvanian of Kansas. Occasional Papers of the Museum of Natural History, University of Kansas 7B, 1–74. REISZ, R. R. ( 1986 ). Handbuch der Paläoherpetologie : Teil 17A, Pelycosauria. Gustav Fischer Verlag, Stuttgart. REISZ, R. R. & LAURIN, M. ( 1991 ). Owenetta and the origin of turtles. Nature 349, 324–326. RIEPPEL, O. ( 1980 ). The edopoid amphibian Cochleosaurus from the Middle Pennsylvanian of Nova Scotia. Palaeontology 23, 143–149. RIEPPEL, O. ( 2000 ). Turtles as diapsid reptiles. Zoologica Scripta 29, 199–212. RIEPPEL, O. & DEBRAGA, M. ( 1996 ). Turtles as diapsid reptiles. Nature 384, 453–455. RIEPPEL, O. & REISZ, R. R. ( 1999 ). The origin and early evolution of turtles. Annual Review of Ecology and Systematics 30, 1–22. ROCEK, Z. ( 2000 ). Mesozoic anurans. In Amphibian Biology, 4 : Palaeontology ( eds. H. Heatwole and R. L. Carroll ), pp. 1295–1331. Surrey Beatty & Sons, Chipping Norton. ROCEK, Z. & RAGE, J.-C. ( 2000a ). Anatomical transformations in the transition from temnospondyl to proanuran stages. In Amphibian Biology, 4 : Palaeontology ( eds. H. Heatwole and R. L. Carroll ), pp. 1274–1282. Surrey Beatty & Sons, Chipping Norton. ROCEK, Z. & RAGE, J.-C. ( 2000b ). Proanuran stages ( Triadobatrachus, Czatkobatrachus ). In Amphibian Biology, 4 : Palaeontology ( eds. H. Heatwole and R. L. Carroll ), pp. 1283–1294. Surrey Beatty & Sons, Chipping Norton. ROLFE, W. D. I., CLARKSON, E. N. K. & PANCHEN, A. L. ( eds. ) ( 1994 ). Volcanism and Early Terrestrial Biotas. Transactions of the Royal Society of Edinburgh : Earth Sciences, 84. ROMER, A. S. ( 1922 ). The locomotor apparatus of certain primitive and mammal-like reptiles. Bulletin of the American Museum of Natural History 46, 517–606. ROMER, A. S. ( 1935 ). Early history of Texas Redbeds vertebrates. Bulletin of the Geological Society of America 46, 1597–1658. ROMER, A. S. ( 1946 ). The primitive reptile Limnoscelis restudied. American Journal of Sciences 244, 149–188. ROMER, A. S. ( 1947 ). Review of the Labyrinthodontia. Bulletin of the Museum of Comparative Zoology, Harvard College 99, 1–368. 302 Marcello Ruta, Michael I. Coates and Donald L. J. Quicke ROMER, A. S. ( 1950 ). The nature and relationships of the Paleozoic microsaurs. American Journal of Science 248, 628–654. ROMER, A. S. ( 1952 ). Fossil vertebrates of the Tri-State area. Art. 2. Late Pennsylvanian and Early Permian vertebrates of the Pittsburgh-West Virginia region. Annals of the Carnegie Museum 33, 1–112. ROMER, A. S. ( 1956 ). Osteology of the Reptiles. University of Chicago Press, Chicago. ROMER, A. S. ( 1957 ). The appendicular skeleton of the Permian embolomerous amphibian Archeria. Contributions of the Museum of Geology, University of Michigan 13, 103–159. ROMER, A. S. ( 1966 ). Vertebrate Paleontology. University of Chicago Press, Chicago. ROMER, A. S. ( 1969 ). A temnospondylous labyrinthodont from the Lower Carboniferous. Kirtlandia 6, 1–20. ROMER, A. S. ( 1970 ). A new anthracosaurian labyrinthodont from the Lower Carboniferous. Kirtlandia 10, 1–16. ROMER, A. S. & WITTER, R. V. ( 1942 ). Edops, a primitive rhachitomous amphibian from the Texas red beds. Journal of Geology 50, 925–960. ROSEN, D. E., FOREY, P. L., GARDINER, B. G. & PATTERSON, C. ( 1981 ). Lungfishes, tetrapods, paleontology, and plesiomorphy. Bulletin of the American Museum of Natural History 167, 159–276. RUTA, M. & COATES, M. I. ( in press ). Bones, molecules and crowntetrapod origins. In Telling the Evolutionary Time : Molecular Clocks and the Fossil Record ( eds. P. C. J. Donoghue and M. P. Smith ). Taylor & Francis, London. RUTA, M., MILNER, A. R. & COATES, M. I. ( 2001 ). The tetrapod Caerorhachis bairdi Holmes and Carroll from the Lower Carboniferous of Scotland. Transactions of the Royal Society of Edinburgh : Earth Sciences 92, 229–261. SALISBURY, B. A. & KIM, J. ( 2001 ). Ancestral state estimation and taxon sampling density. Systematic Biology 50, 557–564. SANDERSON, M. J. & DONOGHUE, M. J. ( 1989 ). Patterns of variation in levels of homoplasy. Evolution 43, 1781–1795. SÄVE-SÖDERBERGH, G. ( 1932 ). Preliminary note on Devonian stegocephalians from East Greenland. Meddelelser om Grønland 94, 1–107. SÄVE-SÖDERBERGH, G. ( 1934 ). Some points of view concerning the evolution of the vertebrates and the classification of this group. Arkiv for Zoologi 26, 1–20. SAWIN, H. J. ( 1941 ). The cranial anatomy of Eryops megacephalus. Bulletin of the Museum of Comparative Zoology, Harvard College 88, 407– 463. SCHOCH, R. R. ( 1992 ). Comparative ontogeny of Early Permian branchiosaurid amphibians from Southwestern Germany. Palaeontographica Abteilung A 222, 43–83. SCHOCH, R. R. ( 1995 ). Heterochrony in the development of the amphibian head. In Evolutionary Change and Heterochrony ( ed. K. J. McNamara ), pp. 107–124. John Wiley & Sons, New York. SCHOCH, R. R. ( 1998 ). Homology of cranial ossifications in urodeles. Neues Jahrbuch für Geologie und Paläontologie, Monatshefte 1998, 1–25. SCHOCH, R. R. & MILNER, A. R. ( 2000 ). Handbuch der Paläoherpetologie : Teil 3B, Stereospondyli. Pfeil, Munich. SCHÖNFELD, G. ( 1911 ). Branchiosaurus tener Schönfeld. Ein neuer Stegocephale aus dem Rotliegenden des nordwestlichen Sachsen. Sitzungsberichte und Abhandlungen Naturwissenschaftliche Gesellschaft Isis 1911, 19– 43. SCHULTZE, H.-P. & ARSENAULT, M. ( 1985 ). The panderichthyid fish Elpistostege : a close relative of tetrapods ? Palaeontology 28, 293–309. SCHULTZE, H.-P. & BOLT, J. R. ( 1996 ). The lungfish Tranodis and the tetrapod fauna from the Upper Mississippian of North America. In Studies on Carboniferous and Permian Vertebrates ( ed. Milner ), pp. 31–54. Special Papers in Palaeontology 52. SCHULTZE, H.-P. & FOREMAN, J. R. ( 1981 ). A new gymnarthrid microsaur from the Lower Permian of Kansas with a review of the tuditanomorph microsaurs ( Amphibia ). Occasional Papers of the Museum of Natural History, University of Kansas 91, 1–25. SCHULTZE, H.-P. & TRUEB, L. ( 1991 ). Origins of the Higher Groups of Tetrapods : Controversy and Consensus. Cornell University Press, Ithaca. SCOTLAND, R. & PENNINGTON, R. T. ( 2000 ). Homology and Systematics : Coding Characters for Phylogenetic Analysis. Taylor & Francis, London. SEQUEIRA, S. E. K. ( 1998 ). The cranial morphology and taxonomy of the saurerpetontid Isodectes obtusus comb. nov. ( Amphibia : Temnospondyli ) from the Lower Permian of Texas. Zoological Journal of the Linnean Society 122, 237–259. SERENO, P. C. ( 1998 ). A rationale on phylogenetic definitions, with application to the higher-level taxonomy of Dinosauria. Neues Jahrbuch für Geologie und Paläontologie, Abhandlungen 210, 41–83. SERENO, P. C. ( 1999 ). Definitions in phylogenetic taxonomy : critique and rationale. Systematic Biology 48, 329–351. SHUBIN, N. H. ( 1995 ). The evolution of paired fins and the origin of tetrapod limbs. Phylogenetic and transformational approaches. Evolutionary Biology 28, 39–86. SHUBIN, N. H. & ALBERCH, P. ( 1986 ). A morphogenetic approach to the origin and basic organisation of the tetrapod limb. Evolutionary Biology 20, 319–387. SHUBIN, N. H. & DAESCHLER, E. B. ( 2001 ). [Abstract] New Devonian tetrapod elements from the Catskill Formation of northcentral Pennsylvania. Journal of Vertebrate Paleontology 21 ( Suppl. 3 ), 101A. SHUBIN, N. H. & JENKINS, F. A. JR. ( 1995 ). An Early Jurassic jumping frog. Nature 377, 49–52. SHUBIN, N. H., TABIN, C. & CARROLL, S. ( 1997 ). Fossils, genes and the evolution of limbs. Nature 388, 639–648. SIDOR, C. A. ( 2001 ). Simplification as a trend in synapsid cranial evolution. Evolution 55, 1419–1442. SMIRNOV, S. ( 1986 ). The evolution of the urodele soundconducting apparatus. In Studies in Herpetology, Proceedings of the European Herpetological Meeting ( ed. Z. Rocek ), pp. 55–58. Charles University, Prague. SMITH, A. B. ( 1994 ). Systematics and the Fossil Record. Blackwell Science, London. SMITHSON, T. R. ( 1982 ). The cranial morphology of Greererpeton burkemorani Romer ( Amphibia : Temnospondyli ). Zoological Journal of the Linnean Society 76, 29–90. SMITHSON, T. R. ( 1985 ). The morphology and relationships of the Carboniferous amphibian Eoherpeton watsoni Panchen. Zoological Journal of the Linnean Society 85, 317– 410. SMITHSON, T. R. ( 1986 ). A new anthracosaur amphibian from the Carboniferous of Scotland. Palaeontology 29, 603–628. SMITHSON, T. R. ( 1989 ). The earliest known reptile. Nature 314, 676–678. SMITHSON, T. R. ( 1994 ). Eldeceeon rolfei, a new reptiliomorph from the Viséan of East Kirkton, West Lothian, Scotland. Transactions of the Royal Society of Edinburgh : Earth Sciences 84, 377–382. SMITHSON, T. R. ( 2000 ). Anthracosaurs. In Amphibian Biology, 4 : Palaeontology ( eds. H. Heatwole and R. L. Carroll ), pp. 1053–1063. Surrey Beatty & Sons, Chipping Norton. SMITHSON, T. R., CARROLL, R. L., PANCHEN, A. L. & ANDREWS, S. M. ( 1994 ). Westlothiana lizziae from the Viséan of East Early tetrapod relationships revisited Kirkton, West Lothian, Scotland, and the amniote stem. Transactions of the Royal Society of Edinburgh : Earth Sciences 84, 383– 412. SMITHSON, T. R. & ROLFE, W. D. I. ( 1990 ). Westlothiana gen. nov. : naming the earliest known reptile. Scottish Journal of Geology 26, 137–138. SORDINO, P. & DUBOULE, D. ( 1995 ). A molecular approach to the evolution of vertebrate paired appendages. Trends in Ecology and Evolution 11, 114–119. SORDINO, P., VAN DER HOEVEN, F. & DUBOULE, D. ( 1995 ). Hox gene expression in teleost fins and the origin of vertebrate digits. Nature 375, 678–681. STEYER, J. S. ( 2002 ). The first articulated trematosaur ‘ amphibian ’ from the Lower Triassic of Madagascar : implications for the phylogeny of the group. Palaeontology 48, 771–793. SUMIDA, S. S. ( 1997 ). Locomotor features of taxa spanning the origin of amniotes. In Amniote Origins : Completing the Transition to Land ( eds. S. S. Sumida and K. L. M. Martin ), pp. 353–398. Academic Press, London. SUMIDA, S. S. & LOMBARD, R. E. ( 1991 ). The atlas–axis complex in the Late Paleozoic diadectomorph amphibian Diadectes and the characteristics of the atlas–axis complex across the amphibian to amniote transition. Journal of Paleontology 65, 973–983. SUMIDA, S. S., LOMBARD, R. E. & BERMAN, D. S. ( 1992 ). Morphology of the atlas–axis complex of the late Palaeozoic tetrapod suborders Diadectomorpha and Seymouriamorpha. Philosophical Transactions of the Royal Society of London 336, 259–273. SWOFFORD, D. L. ( 1998 ). PAUP* : Phylogenetic Analysis Using Parsimony ( *and Other Methods ). Version 4.0b10. Sinauer Associates, Sunderland, MA. THOMSON, K. S. & BOSSY, K. H. ( 1970 ). Adaptive trends and relationships in early Amphibia. Forma et Functio 3, 7–31. THORLEY, J. L. & WILKINSON, M. ( 1999 ). Testing the phylogenetic stability of early tetrapods. Journal of Theoretical Biology 200, 343–344. THOROGOOD, P. ( 1991 ). The development of the teleost fin and implications for our understanding of tetrapod limb evolution. In Developmental Patterns of the Vertebrate Limb ( eds. J. R. Hinchliffe, J. M. Hurle and D. Summerbell ), pp. 347–354. Plenum, New York. THULBORN, R. A., WARREN, A. A., TURNER, S. & HAMLEY, T. ( 1996 ). Carboniferous tetrapods from Australia. Nature 381, 777–780. TICKLE, C. ( 1995 ). Vertebrate limb development. Current Opinions in Genetics and Development 5, 478–484. TRUEB, L. & CLOUTIER, R. ( 1991 ). A phylogenetic investigation of the inter- and intrarelationships of the Lissamphibia ( Amphibia : Temnospondyli ). In Origins of the Higher Groups of Tetrapods : Controversy and Consensus ( eds. H.-P. Schultze and L. Trueb ), pp. 223–313. Cornell University Press, Ithaca. TRUEMAN, J. W. H. ( 1998 ). Reverse successive wighting. Systematic Biology 47, 733–737. VAN TUINEN, M., PORDER, S. & HADLY, E. A. ( 2001 ). [Abstract] Putting confidence limits around molecular and fossil divergence dates. Journal of Vertebrate Paleontology 21 ( Suppl. 3 ), 110A. VAUGHN, P. P. ( 1962 ). The Paleozoic microsaurs as close relatives of reptiles, again. American Midland Naturalist 67, 79–84. VAUGHN, P. P. ( 1964 ). Vertebrates from the Organ Rock Shale of the Cutler Group, Permian of Monument Valley and vicinity, Utah and Arizona. Journal of Paleontology 38, 567–583. VAUGHN, P. P. ( 1966 ). Seymouria from the Lower Permian of Southeastern Utah, and possible sexual dimorphism in that genus. Journal of Paleontology 40, 603–612. 303 VAUGHN, P. P. ( 1969 ). Upper Pennsylvanian vertebrates from the Sangre de Cristo Formation of central Colorado. Los Angeles City Museum Contributions to Sciences 164, 1–27. VOROBYEVA, E. I. ( 1977 ). Morphology and nature of the evolution of crossopterygian fishes. Trudy Paleontologicheskogo Instituta 163, 1–240. VOROBYEVA, E. ( 1992 ). The Problem of the Terrestrial Vertebrate Origin. Nauka, Moscow. VOROBYEVA, E. I. ( 2000 ). Morphology of the humerus in the rhipidistian Crossopterygii and the origin of tetrapods. Paleontological Journal 34, 632–641. VOROBYEVA, E. & SCHULTZE, H.-P. ( 1991 ). Description and systematics of panderichthyid fishes with comments on their relationships to tetrapods. In Origins of the Higher Groups of Tetrapods : Controversy and Consensus ( eds. H.-P. Schultze and L. Trueb ), pp. 68–109. Cornell University Press, Ithaca. WALLISER, J. B. D. ( 1998 ). [Abstract] Analysis of the post-cranial skeleton of Tseajaia campi Vaughn. Journal of Vertebrate Paleontology 18 ( Suppl. 3 ), 84A. WALLISER, J. B. D. ( 1999 ). [Abstract] Phylogenetic relationship of Tseajaia campi Vaughn to other Paleozoic tetrapods based upon post-cranial characteristics. Journal of Vertebrate Paleontology 19 ( Suppl. 3 ), 82A. WATSON, D. M. S. ( 1926 ). Croonian Lecture – The evolution and origin of the Amphibia. Philosophical Transactions of the Royal Society of London, Series B 214, 189–257. WATSON, D. M. S. ( 1929 ). The Carboniferous Amphibia of Scotland. Paleontologia Hungarica 1, 219–252. WATSON, D. M. S. ( 1940 ). The origin of frogs. Transactions of the Royal Society of Edinburgh 60, 195–231. WELLSTEAD, C. F. ( 1982 ). A Lower Carboniferous aı̈stopod amphibian from Scotland. Palaeontology 25, 193–208. WELLSTEAD, C. F. ( 1991 ). Taxonomic revision of the Lysorophia, Permo-Carboniferous lepospondyl amphibians. Bulletin of the American Museum of Natural History 209, 1–90. WHITE, T. E. ( 1939 ). Osteology of Seymouria baylorensis ( Broili ). Bulletin of the Museum of Comparative Zoology, Harvard College 85, 325–409. WHITEAVES, J. F. ( 1881 ). On some remarkable fossil fishes from the Devonian rocks of Scaumenac Bay, P. Q., with descriptions of a new genus and three new species. Canadian Naturalist 2, 27–35. WILKINSON, M. ( 1995 ). Coping with abundant missing entries in phylogenetic inference using parsimony. Systematic Biology 44, 501–514. WILLISTON, S. W. ( 1911 ). A new family of reptiles from the Permian of New Mexico. American Journal of Science 181, 378–398. WILLISTON, S. W. ( 1912 ). Restoration of Limnoscelis, a cotylosaur reptile from New Mexico. American Journal of Science 184, 457– 468. WRIGHT, E. P. & HUXLEY, T. H. ( 1866 ). On a collection of fossils from the Jarrow Colliery, Kilkenny. Geological Magazine 3, 165–171. WYMAN, J. ( 1858 ). On some remains of batrachian reptiles discovered in the Coal formation of Ohio, by Dr J. S. Newberry and C. M. Wheatley. American Journal of Science 2, 158–163. WYSS, A. R. & MENG, J. ( 1996 ). Application of phylogenetic taxonomy to poorly resolved crown clades : a stem-modified node-based definition of Rodentia. Systematic Biology 45, 559–568. XING, X., ZHONG-he, Z. & PRUM, R. O. ( 2001 ). Branched integumental structures in Sinornithosaurus and the origin of feathers. Nature 410, 200–204. 304 Marcello Ruta, Michael I. Coates and Donald L. J. Quicke YATES, A. M. & WARREN, A. A. ( 2000 ). The phylogeny of the ‘ higher ’ temnospondyls ( Vertebrata : Choanata ) and its implications for the monophyly and origins of the Stereospondyli. Zoological Journal of the Linnean Society of London 128, 77–121. ZARDOYA, R. & MEYER, A. ( 2001 ). Vertebrate phylogeny : limits of inference of mitochondrial genome and nuclear rDNA sequence data due to an adverse phylogenetic signal/noise ratio. In Major Events in Early Vertebrate Evolution : Palaeontology, Phylogeny, Genetics and Development ( ed. P. E. Ahlberg ), pp. 135–155. Taylor & Francis, London. ZHANG, F., LI, Y. & WAN, X. ( 1984 ). A new occurrence of Permian seymouriamorphs in Xinjiang, China. Vertebrata PalAsiatica 22, 294–304. ZHU, M. & SCHULTZE, H.-P. ( 2001 ). Interrelationships of basal osteichthyans. In Major Events in Early Vertebrate Evolution : Palaeontology, Phylogeny, Genetics and Development ( ed. P. E. Ahlberg ), pp. 289–314. Taylor & Francis, London. XII. APPENDIX 1. LIST OF SPECIMENS EXAMINED (FOR INSTITUTIONAL ABBREVIATIONS SEE THE RELEVANT LITERATURE ) Acanthostega gunnari: several specimens as listed by Clack (1989, 1994 a, b, 1998b), Coates & Clack (1990, 1991), Coates (1996) and Ahlberg & Clack (1998). Acherontiscus caledoniae: RSM 1967/13/1 (Carroll, 1969b). Adelogyrinus simorhynchus: NMS G 1895/63/7, 1889/101/17 (Andrews & Carroll, 1991). Adelospondylus watsoni: NMS G 1885/57/51 (Andrews & Carroll, 1991). Balanerpeton woodi: BMNH R. 10952-5, 12014-6; NMS G 1985/4/1-2, 6, 1987/7/32-33, 35-36, 1990/79/1, 3-4, 1991/47/2, 1992/14/2, 1992/48/1-2 (Milner & Sequeira, 1994). Baphetes cf. B. kirkbyi: BMNH R. 9663 (Milner & Lindsay, 1998). Batrachiderpeton reticulatum: HM G25-28/30 (Bossy & Milner, 1998). Brachydectes newberryi: BMNH R. 2544 (Wellstead, 1991). Caerorhachis bairdi: cast of MCZ 2271 (Holmes & Carroll, 1977). Colosteus scutellatus: BMNH R. 2547, 2548, 2660, 2664, 9949 (Hook, 1983). Crassigyrinus scoticus: RSM 1859.33.104, BMNH R. 10000 (Clack, 1998c). Dendrerpeton acadianum: BMNH R. 436, 439, 4158, 4163-4165, 4167, 4553, 4555 (Milner, 1980). Discosauriscus austriacus: many of the specimens in the collection of the Zoological Institute, Faculty of Natural Sciences, Comenius University, Bratislava (Klembara, 1997; Klembara & Bartı́k, 2000). Dolichopareias disjectus: NMS G 1950/56/7, 1881/43/37 (Andrews & Carroll, 1991). Greererpeton burkemorani: CMNH 10931, 10939, 11034, 11036, 11068-70, 11072, 11073, 11079, 11082, 11090, 11092, 11093, 11095, 11113, 11129-33, 11219, 11220, 11231-4, 11238, 11240, 11241, 11319, 11320, 11073 (Smithson, 1982; Godfrey, 1989). Hyloplesion longicostatum: RSM 1899.32.3 plus several galvanotypes in the collections of the Grant Museum of Zoology, University College London, UK (Carroll & Gaskill, 1978). Megalocephalus pachycephalus: BMNH R. 2363, 2366 (Beaumont, 1977). Microbrachis pelikani: RSM 1898.105.26 plus several galvanotypes in the collections of the Grant Museum of Zoology, University College London, UK and latex peels in the collections of the Department of Ecology, Comenius University, Bratislava, Slovakian Republic (Carroll & Gaskill, 1978). Oestocephalus amphiuminum: BMNH 2657a, 2673 (Carroll, 1998). Pantylus cordatus: plaster cast of skull of MCZ 2040 in the collections of the Grant Museum of Zoology, University College London, UK (Carroll & Gaskill, 1978). Phlegethontia linearis, Ptyonius marshii, Sauropleura pectinata, Scincosaurus crassus, Urocordylus wandesfordii: several galvanotypes in the collections of the Grant Museum of Zoology, University College London, UK. Proterogyrinus scheelei: CMNH 10938, 10950, 11035, 11067, 11091, 11111, 11112 (Holmes, 1984). Tulerpeton curtum: PIN 2921/7a-f, 16-30, 63, 135, 448, 865, 866 (Lebedev & Coates, 1995). Westlothiana lizziae: NMS G 1990/72/1; 1991/47/1 (Smithson et al., 1994). XIII. APPENDIX 2. CHARACTER LIST (1) Cranial skeleton (a) Skull table Premaxilla 1. PREMAX 1. Absence (0) or presence (1) of alary process. This and the following two characters describe conditions of the backward-pointing, triangular to digitiform processes of the posterodorsal margins of the premaxillae. These processes are found in many derived temnospondyls (e.g. some trimerorhachoids, eryopoids; dissorophoids) as well as in several living and fossil lissamphibians (Milner, 1990; Milner & Sequeira, 1994, 1998; Holmes et al., 1998). They either overlap the nasals (the latter often show impressions for the alary processes), as in several temnospondyls and lissamphibians, or form a butt joint with these, as in albanerpetontids (Gardner, 2001). The alary processes are here considered to be distinct from the posterodorsal nasal rami of the premaxillae, which appear remarkably well developed in several taxa, including lysorophids (Bolt & Wassersug, 1975) and captorhinids (Modesto, 1998), and conjoined along the dorsal skull midline. Ontogenetic shifts of the nasal rami in a mesial direction (and resulting obliteration of the interpremaxillary space) has been documented in Discosauriscus austriacus (Klembara, 1997). Early tetrapod relationships revisited 2. PREMAX 2. Alary process shorter than wide (0) or as long as/longer than wide (1). See character 1 above. In some temnospondyls (e.g. Doleserpeton, Micromelerpeton) and primitive lissamphibians (e.g. Karaurus, Valdotriton), the alary processes are shaped like equilateral or isosceles triangles (Bolt, 1969; Ivakhnenko, 1978; Boy, 1995; Evans & Milner, 1996), whereas in others (e.g. Dendrerpeton), the processes are less well developed and extend backward only for a short distance (Holmes et al., 1998). 3. PREMAX 3. Alary process less than (0) or at least one-third as wide as premaxillae (1). See character 1 above. The alary processes can be rather small relative to the width of the premaxillae (e.g. Eryops; Sawin, 1941), or broad and only slightly less wide than the premaxillae (e.g. Apateon; Schoch, 1992). 4. PREMAX 4. Premaxillae without (0) or with (1) flat, expanded anteromedial dorsal surface and marginal elongation. This is one of the characters used by Milner & Sequeira (1994, 1998) to characterize edopoids (see also Godfrey & Holmes, 1995). In edopoids, the anteromedial portion of the premaxillae is a broad, flat sheet of bone, and the bones show an elongate lateral margin bordering the snout. 5. PREMAX 7. Premaxillae more (0) or less than (1) twothirds as wide as skull. This is a modified version of one of Gauthier et al.’s (1988 b) characters, found ubiquitously among ‘reptiliomorphs’, and which characterizes also several lepospondyls and some primitive lissamphibians (e.g. Eocaecilia; Carroll, 2000). Narrow premaxillae, even in stem-amniotes showing broad and spade-shaped snouts (e.g. Discosauriscus; Klembara, 1997), contrast with the broad premaxillae of temnospondyls and several stem-tetrapods. 6. PREMAX 8. Mouth subterminal so that anteriormost surface of premaxilla faces ventrally: absent (0) or present (1). This character refers to the distinctly oblique anteroventral surface of the premaxillae in several microsaurs (e.g. Batropetes; Cardiocephalus; Euryodus; Micraroter; Pantylus; Pelodosotis; Rhynchonkos; Carroll & Gaskill, 1978; Carroll, 1991a), some diadectomorphs (e.g. Limnoscelis; Williston, 1911; Romer, 1946; Heaton, 1980) and captorhinomorphs (e.g. Captorhinus; Fox & Bowman, 1966; Heaton, 1979). It confers a pointed aspect to the tip of the snout, as described by Laurin (1998 b), and may be related to burrowing habits, at least in some taxa. 7. PREMAX 9. Absence (0) or presence (1) of shelf-like premaxilla-maxilla contact mesial to tooth row on palate. In Ichthyostega ( Jarvik, 1980, 1996), Crassigyrinus (Clack, 1996, 1998 c) and Greererpeton (Smithson, 1982), the ventral surfaces of premaxilla and maxilla form a mesially projecting surface and the two bones contact each other in a mesial position relative to the 305 marginal dentition. The distribution of this character conforms to that of recent cladistic analyses by Clack (1998c, 2001). Anterior tectal 8. TEC 1. Presence (0) or absence (1) of anterior tectal. As pointed out by Clack (1998c), the distinction between the anterior tectal and the septomaxilla is disputed. When the two elements are scored as equivalent, following Clack’s (1998 b) example, the results of a PAUP* analysis are identical to those of the original run. The scoring of both elements conforms to Clack’s (1998c, 2001) analyses, with Acanthostega and Ichthyostega scored as possessing an anterior tectal, but lacking a septomaxilla. Lateral rostral 9. LAT ROS 1. Presence (0) or absence (1) of lateral rostral. As in the case of the anterior tectal (see character 8 above), we follow Clack (1998 c, 2001) in coding for the presence of a lateral rostral in Ichthyostega [see also Jarvik (1980, 1996) and Carroll (1995)]. Septomaxilla 10. SPTMAX 1. Absence (0) or presence (1) of septomaxilla. As explained under character 8 above, the septomaxilla and anterior tectal are here treated as separate elements (Clack, 1998c, 2001), with Acanthostega and Ichthyostega scored as lacking a septomaxilla. 11. SPTMAX 2. Septomaxilla not a detached ossification inside nostril (0) or a detached ossification (1). The occurrence of a detached septomaxilla inside the nostril (i.e. this element is not part of the dermal skull roof) characterizes several derived temnospondyls and certain stemamniotes (e.g. Discosauriscus, Limnoscelis, Seymouria) (White, 1939; Romer, 1946; Milner & Sequeira, 1994; Laurin, 1995, 1996b; Klembara, 1997) Nasal 12. NAS 1. Absence (0) or presence (1) of paired dorsal nasals. At a post-panderichthyid level of organization, paired nasals are widespread in tetrapods (Panchen & Smithson, 1988; Carroll, 1995; Coates, 1996; Ahlberg, 1998; Clack, 1998 c, 2001), except in those taxa (e.g. derived diplocaulid nectrideans) in which only one element is found in the position usually occupied by paired nasals in most other tetrapods (A. C. Milner, 1980; Bossy & Milner, 1998). 13. NAS 2. Nasals more (0) or less than (1) one-third as long as frontals. This is one of several characters describing patterns of elongation in the preorbital region of the skull roof. Its distribution in the taxa surveyed in this study is rather irregular. In adelospondyls, aı̈stopods, colosteids, some microsaurs and certain urocordylids, the nasals are greatly reduced in size relative to the 306 Marcello Ruta, Michael I. Coates and Donald L. J. Quicke frontals. This is especially evident in adelospondyls, aı̈stopods and colosteids. 14. NAS 5. Absence (0) or presence (1) of condition: nasals broad plates delimiting most of the posterodorsal and mesial margins of nostrils and with lateral margins diverging abruptly anteriorly. As defined here, this character describes the snout morphology in several dissorophoids and Karaurus (Boy, 1972, 1974, 1978, 1985, 1986, 1987, 1995; Bolt, 1969, 1977, 1979, 1991; Ivakhnenko, 1978; Schoch, 1992), in which the nasals are not simply expanded and flat sheets of bones, but contribute to most of the posterodorsal and mesial margins of the nostrils. In addition, the anterior part of their lateral margins diverges markedly anterolaterally, so that the nasals increase abruptly in width anteriorly. 15. NAS 6. Parietal/nasal length ratio less than (0) or greater than 1.45 (1). This character refers to a pattern of skull roof bone proportions that matches in part the distribution of character 13 above, although its derived condition occurs sporadically in some lepospondyl groups and temnospondyls. It also represents one of the distinguishing features of adelospondyls and colosteids (Smithson,1982;Hook,1983;Andrews&Carroll,1991). Prefrontal 16. PREFRO 2. Prefrontal less than (0) or more than (1) three times longer than wide. Together with the previous character, the elongation of the prefrontals characterizes adelospondyls and colosteids (Smithson, 1982; Hook, 1983; Andrews & Carroll, 1991; Milner, 1993), but is found also in some representatives of other groups (e.g. aı̈stopods, baphetids; embolomeres; temnospondyls; lysorophids; microsaurs; albanerpetontids). 17. PREFRO 3. Absence (0) or presence (1) of condition: antorbital portion of prefrontal expanded to form a near-equilateral triangular lamina. The presence of a short and broadly triangular preorbital portion of the prefrontal, such as is observed in dissorophoids and Karaurus (Boy, 1972, 1974, 1978, 1985, 1986, 1987, 1995; Bolt, 1969, 1977, 1979, 1991; Ivakhnenko, 1978; Schoch, 1992) is not related simply to paedomorphic shortening of the preorbital skull region. Several short-snouted taxa (e.g. Discosauriscus; Klembara, 1997) do not show the derived condition of this character. 18. PREFRO 6. Prefrontal not sutured with premaxilla (0) or sutured (1). Acherontiscus, colosteids and albanerpetontids show the derived condition of this feature (Carroll, 1969b; Smithson, 1982; Hook, 1983). Although the snout of Acherontiscus is poorly preserved, a prefrontal– premaxilla contact is deduced to have existed, based on the position and morphology of the anteriormost part of the prefrontal, which sends an anteroventral flange lying immediately in front of the lacrimal. 19. PREFRO 7. Prefrontal without (0) or with (1) stout, lateral outgrowth. The outgrowth in question is a stout process marking the posteromesial boundary of the antorbital vacuity in baphetids other than Spathicephalus (Beaumont, 1977; Beaumont & Smithson, 1998), and is introduced here as a distinct character to distinguish between several kinds of irregular orbit outlines (i.e. neither elliptical nor circular) among Palaeozoic tetrapods (see also Clack, 1987 b, 1998a, c, 2001). 20. PREFRO 8. Absence (0) or presence (1) of condition: prefrontal entering nostril margin. The derived state of this character occurs, among others, in some primitive lissamphibians, certain derived temnospondyls, Greererpeton, and some microsaurs, lysorophids, nectrideans and aı̈stopods (Smithson, 1982; Carroll, 1995; Bossy & Milner, 1998; Laurin, 1998b). We code for this character separately instead of as a state in a multistate treatment of the relationships between nostril and surrounding bones, as Carroll (1995) did. This is to prevent the introduction of constraints in the derivation of different bone arrangements in the snout from one another. 21. PREFRO 9. Prefrontal not sutured with maxilla (0) or sutured (1). Some nectrideans and adelospondyls, as well as colosteids and certain primitive lissamphibians, exhibit a lateral contact between prefrontal and maxilla (Smithson, 1982; Carroll, 1995; Bossy & Milner, 1998). As in the case of the previous character, the mutual relationships between these bones are treated separately from the conditions of other elements of the snout. This is also one of the conditions described by Clack (1998c, 2001) regarding the lacrimal contribution (or lack thereof ) to the margin of the nostril. 22. PREFRO 10. Prefrontal contributes to more (0) or less than (1) half of anteromesial orbit margin. The derived condition of this character is observed in various derived temnospondyls and primitive lissamphibians (Boy, 1972, 1974, 1978, 1985, 1986, 1987, 1995; Bolt, 1969, 1977, 1979, 1991; Ivakhnenko, 1978; Schoch, 1992). Lacrimal 23. LAC 1. Presence (0) or absence (1) of lacrimal. Phlegethontia (Anderson, in press) and Valdotriton (Evans & Milner, 1996) lack a lacrimal as a separately ossified element. 24. LAC 2. Lacrimal not allowing (0) or allowing (1) contact between prefrontal and jugal. The prefrontal-jugal contact excludes the lacrimal from the orbit margin, as in Acanthostega (Clack, 1994 a), Archeria (Holmes, 1989), Chenoprosopus (Hook, 1993; Milner & Sequeira, 1994, 1998), Cochleosaurus (Rieppel, 1980; Milner & Sequeira, 1994, 1998), Crassigyrinus (Panchen, 1985; Clack, 1998c), Diplocaulus (A. C. Milner, 1980), Edops Early tetrapod relationships revisited (Romer & Witter, 1942), Eoherpeton (Panchen, 1975; Smithson, 1985), Eryops (Sawin, 1941), Ichthyostega ( Jarvik, 1980, 1996), Micromelerpeton (Boy, 1995), Pholiderpeton attheyi and P. scutigeum (Panchen, 1972; Clack, 1987a),Ventastega (Ahlberg et al., 1994). However, in Isodectes (Sequeira, 1998) the lacrimal separates the prefrontal from the jugal but it fails to contact the orbit margin due to an intervening exposure of the palatine. 25. LAC 4. Lacrimal without (0) or with (1) dorsomesial digitiform process. A dorsomesial digitiform process of the lacrimal is observed in Brachydectes (Wellstead, 1991), such tuditanomorphs as Cardiocephalus, Euryodus, some specimens of Micraroter and Tuditanus (Carroll & Gaskill, 1978), and in urocordylid nectrideans (A. C. Milner, 1980; Bossy & Milner, 1998). 26. LAC 5. Lacrimal without (0) or with (1) V-shaped emargination along its posterior margin. Together with character 19, a deep V-shaped notch of the lacrimal characterizes the anterior portion of the antorbital vacuity of various baphetids (Beaumont, 1977; Beaumont & Smithson, 1998). 27. LAC 6. Absence (0) or presence (1) of condition: portion of lacrimal lying anteroventral to orbit abbreviated (1). As in the case of character 17, an abbreviated preorbital region of the lacrimal is found in several dissorophoids. However, the derived state of this character is also found in urocordylid and diplocaulid nectrideans, Batropetes, Oestocephalus and adelospondyls. Not all shortsnouted tetrapods exhibit such a state, as demonstrated by Discosauriscus (Klembara, 1997). Maxilla 28. MAX 3. Maxilla extending behind level of posterior margin of orbit (0) or terminates anterior to it (1). A rearward extension of the maxilla is a widespread feature of several early tetrapods, and contrasts with the situation of some diplocaulid nectrideans, several microsaurs, embolomeres, primitive crown-amniotes and such stem-amniote groups as gephyrostegids, seymouriamorphs and diadectomorphs (Carroll & Gaskill, 1978; Gauthier et al., 1988b; Carroll, 1991b; Klembara, 1997; Lee & Spencer, 1997; Bossy & Milner, 1998). 29. MAX 5. Maxilla not entering (0) or entering (1) orbit margin. The derived condition of this character is widespread among early tetrapods, and occurs in some adelospondyls (Andrews & Carroll, 1991), some dissorophoids and primitive lissamphibians (Boy, 1972, 1974, 1978, 1985, 1986, 1987, 1995; Bolt, 1969, 1977, 1979, 1991; Ivakhnenko, 1978; Schoch, 1992; Evans & Milner, 1996; Carroll, 2000), aı̈stopods (Wellstead, 1982; Carroll, 1998; Anderson, in press), some urocordylids and diplocaulids (Bossy & Milner, 1998). 307 It is not directly related to small size of the skull and presence of large orbits, as shown by some amphibamids and seymouriamorphs (e.g. Bolt, 1969; Laurin, 1996a, c; Klembara, 1997). 30. MAX 6. Maxillary arcade closed (0) or open (1) posteriorly. A gap between the maxilla and the jugal occurs in primitive caudates (Ivakhnenko, 1978; Evans & Milner, 1996) and branchiosaurids (Boy, 1986). The maxillary arcade is closed in albanerpetontids, as demonstrated by Gardner (1999, 2001). The skull emargination of lysorophids (Wellstead, 1991) is of a different pattern, as explained in the text (see also Carroll, 2001). 31. MAX 7. Dorsal maxillary margin not forming (0) or forming (1) distinct dorsal ‘step’. In pantylids (Carroll & Gaskill, 1978; Berman et al., 1988), the pronounced facial process is separated from the posterior part of the dorsal margin of the maxilla by a sharp dorsal bend. 32. Max 8. Absence (0) or presence (1) of condition: maxillary facial process shaped like a rectangular flange. A (sub)rectangular facial process of the maxilla occurs in primitive lissamphibians and several dissorophoids (Boy, 1972, 1974, 1978, 1985, 1986, 1987, 1995; Bolt, 1969, 1977, 1979, 1991; Ivakhnenko, 1978; Rage & Rocek, 1986, 1989; Schoch, 1992; Evans & Milner, 1996; Carroll, 2000; Rocek & Rage, 2000b). 33. Max 9. Posterior end of maxilla not lying (0) or lying level with (1) posterior end of vomers. The derived condition of this character is one of the synapomorphies uniting scincosaurids and diplocaulids (A. C. Milner, 1980; Bossy & Milner, 1998). Frontal 34. FRO 1. Frontal unpaired (0) or paired (1). Paired frontals occur commonly among early tetrapods including panderichthyids (Carroll, 1995; Laurin, 1998b), except in derived diplocaulids, albanerpetontids, Phlegethontia and Sauropleura (A. C. Milner, 1980; Bossy & Milner, 1998; Gardner, 1999, 2001; Anderson, in press). 35. FRO 2. Frontal shorter than (0), longer than (1), or subequal to (2) parietals. This is one of several characters describing the relative proportions of the bones in the skull roof and the pattern of elongation of its inter-, pre- and postorbital regions (see also Milner, 1993). 36. FRO 4. Frontal excluded from (0) or contributing to (1) margin of orbit. The coding for this character follows mainly Laurin (1998 b). The derived condition is acquired in parallel in several dissorophoids and primitive lissamphibians, as well as in gephyrostegids, crownamniotes, derived diplocaulid nectrideans and some microsaurs. 308 Marcello Ruta, Michael I. Coates and Donald L. J. Quicke 37. FRO 5. Frontals and parietals not co-ossified as frontoparietal (0) or co-ossified (1). The derived state characterizes salientians only (review in Milner, 1988). 38. FRO 6. Absence (0) or presence (1) of condition: anterior margin of frontals deeply wedged between posterolateral margins of nasals. In its derived state, this character occurs in adelospondyls, colosteids, Gephyrostegus, urocordylid nectrideans and Whatcheeria (Carroll, 1970; A. C. Milner, 1980; Smithson, 1982; Hook, 1983; Andrews & Carroll, 1991; Lombard & Bolt, 1995; Bossy & Milner, 1998). In these taxa, the posteror margins of the nasals are more or less deeply excavated to accommodate the anteriormost part of the frontals. Parietal 39. PAR 1. Absence (0) or presence (1) of parietal/tabular suture. Panchen & Smithson (1987, 1988), Gauthier et al. (1988 b) and most subsequent authors have placed emphasis on the mutual spatial relationships of the bones in the posterior and posterolateral parts of the skull table, and have considered them to be indicators of a deep dichotomy among primitive tetrapods. A parietal-tabular suture appears on the tetrapod stem and is maintained in crown-amniotes in Laurin’s (1998 b) analysis. Conversely, our study shows that the contact between these two bones characterizes a large portion of the ‘reptiliomorph’ branch of the tetrapod tree (see discussion above). 40. PAR 2. Absence (0) or presence (1) of parietal/postorbital suture. The derived state of this character appears as a transitional stem-tetrapod feature, and is also acquired in a large assemblage of stem-amniotes (with reversals among most tuditanomorphs, seymouriamorphs and Scincosaurus (Carroll & Gaskill, 1978; A. C. Milner, 1980; Laurin, 1995, 1996b; Klembara, 1997; Bossy & Milner, 1998). 41. PAR 4. Anterior margin of parietal lying in front of (0), level with (1), or behind (2) orbit midlength. Together with character 35, this character relates the proportions of various skull elements to each other. Its distribution is less random than that of character 35. The derived state 1 is found in some Devonian tetrapods, Whatcheeria and Kotlassia. The derived state 2 is found in some stemtetrapods (Tulerpeton, colosteids, Crassigyrinus and baphetids) as well as in almost all crown-tetrapods. 42. PAR 5. Anteriormost third of parietals not wider than frontals (0) or at least marginally wider (1). With the exception of Eusthenopteron, Panderichthys, Platyrhinops, Eocaecilia and Triadobatrachus, all other taxa show the derived condition of this feature. 43. PAR 6. Parietals more than two and a half times as long as wide (0) or less (1). Eusthenopteron, Panderichthys, colosteids, trimerorhachoids, lysorophids, adelospondyls, aı̈stopods and urocordylids show the plesiomorphic state for this character (see also characters 35 and 41 above). 44. PAR 7. Absence (0) or presence (1) of parietal/squamosal suture on skull roof. Crown-lissamphibians, crown-amniotes, Batropetes, microbrachomorphs and diplocaulids display the derived condition of this character. This is one of several characters describing the contact between the skull table and the cheek (Carroll, 1995; Coates, 1996; Clack, 1998 c, 2001; Laurin, 1998b). 45. PAR 8. Parietal-frontal suture not strongly interdigitating (0) or strongly interdigitating (1). The derived state of this character is not a simple function of the size or degree of ossification of the skull roof bones, and its distribution is rather discontinuous (e.g. most lepospondyls; some ‘reptiliomorphs’, some temnospondyls, colosteids). 46. PAR 9. Parietal-postparietal suture not strongly interdigitating (0) or strongly interdigitating (1). The occurrences of the derived condition of this character match closely those of the previous character. Postparietal 47. POSPAR 1. Presence (0) or absence (1) of postparietals. Salientians, caudates, Batropetes, Scincosaurus and Phlegethontia all lack ossified postparietals (A. C. Milner, 1980; Milner, 1988, 2000; Carroll, 1991 a; Bossy & Milner, 1998; Rocek & Rage, 2000b; Anderson, in press). 48. POSPAR 2. Postparietals paired (0) or unpaired (1). The derived condition is found only in Ichthyostega (Jarvik, 1980, 1996), diadectomorphs (Romer, 1946; Berman et al., 1998; Berman, 2000) and Odonterpeton ( fide Carroll & Gaskill, 1978). 49. POSPAR 3. Postparietal less than (0) or more than (1) four times wider than long. In its derived state, this character is observed in several post-embolomere ‘reptiliomorphs’, in Microbrachis and Hyloplesion and in several of the temnospondyls that lie crownward of Balanerpeton [see Milner (1990) and Yates & Warren (2000) for analysis of this character in temnospondyls]. 50. POSPAR 4. Postparietals without (0) or with (1) median lappets. The median posterior lappets of the postparietals occur in Crassigyrinus, Whatcheeria, embolomeres and Ptyonius. The distribution of this character follows Clack (1998 c), except for the coding of Dendrerpeton (see Holmes et al., 1998). 51. POSPAR 5. Absence (0) or presence (1) of postparietal/ exoccipital suture. The relationships between the occiput and the skull table have been dealt with extensively in the literature on early tetrapods (e.g. Smithson, 1985; Panchen & Smithson, 1987, 1988) Carroll (1995) recognized no fewer than 12 states describing the nature of the skull table-occiput contact. Berman’s (2000) Early tetrapod relationships revisited analysis of occipital characters is followed in part here, since it simplifies to a greater degree the known spatial relationships of the otic capsules and supraoccipital (where present). The derived condition of character 51 is found among most lepospondyls, temnospondyls, a few ‘reptiliomorphs’ and colosteids. 52. POSPAR 6. Postparietals not entirely on occipital surface (0) or entirely on this surface (1). The derived state of this character is shared by diadectomorphs and crownamniotes (Berman, 2000). 53. POSPAR 7. Postparietals without (0) or with (1) posteroventrally sloping occipital exposure. A gently sloping surface at the back of the postparietals characterizes some tuditanomorphs and lysorophids (Carroll & Gaskill, 1978; Wellstead, 1991). 54. POSPAR 8. Postparietals without (0) or with (1) sinuous posterior ridge. A sinuous ridge runs across the posterior part of the postparietals in ostodolepids (Carroll & Gaskill, 1978). 55. POSPAR 9. Postparietals without (0) or with (1) broad, concave posterior emargination. This character is taken from A. C. Milner’s (1980) analysis of nectridean interrelationships and is shared by some diplocaulids. 56. POSPAR 10. Nasals not smaller than postparietals (0) or smaller (1). Nasals which are comparatively much smaller than the postparietals occur in colosteids, aı̈stopods, nectrideans (except urocordylids) and adelospondyls (A. C. Milner, 1980; Wellstead, 1982; Andrews & Carroll, 1991; Bossy & Milner, 1998; Carroll, 1998; Anderson, in press). Postfrontal 57. POSFRO 1. Presence (0) or absence (1) of postfrontal. Albanerpetontids (Gardner, 2001), salientians (Rocek & Rage, 2000b), caudates (Evans & Milner, 1996; Milner, 2000) and lysorophids all lack ossified postfrontals. 58. POSFRO 3. Postfrontal not contacting tabular (0) or contacting it (1). The derived condition of this character describes the dermal skull roof configuration of tuditanomorph microsaurs (Carroll & Gaskill, 1978), and is acquired in parallel by Scincosaurus (A. C. Milner, 1980; Bossy & Milner, 1998). 59. POSFRO 4. Absence (0) or presence (1) of condition: posterior margin of postfrontal lying flush with posterior jugal margin. The posterior margins of the postfrontal and jugal lie approximately at the same transverse level in Leptorophus (Boy, 1972), Paleothyris and Petrolacosaurus (Clark & Carroll, 1973; Reisz, 1977, 1981), Diplocaulus (A. C. Milner, 1980; Bossy & Milner, 1998), Phlegethontia (Anderson, in press), as well as in ostodolepid, gymnarthrid and rhynchonkid microsaurs (Gregory et al., 1956; Carroll & Gaskill, 1978). 309 Intertemporal 60. INTEMP 1. Intertemporal present (0) or absent (1) as a separate ossification. The distribution of the intertemporal is problematic, as recognized by Clack (1998c). Intertemporal presence has been considered to be primitive for tetrapods, because it has been homologized with the intertemporal of ‘osteolepiforms’. Panderichthys (Vorobyeva & Schultze, 1991) and Greererpeton (Smithson, 1982) have been coded as polymorphic for this character. Under ACCTRAN and, partially, under DELTRAN optimizations, intertemporal absence appears as a transitional feature encompassing a series of crownward stem-tetrapods (Acanthostega, Ichthyostega and Colosteus; Clack, 1994a; Jarvik, 1980, 1996; Hook, 1983). An intertemporal is present in the apical part of the tetrapod stem-group, as well as the basal portions of the lissamphibian and amniote stem-groups. However, we note that Megalocephalus lacks an intertemporal (Beaumont, 1977). The bone is lost again in the clade including crown-lissamphibians and temnospondyls more derived than Dendrerpeton (Milner, 1988, 1990). It disappears also in Solenodonsaurus (Carroll, 1970; Laurin & Reisz, 1999), diadectomorphs (Romer, 1946; Berman et al., 1998), crown-amniotes (Fox & Bowman, 1966; Clark & Carroll, 1973; Reisz, 1977, 1981), Westlothiana (Smithson et al., 1994) and lepospondyls (Carroll & Gaskill, 1978; A. C. Milner, 1980; Andrews & Carroll, 1991; Wellstead, 1991; Bossy & Milner, 1998). Carroll & Gaskill (1978) discussed the possibility that the intertemporal may have been incorporated in surrounding skull roof bones in the ancestry of the two major groups of microsaurs. Putative differing fusion patterns provide the basis for the distinction between tuditanomorphs (intertemporal-postfrontal fusion) and microbrachomorphs (intertemporal-parietal fusion). We question the homology of the intertemporals of limbed tetrapods, which seem to be anamestic, and the canalbearing intertemporals which are widespread among basal osteichthyans. Discrete coding for these alternative conditions is likely to provide a more informative signal. 61. INTEMP 2. Intertemporal not interdigitating with cheek (0) or interdigitating (1). The plesiomorphic condition, as found in Crassigyrinus (Clack, 1998c, 2001; Paton et al., 1999), is acquired in parallel in Trimerorhachis (Case, 1935), some embolomeres (Holmes, 1984, 1989; Smithson, 1985; Clack, 1987a), gephyrostegids (Carroll, 1970; Boy & Bandel, 1973) and at least one seymouriamorph (Klembara, 1997). 62. INTEMP 3. Intertemporal not contacting squamosal (0) or contacting it (1). Whatcheeria (Lombard & Bolt, 1995) and seymouriamorphs (White, 1939; Bystrow, 1944; 310 Marcello Ruta, Michael I. Coates and Donald L. J. Quicke Laurin, 1996b; Klembara, 1997) show the derived condition of an intertemporal-squamosal suture. Supratemporal 63. SUTEMP 1. Presence (0) or absence (1) of supratemporal. A separately ossified supratemporal is missing in crown-lissamphibians (Carroll, 2000; Milner, 2000; Rocek & Rage, 2000a), as well as in microsaurs (Carroll & Gaskill, 1978), lysorophids (Wellstead, 1991), adelospondyls (Andrews & Carroll, 1991), scincosaurids and diplocaulids (A. C. Milner, 1980; Bossy & Milner, 1998), and Phlegethontia (Anderson, in press). 64. SUTEMP 2. Absence (0) or presence (1) of condition: supratemoral forming anterior edge of temporal notch. This character is considered by Clack (1998 a, 2001) as a potential synapomorphy of Eucritta and baphetids [see also Beaumont (1977), Beaumont & Smithson (1998) and Milner & Lindsay (1998)]. 65. SUTEMP 3. Absence (0) or presence (1) of condition: supratemoral narrow and strap-like, at least three times as long as wide. The derived configuration of the supratemporal is observed in some aı̈stopods (Wellstead, 1982; Carroll, 1998) and urocordylids (A. C. Milner, 1980; Bossy & Milner, 1998). 66. SUTEMP 4. Supratemoral contact with squamosal smooth (0) or interdigitating (1). This is one of several characters describing the nature of the contact between the skull table and the cheek (Clack, 1998c, 2001; Paton et al., 1999). The derived state appears in the tetrapod stem-group (Beaumont, 1977; Smithson, 1982; Hook, 1983; Jarvik, 1980, 1996; Clack, 1994 a; Lombard & Bolt, 1995), although Crassigyrinus shows a reversal to the primitive state, and is maintained in most temnospondyls; exceptions are Isodectes (Sequeira, 1998), Broiliellus (Carroll, 1964), and some amphibamids (Watson, 1940; Bolt, 1969, 1991). On the amniote branch, the derived state appears in Anthracosaurus (Panchen, 1977; Clack, 1987b), Solenodonsaurus (Carroll, 1970; Laurin & Reisz, 1999) and some seymouriamorphs (White, 1939; Bystrow, 1944; Laurin, 1996a). Tabular 67. TAB 1. Tabular present (0) or absent (1) as separate ossification. A separately ossified tabular is absent in crown-lissamphibians (except perhaps Eocaecilia; Evans & Milner, 1996; Carroll, 2000; Milner, 2000; Rocek & Rage, 2000b), Captorhinus (Fox & Bowman, 1966), Odonterpeton (Carroll & Gaskill, 1978) and adelospondyls (Andrews & Carroll, 1991). In the latter group, a ‘squamosotabular’ ossification is present in the position normally occupied by the squamosal and tabular in other tetrapods (see also character 86 below). 68. TAB 2. Absence (0) or presence (1) of subdermal, blade-like postero-lateral horn of tabular. As defined here [see also Smithson (1985), Clack (1987a, 1998 c, 2001) and Paton et al. (1999)], a subdermal, blade-like posterolateral tabular horn characterizes Acanthostega (Clack, 1994 a), Crassigyrinus (Clack, 1998c), Whatcheeria (Lombard & Bolt, 1995), Caerorhachis (Holmes & Carroll, 1977; Ruta et al., 2001), embolomeres (Panchen, 1972, 1977; Holmes, 1984, 1989; Smithson, 1985; Clack, 1987 a, 1998c, 2001) and gephyrostegids (Carroll, 1970; Boy & Bandel, 1973). See also Klembara (1997) for a discussion of this character. 69. TAB 3. Absence (0) or presence (1) of rounded, subdermal, button-like posterior process of tabular. This character is found in colosteids (Smithson, 1982), baphetids (Beaumont, 1977; Beaumont & Smithson, 1998), Edops (Romer & Witter, 1942) and, possibly, Dendrerpeton (Holmes et al., 1998). 70. TAB 4. Tabular contacts squamosal on dorsal surface of skull table (0) or not (1). In its derived state, this character is found in stem-tetrapods more derived than colosteids (Beaumont, 1977; Lombard & Bolt, 1995; Clack, 1998 c), Eucritta (Clack, 2001), various temnospondyls (except some trimerorhachoids, Broiliellus, Ecolsonia and Doleserpeton; Chase, 1965; Carroll, 1964; Bolt, 1969, 1991; Berman et al., 1985; Sequeira, 1998), Eocaecilia (Carroll, 2001), embolomeres (Panchen, 1972, 1977; Holmes, 1984, 1989; Smithson, 1985; Clack, 1987a, 1998 c, 2001), gephyrostegids (Carroll, 1970), seymouriamorphs (White, 1939; Bystrow, 1944; Laurin, 1996 b; Klembara, 1997), diadectomorphs (Romer, 1946; Berman et al., 1998), Paleothyris and Petrolacosaurus (Clark & Carroll, 1973; Reisz, 1977, 1981), and Westlothiana (Smithson et al., 1994). It is present also in Ptyonius (A. C. Milner, 1980; Bossy & Milner, 1998). 71. TAB 5. Tabular contact with squamosal smooth (0) or interdigitating (1). This is one of the characters describing the nature of the contact between the cheek and the skull table (Clack, 1998c, 2001; Paton et al., 1999). An interdigitating tabular-squamosal suture occurs in Ichthyostega (Jarvik, 1980, 1996), Greererpeton (Smithson, 1982), Neldasaurus (Chase, 1965), Ecolsonia (Berman et al., 1985), Pantylus and ostodolepids (Carroll & Gaskill, 1978). 72. TAB 6. Absence (0) or presence (1) of condition: tabular elongate posteriorly or posterolaterally in the form of massive horn. This character is a shared derived feature of diplocaulidnectrideans(Beerbower,1963;A. C.Milner, 1980; Bossy & Milner, 1998). 73. TAB 7. Parietal-parietal width smaller than (0) or greater than (1) distance between posterior margin of skull table and posterior margin of orbits measured along the skull midline. The apomorphic condition of this character shows no simple Early tetrapod relationships revisited distribution among crown-tetrapods. It is recorded in several dissorophoids and crown-lissamphibians (Bolt, 1969, 1991; Schoch, 1992; Daly, 1994; Milner, 2000; Rocek & Rage, 2000 a), as well as seymouriamorphs (White, 1939; Bystrow, 1944; Laurin, 1996b; Klembara, 1997), diadectomorphs (Romer, 1946; Berman et al., 1998), Paleothyris and Petrolacosaurus (Clark & Carroll, 1973; Reisz, 1977, 1981), Westlothiana (Smithson et al., 1994), various tuditanomorphs, some microbrachomorphs (Carrol & Gaskill, 1978) and some diplocaulids (Beerbower, 1963; A. C. Milner, 1980; Bossy & Milner, 1998). 74. TAB 8. Tabular without (0) or with (1) posteroventrally sloping occipital exposure. The posterior region of the tabulars of Stegotretus (Berman et al., 1988), ostodolepids, rhynchonkids and gymnarthrids (Gregory et al., 1956; Carroll & Gaskill, 1978) slopes obliquely posteroventrally and is distinctly separated from its anterior region. Postorbital 75. POSORB 1. Postorbital present (0) or absent (1) as a separate ossification. Absence of separately ossified postorbitals is shared by crown-lisamphibians (Carroll, 2000; Milner, 2000; Rocek & Rage, 2000 b), lysorophids (Wellstead, 1991), Adelospondylus (Andrews & Carroll, 1991) and the aı̈stopods Oestocephalus and Phlegethontia (Carroll, 1998; Anderson, in press). 76. POSORB 2. Postorbital without (0) or with (1) ventrolateral digitiform process fitting into deep vertical jugal groove. The presence of a distinct, ventrolateral digitiform process of the postorbital is a shared derived feature of urocordylid nectrideans (A. C. Milner, 1980; Bossy & Milner, 1998). 77. POSORB 3. Postorbital contributing to (0) or excluded from (1) margin of orbit. The postorbital is excluded from the orbit margin in Colosteus (Hook, 1983), Acherontiscus (Carroll, 1969 b), Adelogyrinus and Dolichopareias (Andrews & Carroll, 1991) and the diplocaulids Diplocaulus and Diploceraspis (Beerbower, 1963; A. C. Milner, 1980; Bossy & Milner, 1998). 78. POSORB 4. Postorbital irregularly polygonal (0) or broadly crescentic and narrowing to a posterior point (1). The derived state of this character is recorded in Eucritta (Clack, 1998a, 2001), most temnospondyls (Chenoprosopus, Trimerorhachis and Phonerpeton are exceptions; Hook, 1993; Case, 1935; Dilkes, 1990), several stem-amniotes (except the embolomere Anthracosaurus; Panchen, 1977), Paleothyris (Clark & Carroll, 1973) and Westlothiana (Smithson et al., 1994). 79. POSORB 5. Postorbital not contacting tabular (0) or contacting it (1). A postorbital-tabular suture occurs in Scincosaurus (A. C. Milner, 1980; Bossy & Milner, 1998) 311 and tuditanomorphs (except Hapsidopareion; Carroll & Gaskill, 1978). 80. POSORB 6. Postorbital not wider than orbit (0) or wider (1). Under ACCTRAN, the apomorphic condition of this character is acquired in parallel by Acanthostega (Clack, 1994a), colosteids (Smithson, 1982; Hook, 1983), Edops (Romer & Witter, 1942), Trimerorhachis (Case, 1935), Pantylus (Carroll & Gaskill, 1978), and a diverse assemblage consisting of adelospondyls (except Adelogyrinus; Andrews & Carroll, 1991), Lethiscus (Wellstead, 1982) and nectrideans (except Keraterpeton and Diploceraspis; Beerbower, 1963; A. C. Milner, 1980; Bossy & Milner, 1998). 81. POSORB 7. Absence (0) or presence (1) of condition: postorbital at least one-fourth the width of the skull table at the same transverse level. The distribution of this character is almost identical to that of the previous character, except that all temnospondyls exhibit the plesiomorphic state which is also observed, among nectrideans, in Scincosaurus and Diploceraspis. 82. POSORB 8. Anteriormost part of dorsal margin of postorbital with sigmoid profile absent (0) or present (1). Characteristic ‘kink’ is observed in the anteromedial margin of the postorbital in most microsaurs (ostodolepids are a notable exception; Carroll & Gaskill, 1978). Squamosal 83. SQU 1. Anterior part of squamosal lying behind (0) or in front (1) of parietal midlength. The derived condition is found in almost all tetrapods more derived than Ichthyostega. Exceptions are Greererpeton (Smithson, 1982), Eucritta (Clack, 2001), Trimerorhachis (Case, 1935), some crown-lissamphibians (Evans & Milner, 1996; Milner, 2000), such microsaurs as Batropetes (Carroll, 1991 a), Tuditanus, Asaphestera, hapsidopareiontids and ostodolepids (Carroll & Gaskill, 1978) and aı̈stopods (Wellstead, 1982; Carroll, 1998). 84. SQU 2. Absence (0) or presence (1) of condition: posterior margin of squamosal sloping anteroventrally. Despite similarities in the suspensorium configuration, an anteroventrally sloping squamosal does not identify a clade. The character appears to have been developed in albanerpetontids (Gardner, 2001), Batropetes (Carroll, 1991a), Cardiocephalus, Odonterpeton (Carroll & Gaskill, 1978) and lysorophids (Wellstead, 1991). 85. SQU 3. Squamosal without (0) or with (1) broad, concave semicircular embayment. A squamosal embayment is observed in Adelospondylus (Andrews & Carroll, 1991), some seymouriamorphs (Bystrow, 1944; Klembara, 1997), most temnospondyls (Isodectes is an exception; Sequeira, 1998) and Triadobatrachus (Rocek & Rage, 2000b). 312 Marcello Ruta, Michael I. Coates and Donald L. J. Quicke 86. SQU 4. Absence (0) or presence (1) of single ‘squamosotabular’ in the position of squamosal and tabular. This is one of the features employed by Andrews & Carroll (1991) to diagnose adelospondyls. 87. SQU 5. Squamosal without (0) or with (1) internal shelf bracing quadrate from behind. The derived configuration of the squamosal is shared by scincosaurid and diplocaulid nectrideans (Beerbower, 1963; A. C. Milner, 1980; Bossy & Milner, 1998). Jugal 88. JUG 1. Presence (0) or absence (1) of jugal. A separately ossified jugal is absent in salientians, caudates and lysorophids (Wellstead, 1991; Evans & Milner, 1996; Milner, 2000; Rocek & Rage, 2000 b). 89. JUG 2. Jugal not contributing (0) or contributing (1) to ventral margin of skull roof. The jugal enters the ventral skull roof margin in Baphetes (Beaumont, 1977), cochleosaurid edopoids (Rieppel, 1980; Hook, 1993; Godfrey & Holmes, 1995; Milner & Sequeira, 1998), some trimerorhachoids (Chase, 1965; Sequeira, 1998), Schoenfelderpeton (Boy, 1972), albanerpetontids (Gardner, 2001), Eocaecilia (Carroll, 2000), several embolomeres (Panchen, 1972; Holmes, 1984, 1989) and all amniotes more derived than gephyrostegids. Among lepospondyls, the plesiomorphic state is observed in Oestocephalus (Carroll, 1998) and the urocordylids Sauropleura and, possibly, Urocordylus (A. C. Milner, 1980; Bossy & Milner, 1998). 90. JUG 3. Jugal not contacting (0) or contacting (1) pterygoid. A jugal-pterygoid suture is found in Megalocephalus (Beaumont, 1977), cochleosaurid edopoids (Rieppel, 1980; Hook, 1993; Godfrey & Holmes, 1995; Milner & Sequeira, 1998), some embolomeres (Holmes, 1984, 1989; Clack, 1987 a) and Captorhinus (Fox & Bowman, 1966). 91. JUG 4. Jugal depth below orbit greater (0) or smaller (1) than half orbit diameter. Primitively, the jugal forms a broad area ventral to the orbit (Whatcheeria is an exception among stem-tetrapods; Lombard & Bolt, 1995). Among crown-tetrapods, the plesiomorphic state is observed in edopoids (Milner & Sequeira, 1998), Eryops (Sawin, 1941), Pholiderpeton scutigerum (Clack, 1987a) and Diplocaulus (A. C. Milner, 1980; Bossy & Milner, 1998). 92. JUG 6. Absence (0) or presence (1) of condition: jugal ventrally expanded to form flange overlapping posterior end of maxilla. This particular configuration of the jugal is a synapomorphy of pantylids (Carrol & Gaskill, 1978; Berman et al., 1988). 93. JUG 7. Jugal without (0) or with (1) V-shaped indentation of dorsal margin. A dorsal indentation of the jugal occurs in Crassigyrinus (Clack, 1998c), Whatcheeria (Lombard & Bolt, 1995), baphetids (Beaumont, 1977), and a clade comprising Anthracosaurus (Panchen, 1977; Clack, 1987 b) and Pholiderpeton attheyi (Panchen, 1972). This is one of the characters describing irregular orbit outlines in early tetrapods, and appears as a transitional stemtetrapod feature under both ACCTRAN and DELTRAN. 94. JUG 8. Jugal not extending (0) or extending (1) anterior to anterior orbit margin. The jugal extends anterior to the anterior orbit margin in Acanthostega (Clack, 1994a), edopoids (Milner & Sequeira, 1998), Eryops (Sawin, 1941), most embolomeres (Panchen, 1972, 1977; Holmes, 1989; Smithson, 1985), Seymouria (White, 1939; Laurin, 1996b), Captorhinus (Fox & Bowman, 1966), Pantylus (Carroll & Gaskill, 1978), Scincosaurus and derived diplocaulids (Beerbower, 1963; A. C. Milner, 1980; Bossy & Milner, 1998). Quadratojugal 95. QUAJUG 1. Presence (0) or absence (1) of quadratojugal. Valdotriton (Evans & Milner, 1996) and lysorophids (Wellstead, 1991) do not show an ossified quadratojugal. 96. QUAJUG 2. Absence (0) or presence (1) of condition: quadratojugal much smaller than squamosal. The derived condition of this character occurs in Karaurus (Milner, 2000), and several microsaurs, such as Odonterpeton and a clade composed of Asaphestera, hapsidopareiontids, ostodolepids, rhynchonkids and gymnarthrids (Carroll & Gaskill, 1978). 97. QUAJUG 3. Absence (0) or presence (1) of condition: quadratojugal an anteroposteriorly elongate and dorsoventrally narrow splinter of bone. The distribution of the derived condition of this character overlaps that of the preceding character, except for its absence in Asaphestera. Quadrate 98. QUA 1. Quadrate without (0) or with (1) dorsal process. The dorsal process of the quadrate was discussed by Bolt (1969, 1991), Bolt & Lombard (1985) and Milner (1988, 1993). It is regarded as the homologue of the tympanic annulus which suspends the tympanum in several salientians. It occurs in Dendrerpeton (Holmes et al., 1998), trematopids (Olson, 1941; Dilkes, 1990), dissorophids and amphibamids (Bolt, 1969, 1991; Daly, 1994). A similar process has been documented in other taxa, such as certain seymouriamorphs (White, 1939; Bystrow, 1944; Laurin, 1996b; Klembara, 1997) and diadectomorphs (Romer, 1946; Berman et al., 1998), but it is not certain whether it had the same function. Preopercular 99. PREOPE 1. Absence (1) or presence (0) of preopercular. A preopercular is lost in all tetrapods more derived than Early tetrapod relationships revisited Ichthyostega (Clack, 1998 c, 2001; see also discussion in Panchen, 1991). Nostrils 100. NOS 3. Absence (0) or presence (1) of condition: nostrils elongate and key-hole shaped. Acheloma, Phonerpeton and Ecolsonia share the derived state (Olson, 1941; Berman et al., 1985; Dilkes, 1990). See also Dilkes (1990) and Daly (1994) for a discussion of this character. 101. NOS 4. Absence (0) or presence (1) of condition: nostrils elliptical, with greater axis oriented obliquely in anteromedial to posterolateral direction, and at least 70% the length of the internasal suture. The derived condition of the nostril is found in branchiosaurids (Boy, 1972; Schoch, 1992; Boy & Sues, 2000), some amphibamids (Watson, 1940; Bolt, 1969, 1977, 1979, 1991) and certain crownlissamphibians (Evans & Milner, 1996; Milner, 2000). Internarial fenestra 102. INT FEN 1. Absence (0) or presence (1) of internarial fenestra. The presence of an internarial fenestra (see Clack, 1998 c, 2001) characterizes Acanthostega (Clack, 1994a), Ichthyostega ( Jarvik, 1980, 1996), Greererpeton (Smithson, 1982), Crassigyrinus (Clack, 1996, 1998c), baphetids (Beaumont, 1977), some dissorophoids (Olson, 1941; Bolt, 1969, 1991; Boy, 1972, 1987, 1995; Dilkes, 1990; Schoch, 1992; Daly, 1994) and Karaurus (Milner, 2000). Orbits 103. ORB 1. Interorbital distance greater than (0), smaller than (1), or subequal to (2) half skull table width. The plesiomorphic state is found in Eocaecilia (Carroll, 2000), most tuditanomorphs and one microbrachomorph (Carroll & Gaskill, 1978) and in Batrachiderpeton (A. C. Milner, 1980; Bossy & Milner, 1998). State 2 appears sporadically on the tree, in Colosteus (Hook, 1983), Chenoprosopus (Hook, 1993; Milner & Sequeira, 1998), Eoherpeton (Smithson, 1985), some seymouriamorphs (White, 1939; Bystrow, 1944; Laurin, 1996b), Diadectes (Berman et al., 1998), Westlothiana (Smithson et al., 1994), Batropetes, Tuditanus, Microbrachis and Hyloplesion (Carroll & Gaskill, 1978; Carroll, 1991a), Scincosaurus, some diplocaulids (A. C. Milner, 1980; Bossy & Milner, 1998) and Oestocephalus (Carroll, 1998). 104. ORB 2. Interorbital distance greater than (0), smaller than (1) or subequal to (2) maximum orbit diameter. The different states of this character show a more complicated distribution than those of the preceding character. State 1, found in Acanthostega (Clack, 1994a), Crassigyrinus (Clack, 1998 c) and Whatcheeria (Lombard & Bolt, 1995), is widespread among crown-tetrapods. It is observed in Eucritta (Clack, 2001), Trimerorhachis (Case, 1935), Balanerpeton (Milner & Sequeira, 1994), 313 and most dissorophoids and crown-lissamphibians (Bolt, 1969, 1991; Boy, 1972, 1987, 1995; Schoch, 1992; Daly, 1994; Evans & Milner, 1996; Milner, 2000; Rocek & Rage, 2000b). It is also found in many embolomeres (Panchen, 1972, 1977; Holmes, 1984, 1989; Clack, 1987a), gephyrostegids (Carroll, 1970), Discosauriscus (Klembara, 1997), basal crownamniotes (Fox & Bowman, 1966; Clark & Carroll, 1973; Reisz, 1977, 1981) and a clade of aı̈stopods plus urocordylids (Wellstead, 1982; A. C. Milner, 1980; Bossy & Milner, 1998; Carroll, 1998; Anderson, in press). State 2 characterizes baphetids (Beaumont, 1977), such dissorophoids as Ecolsonia and Broiliellus (Carroll, 1964; Berman et al., 1985), Batropetes (Carroll, 1991a), adelospondyls (Andrews & Carroll, 1991) and derived diplocaulids (Beerbower, 1963; A. C. Milner, 1980; Bossy & Milner, 1998). The plesiomorphic condition, exhibited by Ichthyostega (Jarvik, 1980, 1996) and colosteids (Smithson, 1982; Hook, 1983), occurs also in edopoids (Milner & Sequeira, 1998), some trimerorhachoids (Chase, 1965; Sequeira, 1998), Dendrerpeton (Holmes et al., 1998), Eryops (Sawin, 1941), Platyrhinops (Milner, 2000), Eocaecilia (Carroll, 2000), Eoherpeton (Smithson, 1985), Solenodonsaurus (Laurin & Reisz, 1999), some seymouriamorphs (White, 1939; Bystrow, 1944; Laurin, 1996b), diadectomorphs (Romer, 1946; Berman et al., 1998), most microsaurs (Carroll & Gaskill, 1978) scincosaurids and several diplocaulids (A. C. Milner, 1980; Bossy & Milner, 1998). 105. ORB 3. Absence (0) or presence (1) of angle at anteroventral orbit corner. The derived state of this character (coding follows Clack, 1998 c, 2001) is observed in Crassigyrinus (Clack, 1998 c), Whatcheeria (Lombard & Bolt, 1995) and Eucritta (Clack, 2001). It appears as a transitional feature of stem-tetrapods under ACCTRAN, but is developed independently in Crassigyrinus and Whatcheeria under DELTRAN. Pineal foramen 106. PIN FOR 1. Presence (0) or absence (1) of pineal foramen. A pineal foramen is absent in cochleosaurid edopoids (Milner & Sequeira, 1998), Pantylus, ostodolepids (Carroll & Gaskill, 1978), lysorophids (Wellstead, 1991), albanerpetontids and most crownlissamphibians (Evans & Milner, 1996; Carroll, 2000; Milner, 2000; Gardner, 2001). Milner & Sequeira (1998) discussed in detail the occurrence of this character in cochleosaurids. Several species show progressive obliteration of the foramen during growth. As a general condition, closure of the foramen is a diagnostic feature of cochleosaurids with a skull length of 120 mm or more. 314 Marcello Ruta, Michael I. Coates and Donald L. J. Quicke 107. PIN FOR 2. Position of pineal foramen behind (0), at the level of (1) or anterior to (2) interparietal suture mid length. The plesiomorphic state of this character occurs in stemtetrapods (Beaumont, 1977; Jarvik, 1980, 1996; Smithson, 1982; Hook, 1983; Clack, 1994a, 1998c; Lombard & Bolt, 1995), as well as Eucritta (Clack, 2001) and temnospondyls (Milner & Sequeira, 1998). Reversal to this condition occurs sporadically, in some temnospondyls (Trimerorhachis; Eryops; Case, 1935; Sawin 1941), stem-amniotes (Pholiderpeton scutigerum; Clack, 1987a), crown-amniotes (Paleothyris; Clark & Carroll, 1973), microsaurs (Asaphestera; Carroll & Gaskill, 1978), adelospondyls (Adelogyrinus; Dolichopareias; Andrews & Carroll, 1991) and aı̈stopods (Oestocephalus; Carroll, 1998). State 1 is observed in Kotlassia, Diplocaulus, Diploceraspis, Stegotretus, Ecolsonia, Adelospondylus, Isodectes (Bystrow, 1944; Beerbower, 1963; A. C. Milner, 1980; Berman et al., 1988; Andrews & Carroll, 1991; Bossy & Milner, 1998; Sequeira, 1998). State 2 is found in most crown-tetrapods. Central, elongate and lightly sculptured area of skull table 108. L SC SKU 1. Absence (0) or presence (1) of lightly sculptured area. Milner & Sequeira (1998, p. 279) describe ‘Depressed areas with subdued sculpture between parallel anteroposterior sculpture ridges on either side of the skull table’ as a shared derived feature of cochleosaurid edopoids (see also Godfrey & Holmes, 1995). Accordingly, the derived state of this character is assigned to Cochleosaurus and Chenoprosopus. Posttemporal fossae 109. PTF 1. Fossa at dorsolateral corner of occiput, not bordered laterally, roofed over by skull table and floored by dorsolateral extension of opisthotic (0); fossa near dorsolateral corner of occiput, roofed over by occipital flanges of tabular and postparietal and bordered laterally and ventrally by dorsolateral extension of opisthotic meeting ventromedial flange of tabular (1); small fossa near ventrolateral corner of occiput bordered laterally by ventromedial flange of tabular, roofed over by dorsal portion of lateral margin of supraoccipital–opisthotic complex and floored by lateral extension of opisthotic (2); absence of fossa (3). The different conditions of the posttemporal fossae and associated codings are based on Berman’s (2000) recent study of the occipital region in early tetrapods. State 3 is acquired in parallel in Ichthyostega (Jarvik, 1980, 1996) and embolomeres (Panchen, 1977; Holmes, 1984, 1989; Smithson, 1985; Clack, 1987a). State 2 characterizes diadectomorphs and crown-amniotes (Romer, 1946; Fox & Bowman, 1966; Clark & Carroll, 1973; Reisz, 1977, 1981; Berman et al., 1998). State 1 is widespread among derived stem-tetrapods (Beaumont, 1977; Smithson, 1982, 1985). Proportions of skull table 110. SKU TAB 1. Absence (0) or presence (1) of condition: postorbital region of skull roof abbreviated. Although shortening of the postorbital skull roof region occurs in various degrees, several crown-group taxa are distinctly different from the outgroup and from stem-tetrapods in that their skull roof is usually wider than long, regardless of the morphology and proportions of its various constituent bones. Several temnospondyls more crownward than trimerorhachoids (Carroll, 1964; Bolt, 1969, 1991; Boy, 1972, 1987, 1995; Berman et al., 1985; Schoch, 1992; Daly, 1994; Milner & Sequeira, 1994) and crown-lissamphibians (Carroll, 2000; Milner, 2000; Rocek & Rage, 2000 b) have acquired the derived state of this character independent of several basal and crown-amniotes, Westlothiana, microsaurs, Scincosaurus and derived diplocaulids (White, 1939; Bystrow, 1944; Romer, 1946; Beerbower, 1963; Fox & Bowman, 1966; Carroll, 1970; Boy & Bandel, 1973; Panchen, 1977; Reisz, 1977, 1981; A. C. Milner, 1980; Smithson et al., 1994; Laurin, 1996 b; Klembara, 1997; Berman et al., 1998; Bossy & Milner, 1998). Temporal fenestra 111. TEM FEN 1. Absence (0) or presence (1) of broad postorbital opening (aı̈stopod pattern). In all aı̈stopods, a large temporal fenestra, not confluent with the orbit, occupies more than half of the skull length (Wellstead, 1982; Carroll, 1998; Anderson, in press). Repatterning of the postorbital region of the skull involves loss of some bones. Cheek emargination 112. CHE EMA 1. Absence (0) or presence (1) of ventral cheek emargination ( pattern of certain tuditanomorph microsaurs). In hapsidopareiontids, ostodolepids and some gymnarthrids (Gregory et al., 1956; Carroll & Gaskill, 1978), the ventral cheek margin is excavated to various degrees without involving loss of cheek bones. This excavation, particularly evident in hapsidopareiontids and ostodolepids, confers a strongly arched profile to the posteroventral part of the skull table. Interfrontonasal 113. IFN 1. Absence (0) or presence (1) of interfrontonasal. An interfrontonasal appears to be diagnostic of Eryops (Sawin, 1941), and is present in other eryopoids, such as Clamorosaurus Gubin, 1983. Suspensorium 114. SUS 1. Absence (0) or presence (1) of anteroposteriorly narrow, bar-like squamosal. The derived condition of the squamosal results in a shortened, oblique configuration of the posterior, external surface of the suspensorium in Early tetrapod relationships revisited lysorophids (Wellstead, 1991), aı̈stopods (Wellstead, 1982; Carroll, 1998; Anderson, in press), as well as in the microsaurs Batropetes (Carroll, 1991 a), Hapsidopareion and, possibly, Cardiocephalus (Carroll & Gaskill, 1978). Sensory canals 115. SC 1. Lateral line system on skull roof totally enclosed (0), mostly enclosed with short sections in grooves (1), mostly in grooves with short sections enclosed (2), entirely in grooves (3), absent (4). The codings for this and the following character are based on data collated from analyses by Clack (1998c, 2001), Ahlberg & Clack (1998) and Paton et al. (1999). Inspection of the data set reveals that, in the case of both characters, state 4 is widespread in the crown-group, and that states 1, 2 and 3 are acquired independently (and, often, coexist) in different clades. 116. SC 2. Mandibular canal totally enclosed (0), mostly enclosed with short sections in grooves (1), mostly in grooves with short sections enclosed (2), entirely in grooves (3), absent (4). See character 115 above. (b) Palate Vomer 117. VOM 1. Absence (0) or presence (1) of condition: ventral, exposed surface of vomers narrow, elongate and strip-like, without extensions anterolateral or posterolateral to choana and two and a half to three times longer than wide. The derived state of this character is widespread on the amniote branch of the tetrapod tree (Gauthier et al., 1988b; Lee & Spencer, 1997). Importantly, its occurrence does not depend upon the overall morphology of the preorbital skull region. It is found in broad-snouted stemamniotes, such as Discosauriscus (Klembara, 1997), as well as in long-snouted forms, such as embolomeres (Panchen 1977; Holmes, 1984, 1989; Smithson, 1985; Clack, 1987 a), gephyrostegids (Carroll, 1970), diadectomorphs (Romer, 1946; Fracasso, 1987; Berman et al., 1998), crown-amniotes (Fox & Bowman, 1966; Clark & Carroll, 1973; Reisz, 1977, 1981) and Westlothiana (Smithson et al., 1994). Microsaurs (Carroll & Gaskill, 1978), lysorophids (Wellstead, 1991), Sauropleura and Urocordylus (Bossy & Milner, 1998) also show elongate vomers. The plesiomorphic condition of this character is widespread in stem-tetrapods and in the temnospondyl-lissamphibian clade, regardless of the degree of elongation of the snout (several longirostrine temnospondyls represent an exception; Schoch & Milner, 2000). 118. VOM 3. Vomer with (0) or without (1) fang pair. The coding of this character follows Gauthier et al. (1988b), Clack (1998 c, 2001) and Laurin (1998 b). Vomerine fangs are absent in Ichthyostega ( Jarvik, 1980, 1996), 315 certain dissorophoids and crown-lissamphibians (Watson, 1940; Carroll, 1964, 2000; Bolt, 1969, 1977, 1979, 1991; Boy, 1972; Evans & Milner, 1996; Milner, 2000), Pholiderpeton attheyi (Panchen, 1972), Kotlassia (Bystrow, 1944), diadectomorphs (Romer, 1946; Fracasso, 1987; Berman et al., 1998), crown-amniotes (Fox & Bowman, 1966; Clark & Carroll, 1973; Reisz, 1977, 1981), and all lepospondyls in which the palate is visible (Euryodus is, however, an exception, since fang pairs are present; also, Micraroter is polymorphic for this character; Carroll & Gaskill, 1978). 119. VOM 4. Vomer without (0) or with (1) denticles. The presence of a denticle shagreen patch on the vomer shows a rather uneven distribution. Among stemtetrapods, it is observed in Tulerpeton (Lebedev & Clack, 1993), Whatcheeria (Lombard & Bolt, 1995) and baphetids (Beaumont, 1977). Among crown-tetrapods, it is found in Eucritta (Clack, 2001), several temnospondyls (Sawin, 1941; Romer & Witter, 1942; Carroll, 1964; Bolt, 1969, 1991; Berman et al., 1985; Milner & Sequeira, 1994, 1998; Holmes et al., 1998; Dilkes, 1990; Daly, 1994), Caerorhachis (Holmes & Carroll, 1977; Ruta et al., 2001), gephyrostegids (Carroll, 1970), Discosauriscus (Klembara, 1997), Seymouria (White, 1939; Laurin, 1996b), Petrolacosaurus (Reisz, 1977, 1981), some tuditanomorphs (e.g. Saxonerpeton; Micraroter; Euryodus) and all microbrachomorphs (Carroll & Gaskill, 1978). 120. VOM 5. Vomer excluded from (0) or contributing to (1) interpterygoid vacuities. The vomers enter the margins of the palatal vacuities in post-edopoid temnospondyls, albanerpetontids and crown-lissamphibians (Milner, 1988, 1990, 2000; Carroll, 2000; Rocek & Rage, 2000b; Gardner, 2001), as well as in Hyloplesion (Carroll & Gaskill, 1978) and derived diplocaulids (Beerbower, 1963; A. C. Milner, 1980; Bossy & Milner, 1998). 121. VOM 7. Vomer not forming (0) or forming (1) suture with maxilla anterior to choana. The derived condition of this character is found in Ichthyostega ( Jarvik, 1980, 1996), Crassigyrinus (Clack, 1996, 1998c), Cochleosaurus (Godfrey & Holmes, 1995; Milner & Sequeira, 1998), certain trimerorhachoids (Case, 1935; Chase, 1965), Eoscopus (Daly, 1994) and primitive crown-lissamphibians, where observed (Evans & Milner, 1996; Carroll, 2001). 122. VOM 8. Vomer with (0) or without (1) toothed lateral crest. With the exception of Crassigyrinus (Clack, 1998c) and Eoscopus (Daly, 1994), all tetrapods more derived than Tulerpeton lack a lateral crest on the ventral surface of the vomer (for descriptions, see Lebedev & Clack, 1993). 123. VOM 9. Vomer with (0) or without (1) anterior crest. The distribution of this character is almost identical to that of the previous character, except that the derived condition is also present in Ichthyostega 316 Marcello Ruta, Michael I. Coates and Donald L. J. Quicke ( Jarvik, 1980, 1996) and Tulerpeton (Lebedev & Clack, 1993). 124. VOM 10. Vomer in contact with anterior ramus of pterygoid (0) or not (1). Absence of a vomer-pterygoid contact is found in Isodectes (Sequeira, 1998), all temnospondyls more derived than Ecolsonia (Milner, 1988, 1990), crown-lissamphibians (Evans & Milner, 1996; Carroll, 2000; Milner, 2000; Rocek & Rage, 2000 b) and derived diplocaulids (Beerbower, 1963; A. C. Milner, 1980; Bossy & Milner, 1998). 125. VOM 11. Vomer without (0) or with (1) transverse patch of small teeth posteromesial to choana. Several crownlissamphibians share the derived condition of this feature with Doleserpeton (Bolt, 1969, 1977, 1979, 1991; Milner, 1988, 1990, 1993, 2000). 126. VOM 12. Absence (0) or presence (1) of distinct posterolateral vomer process bordering more than half of posterior margin of choana. The derived state of this character occurs in some amphibamids and branchiosaurids (Bolt, 1969, 1977, 1979, 1991; Boy, 1972; Milner, 1988, 1990, 1993, 2000; Schoch, 1992), as well as in Karaurus (Milner, 2000). Palatines 127. PAL 1. Palatine with (0) or without (1) fangs. Loss of palatine fangs occurs independently in Ichthyostega ( Jarvik, 1980, 1996), some branchiosaurids (Boy, 1972), some amphibamids and crown-lissamphibians (Bolt, 1969, 1977, 1979, 1991; Milner, 1988, 1990, 1993, 2000; Carroll, 2000; Rocek & Rage, 2000 b), and in a large portion of the amniote tree including Kotlassia (Bystrow, 1944), diadectomorphs (Romer, 1946; Berman et al., 1998), crown-amniotes (Fox & Bowman, 1966; Clark & Carroll, 1973; Reisz, 1977, 1981) and most lepospondyls except pantylids (Beerbower, 1963; Carroll & Gaskill, 1978; A. C. Milner, 1980; Berman et al., 1988; Wellstead, 1991; Bossy & Milner, 1998). 128. PAL 2. Palatine without (0) or with (1) denticles. Presence of a denticle shagreen patch on the palatines is recorded in Whatcheeria (Lombard & Bolt, 1995), baphetids (Beaumont, 1977), Eucritta (Clack, 2001), some temnospondyls (Watson, 1940; Sawin, 1941; Romer & Witter, 1942; Berman et al., 1985; Milner & Sequeira, 1994, 1998; Godfrey & Holmes, 1995; Holmes et al., 1998), Caerorhachis (Holmes & Carroll, 1977; Ruta et al., 2001), gephyrostegids (Carroll, 1970), some seymouriamorphs (White, 1939; Laurin, 1996b; Klembara, 1997), certain crown-amniotes (Clark & Carroll, 1973; Reisz, 1977, 1981), Westlothiana (Smithson et al., 1994), microbrachomorphs, some tuditanomorphs (Carroll & Gaskill, 1978), at least one adelospondyl (Adelospondylus; Andrews & Carroll, 1991) and one nectridean (Scincosaurus; A. C. Milner, 1980; Bossy & Milner, 1998). 129. PAL 3. Palatine excluded from (0) or contributing to (1) interpterygoid vacuities. The derived state of this character is shared by scincosaurids and diplocaulids (Beerbower, 1963; A. C. Milner, 1980; Bossy & Milner, 1998), as well as by derived dissorophoids and crown-lissamphibians (Carroll, 1964; Bolt, 1969, 1977, 1979, 1991; Boy, 1972, 1987, 1995; Milner, 1988, 1990, 1993, 2000; Carroll, 2000; Rocek & Rage, 2000 b). It is also documented in Isodectes (Sequeira, 1998). 130. PAL 4. Palatine with (0) or without (1) tooth row (3+). A palatine tooth row is present in all stemtetrapods less derived than baphetids (Clack, 1998c, 2001; Paton et al., 1999), as well as in trimerorhachoids (Case, 1935; Chase, 1965), some dissorophoids and Eocaecilia (Bolt, 1969, 1977, 1979, 1991; Boy, 1972, 1987, 1995; Milner, 1988, 1990, 1993, 2000; Schoch, 1992; Carroll, 2000; Rocek & Rage, 2000 b), Kotlassia (Bystrow, 1944), some tuditanomorphs and microbrachomorphs (Carroll & Gaskill, 1978) and nectrideans other than Scincosaurus (Beerbower, 1963; A. C. Milner, 1980; Bossy & Milner, 1998). 131. PAL 5. Palatine without (0) or with (1) lateral exposure in anteroventral orbit margin. Among the taxa included in the present analysis, a lateral exposure of the palatine contributing to the orbit margin is present in Isodectes (Sequeira, 1998); Ecolsonia (Berman et al., 1985), Micromelerpeton (Boy, 1995) and Doleserpeton (Bolt, 1969, 1977, 1979, 1991). 132. PAL 6. Absence (0) or presence (1) of condition: palatine articulates with maxilla only at its anterior end. Apateon (Schoch, 1992), Leptorophus and Schoenfelderpeton (Boy, 1972) share the derived condition (Milner, 1990, 1993; Trueb & Cloutier, 1991). This is also present in Petrolacosaurus (Reisz, 1977, 1981). 133. PAL 7. Palatine not reduced (0) or reduced (1) to slender, strut-like bone. As described by Milner (1990, 1993), this condition is found in micromelerpetontids, branchiosaurids, some amphibamids and some crownlissamphibians(Boy,1972,1987,1995;Bolt,1969,1977, 1979, 1991; Milner, 1990, 1993; Rocek & Rage, 2000b). Ectopterygoids 134. ECT 1. Presence (0) or absence (1) of ectopterygoid. The ectopterygoid is absent in Doleserpeton (Bolt, 1969, 1977, 1979, 1991), crown-lissamphibians (Evans & Milner, 1996; Carroll, 2000; Milner, 2000; Rocek & Rage, 2000 b), Captorhinus (Fox & Bowman, 1966; Laurin, 1998 b), pantylids (Carroll & Gaskill, 1978; Berman et al., 1988), lysorophids (Wellstead, 1991), scincosaurids and diplocaulids (A. C. Milner, 1980; Bossy & Milner, 1998). Early tetrapod relationships revisited 135. ECT 2. Ectopterygoid with (0) or without (1) fangs. Ectopterygoid fangs [see coding in Clack (1998c, 2001) and Paton et al. (1999)] are absent in Acanthostega (Clack, 1994a), Ichthyostega ( Jarvik, 1980, 1996), Trimerorhachis (Case, 1935), micromelerpetontids, branchiosaurids (Boy, 1972, 1987, 1995; Milner, 1990, 1993; Schoch, 1992), Bruktererpeton (Boy & Bandel, 1973), seymouriamorphs and all more crownward stem-amniotes, as well as crown-amniotes (White, 1939; Bystrow, 1944; Romer, 1946; Fox & Bowman, 1966; Clark & Carroll, 1973; Reisz, 1977, 1981; Smithson et al., 1994; Laurin, 1996b; Klembara, 1997; Berman et al., 1998), as well as several lepospondyls (Carroll & Gaskill, 1978; A. C. Milner, 1980; Wellstead, 1991; Bossy & Milner, 1998). 136. ECT 3. Ectopterygoid without (0) or with (1) denticles. A denticle shagreen patch on the ectopterygoid is observed in Whatcheeria (Lombard & Bolt, 1995), baphetids (Beaumont, 1977), some temnospondyls (Sawin, 1941; Romer & Witter, 1942; Milner & Sequeira, 1994, 1998; Godfrey & Holmes, 1995; Holmes et al., 1998), Caerorhachis (Holmes & Carroll, 1977; Ruta et al., 2001), Proterogyrinus (Holmes, 1984), gephyrostegids (Carroll, 1970; Boy & Bandel, 1973), some seymouriamorphs (White, 1939; Laurin, 1996b; Klembara, 1997), Westlothiana (Smithson et al., 1994), some microbrachomorphs, few tuditanomorphs (Carroll & Gaskill, 1978) and at least one adelospondyl (Adelospondylus; Andrews & Carroll, 1991). 137. ECT 4. Ectopterygoid longer than/as long as palatines (0) or not (1). Primitively, the ectopterygoid is an elongate and subrectangular bone, as found in the tetrapod stemgroup, in temnospondyls up to the level of trematopids (e.g. Olson, 1941; Sawin, 1941; Romer & Witter, 1942; Milner & Sequeira, 1994, 1998; Godfrey & Holmes, 1995; Holmes et al., 1998), in Caerorhachis and embolomeres (Panchen, 1972, 1977; Holmes & Carroll, 1977; Smithson, 1985; Clack, 1987 a; Ruta et al., 2001). A reversal to the plesiomorphic state is documented in Ptyonius (A. C. Milner, 1980; Bossy & Milner, 1998) and some seymouriamorphs (Bystrow, 1944; Klembara, 1997). 138. ECT 5. Ectopterygoid with (0) or without (1) tooth row (3+). Absence of an ectopterygoid tooth row [see coding in Clack (1998c, 2001) and Paton et al. (1999)] is a shared derived feature of baphetids and crowntetrapods, but reversals to the primitive state are widespread. Such reversals are documented in trimerorhachoids (Case, 1935; Chase, 1965), micromelerpetontids and branchiosaurids (Boy, 1972, 1987, 1995; Milner, 1990, 1993; Schoch, 1992), embolomeres (Panchen, 1972, 1977; Holmes, 1984, 1989; Smithson, 1985; Clack, 1987 a), Kotlassia (Bystrow, 1944), ostodolepid, gymnarthrid and rhynchonkid 317 tuditanomorphs (Gregory et al., 1956; Carroll & Gaskill, 1978) and Ptyonius (A. C. Milner, 1980; Bossy & Milner, 1998; optimized as present in remaining nectrideans under ACCTRAN). 139. ECT 6. Ectopterygoid contacting maxilla (0) or not (1). A separation between ectopterygoid and maxilla is found in micromelerpetontids and branchiosaurids (Boy, 1972, 1987, 1995; Milner, 1990, 1993; Schoch, 1992), in Petrolacosaurus (Reisz, 1977, 1981) and in the microbrachomorphs Hyloplesion and Odonterpeton (Carroll & Gaskill, 1978). 140. ECT 7. Absence (0) or presence (1) of condition: ectopterygoid narrowly wedged between palatine and pterygoid. The derived state is shared by Hyloplesion (in which it is more pronounced) and Odonterpeton (Carroll & Gaskill, 1978). Pterygoids 141. PTE 3. Absence (0) or presence (1) of pterygoid flange oriented transversely. There is as yet no consensus on what counts as a transverse pterygoid flange, although this is one of the most widely discussed apomorphies of amniotes (Heaton, 1980; Gauthier et al., 1988b; Carroll, 1991 b; Lee & Spencer, 1997). As pointed out by Laurin (1998 b), this character is more widespread among tetrapods than previously assumed, and is certainy present in some temnospondyls (Yates & Warren, 2000). However, we point out that the flanges of Eryops, Amphibamus and Ecolsonia (Watson, 1940; Sawin, 1941; Berman et al., 1985) are neither as developed as, nor conform to the pattern (e.g. transverse orientation) of, those of gephyrostegids (Carroll, 1970), seymouriamorphs (White, 1939; Bystrow, 1944; Laurin, 1996 a; Klembara, 1997), diadectomorphs (Romer, 1946; Berman et al., 1998) and crownamniotes (Fox & Bowman, 1966; Clark & Carroll, 1973; Reisz, 1977, 1981). This character should be read in conjunction with character 144 below. It refers to the presence of a transverse, ventrally directed thickening of the posterior margin of the pterygoid region lying immediately posterolateral to the recess for the basipterygoid process. It may coexist with a posterolateral flange (character 144 below), as in seymouriamorphs. 142. PTE 4. Absence (0) or presence (1) of teeth on transverse pterygoid flange. Limnoscelis shares the presence of pterygoid teeth on the transverse flange with Paleothyris and Petrolacosaurus (Romer, 1946; Clark & Carroll, 1973; Reisz, 1977, 1981). 143. PTE 7. Absence (0) or presence (1) of condition: quadrate ramus of pterygoid laterally oriented. In caudates, the laterally directed quadrate ramus of the pterygoid 318 Marcello Ruta, Michael I. Coates and Donald L. J. Quicke extends almost perpendicularly to the anteroposterior axis (Evans & Milner, 1996; Milner, 2000). 144. PTE 9. Pterygoid without (0) or with (1) posterolateral flange. A posterolateral flange (not transversely oriented) is found in several primitive and derived temnospondyls (Olson, 1941; Sawin, 1941; Romer & Witter, 1942; Carroll, 1964; Chase, 1965; Bolt, 1969, 1977, 1979, 1991; Berman et al., 1985; Dilkes, 1990; Daly, 1994; Milner & Sequeira, 1994, 1998; Godfrey & Holmes, 1995; Holmes et al., 1998; Carroll, 2000; Milner, 2000), as well as seymouriamorphs (see also character 141 above), Limnoscelis (Romer, 1946; the situation of Diadectes is unclear), some tuditanomorphs and few microbrachomorphs (Carroll & Gaskill, 1978). 145. PTE 10. Pterygoids not sutured with each other (0) or sutured (1). Pterygoids that fail to meet in the midline are documented in temnospondyls (except Edops; Romer & Witter, 1942; Milner & Sequeira, 1998) and crown-lissamphibians, as well as Kotlassia (Bystrow, 1944), Pelodosotis, Rhynchonkos and Hyloplesion (Carroll & Gaskill, 1978), lysorophids (Wellstead, 1991) and Scincosaurus (A. C. Milner, 1980; Bossy & Milner, 1998). 146. PTE 11. Pterygoid not sutured with maxilla (0) or sutured (1). A maxilla-pterygoid contact is documented in Doleserpeton (Bolt, 1969, 1991), Valdotriton (Evans & Milner, 1996) and Triadobatrachus (Rocek & Rage 2000 b). 147. PTE 12. Pterygoid not sutured with posterior end of palatine (0) or sutured (1). Platyrhinops (Milner, 2000), Doleserpeton (Bolt, 1969, 1991) and Eocaecilia (Carroll, 2000) share the derived condition of this character. 148. PTE 13. Pterygoid without (0) or with (1) distinct, mesially directed process for the basipterygoid recess. The derived condition characterizes several temnospondyls and some crown-lissamphibians (Olson, 1941; Sawin, 1941; Carroll, 1964; Bolt, 1969, 1977, 1979, 1991; Boy, 1972, 1987, 1995; Berman et al., 1985; Holmes et al., 1998; Schoch, 1992; Daly, 1994; Boy & Sues, 2000; Milner, 2000; Rocek & Rage, 2000 b). 149. PTE 14. Absence (0) or presence (1) of condition: quadrate ramus of pterygoid robust, indistinctly merging into basal and palatal processes. A quadrate ramus merging indistinctly into the posterior part of the palatal ramus is observed in micromelerpetontids, branchiosaurids and crown-lissamphibians (Boy, 1972, 1987, 1995; Schoch, 1992; Boy & Sues, 2000; Carroll, 2000; Milner, 2000; Rocek & Rage, 2000 b). 150. PTE 15. Absence (0) or presence (1) of condition: quadrate ramus of pterygoid straight, rod-like and gently tapering distally. The derived condition occurs in micromelerpetontids, branchiosaurids, some amphibamids and Eocaecilia (Bolt, 1969, 1977, 1979, 1991; Boy, 1972, 1987, 1995; Schoch, 1992; Boy & Sues, 2000; Carroll, 2000; Milner, 2000). 151. PTE 16. Palatal ramus of pterygoid without (0) or with (1) distinct, anterior, unornamented digitiform process. The process in question, at the anterior end of the palatal ramus of the pterygoid, is visible in colosteids (Smithson, 1982; Hook, 1983), Euryodus and Microbrachis (Carroll & Gaskill, 1978) and in Adelospondylus (Andrews & Carroll, 1991). 152. PTE 17. Basal region of pterygoid immediately anterior to quadrate ramus without (0) or with (1) sharply defined, elongate longitudinal groove. An anteroposteriorly elongate sulcus, marking a deflection between two parts of the basal region of the pterygoid is found in hapsidopareiontids, ostodolepids, rhynchonkids and gymnarthrids (Carroll & Gaskill, 1978). Interpterygoid vacuities 153. INT VAC 1. Presence (0) or absence (1) of interpterygoid vacuities. Regardless of their outline and extension, vacuities are widespread among tetrapods. Entirely closed (‘sealed off’) palates are documented in baphetids (Beaumont, 1977), Eucritta (Clack, 2001), Discosauriscus (discussion in Klembara, 1997), Diadectes (Berman et al., 1998), lysorophids (Wellstead, 1991) and Batrachiderpeton (A. C. Milner, 1980; Bossy & Milner, 1998). 154. INT VAC 2. Absence (0) or presence (1) of condition: interpterygoid vacuities occupying at least half of palatal width. The derived state of this character occurs in most temnospondyls (Edops is a notable exception; Romer & Witter, 1942; Milner & Sequeira, 1994, 1998; Holmes, 2000; Ruta et al., 2001), crownlissamphibians (Carroll, 2000; Milner, 2000; Rocek & Rage, 2000 b), Ptyonius and several derived diplocaulids (Beerbower, 1963; A. C. Milner, 1980; Bossy & Milner, 1998). See also discussion in Anderson (2001). 155. INT VAC 3. Absence (0) or presence (1) of condition: interpterygoid vacuities concave along their whole margins. Except for Ptyonius, the distribution of this character is identical to that of the previous character (see also Ruta et al., 2001). 156. INT VAC 4. Absence (0) or presence (1) of condition: interpterygoid vacuities together broader than long. The distribution of this character is identical to that of character 155, except for Eocaecilia (Carroll, 2000), Chenoprosopus (Hook, 1993; Milner & Sequeira, 1998), trimerorhachoids (Case, 1935; Chase, 1965; Sequeira, 1998), Balanerpeton (Milner & Sequeira, 1994), Dendrerpeton (Holmes et al., 1998), Eryops (Sawin, 1941) and dissorophids (Olson, 1941; Dilkes, 1990). Choanae 157. CHO 1. Absence (0) or presence (1) of condition: choanae wider anteriorly than posteriorly. The derived condition Early tetrapod relationships revisited is shared by Chenoprosopus and Cochleosaurus (Hook, 1993; Godfrey & Holmes, 1995; Milner & Sequeira, 1998). Anterior palatal vacuity 158. ANT VAC 1. Presence (0) or absence (1) of anterior palatal vacuity. Absence of an anterior palatal vacuity characterizes all tetrapods more crownward than Crassigyrinus (Megalocephalus, however, is an exception; Beaumont, 1977). A reversal to the plesiomorphic condition is documented in trimerorhachoids (Case, 1935; Chase, 1965; Sequeira, 1998), Acheloma (Olson, 1941) and Micromelerpeton (Boy, 1995). 159. ANT VAC 2. Anterior palatal vacuity single (0) or double (1). A double palatal vacuity occurs in Acanthostega (Clack, 1994a), Greererpeton (Smithson, 1982), Crassigyrinus (Clack, 1996, 1998c) and trimerorhachoids (Case, 1935; Chase, 1965; Sequeira, 1998). (c) Occiput and braincase Supraoccipital 160. SUPOCC 1. Supraoccipital absent (0) or present (1) as separate ossification. The derived state of this character is found in Limnoscelis (Berman et al., 1992; Berman, 2000), basal crown-amniotes (Fox & Bowman, 1966; Clark & Carroll, 1973; Reisz, 1977, 1981), Westlothiana (Smithson et al., 1994), lysorophids and microsaurs (Carroll & Gaskill, 1978; Wellstead, 1991). However, Berman (2000) postulated that the microsaur supraoccipital is not homologous with that of amniotes. Our treatment of this element is more conservative, but we acknowledge several merits in Beman’s (2000) proposal (see also discussion of Cardiocephalus therein). According to Berman (2000), the bone conventionally referred to as a supraoccipital in many microsaurs and lysorophids derives from the tectum posterius, and not from the tectum synoticum as in other taxa. Recoding the occurrence of a supraoccipital according to Berman’s (2000) suggestion (and imposing an unknown condition for Westlothiana) does not affect the results of the analysis, but restricts the presence of a separately ossified supraoccipital to crown-amniotes and Limnoscelis only. Exoccipitals 161. EXOCC 2. Absence (0) or presence (1) of condition: exoccipitals enlarged to form flattened, widely spaced double occipital condyles. This character is treated separately from characters 162 and 164 below. It is observed in scincosaurids and diplocaulids (Beerbower, 1963; A. C. Milner, 1980; Bossy & Milner, 1998), in which the condylar surfaces are transversely expanded and extremely flattened dorsoventrally. 319 162. EXOCC 3. Absence (0) or presence (1) of condition: exoccipitals enlarged, about as broad as high and forming stout, double occipital condyles. Enlarged exoccipital condyles (although not necessarily appressed and obliterating the basioccipital; see character 164 below) are observed in Dendrerpeton and more derived temnospondyls (Olson, 1941; Bolt, 1969, 1991; Berman et al., 1985; Milner, 1988, 1990, 1993, 2000; Dilkes, 1990; Holmes et al., 1998; Carroll, 2000; Rocek & Rage, 2000 b; Gardner, 2001). 163. EXOCC 4. Absence (0) or presence (1) of condition: exoccipitals forming continuous, concave, strap-shaped articular surfaces with basioccipital. A strap-shaped, transversely concave articular surface of the occiput is found exclusively in microsaurs and lysorophids (Carroll & Gaskill, 1978; Wellstead, 1991). 164. EXOCC 5. Absence (0) or presence (1) of condition: exoccipitals expanded and appressed to each other, so as to obliterate basioccipital posterior surface. The derived condition characterizes a more restricted set of taxa than that implied by character 162 above, including dissorophoids, albanerpetontids and crown-lissamphibians. Basioccipital 165. BASOCC 1. Basioccipital notochordal (0) or not (1). Following Clack (1998c, 2001), a notochordal basioccipital is primitively preset in Acanthostega, Ichthyostega and Crassigyrinus. Under DELTRAN, the plesiomorphic condition appears to be a transitional feature of stem-tetrapods, implying parallel acquisitions of the derived state in Greererpeton (Smithson, 1982) and in a clade consisting of baphetids plus crown-tetrapods. Under ACCTRAN, the plesiomorphic state of Crassigyrinus appears as a reversal. 166. BASOCC 5. Articular surface of basioccipital not convex (0) or convex (1). A convex basioccipital is usually considered to be a shared derived character of crownamniotes (Fox & Bowman, 1966; Clark & Carroll, 1973; Reisz, 1977, 1981), but a less pronounced version of their ‘bulbous’ basioccipital articular surface is present in diadectomorphs (Romer, 1946; Fracasso, 1987; Berman et al., 1998; Berman, 2000) [see also Gauthier et al. (1988b) and Carroll (1991 b)]. 167. BASOCC 6. Absence (0) or presence (1) of condition: basioccipital circular and recessed. Although usually regarded as an aı̈stopod synapomorphy (Carroll, 1998; see also Anderson (in press)), a basioccipital with a recessed posterior surface for a condylar process of the first cervical vertebra is also reported in Adelogyrinus (Andrews & Carroll, 1991). Opisthotic 168. OPI 2. Absence (0) or presence (1) of condition: opisthotic forming thick plate with supraoccipital, separating exoccipitals 320 Marcello Ruta, Michael I. Coates and Donald L. J. Quicke from skull table. Based on Berman’s (2000) recent investigation, the derived condition of the character in question is found in baphetids (Beaumont, 1977), embolomeres (Panchen, 1977; Holmes, 1984, 1989; Smithson, 1985; Clack, 1987a), diadectomorphs (Romer, 1946; Berman et al., 1998; Berman, 2000) and basal crown-amniotes (Fox & Bowman, 1966; Clark & Carroll, 1973; Reisz, 1977, 1981). Parasphenoid 169. PASPHE 1. Parasphenoid without (0) or with (1) elongate, strut-like cultriform process. A parallel-sided, strutlike cultriform process is a temnospondyl feature (Milner, 1988, 1990, 1993, 2000), although its presence is also recorded in colosteids (Smithson, 1982; Hook, 1983), Microbrachis (Carroll & Gaskill, 1978) and Ptyonius (Bossy & Milner, 1998). 170. PASPHE 3. Parasphenoid without (0) or with (1) posterolaterally directed, ventral thickenings (ridges ending in basal tubera). The occurrence of elongate ridges on the parasphenoid is documented in Crassigyrinus (Clack, 1998 c), Eucritta (Clack, 2001), embolomeres (Panchen, 1972, 1977; Holmes, 1984, 1989; Smithson, 1985; Clack, 1987a), seymouriamorphs (White, 1939; Bystrow, 1944; Laurin, 1996b; Klembara, 1997), diadectomorphs (Romer, 1946; Fracasso, 1987; Berman et al., 1998) and crown-amniotes (Fox & Bowman, 1966; Clark & Carroll, 1973; Reisz, 1977, 1981). 171. PASPHE 6. Parasphenoid without (0) or with (1) single median depression. A shallow depressed area occupies a subcentral position in the posterior plate of the parasphenoid of Crassigyrinus (Clack, 1998 c, 2001), Whatcheeria (Lombard & Bolt, 1995), Eucritta (Clack, 2001) and several stem-amniotes, including Caerorhachis (Holmes & Carrol, 1977; Ruta et al., 2001), embolomeres (Panchen, 1972, 1977; Holmes, 1984, 1989; Smithson, 1985; Clack, 1987a), Kotlassia (Bystrow, 1944), Diadectes (Berman et al., 1998), basal crownamniotes (Fox & Bowman, 1966; Clark & Carroll, 1973; Reisz, 1977, 1981) and Westlothiana (Smithson et al., 1994). 172. PASPHE 7. Parasphenoid without (0) or with (1) paired lateral depressions. In Greererpeton (Smithson, 1982), baphetids (Beaumont, 1977), Cochleosaurus (Godfrey & Holmes, 1995) and Micraroter (Carroll & Gaskill, 1978), the posterior plate of the parasphenoid shows two shallow, anteroposteriorly elongate depressions (Coates, 1996; Clack, 1998c, 2001). 173. PASPHE 9. Ventral cranial fissure not sutured (0), sutured but traceable (1), or eliminated (2). Coding of this character follows Clack (1998c, 2001). Among ingroup taxa, Ichthyostega shows the plesiomorphic condition, whereas state 1 occurs in Acanthostega and Crassigyrinus. 174. PASPHE 11. Parasphenoid without (0) or with (1) anterolateral wings projecting anterior to cultriform insertion. In Apateon (Schoch, 1992), Amphibamus (Watson, 1940), Doleserpeton (Bolt, 1969), Karaurus (Milner, 2000) and Triadobatrachus (Rocek & Rage, 2000b), the basipterygoid processes extend considerably laterally and slightly anterior to the proximal insertion of the cultriform process, so that the anterior margin of the parabasisphenoid appears shallowly concave. 175. PASPHE 12. Parasphenoid without (0) or with (1) triangular denticle patch with raised margins at base of cultriform process. A triangular patch of denticles with distinct, raised margins is observed in Cochleosaurus (Godfrey & Holmes, 1995), Dendrerpeton (Holmes et al., 1998), Eoscopus (Daly, 1994) and Doleserpeton (Bolt, 1969). 176. PASPHE 13. Absence (0) or presence (1) of condition: parasphenoid much wider than long immediately behind basal articulation. The derived condition of this feature is shared by derived dissorophoids (Bolt, 1969, 1977, 1979, 1991; Boy, 1972, 1986, 1995), albanerpetontids (Gardner, 2001) and crown-lissamphibians (Schoch, 1992, 1995, 1998; Carroll, 2000; Milner, 2000; Rocek and Rage, 2000b), but it is found also in Isodectes (Sequeira, 1998) and Batropetes (Carroll, 1991a). In all these taxa, the posterior plate of the parasphenoid is at least 25% wider than long and subrectangular or subtrapezoidal in outline. (d ) Lower jaw Jaw articulation 177. JAW ART 1. Jaw articulation lying behind (0), level with (1) or anterior to (2) occiput. No coherent set of internested state changes can be detected for this character. The plesiomorphic state, related or not to the presence of an elongate suspensorium, is widespread among stemgroup taxa and several temnospondyls, but the two derived states occur among trimerorhachoids and some dissorophoids. State 2 is found in several primitive crown-lissamphibians, as well as in diplocaulids, some aı̈stopods, lysorophids, various tuditanomorphs and few microbrachomorphs. The plesiomorphic condition characterizes also the basal portion of the amniote stemgroup, few tuditanomorphs and some urocordylids. State 1 occurs in seymouriamorphs, diadectomorphs, crown-amniotes, Westlothiana and some lepospondyls. Coding follows Clack (1998c, 2001), Paton et al. (1999) and Laurin (1998b). Parasymphysial plate 178. PSYM 1. Presence (0) or absence (1) of parasymphysial plate. A parasymphysial plate is ubiquitous among stem-tetrapods, such as Ventastega (Ahlberg et al., 1994), Early tetrapod relationships revisited Acanthostega (Ahlberg & Clack, 1998), Ichthyostega ( Jarvik, 1980, 1996), Greererpeton (Bolt & Lombard, 2001), Crassigyrinus (Ahlberg & Clack, 1998), Whatcheeria (Lombard & Bolt, 1995) and baphetids (Beaumont, 1977; Ahlberg & Clack, 1998). It is also present in the basal part of the amniote stem-group, in Caerorhachis (Holmes & Carroll, 1977; Ruta et al., 2001), Archeria (Holmes, 1989) and Pholiderpeton scutigerum (Clack, 1987a). Some mandibular fragments of Proterogyrinus, originally figured by Holmes (1984), show a disrupted, denticle-covered area of bone near the symphysis; the latter may represent a parasymphysial plate, although evidence is ambiguous. Whether a parasymphysial plate is present in Anthracosaurus is uncertain (Panchen, 1977). 179. PSYM 2. Parasymphysial plate without (0) or with (1) paired fangs. Parasymphysial plate fangs occur in Acanthostega (Ahlberg & Clack, 1998), Ichthyostega (Jarvik, 1980, 1996), Greererpeton (Bolt & Lombard, 2001), Crassigyrinus (Ahlberg & Clack, 1998), baphetids (Beaumont, 1977; Ahlberg & Clack, 1998) and, possibly, Caerorhachis (Holmes & Carroll, 1977; Ruta et al., 2001). 180. PSYM 3. Parasymphysial plate without (0) or with (1) tooth row. A tooth row on the parasymphysial plate is observed in Ventastega (Ahlberg et al., 1994), Acanthostega (Ahlberg & Clack, 1998) and Whatcheeria (Lombard & Bolt, 1995). 181. PSYM 4. Parasymphysial plate with (0) or without (1) denticles. Among the ingroup taxa, clusters of denticles on the parasymphysial plate are observed in Acanthostega (Ahlberg & Clack, 1998) and, possibly, Caerorhachis (Holmes & Carroll, 1977; Ruta et al., 2001). Dentary 182. DEN 1. Dentary with (0) or without (1) accessory tooth row. Within the crown-group, acces sory tooth rows are recorded in Pantylus and Captorhinus (Fox & Bowman, 1966; Carroll & Gaskill, 1978). 183. DEN 2. Dentary with (0) or without (1) anterior fang pair. The loss of anterior dentary fangs is observed in Acheloma (Olson, 1941), some derived amphibamids (Bolt, 1969), albanerpetontids (Gardner, 2001), crownlissamphibians (Carroll, 2000; Milner, 2000), several embolomeres such as Anthracosaurus (Panchen, 1977), Pholiderpeton attheyi (Panchen, 1972) and Archeria (Holmes, 1989), gephyrostegids (Carroll, 1970), seymouriamorphs (White, 1939; Bystrow, 1944; Laurin, 1996b; Klembara, 1997), Diadectes (Romer, 1946; Berman et al., 1998), Paleothyris and Petrolacosaurus (Clark & Carroll, 1973; Reisz, 1977, 1981), Westlothiana (Smithson et al., 1994) and the majority of lepospondyls, except pantylids, Microbrachis and, possibly, Acherontiscus (Carroll, 1969b; Carroll & Gaskill, 1978; Berman et al., 1988). 321 184. DEN 3. Dentary with (1) or without (0) chamfered ventral margin. A chamfered ventral margin of the dentary has been documented only in Metaxygnathus (not included in our analysis), Ventastega and Acanthostega (Ahlberg et al., 1994; Ahlberg & Clack, 1998). 185. DEN 4. Dentary without (0) or with (1) U-shaped notch for premaxillary tusks. The occurrence of a deep, smoothsurfaced notch near the anterior end of the lateral surface of the dentary is an apomorphy of colosteids (Smithson, 1982; Godfrey, 1989; Bolt & Lombard, 2001). 186. DEN 7. Dentary toothed (0) or toothless (1). A toothless dentary is autapomorphic for salientians (Milner, 1988; Rocek & Rage, 2000 b). 187. DEN 8. Dentary length greater (0) or smaller (1) than half the length between snout and occiput. The derived condition of an abbreviated, stout dentary is found in White, 1939; Batropetes (Carroll, 1991 a) and Brachydectes (Wellstead, 1991). Splenial 188. SPL 2. Absence (0) or presence (1) of condition: rearmost extension of mesial lamina of splenial closer to anterior margin of adductor fossa than to anterior end of jaw. The derived state of this character (see also Ruta et al., 2001) is widespread in the amniote branch of the tetrapod tree [notable exceptions are Phlegethontia (Anderson, in press), Pholiderpeton attheyi (Panchen, 1972) and Rhynchonkos (Carroll & Gaskill, 1978)]. It is also observed in Greererpeton (Bolt & Lombard, 2001), Crassigyrinus (Ahlberg & Clack, 1998) and Megalocephalus (Beaumont, 1977; Ahlberg & Clack, 1998). 189. SPL 3. Absence (0) or presence (1) of suture between splenial and anterior coronoid. The plesiomorphic state of this character, as found in the outgroups, is also present in Acanthostega and Crassigyrinus (see Ahlberg & Clack, 1998). 190. SPL 4. Absence (0) or presence (1) of suture between splenial and middle coronoid. The contact between the splenial and the middle coronoid occurs, under ACCTRAN optimization, in edopoids (Godfrey & Holmes, 1995), trimerorhachoids (Case, 1935), Phonerpeton (Dilkes, 1990), embolomeres other than Anthracosaurus (Panchen, 1972, 1977; Holmes, 1984, 1989; Smithson, 1985; Clack, 1987a), Gephyrostegus (Carroll, 1970; Ahlberg & Clack, 1998), Discosauriscus (Klembara, 1997), Seymouria (White, 1939; Laurin, 1996b) and Rhynchonkos (Carroll & Gaskill, 1978). Postsplenial 191. POSPL 1. Presence (0) or absence (1) of postsplenial. A separately ossified postsplenial is absent in albanerpetontids (Gardner, 2001), crown-lissamphibians (Schoch, 1998; Carroll, 2000; Milner, 2000; Rocek & 322 Marcello Ruta, Michael I. Coates and Donald L. J. Quicke Rage, 2000b), diadectomorphs (Romer, 1946; Berman et al., 1998), crown-amniotes except Petrolacosaurus (Fox & Bowman, 1966; Clark & Carroll, 1973; Reisz, 1977, 1981), Hapsidopareion, Euryodus (Carroll & Gaskill, 1978), lysorophids (Wellstead, 1991), aı̈stopods (Carroll, 1998; Anderson, in press) and nectrideans (Beerbower, 1963; Bossy & Milner, 1998). 192. POSPL 2. Postsplenial without (0) or with (1) mesial lamina. Under ACCTRAN, a mesial lamina of the postsplenial is recorded in all tetrapods more crownward than Crassigyrinus (Ahlberg & Clack, 1998), although a reversal to the plesiomorphic condition characterizes Gephyrostegus (Carroll, 1970) as well as lepospondyls other than microsaurs. 193. POSPL 3. Postsplenial with (0) or without (1) pit line. Following Ahlberg et al. (1994) and Ahlberg & Clack (1998), absence of a postsplenial pit line is recorded in all tetrapods more crownward than Ventastega (however, Ichthyostega is an exception; Jarvik, 1980, 1996). 199. SURANG 3. Surangular with (0) or without (1) pit line. Following Ahlberg & Clack (1998), absence of a surangular pit line is recorded in all tetrapods more crownward than Ventastega (see also Ahlberg et al., 1994). 200. SURANG 5. Absence (0) or presence (1) of condition: lateral exposure of the surangular smaller than that of the angular. In the tuditanomorphs Hapsidopareion, Cardiocephalus and Euryodus (Gregory et al., 1956; Carroll & Gaskill, 1978), the lateral surface of the surangular is a slender, dorsoventrally narrow, elongate and oblique splinter of bone situated at the posterodorsal corner of the angular. Angular 194. ANG 1. Presence (0) or absence (1) of angular. A separately ossified angular is absent in Phlegethontia (Anderson, in press), and crown-lissamphibians except Karaurus (Milner, 1988, 2000). 195. ANG 2. Angular without (0) or with (1) mesial lamina. Following Ahlberg & Clack (1998), the occurrence of an angular mesial lamina characterizes all tetrapods more crownward than Tulerpeton, although Gephyrostegus (Carroll, 1970) shows a reversal to the plesiomorphic condition. 196. ANG 3. Angular contacting prearticular (0) or not (1). The derived feature of this character is apparently autapomorphic for Acanthostega (Ahlberg & Clack, 1998). 197. ANG 4. Angular not reaching (0) or reaching (1) posterior end of lower jaw. The posteriormost part of the external surface of the angular reaches the rear end of the lower jaw in Isodectes (Sequeira, 1998), dissorophoids (Olson, 1941; Boy, 1972, 1987, 1995; Berman et al., 1985; Dilkes, 1990; Boy & Sues, 2000), Discosauriscus (Klembara, 1997), Seymouria (White, 1939; Laurin, 1996b), diadectomorphs (Romer, 1946; Berman et al., 1998), crown-amniotes (Fox & Bowman, 1966; Clark & Carroll, 1973; Reisz, 1977, 1981), Westlothiana (Smithson et al., 1994) and lepospondyls (Carroll & Gaskill, 1978; Andrews & Carroll, 1991; Wellstead, 1991; Bossy & Milner, 1998; Carroll, 1998). Anterior coronoid 202. ANT COR 1. Anterior coronoid present (0) or absent (1). The anterior coronoid is either absent as a separate ossification or of questionable identification in albanerpetontids (Gardner, 2001), crown-lissamphibians (Schoch, 1998; Carroll, 2000; Milner, 2000; Rocek & Rage, 2000b), Diadectes (Berman et al., 1998), crownamniotes (Fox & Bowman, 1966; Clark & Carroll, 1973; Reisz, 1977, 1981), Pantylus, Rhynchonkos (Carroll & Gaskill, 1978), lysorophids (Wellstead, 1991), Batrachiderpeton, Diploceraspis (Beerbower, 1963; Bossy & Milner, 1998) and Phlegethontia (Anderson, in press). 203. ANT COR 2. Anterior coronoid with (0) or without (1) fangs. Absence of fangs on the anterior coronoid is a character of all tetrapods more derived than Ventastega (Ahlberg & Clack, 1998); exceptions are Greererpeton (Bolt & Lombard, 2001) and Gephyrostegus (Carroll, 1970). 204. ANT COR 3. Anterior coronoid with (0) or without (1) denticles. A patch of denticles on the anterior coronoid is documented in Crassigyrinus, Whatcheeria and crowntetrapods (where observed) (Ahlberg & Clack, 1998). 205. ANT COR 4. Anterior coronoid with (0) or without (1) tooth row. All tetrapods more crownward than colosteids (except Whatcheeria; Lombard & Bolt, 1995) lack a tooth row on the anterior coronoid (Ahlberg & Clack, 1998). Surangular 198. SURANG 1. Presence (0) or absence (1) of surangular. Absence of a separately ossified surangular is recorded in albanerpetontids (Gardner, 2001), crownlissamphibians (Carroll, 2000; Milner, 2000; Rocek & Rage, 2000 b) and Phlegethontia (Anderson, in press). Middle coronoid 206. MID COR 1. Middle coronoid present (0) or absent (1). The middle coronoid is either absent as a separate ossification or cannot be identified unambiguously in albanerpetontids (Gardner, 2001), crown-lissamphibians (Schoch, 1998; Carroll, 2000; Milner, 2000; Prearticular 201. PREART 5. Prearticular sutured with splenial (0) or not (1). In post-edopoid temnospondyls, Anthracosaurus and Pholiderpeton attheyi (Panchen, 1972, 1977), the prearticular fails to contact the splenial (Ahlberg & Clack, 1998). Early tetrapod relationships revisited Rocek & Rage, 2000 b), Diadectes (see Berman et al., 1998), crown-amniotes (Fox & Bowman, 1966; Clark & Carroll, 1973; Reisz, 1977, 1981), Pantylus (Carroll & Gaskill, 1978), lysorophids (Wellstead, 1991), Batrachiderpeton, Diploceraspis (Beerbower, 1963; Bossy & Milner, 1998) and Phlegethontia (Anderson, in press). 207. MID COR 2. Middle coronoid with (0) or without (1) fangs. Middle coronoid fangs are absent in all tetrapods more derived than Ventastega (Ahlberg & Clack, 1998), but Gephyrostegus shows a reversal to the plesiomorphic condition (Carroll, 1970). 208. MID COR 3. Middle coronoid with (0) or without (1) denticles. A denticle patch on the middle coronoid characterizes most Devonian taxa, Whatcheeria, baphetids and tuditanomorphs (although only Rhynchonkos can be scored for this character) (Beaumont, 1977; Carroll & Gaskill, 1978; Jarvik, 1980, 1996; Ahlberg et al., 1994; Lombard & Bolt, 1995; Ahlberg & Clack, 1998). 209. MID COR 4. Middle coronoid with (0) or without (1) marginal tooth row. Ventastega (Ahlberg et al., 1994), Acanthostega (Ahlberg & Clack, 1998), Ichthyostega ( Jarvik, 1980, 1996), Whatcheeria (Lombard & Bolt, 1995), Trimerorhachis (Case, 1935) and Rhynchonkos (as well as remaining tuditanomorphs under A) show a tooth row on the middle coronoid (Carroll & Gaskill, 1978; Ahlberg & Clack, 1998). Posterior coronoid 210. POST COR 1. Posterior coronoid present (0) or absent (1). The posterior coronoid is absent as a separate ossification, or cannot be identified unambiguously, in lysorophids ( fide Wellstead, 1991), Sauropleura (Bossy & Milner,1998), Phlegethontia(Anderson,inpress), albanerpetontids (Gardner, 2001) and crown-lissamphibians (Carroll, 2000; Milner, 2000; Rocek & Rage, 2000b). However, we do point out recent contributions by Schoch (1998) bearing on the issue of identification of the caudate coronoid as the posterior coronoid. 211. POST COR 2. Posterior coronoid with (0) or without (1) fangs. Following Ahlberg & Clack (1998), absence of fangs on the posterior coronoid characterizes all postpanderichthyid tetrapods. 212. POST COR 3. Posterior coronoid with (0) or without (1) denticles. A patch of denticles occurs in colosteids, all temnospondyls in which the third coronoid is observed, most stem-amniotes and several lepospondyls (it is absent, however, in Pantylus and Diploceraspis; Beerbower, 1963; Carroll & Gaskill, 1978). Panderichthys, Devonian tetrapods, Whatcheeria, baphetids and crown-amniotes exhibit a denticle-less posterior coronoid (Ahlberg & Clack, 1998). 213. POST COR 4. Posterior coronoid with (0) or without (1) tooth row. A tooth row is primitively present in several 323 Devonian taxa, such as Ventastega (Ahlberg et al., 1994), Acanthostega (Ahlberg & Clack, 1998) and Ichthyostega ( Jarvik, 1980, 1996), but is also documented in Whatcheeria (Lombard & Bolt, 1995), Trimerorhachis (Case, 1935), Diploceraspis (Beerbower, 1963) and Anthracosaurus (Panchen, 1977). 214. POST COR 5. Posterior coronoid without (0) or with (1) posterodorsal process. In its derived condition, this character is present in the majority of crown-tetrapods (Ahlberg & Clack, 1998), except in the diplocaulid Batrachiderpeton (Bossy & Milner, 1998) and in the embolomeres Anthracosaurus (Panchen, 1977), Pholiderpeton attheyi (Panchen, 1972) and Proterogyrinus (Holmes, 1984). 215. POST COR 6. Posterior coronoid not exposed (0) or exposed (1) in lateral view. The posterior coronoid is visible in lateral aspect, immediately posterodorsal to the rearmost end of the dentary, in Greererpeton (Bolt & Lombard, 2001), Whatcheeria (Lombard & Bolt, 1995), some temnospondyls (e.g. Isodectes, Eryops, Micromelerpeton, Leptorophus, Schoenfelderpeton; Sawin, 1941; Boy, 1972, 1987, 1995; Sequeira, 1998; Boy & Sues, 2000), such embolomeres as Archeria (Holmes, 1989) and Pholiderpeton scutigerum (Clack, 1987a; Ahlberg & Clack, 1998), gephyrostegids (Carroll, 1970), seymouriamorphs (White, 1939; Bystrow, 1944; Laurin, 1996b; Klembara, 1997; Ahlberg & Clack, 1998), diadectomorphs (Berman et al., 1998) and primitive crownamniotes (Fox & Bowman, 1966; Clark & Carroll, 1973; Reisz, 1977, 1981). 216. POST COR 7. Posterodorsal process of posterior coronoid not contributing (0) or contributing (1) to tallest point of lateral margin of adductor fossa (‘surangular’ crest). Where present, the posterodorsal process of the posterior coronoid may extend rearward and dorsalward to the point of maximum elevation of the ‘surangular’ crest. Under ACCTRAN, the derived state of this character is found in temnospondyls more derived than trimerorachoids (Sawin, 1941; Boy, 1972, 1987, 1995; Dilkes, 1990; Boy & Sues, 2000), in stem-amniotes more derived than gephyrostegids (White, 1939; Bystrow, 1944; Laurin, 1996b; Klembara, 1997; Ahlberg & Clack, 1998; Berman et al., 1998), and in some basal crown-amniotes (Fox & Bowman, 1966; Clark & Carroll, 1973), although not in Petrolacosaurus (Reisz, 1977, 1981). It is also recorded in Microbrachis and Pantylus, although the situation of other microsaurs is uncertain (Carroll & Gaskill, 1978). Adductor fossa 217. ADD FOS 1. Adductor fossa facing dorsally (0) or mesially (1). A mesially facing adductor fossa occurs in baphetids and all crown-tetrapods in which the 324 Marcello Ruta, Michael I. Coates and Donald L. J. Quicke posterior part of the mesial surface of the lower jaw is observed. Coding for this character follows Ahlberg & Clack (1998). (e) Teeth 218. TEETH 1. Absence (0) or presence (1) of pedicely on marginal teeth. The derived condition of this character is present in Amphibamus and Doleserpeton (Bolt, 1969, 1977, 1979, 1991; Milner, 1988, 1990, 1993, 2000; Trueb & Cloutier, 1991), and is shared with crownlissamphibians, although only Eocaecilia (Carroll, 2000) and Valdotriton (Evans & Milner, 1996) could be scored for the presence of pedicely. 219. TEETH 2. Marginal teeth monocuspid (0) or multicuspid (1). Multicuspid teeth characterize different tetrapod groups, but the position and number of the cusps vary (Carroll, 1991 a; Gardner, 1999, 2001). For this reason, we coded for the ‘multicuspid’ condition separately from the condition ‘two cuspules labiolingually arranged’, detailed in the following character. We also note that in several Permo-Carboniferous and Triassic stem-amniotes, the marginal teeth show several cusps (Ivakhnenko, 1987). Laurin (1998 b) did not distinguish between the mesiolateral cusps shown by brachystelechids and the labiolingual cusps of dissorophoids/lissamphibians, thus conflating under the same state (multicuspid) different morphological patterns. 220. TEETH 3. Marginal teeth without (0) or with (1) two cuspules labiolingually arranged. This character is present in some dissorophoids, notably Amphibamus and Doleserpeton (Bolt, 1969, 1977, 1979, 1991; Milner, 1988, 1990, 1993, 2000; Trueb & Cloutier, 1991), and is shared with crown-lissamphibians, although only Eocaecilia could be scored for this condition (Carroll, 2000). 221. TEETH 5. Dentary teeth not larger (0) or larger (1) than maxillary teeth. Dentary teeth are distinctly larger than maxillary teeth in colosteids (Smithson, 1982; Hook, 1983), but this condition is achieved in parallel by the temnospondyls Isodectes (Sequeira, 1998) and Balanerpeton (Milner & Sequeira, 1994). 222. TEETH 6. Marginal tooth crowns not chisel-tipped (0) or chisel-tipped (1). The crown tips of the marginal teeth are chisel-shaped in adelospondyls (Andrews & Carroll, 1991), as well as in some embolomeres, notably Proterogyrinus (Holmes, 1984), Pholiderpeton attheyi (Panchen, 1972, 1980), P. scutigerum (Clack, 1987 a) and Archeria (Holmes, 1989). 223. TEETH 7. Marginal tooth crowns without (0) or with (1) dimple. According to Andrews & Carroll (1991), the presence of an anteroposteriorly elongate depression, or dimple, on the proximal half of the labial and lingual surfaces of the tooth crowns is regarded as a diagnostic feature of adelospondyls (see also character 222). 224. TEETH 8. Absence (0) or presence (1) of condition: marginal tooth crowns robust, conical structures. Pantylid and gymnarthrid microsaurs possess blunt and massive tooth crowns the height of which is only slightly greater than or as great as their basal width (Gregory et al., 1956; Carroll & Gaskill, 1978). (2) Postcranial skeleton (a) Pectoral girdle Cleithrum 225. CLE 1. Absence (0) or presence (1) of T-shaped dorsal expansion of cleithrum. A robust, T-shaped expansion of the dorsal portion of the cleithrum is an apomorphy of diplocaulid nectrideans (A. C. Milner, 1980; Bossy & Milner, 1998). 226. CLE 2. Cleithrum with (0) or without (1) postbranchial lamina. As discussed by Coates (1996), a postbranchial lamina is primitively retained in Devonian taxa (Acanthostega, Ichthyostega) and Greererpeton, and may be present also in Whatcheeria [but see Coates (1996), Lombard & Bolt (1995) and Bolt & Lombard (2000)]. 227. CLE 3. Cleithrum co-ossified with (0) or separate from (1) scapulocoracoid. The derived state of this character [coding based on Carroll (1995), Coates (1996), Clack (1998 c, 2001) and Laurin (1998 b)] is found in Tulerpeton and all post-Devonian tetrapods (Lebedev & Coates, 1995). Clavicle 228. CLA 3. Clavicles meet anteriorly (0) or not (1). The condition of anteriorly separated clavicles is widespread among early tetrapods; it is found in Acanthostega, Ichthyostega, baphetids, most temnospondyls and the vast majority of ‘reptiliomorphs’, except seymouriamorphs (White, 1939; Bystrow, 1944; Laurin, 1995, 1996 b; Klembara & Bartı́k, 2000), some tuditanomorphs (Carroll & Gaskill, 1978), Scincosaurus and diplocaulid nectrideans (A. C. Milner, 1980; Bossy & Milner, 1998). The present character has been scored as unknown in Crassigyrinus (but see Clack, 1998c, 2001). Interclavicle 229. INTCLA 1. Absence (0) or presence (1) of condition: posterior margin of interclavicle drawn out into parasternal process. This character is widespread among stemtetrapods (Acanthostega, Ichthyostega, Tulerpeton, Crassigyrinus, Whatcheeria) and several ‘reptiliomorphs’ (except for lepospondyls other than microsaurs) (White, 1939; Bystrow, 1944; Carroll & Gaskill, 1978; Jarvik, 1980, 1996; Carroll, 1995; Laurin, 1995, 1996 b; Lebedev & Early tetrapod relationships revisited Coates, 1995; Lombard & Bolt, 1995; Coates, 1996; Klembara & Bartı́k, 2000; Smithson, 2000). 230. INTCLA 2. Absence (0) or presence (1) of condition: parasternal process elongate and parallel-sided for most of its length. Ichthyostega, Whatcheeria and the vast majority of stem- and crown-amniotes display an elongate parasternal process (White, 1939; Bystrow, 1944; Carroll & Gaskill, 1978; Jarvik, 1980, 1996; Carroll, 1995; Laurin, 1995, 1996 b; Lombard & Bolt, 1995; Klembara & Bartı́k, 2000; Smithson, 2000). 231. INTCLA 3. Absence (0) or presence (1) of condition: interclavicle wider than long. This is one of several characters (e.g. see Clack, 1998c, 2001) describing the overall shape of the interclavicle. The occurrence of the derived state matches that of the previous character to a large extent; however, it is not found in the majority of lepospondyls, basal ‘reptiliomorphs’ (embolomeres and gephyrostegids) and most temnospondyls (Ecolsonia, Apateon and Schoenfelderpeton are notable exceptions). 232. INTCLA 4. Interclavicle rhomboidal with posterior half longer (0) or shorter (1) than anterior half. See also Clack (1998c, 2001). The derived state is shown by colosteids, some trimerorhachoids and lepospondyls other than microsaurs and lysorophids. Scapulocoracoid 233. SCACOR 1. Absence (0) or presence (1) of separate scapular ossifications. Based on Carroll (1995), Lebedev & Coates (1995), Coates (1996), Clack (1998 c) and Laurin (1998c), the derived state of this character is found sporadically among tetrapods, and does not identify monophyletic groups with the exception of the clade encompassing Discosauriscus and Seymouria. 234. SCACOR 2. Glenoid subterminal (0) or not (1) (scapulocoracoid extending ventral to posteroventral margin of glenoid). Reisz (1981) noted this feature in several basal crownamniotes, but its occurrence is more widespread (e.g. Ichthyostega, Tulerpeton, several lepospondyls, derived temnospondyls). 235. SCACOR 3. Presence (0) or absence (1) of enlarged glenoid foramen. The derived condition of this character is found in Acanthostega, derived temnospondyls and most lepospondyls (Carroll & Gaskill, 1978; Coates, 1996). 236. SCACOR 4. Absence (0) or presence (1) of ventromesially extended infraglenoid buttress. The derived state (where observed) is found in all tetrapods more derived than Acanthostega (Lebedev & Coates, 1995; Coates, 1996). Anocleithrum 237. ANOCLE 1. Presence (0) or absence (1) of anocleithrum. The distribution of the anocleithrum among 325 early tetrapods is rather sparse. Among Devonian post-panderichthyid tetrapods, it is found in Acanthostega and Tulerpeton (Lebedev & Coates, 1995; Coates, 1996), whereas among post-Devonian taxa, it has been recorded so far in Pholiderpeton scutigerum (Clack, 1987b) and Discosauriscus austriacus (Klembara & Bartı́k, 2000). J. Klembara and M. Ruta (personal observations) have identified a possible anocleithrum in a small specimen of the Upper Carboniferous-Lower Permian seymouriamorph Utegenia. (b) Forelimb Humerus 238. HUM 1. Latissimus dorsi process offset anteriorly (0) or aligned with ectepicondyle (1). Coding for this character is based on Coates (1996; see also Clack, 1998c, 2001). The primitive condition is found in Acanthostega (Coates, 1996), Whatcheeria (Lombard & Bolt, 1995), baphetids (Milner & Lindsay, 1998), Discosauriscus [Klembara, 1997; but see also Klembara et al. (2001) for an alternative interpretation of this process in Seymouria], Pantylus (Carroll & Gaskill, 1978) and diadectomorphs (Heaton, 1980; Berman & Sumida, 1990; Sumida, 1997; Berman et al., 1998). 239. HUM 2. Absence (0) or presence (1) of distinct supinator process projecting anteriorly. A distinct, robust and anteriorly projecting supinator process occurs in some temnospondyls (especially heavily built and terrestrial forms), someseymouriamorphs,diadectomorphs,severalprimitive crown-amniotes and some nectrideans (Miner, 1925; Olson, 1941; Bystrow, 1944; Heaton, 1980; Reisz, 1977, 1981; Sumida, 1997; Berman et al., 1998; Bossy & Milner, 1998). 240. HUM 3. Presence (0) or absence (1) of ventral humeral ridge. Embolomeres are the only group among crowntetrapods that retain such a ridge, which is otherwise found in some stem-tetrapods ( Jarvik, 1980, 1996; Panchen, 1985; Godfrey, 1989; Lebedev & Coates, 1995; Coates, 1996; Milner & Lindsay, 1998). 241. HUM 4. Latissimus dorsi process confluent with (0) or distinct from (1) deltopectoral crest. The description of this character is based on data from Lebedev & Coates (1995) and Coates (1996). The derived state is observed in all tetrapods more derived than Tulerpeton. 242. HUM 5. Presence (0) or absence (1) of entepicondylar foramen. See Carroll (1995), Lebedev & Coates (1995), Coates (1996), Clack (2001) and Laurin (1998b) for an analysis of the distribution of this character. The entepicondylar foramen is absent in some tuditanomorphs, most nectrideans and most temnospondyls [for exceptions, see Carroll & Gaskill (1978), A. C. Milner 326 Marcello Ruta, Michael I. Coates and Donald L. J. Quicke (1980), Milner (1988, 1990, 1993) and Bossy & Milner (1998)]. 243. HUM 6. Presence (0) or absence (1) of ectepicondylar foramen. The ectepicondylar foramen is absent in crown-tetrapods as well as in Tulerpeton (Lebedev & Coates, 1995), Greererpeton (Godfrey, 1989), Whatcheeria (Lonbard & Bolt, 1995) and baphetids (Milner & Lindsay, 1998). 244. HUM 7. Presence (0) or absence (1) of distinct ectepicondyle. This character is slightly modified from Laurin (1998 b). Absence of a distinct ectepicondyle characterizes several derived temnospondyls, crownlissamphibians, certain tuditanomorphs, microbrachomorphs and lysorophids (Boy, 1972, 1974, 1978, 1985, 1986, 1987, 1995; Bolt, 1969, 1977, 1979, 1991; Carroll & Gaskill, 1978; Ivakhnenko, 1978; Wellstead, 1991; Schoch, 1992; Evans & Milner, 1996; Gardner, 1999, 2001; Rocek & Rage, 2000 b). 245. HUM 8. Absence (0) or presence (1) of condition: ectepicondylar ridge extending distally to reach distal humeral end. The derived state of this character is found in panderichthyids and all more crownward tetrapods (Vorobyeva, 1977, 1992, 2000; Vorobyeva & Schultze, 1991). 246. HUM 9. Distal extremity of ectepicondylar ridge aligned with ulnar condyle (0), between ulnar and radial condyles (1), or aligned with radial condyle (2). Tulerpeton (Lebedev & Coates, 1995), colosteids (Godfrey, 1989), Whatcheeria (Lombard & Bolt, 1995) and Eoherpeton (Smithson, 1985) possess state 1. State 2 characterizes most crowntetrapods. 247. HUM 10. Humerus without (0) or with (1) expanded extremities (waisted). Regardless of the degree of torsion along the axis of the shaft, expansion of the humeral head and humeral condylar extremity occurs in temnospondyls, crown-lissamphibians, several ‘reptiliomorphs’ (but not embolomeres and gephyrostegids), crown-amniotes, most microsaurs (except Microbrachis), lysorophids and Scincosaurus (Wellstead, 1991; Coates, 1996; Bossy & Milner, 1998; Clack, 1998 c, 2001). 248. HUM 11. Radial condyle terminal (0) or ventral (1). Except for Ichthyostega ( Jarvik, 1980, 1996), stemtetrapods show a terminal radial condyle, as do some temnospondyls and basal ‘reptiliomorphs’. 249. HUM 12. Humerus slender and elongate, with length less (0) or more (1) than three times the diameter of its distal end. The derived state of this character applies to several derived dissorophoids and crown-lissamphibians (Milner, 1988, 1990, 1993), but it is known also in some crown-amniotes (Clark & Carroll, 1973; Reisz, 1977, 1981; Carroll, 1991 b). 250. HUM 13. Posterolateral margin of entepicondyle lying distal with respect to plane of radial-ulnar facets (0) or not (1). The primitive condition of the entepicondyle is found in the outgroups (Andrews & Westoll, 1970; Vorobyeva, 1977, 1992, 2000), but occurs also in some temnospondyls, such as Eryops and trematopids (Miner, 1925; Olson, 1941). 251. HUM 14. Posterolateral margin of entepicondyle markedly concave (0) or not (1). The derived state is found in Eusthenopteron and Acheloma only (Andrews & Westoll, 1970; Olson, 1941). 252. HUM 15. Width of entepicondyle greater (0) or smaller (1) than half humeral length. In its derived state, this character is acquired in parallel by the clade encompassing derived dissorophoids plus crown-lissamphibians (Milner, 1988, 1990, 1993; Boy & Sues, 2000), Pholiderpeton scutigerum (Clack, 1987a) among embolomeres, crown-amniotes, most lepospondyls, except pantylid and ostodolepid tuditanomorphs, diplocaulids and Urocordylus (Carroll & Gaskill, 1978; A. C. Milner, 1980; Bossy & Milner, 1998). 253. HUM 16. Portion of humeral shaft length proximal to entepicondyle smaller (0) or greater (1) than humeral head width. The derived condition relates to elongation of the humerus in various stem-amniotes (exceptions are Proterogyrinus, seymouriamorphs and diadectomorphs), crown-amniotes and most lepospondyls other than diplocaulids and Urocordylus (White, 1939; Bystrow, 1944; Fox&Bowman,1966;Reisz,1977,1981; Heaton, 1980; A. C. Milner, 1980; Holmes, 1984; Laurin, 1995, 1996 b; Sumida, 1997; Bossy & Milner, 1998; Klembara & Bartı́k, 2000). It is also known in trimerorhachoids, dissorophoids and crown-lissamphibians. 254. HUM 17. Presence (0) or absence (1) of accessory foramina on humerus. Following Lebedev & Coates (1995), Coates (1996) and Clack (1998 c, 2001), the primitive state of this character is observed in all tetrapods more derived than Tulerpeton. Crassigyrinus is, however, a notable exception (Panchen, 1985). 255. HUM 18. Humerus length greater (0) or smaller (1) than the length of two and a half mid-trunk vertebrae. Tulerpeton, colosteids and Crassigyrinus exhibit the plesiomorphic condition, which also appears as a reversal in Westlothiana and lepospondyls (except for tuditanomorphs) (Hook, 1983; Godfrey, 1989; Smithson et al., 1994; Lebedev & Coates, 1995; Coates, 1996). Radius 256. RAD 1. Radius longer (0) or shorter (1) than humerus. Most post-panderichthyid tetrapods exhibit the derived condition (Coates, 1996). The plesiomorphic state appears in some diplocaulids (A. C. Milner, 1980; Bossy & Milner, 1998). 257. RAD 2. Radius longer than (0), as long as (1), or shorter than (2) ulna. The derived conditions expressed by states 1 and 2 overlap each other in several regions of the tree, Early tetrapod relationships revisited including the lissamphibian stem (state 1 occurs in crown-lisamphibians and some dissorophoids as well as among basal temnospondyls) and the lepospondyl branch (in some microbrachomorphs and most nectrideans). Ulna 258. ULNA 1. Absence (0) or presence (1) of olecranon process. An olecranon process [see coding in Coates (1996), Clack (1998 c, 2001) and Laurin (1998b)] occurs in some Devonian and most post-Devonian taxa. It is absent in some tetrapods that possess poorly developed and/or miniaturized limbs. (c) Pelvic girdle Ilium 259. ILI 3. Absence (0) or presence (1) of dorsal iliac process. Following Lebedev & Coates (1995), Coates (1996), Clack (1998 c, 2001) and Laurin (1998b), the presence of a dorsal iliac process is primitive for tetrapods (Devonian and various basal Carboniferous forms). Its loss or drastical reduction is documented in temnospondyls, crown-amniotes and most lepospondyls (some microsaurs represent exceptions; Carroll & Gaskill, 1978). 260. ILI 6. Supraacetabular iliac buttress less (0) or more (1) prominent than postacetabular buttress. The coding for this character derives from Coates (1996). Its derived state characterizes all tetrapods more derived than Acanthostega. 261. ILI 7. Absence (0) or presence (1) of transverse pelvic ridge. A transverse pelvic ridge appears in some temnospondyls (notably Eryops and Dendrerpeton; Romer, 1947; Holmes et al., 1998). It also represents a transient feature of basal stem-amniotes, in agreement with the conclusions of Coates (1996) and Ruta et al. (2001). For a discussion of the nature of the ridge and its possible homology with the iliac shelf of seymouriamorphs and diadectomorphs, see Sumida (1997) and Klembara & Bartı́k (2000). 262. ILI 9. Absence (0) or presence (1) of condition: ilium an elongate rod directed anteriorly. This character applies exclusively to salientians (Milner, 1988; Rage & Rocek, 2000b). 263. ILI 10. Acetabulum directed posteriorly (0) or laterally (1). The coding for this character is from Coates (1996) and characterizes all post-Eusthenopteron tetrapods (Andrews & Westoll, 1970; Vorobyeva, 1977, 1992, 2000). Ischium 264. ISC 1. Ischium not contributing (0) or contributing (1) to pelvic symphysis. The distribution of this character is identical to that of the previous character. 327 (d ) Hindlimb Femur 265. FEM 1. Absence (0) or presence (1) of condition: internal trocanter with a distinct process. Whatcheeria, some primitive crown-lissamphibians, certain embolomeres, Seymouria and Limnoscelis display the plesiomorphic condition (White, 1939; Romer, 1946; Panchen, 1972; Lombard & Bolt, 1995). 266. FEM 2. Absence (0) or presence (1) of condition: internal trocanter separated from femur by distinct trough-like space. This character is based on data from Coates (1996). It appears in its derived state in certain dissorophoids and microsaurs, some ‘reptiliomorphs’ (e.g. Caerorhachis, Kotlassia and Westlothiana; Bystrow, 1944; Holmes & Carroll, 1977; Smithson et al., 1994; Ruta et al., 2001) and several stem-tetrapods, including Acanthostega, Tulerpeton, Crassigyrinus, Whatcheeria and colosteids (Panchen, 1985; Godfrey, 1989; Lebedev & Coates, 1995; Lombard & Bolt, 1995). 267. FEM 3. Absence (0) or presence (1) of condition: fourth trocanter with a distinct rugose area. The character appears in stem-tetrapods and is maintained in the basal part of the stem-lissamphibian and stem-amniote trees. It is found also in some primitive crown-amniotes, such as Captorhinus (Fox & Bowman, 1966). It is lost in most lepospondyls, some seymouriamorphs, gephyrostegids and various primitive crown-amniotes. 268. FEM 4. Proximal end of adductor crest of femur not reaching (0) or reaching (1) midshaft length. Coates (1996) examined patterns of proximal displacement of several processes of the femur in several lineages within the tetrapod crown-group. Whatcheeria, some seymouriamorphs (White, 1939; Klembara & Bartı́k, 2000), diadectomorphs (Romer, 1946; Berman & Sumida, 1990) and Balanerpeton (Milner & Sequeira, 1994) appear to reverse to the plesiomorphic state, whereas Greererpeton (Godfrey, 1989) and Crassigyrinus (Panchen, 1985) show the derived condition. 269. FEM 5. Femur shorter than (0), as long as (1), or longer than humerus (2). Acanthostega (Coates, 1996), trimerorhachoids (Case, 1935), Ecolsonia (Berman et al., 1985) and pantylids (Berman et al., 1988) exhibit state 1, whereas state 2 is ubiquitous among remaining tetrapods. We coded Ichthyostega as unknown, pending redescription of postcranial material showing association of anterior and posterior limbs. Tibia 270. TIB 7. Without (0) or with (1) flange on posterior edge. The coding for this character follows Lebedev & Coates (1995) and Coates (1996). The derived condition is shown by Tulerpeton, Whatcheeria and Westlothiana 328 Marcello Ruta, Michael I. Coates and Donald L. J. Quicke (Smithson et al., 1994; Lebedev & Coates, 1995; Lombard & Bolt, 1995). Fibula 271. FIB 1. Fibula not waisted (0) or waisted (1). Based on Lebedev & Coates (1995), Coates (1996) and Clack (1998 c, 2001), a waisted fibula occurs in all tetrapods more crownward than Ichthyostega. 272. FIB 3. Absence (0) or presence (1) of ridge near posterior edge of flexor surface of fibula. This and the following characters describe conditions of the flexor surface of the fibula [see Lebedev & Coates (1995) and Coates (1996)]. Character 272 shows its derived state in Acanthostega, Crassigyrinus, Baphetes, Eryops, embolomeres, Gephyrostegus, Seymouria and Limnoscelis (White, 1939; Romer, 1946, 1947; Carroll, 1970; Panchen, 1980, 1985; Holmes, 1984, 1989; Smithson, 1985; Milner & Lindsay, 1998). 273. FIB 4. Absence (0) or presence (1) of rows of tubercles near posterior edge of flexor surface of fibula. The derived condition is present only in Tulerpeton (Lebedev & Coates, 1995) and Whatcheeria (Lombard & Bolt, 1995). Tarsus 274. TAR 1. Absence (0) or presence (1) of ossified tarsus. All post-panderichthyid tetrapods show ossified elements in the tarsus (discussion in Coates, 1996). 275. TAR 2. Absence (0) or presence (1) of one proximal tarsal ossification, or presence of more than two ossifications (2). Proximal tarsal elements are observed in Acanthostega and all more derived tetrapods ( Jarvik, 1980, 1996; Lebedev & Coates, 1995; Coates, 1996). 276. TAR 3. Tarsus without (0) or with (1) L-shaped proximal tarsal element. A proximal tarsal element with an indentation along its proximal margin (Lebedev & Coates, 1995; Coates, 1996) is found in Tulerpeton (Lebedev & Coates, 1995), several basal stem-amniotes including embolomeres and gephyrostegids (but not seymouriamorphs; White, 1939; Carroll, 1970; Boy & Bandel, 1973; Holmes, 1984, 1989; Sumida, 1997; Berman et al., 2000; Ruta et al., 2001), Westlothiana (Smithson et al., 1994), Diadectes (Heaton, 1980; Sumida, 1997) and several primitive crown-amniotes (Fox & Bowman, 1966; Clark & Carroll, 1973; Reisz, 1977, 1981). 277. TAR 4. Absence (0) or presence (1) of distal tarsals between fibulare and digits. The derived condition of this character is found in Tulerpeton and all more crownward Letrapods [data from Lebedev & Coates (1995) and Coates (1996)]. 278. TAR 5. Absence (0) or presence (1) of distal tarsals between tibiale and digits. The distribution of this character is identical to that of the previous character. (e) Axial skeleton Ribs 279. RIB 2. Cervical ribs with (0) or without (1) flattened distal ends. The coding for this character follows in part Clack (1998 c, 2001). The primitive state appears to be widespread among tetrapods. The derived condition is optimized as transitional for at least part of the tetrapod stem-group (from Ichthyostega to Whatcheeria) under ACCTRAN. 280. RIB 3. Ribs mostly straight (0) or ventrally curved (1) in at least part of the trunk. Markedly curved trunk ribs (curvature extending from proximal head to distal tip of the ribs) are found in stem-tetrapods more derived than Acanthostega [implying reversal in colosteids under ACCTRAN (see Godfrey, 1989)], as well as on the amniote branch of the tetrapod tree, including the vast majority of lepospondyls. Poorly pronounced curvature characterizes the ribs of Discosauriscus and Seymouria (White, 1939; Klembara & Bartı́k, 2000), derived diplocaulids (Bossy & Milner, 1998) and aı̈stopods (McGinnis, 1967; Wellstead, 1982; Carroll, 1998; Anderson, in press). As noted by A. R. Milner (1990), a slight curvature is observed in some of the largest temnospondyls (see also Schoch & Milner, 2000). 281. RIB 5. Absence (0) or presence (1) of triangular spur-like posterodorsal process in at least some trunk ribs. Such a process is found uniquely in the ribs of colosteids and adelospondyls (Godfrey, 1989; Andrews & Carroll, 1991). It differs from the slender, needle-like process of certain aı̈stopods (McGinnis, 1967). 282. RIB 6. Absence (0) or presence (1) of condition: elongate posterodorsal flange in midtrunk ribs. The distribution of this character overlaps that of the previous character, but it is not identical to the latter. It describes the occurrence of a sheet-like flange stretching along part of the posterodorsal margin of at least some trunk ribs, and is observed in lysorophids (Wellstead, 1991) as well as colosteids and adelospondyls (Godfrey, 1989; Andrews & Carroll, 1991). 283. RIB 7. Absence (0) or presence (1) of condition: longest trunk ribs poorly ossified, slender rods, the length of which is smaller than the length of three trunk vertebrae. This is one of the characters used by Milner (1988, 1990, 1993, 2000) to unite derived dissorophoids with crown-lissamphibians (see also Boy & Sues, 2000). Cervical vertebrae 284. CER VER 1. Atlas neural arch halves unfused (0) or fused (1). Albanerpetontids (Gardner, 1999, 2001), crownlissamphibians (Bolt, 1991), pantylids, Rhynchonkos (Carroll & Gaskill, 1978), Scincosaurus and diplocaulids (Bossy & Milner, 1998) display the derived condition of this feature [data from Sumida & Lombard (1991), Early tetrapod relationships revisited Sumida et al. (1992), Carroll (1995) and Laurin (1998b)]. 285. CER VER 3. Axial arch not fused (0) or fused (1) to axial (pleuro)centrum. Fusion between axial arch and centrum occurs in crown-lissamphibians, crown-amniotes, diadectomorphs, Westlothiana and most lepospondyls [data from Carroll & Gaskill (1978), Sumida & Lombard (1991), Sumida et al. (1992), Smithson et al. (1994), Carroll (1995) and Laurin (1998b)]. Trunk and tail vertebrae 286. TRU VER 1. Absence (0) or presence (1) of extra articulations above zygapophyses in at least some trunk and caudal vertebrae. The derived condition of this character unites nectrideans and aı̈stopods, although it may not be present in all members of the latter group (A. C. Milner, 1980; Wellstead, 1982; Milner, 1993; Bossy & Milner, 1998; Carroll, 1998; Anderson, in press). 287. TRU VER 2. Absence (0) or presence (1) of condition: neural and haemal spines rectangular to fan-shaped in lateral view. This and the following two characters were used by A. C. Milner (1980), Milner (1993) and Bossy & Milner (1998) to characterize nectrideans. 288. TRU VER 3. Absence (0) or presence (1) of condition: neural and haemal spines facing each other dorsoventrally. See character 287 above. 289. TRU VER 4. Haemal spines not fused (0) or fused (1) to caudal centra. See character 287 above. The derived state is present also in Valdotriton (Evans & Milner, 1996). 290. TRU VER 5. Absence (0) or presence (1) of extra articulations on haemal spines. The derived condition is a shared feature of urocordylids [data from A. C. Milner (1980), Milner (1993) and Bossy & Milner (1998)]. 291. TRU VER 6. Absence (0) or presence (1) of long, distally bifurcated transverse processes on trunk centra. The derived condition is a shared feature of Diplocaulus and Diploceraspis [data from A. C. Milner (1980), Milner (1993) and Bossy & Milner (1998)]. See also character 297 below. 292. TRU VER 7. Absence (0) or presence (1) of ossified pleurocentra. Absence of ossified pleurocentra is only documented in Panderichthys (Vorobyeva, 1992; Vorobyeva & Schultze, 1991) and Crassigyrinus (Panchen, 1985). 293. TRU VER 8. Trunk pleurocentra not fused midventrally (0) or fused (1). The derived condition characterizes Doleserpeton (Bolt, 1969; Daly, 1994), albanerpetontids (Gardner, 1999, 2001), crown-lissamphibians (Bolt, 1991) and the entire amniote branch of the tetrapod tree. 294. TRU VER 9. Trunk pleurocentra not fused middorsally (0) or fused (1). Dorsal fusion of pleurocentra has been 329 documented by Lombard & Bolt (1995) in some specimens of Whatcheeria, and is also found in albanerpetontids (Gardner, 1999, 2001), crown-lissamphibians (Bolt, 1991), some embolomeres (Panchen, 1972; Clack, 1987a; Holmes, 1989), Solenodonsaurus (Laurin & Reisz, 1999), seymouriamorphs (White, 1939; Bystrow, 1944; Klembara & Bartı́k, 2000), diadectomorphs (Heaton, 1980; Sumida, 1997; Berman et al., 1998), crown-amniotes (Fox & Bowman, 1966; Clark & Carroll, 1973; Reisz, 1977, 1981; Carroll, 1991b), Westlothiana (Smithson et al., 1994) and lepospondyls (Carroll, 1999). 295. TRU VER 10. Neural spines without (0) or with (1) distinct convex lateral surfaces. ‘Swollen’ neural arches are present in seymouriamorphs (White, 1939; Bystrow, 1944; Klembara & Bartı́k, 2000), diadectomorphs (Heaton, 1980; Sumida, 1997; Berman et al., 1998), various basal amniotes (the condition is polymorphic for Petrolacosaurus; Reisz, 1977, 1981) and Westlothiana (Smithson et al., 1994). 296. TRU VER 11. Neural spines of trunk vertebrae not fused to centra (0) or fused (1). This character has a nonhomogeneous distribution. It is observed in Doleserpeton (Bolt, 1969; Daly, 1994), albanerpetontids (Gardner, 1999, 2001), crown-lissamphibians (Bolt, 1991), Solenodonsaurus (Laurin & Reisz, 1999), some seymouriamorphs (White, 1939; Bystrow, 1944), diadectomorphs (Heaton, 1980; Sumida, 1997; Berman et al., 1998), crown-amniotes (Fox & Bowman, 1966; Clark & Carroll, 1973; Reisz, 1977, 1981; Carroll, 1991b), Westlothiana (Smithson et al., 1994), microsaurs (except pantylids, Pelodosotis and microbrachomorphs; Carroll & Gaskill, 1978), nectrideans (A. C. Milner, 1980; Bossy & Milner, 1998) and aı̈stopods (McGinnis, 1967; Wellstead, 1982; Carroll, 1998, 1999; Anderson, in press). 297. TRU VER 12. Absence (0) or presence (1) of bicipital rib bearers on trunk centra. Caudates and derived diplocaulids possess this character (A. C. Milner, 1980; Milner, 1988; Evans & Milner, 1996; Bossy & Milner, 1998). A specialized condition of rib bearers is detailed under character 291 above. 298. TRU VER 13. Presence (0) or absence (1) of trunk intercentra. Loss of ossified intercentra characterizes albanerpetontids (Gardner, 1999, 2001) and several lepospondyls (Carroll & Gaskill, 1978; A. C. Milner, 1980; Bossy & Milner, 1998; Andrews & Carroll, 1991; Carroll, 1998, 1999), except some tuditanomorphs, microbrachomorphs and Acherontiscus (fide Carroll, 1969b). 299. TRU VER 14. Trunk intercentra not fused middorsally (0) or fused (1). The derived condition occurs in some embolomeres (Panchen, 1972; Clack, 1987a; Holmes, 1989). 330 Marcello Ruta, Michael I. Coates and Donald L. J. Quicke 300. TRU VER 15. Absence (0) or presence (1) of lateral and ventral carinae on trunk centra. Lateral and ventral elongate keels are present in the centra of lysorophids and adelospondyls (Andrews & Carroll, 1991; Wellstead, 1991). 301. TRU VER 16. Absence (0) or presence (1) of strong proximal emargination in haemal spines of posterior tail vertebrae. This character describes the proximal constriction of posterior haemal spines in the diplocaulids Keraterpeton, Diceratosaurus, Diplocaulus and Diploceraspis (A. C. Milner, 1980; Bossy & Milner, 1998). 302. TRU VER 17. Absence (0) or presence (1) of strong proximal emargination in haemal spines of anterior tail vertebrae. See also preceding character. The proximal constriction of anterior haemal spines is found in Diceratosaurus, Diplocaulus and Diploceraspis (A. C. Milner, 1980; Bossy & Milner, 1998). 303. TRU VER 18. Absence (0) or presence (1) of striate ornament on vertebral centra. As described by A. C. Milner (1980) and Bossy & Milner (1998), striations are present on the vertebral centra of Diceratosaurus, Diplocaulus and Diploceraspis. 304. TRU VER 19. Absence (0) or presence (1) of condition: tallest ossified part of neural arch in posterior trunk vertebrae lying above posterior half of vertebral centrum. As described here, this feature occurs in Doleserpeton (Bolt, 1991; Daly, 1994), some seymouriamorphs (White, 1939; Bystrow, 1944; Sumida, 1997), Limnoscelis (Romer, 1946), primitive crown-amniotes (Fox & Bowman, 1966; Clark & Carroll, 1973; Reisz, 1977, 1981; Carroll, 1991 b), Westlothiana (Smithson et al., 1994), microsaurs (Carroll & Gaskill, 1978) and lysorophids (Wellstead, 1991). 305. TRU VER 20. Absence (0) or presence (1) of prezygapophyses on trunk vertebrae. The derived condition of this character occurs in all post-panderichthyid tetrapods (Coates, 1996). 306. TRU VER 21. Absence (0) or presence (1) of postzygapophyses on trunk vertebrae. The distribution of this character overlaps that of the preceding character (Coates, 1996), except for the occurrence of the plesiomorphic state in Crassigyrinus (Panchen, 1985) and, possibly, Trimerorhachis (Case, 1935). 307. TRU VER 22. Absence (0) or presence (1) of prezygapophyses on proximal tail vertebrae. The distribution of this character overlaps that of character 305 above (Coates, 1996). 308. TRU VER 23. Absence (0) or presence (1) of postzygapophyses on proximal tail vertebrae. The distribution of this character overlaps that of character 305 above (Coates, 1996). 309. TRU VER 24. Absence (0) or presence (1) of prezygapophyses on distal tail vertebrae. The derived state of this character is found in Tulerpeton and all more derived tetrapods (Lebedev & Coates, 1995; Coates, 1996). 310. TRU VER 25. Absence (0) or presence (1) of postzygapophyses on distal tail vertebrae. The distribution of this character overlaps that of character 309 above (Coates, 1996). 311. TRU VER 26. Absence (0) or presence (1) of capitular facets on posterior rim of vertebral midtrunk centra. A capitular facet on the posterior rim of vertebral midtrunk centra is found in some tuditanomorphs, microbrachomorphs, lysorophids and adelospondyls (Carroll & Gaskill, 1978; Carroll, 1991a; Andrews & Carroll, 1991; Wellstead, 1991). 312. TRU VER 27. Height of neural arch in midtrunk vertebrae greater (0) or smaller (1) than distance between pre- and postzygapophyses. The derived state is present in microsaurs (Carroll & Gaskill, 1978; Carroll, 1991a), lysorophids (Wellstead, 1991), Westlothiana (Smithson et al., 1994), Kotlassia (Bystrow, 1944) and Captorhinus (Fox & Bowman, 1966). ( f ) Digits 313. DIG 1. Absence (0) or presence (1) of digits. Dactyly (Coates, 1996) is a feature of all post-panderichthyid tetrapods, except where secondary loss of limbs occurs (aı̈stopods). 314. DIG 2. Absence (0) or presence (1) of no more than four digits in manus. A tetradactyl manus characterizes Colosteus ( fide Hook, 1983), the temnospondyllissamphbian clade (Milner, 1988) and lepospondyls other than microbrachomorphs (Carroll et al., 1998). 315. DIG 3. Absence (0) or presence (1) of no more than five digits in manus. A pentadactyl manus characterizes Greererpeton (Coates, 1996) and the ‘reptiliomorph’ branch of the tetrapod tree (excluding lepospondyls). 316. DIG 4. Absence (0) or presence (1) of no more than three digits in manus. A tridactyl manus is observed in microbrachomorphs (Carroll & Gaskill, 1978). (g) Fins 317. DOR FIN 1. Presence (0) or absence (1) of dorsal fin. A dorsal fin is lost in Panderichthys and all more crownward tetrapods ( Jarvik, 1980, 1996; Lebedev & Coates, 1995; Cloutier & Ahlberg, 1996; Coates, 1996; Ahlberg & Johanson, 1998). 318. CAU FIN 1. Presence (0) or absence (1) of caudal fin. A caudal fin is lost in Tulerpeton (under ACCTRAN optimization) and all more crownward tetrapods ( Jarvik, 1980, 1996; Lebedev & Coates, 1995; Cloutier & Ahlberg, 1996; Coates, 1996; Ahlberg & Johanson, 1998). 319. BAS SCU 1. Presence (0) or absence (1) of basal scutes. Basal fin scutes are lost in Panderichthys and all more crownward tetrapods (Cloutier & Ahlberg, 1996; Ahlberg & Johanson, 1998). Early tetrapod relationships revisited 331 XIV. APPENDIX 3. DATA MATRIX Characters are divided into groups of five, arranged in horizontal rows and numbered from left to right; bold numbers refer to characters in the leftmost position in each row; parentheses ( ) indicate polymorphism, whereas braces { } indicate partial uncertainty; question marks denote unknown or inapplicable characters. Acanthostega gunnari 1 51 101 151 201 251 301 0??00 00000 01110 00000 00110 10001 00001 00010 00001 00000 00010 01100 10010 11100 ?1000 ??000 00001 00000 11000 00111 00100 00000 10100 10000 00000 ?0000 11010 0101 00010 00000 00000 00100 00000 01010 00000 00001 00000 00011 00110 00000 00012 10000 00001 01010 00001 00000 00001 00000 00000 00000 00000 00000 11100 00010 0?001 00100 00001 01000 00000 00000 00000 10010 00001 00000 Acherontiscus caledoniae 1 51 101 151 201 251 301 ????? ????? 0???0 ????? ????? ????? 000?? ??11? ????? ??0?0 ????? ????? ????? ????? ?1101 1010? ?0000 01000 00?1? 00??1 ????? ????? ????? ????? ????0 010?1 10?00 ??0?? 1?00? ???1? 0000{12} ????? ????? ????? 0???? ????? ????? ????? ????? ????? ????? ?1??? ??000 00??? ???0? ?1??? ????? ??000 ?000? ???0? 01??? ????? ????? ????? ????? ????? ????? ????1 ??0?? 00000 01110 00010 ????? ?11? Adelogyrinus simorhynchus 1 51 101 151 201 251 301 0??00 ?0000 0?120 ????? ????? ????? ????1 0?11? 10?01 00??0 ????? ????? ????? 1???? ?1101 ??1?? 0000? 00001 ????? ????? 10??? 0000? ?1??? 3???? 01??? ??000 ????? ?1?? 100?0 ??0?0 ????? ????? 01100 ????? 01010 010?0 ????? ?1??? 11?0? ????1 00?1? 00?00 0???? ??100 01??? 1100? ?01?1 1?0?? ????? 00??? ?1??? 0???? ?10?1 10000 ????? 0010? ????? 0111? 10000 0001? ????? ?1010 ????? 001?1 Adelospondylus watsoni 1 51 101 151 201 251 301 ????? ?0000 ???20 10000 ????? ????? ????1 ????? 10?01 01?10 0???? ????? ????? 1???? ?11?1 1000? 00000 010?0 00012 001?? 210?1 10000 ??1?? ?1??? ??0?1 ????? ???01 10010 10000 0001? 0000{12} 3???? ????? ??10? 00001 11100 0?00? 00000 ????? ???0? ??2?? ?1??? ??100 00??? 0??0? ?1010 ????? ??000 0110? ???0? ????? ????? ????? ????? ????? ????? ????? ????1 110?? 0???? 01110 001?1 1???? ???? ALBANERPETONTIDAE 1 51 101 151 201 251 301 11100 ????? 001?0 ????? ?1??? 11111 0000? 0?11? ?1??? 1?0?1 ????? 1???1 120?? ????? ?1101 ????? 0000? 01011 ????? ???1? 00110 10101 ????? ????? 00??? ?1010 ???2? 0111 01000 ????? ????? ????? 0000? 1???? 00?10 ????? ????? ??1?? ????? ???00 01?02 ??110 ????? ?1100 ????? 00110 1000? 0?01? ????? 000?? ???0? 00000 ?1100 ???0? ????? 1??01 ?111? 01110 ????? ??010 ????? ??1?? ?1111 ?01?0 01000 00??? ?0111 00?10 ?1111 ???20 0111 01000 ?0100 01110 00210 00000 100?? 00010 00100 11111 10??? 1110? ???00 01011 00101 ?0100 ???00 000?? 001?0 00000 00000 01100 00??? 11?01 00000 21100 10000 0?000 0??0? ?111? 01000 00010 00110 00101 ??010 ?1?11 00000 Amphibamus grandiceps 1 51 101 151 201 251 301 11000 ?0000 10110 00011 ?0??? 11111 00001 0?11? 00001 020?1 101?? 0???0 1??01 11111 ?1010 ??000 0000? 01011 ????? 0011? 00110 Marcello Ruta, Michael I. Coates and Donald L. J. Quicke 332 Anthracosaurus russelli 1 51 101 151 201 251 301 0??01 00000 00110 00000 101?? ????? ????? 0011? 00000 02031 001?0 01010 ????? ????? ?1000 10000 00000 00?00 00011 00010 21100 00001 10000 10101 ?0100 00000 00100 00000 10110 00010 0000{12} 3100? 01100 00001 00000 00000 0?00? 00000 00001 00101 10200 00??? ?1100 00?10 01101 0?010 10000 ??000 0000? ????? ????? ????? ????? ????? ????? ????? ????? ????? ????? ????? ????? ????? ????? ???? Apateon pedestris 1 51 101 151 201 251 301 10100 ?0000 11110 00011 ????? 11111 000?1 0011? 00001 02011 101?? ????? 1100? 11111 ?1011 ??000 0000? ????? ????? 0011? 00110 01000 100?1 ?0001 ???10 ?1?00 ???20 0111 01000 ?0100 01110 00210 00000 1???? 00010 00100 10010 11??? 1110? ???00 01011 00101 01101 ???00 100?? 001?? 10000 00000 01010 00??? 11?01 00000 21100 10000 0?000 0??0? ???1? 0???0 00010 00?10 00111 ??010 ?1?11 ?0000 Archeria crassidisca 1 51 101 151 201 251 301 0??01 00000 00110 00000 00101 10111 00001 00110 00000 02030 0???0 01010 12111 11111 ?1000 10000 00010 00100 00011 00010 21100 10001 00000 00101 ?0000 00100 00100 00011 10010 00010 0000{23} ????0 ????0 ???0? 0?00? 0?000 0?001 00000 00001 00101 10200 00000 11100 00111 01101 00010 10111 01000 01000 11111 00000 11100 10101 20001 10111 01120 11012 11101 ??000 00000 01110 00010 00101 0111 Asaphestera intermedia 1 51 101 151 201 251 301 0??01 10000 00000 00000 ????? ???11 ???11 0011? 00101 00011 0???1 ????? ???1? 1???? ?1000 ??1?? 00004 00101 ????? 0011? ?11?? 00000 00000 00100 00011 ?0010 21100 00000 ?0??0 00100 00010 01000 00010 10000 10?10 {01}???? ????? ??100 000?? ????? 0?0?? 00000 000?? ????? ????? ??100 00??? ???0? ?1010 ??000 0000? ??110 10011 1??0? ?0101 21?0? ???{12}? ????? ????1 000?? 00??? 0???0 10??0 ?1?1 ?1010 10000 00004 00001 101?? 00111 00110 00000 10001 40011 00?10 ?1000 1?020 0111 00000 ?0000 01100 00200 10000 10012 00000 00100 00101 001?? 1110? 01100 00011 00101 00000 ?1000 000?? 00000 00000 00000 10100 00010 ?1?01 00000 2110(01) (01)0000 10000 00010 0?010 00000 01101 00010 ?011? ?1001 01000 00000 01000 10010 00003 0000? ????? 0011? ????? 00010 10011 ?0010 00100 ??000 ????0 ???1 00000 ?0000 01100 01200 0000? 110?? 10000 00000 00101 00010 ????? ????? 00011 00100 00000 1??00 ????? ????? 00000 00010 10100 00??? ??000 00000 2110? 00100 0?001 0?1?? 10101 0???? Balanerpeton woodi 1 51 101 151 201 251 301 10100 ?0000 00110 00011 10101 10011 00001 0011? 00000 020?1 101?? 01010 11101 111?? Baphetes kirkbyi 1 51 101 151 201 251 301 0??00 00000 01120 001?? ????? 1001? 000?? 00111 00000 00010 ?01?0 ????? ???11 ????? ?0000 00010 00000 ????? 20001 ??00? Early tetrapod relationships revisited 333 Batrachiderpeton reticulatum 1 51 101 151 201 251 301 0??00 10001 00000 001?? 01??? ????? 00001 00110 10001 020?0 ?01?? 1???0 ????? 11111 ?1001 ??1?? 00004 10001 10100 ???1? 00110 00000 ?0000 40100 00?00 ??000 ????? 0??? 00000 01100 ?1100 00200 00001 ????? 01110 00001 010?0 021?? 1100? ????? 00010 10100 0001? ?1100 01??? ????? 00011 01010 ????? 001?? ????? 11110 21111 10000 0?001 1??01 ????? 0???0 ?1000 ??1?? 00011 00101 ????? 001?1 11110 00001 00000 01100 00?12 10011 21110 ?0000 00100 00010 0?010 0001? 1000? ????? ????? ????? ????? ????? 0?0?? 00000 00200 12??? ?1100 01??? ???0? ??010 00000 ?1111 10011 11?01 ?01?? 101(12)0 1001? 01?01 000?? 000(01)0 0111 00000 00010 00000 01010 ????? ????0 Batropetes fritschia 1 51 101 151 201 251 301 0??01 ????? 00220 000?? ????? 11111 00011 1011? ?0001 020?1 ????1 ????? 121?? 11111 ?1??? ??010 ??000 ????? ?1101 0111? (01)01?0 Brachydectes elongatus / newberryi 1 51 101 151 201 251 301 0??00 10100 00??0 001?? ?1??? 11110 00011 00110 01??1 1?010 ?01?1 1???1 1100? 111?? ?1000 ??1?? 00014 00101 ????? 00111 111?? 10001 ?0000 4110? 00000 ??000 00120 0111 00001 00?01 01100 00200 00000 1???? 01?10 ????? 010?1 021?? 1110? ???11 00110 ??110 ?001? ?1100 101?? 0100? 0001? 001?? ????? 01??? ?1?0? 00000 ?1001 ????1 0?000 1??01 ?111? 0???0 10000 ??010 00000 0101? ?1?01 00??1 ?1000 ??000 0000? ????? ????? 00111 ?0??? 01000 00000 ?01?1 ???10 ?1011 0???? ?111 01000 ?0000 01110 00200 00000 ????? 00000 00100 00?1? 10??? 11??? ????0 0?012 00101 000?? ???00 ??0?? 000?? 10000 00000 ????? 00??? ?1?0? 00000 21100 10000 0?010 ???0? ?111? 0???0 00000 00?10 00100 ?1??? ?1?01 0?000 ?00?0 001?? ??0?? ????? ??000 10120 ?111 000?? ?000? ????? ????? ?000? 10012 ?0??0 ????? ?0?0? ????? 11??? 1??01 00011 ????? ??001 ????? ??1?0 000?? 10010 00??? 1??00 ????? 1?101 00000 21100 ????? ????1 ????? 10101 01100 00010 ????0 ?0??? ????? 20101 00000 Broiliellus brevis 1 51 101 151 201 251 301 ???00 ?0000 00120 00011 ????? 10111 000?1 0?11? 00001 020?1 101?? ????? ????1 1???? Bruktererpeton fiebigi 1 51 101 151 201 251 301 0???? ?0000 00??? ????? ????? 10111 00001 ??11? 0???0 02??1 ????? ????? 121?? 11111 ?1000 0?000 ?00?? ????? ????? ????1 001?? Caerorhachis bairdi 1 51 101 151 201 251 301 0???? ?0000 00??? 00000 00101 ????? 00001 ????? ?1000 ????? ????? ??0?0 0001? ?0000 21101 10000 000?0 1?000 ?0100 ?0000 00100 00100 00??0 ????0 00?1? 0{12}0?0 0000? ??010 ?1100 00101 ?0000 10100 0?001 00000 001?? 0000? ????0 10200 00010 01000 00110 01?01 00010 01010 10110 01000 0000? ????? ????? ????? ????? ????? ??111 10111 1?1?0 1??12 11101 000?0 00000 01100 00000 11111 001?? ?111 Marcello Ruta, Michael I. Coates and Donald L. J. Quicke 334 Captorhinus aguti 1 51 101 151 201 251 301 0??01 01000 00110 00000 01??? 11111 00011 10111 00?01 02020 001?1 1???0 12101 11111 01000 ??000 00004 00001 11111 ?0111 01101 00000 01??? 41100 10101 11000 11120 0111 00000 ??0?0 01100 10200 0000? 1??12 00100 000?0 01001 011?? ??111 11101 00011 00100 0001? ?0000 10010 00001 100?1 00011 ????? 00??? 11101 00000 21110 10010 10001 1??01 10101 01111 00010 00010 00000 01010 21101 10000 Cardiocephalus sternbergi 1 51 101 151 201 251 301 0??01 10100 00000 01000 ????? ????? ????? 10111 00111 020?1 0?1?0 ????? ????? ????? 01000 ??1?? 01014 00101 ????? ????? ????? 10000 00001 00110 00010 00010 21100 00000 ?0000 00010 00010 01?1? 00010 1000? ??010 {01}1100 01100 01000 00001 01000 0?001 00000 00000 00200 02??? ?1100 001?? 01101 ?1011 ?1000 0001? ????? ????? ????? ????? ????? ????? ????? ????? ????? ????? ????? ????? ???? 01000 10000 00004 00001 ????? ????? ????? 10000 100?1 400?0 00?10 ??000 ????? ???? 00010 ?0000 01100 0?200 0000? ????? 00000 00000 00?01 00??? ????? ????? 00011 00101 00000 ???00 ????? ????? 00000 00011 ?0100 ?0??? ????? ????? 21100 00010 0?010 0?101 ????? ????? 00000 00010 00000 ??010 ????? ????? 01000 10000 00004 ????? 1011? ????? ????? 00000 100?1 40010 ???10 ?1000 ????? ???? 00010 ?0000 11100 01201 0000? ????? 00100 00100 00101 001?? ????? ????? 00011 00101 00000 ?1000 ????? ????? 00000 00011 10100 00011 ????? ????? 21100 00010 0?010 01101 ????? ????? 00000 00010 00100 00010 ????? ????? ?1101 ??000 00002 ????? 10100 ????? ??110 10100 100?0 2???? ???1? ??000 ????? 0??? 1?000 00000 ????? ??2?? 1000? ????? 00000 01001 ?0?00 ?0??? ??000 ????? 00011 10100 00000 ?1001 01??? ????? 00101 000?? ??000 00??? ????? 00000 21001 00000 ??00? ????? ????? 01000 (01)0000 00?10 00000 ????? ????? 0?000 00001 00101 30000 00?01 ?0000 11121 ?1?1 00010 ?0000 10000 10100 00000 110?? 00000 00000 00000 00010 ?1?10 ????1 00011 00100 00000 11000 00??? 0000? 00000 00000 00000 00100 ??100 00000 2110? 00100 0?001 00101 10001 00??0 ?0001 00010 ?0000 00010 {12}0001 ?0000 Chenoprosopus lewisi 1 51 101 151 201 251 301 0??10 ?0000 00200 00011 ?0??? ????? ????? 00111 00000 1?1?0 011?? 0???0 ????? ????? Cochleosaurus florensis 1 51 101 151 201 251 301 ???10 ?0000 00100 00011 00101 ????? ????? 00111 00000 1?1?0 111?? 01010 ????? ????? Colosteus scutellatus 1 51 101 151 201 251 301 0??00 ?0000 0?200 10000 ?0110 ????0 00001 0?110 10001 000?0 00??? 0???0 1???? 1???? Crassigyrinus scoticus 1 51 101 151 201 251 301 0??00 ?0000 01111 00?00 00101 10000 000?1 01111 00000 000?0 0001? 01010 1001? 0???? 11000 0?000 00001 ????0 ????? 00111 001?? Early tetrapod relationships revisited 335 Dendrerpeton acadianum 1 51 101 151 201 251 301 10000 ?0000 00200 00011 ?0??? 10011 00001 00111 00000 020?0 001?? 0???0 11101 1???? 11000 10000 00004 01001 ????? 10111 00110 00000 10011 ?0011 00?10 ?1000 1?120 0111 00000 ?0000 ?1100 00201 00000 1??12 00000 00100 00101 00??? 1110? 01100 00010 00101 00000 ???00 00000 000?? 00000 00000 10100 00??? 11?01 00000 21101 10010 100(01)0 00110 0?010 00100 01101 00010 ?0101 21001 01000 00000 01000 ??000 00004 00001 1?111 10111 00101 00000 00001 41100 10101 11000 01020 0111 00000 00100 00011 00011 21100 00110 ?0100 00100 00100 00010 10000 00110 01100 01001 00001 01100 100?1 00000 10200 011?? ?1100 001?? 1??01 01010 00000 11111 10110 11011 10101 21101 1001{12} 11?01 00001 00000 01111 10000 10001 ?00?0 4???? ???0? ??000 ???20 0111 00000 01000 ????? ????? ?0001 1???? 01110 00001 ????? ????? 1100? ????1 00?12 10100 0???? ??100 01??? 000?1 00011 0101? ????? 00??? ?1?0? 11110 21111 10000 0?00? ????? ?11?? 0???0 00000 00?10 0?000 ????? ???01 10??0 ?0??? ??1?? 00004 10001 ????? ???1? 00110 00000 ?0000 40101 00?00 ??000 ???20 0111 10010 01100 01110 00200 00001 1???? 01100 01001 01010 02??? 1100? ???00 00101 10100 000?? ??100 01??? 000?1 10?11 01010 ????? 00??? ?1?0? 11110 21110 00010 0?001 ????? ?1101 1???0 00000 00010 00000 ????? ?0?01 11??0 ?0??? ??1?? 00004 10001 11010 ????? 00110 ????? ?0000 40101 00000 01000 ????? 011? ????? 01100 01110 00200 0000? ????? ??100 01000 01010 021?? ??00? ????? 00101 00100 000?? ?1100 01??? ???11 10?11 01010 ?1??? 001?? ?1??? 11110 21110 10010 0?001 1??01 ????? 1???0 00000 00010 01000 01010 ????? 11??0 00000 00001 31010 00?01 11000 00020 01?1 00000 ?0100 01100 00200 00000 100?? 00100 00100 00101 011?? 11011 ???00 00012 00101 00001 ?1100 10100 00000 00010 00010 10100 00111 10001 00000 21100 10000 10011 01101 10101 01111 00010 00110 00000 01010 21001 00000 Diadectes absitus 1 51 101 151 201 251 301 0??01 01000 00200 001?? 01??? 10011 00001 00111 00001 02021 ?01?? 1???0 12111 11111 Diceratosaurus brevirostris 1 51 101 151 201 251 301 0??00 ?0001 00200 00011 ????? 10010 11101 0?110 100?1 020?0 1???? ????? 0110? 11111 ?1001 ??1?? 00004 ????? ????? 0011? 00110 Diplocaulus magnicornis 1 51 101 151 201 251 301 0??01 10001 ?0120 00011 ????? 10010 11101 00110 ?0011 010?1 101?? ????? 110?? 11111 Diploceraspis burkei 1 51 101 151 201 251 301 0??01 10001 ?0120 00011 01??? ????? 11101 00110 ?0001 01011 101?0 1???0 ????? 11111 Discosauriscus austriacus 1 51 101 151 201 251 301 0??01 ?0000 00110 001?? 00101 10011 00001 00111 00000 02011 ?01?? 01010 1101? 11111 11000 01000 00003 0000? 10111 10111 00101 Marcello Ruta, Michael I. Coates and Donald L. J. Quicke 336 Doleserpeton annectens 1 51 101 151 201 251 301 11000 ?0000 11110 00011 ?0??? 1111? 00011 0011? 00001 020?1 101?? 0???0 ??00? 11111 ?1010 ??000 00004 01011 ????? ?01?1 00110 01000 00000 40111 00?10 ??111 1?1?? 0??1 01000 ?0100 01111 00211 0000? ????? 00000 00100 11010 101?? ?1??? ????0 01011 00101 1011? ?1100 ??0?? 0010? 10000 00000 ????? 00??? ?1??? 00000 21100 00010 10000 00110 0?010 11101 0??0? ?1010 ?111? ?1?11 011(01)0 10000 ?1101 ??1?? 00001 ????? ????? ????? ????? 10001 ?1??? ????? ????? ??000 ????? ???? 10000 ??0?0 ????? ????? ?110? ????? 010?0 010?1 ????? ????? ????? ????? 00?10 10?0? ????? ????? ????? ????? 001?1 1?0?? ????? ????? ????? ????? 210?1 1000? ????? ????? ????? ????? 0?111 00001 01011 10??0 01010 ??101 11111 11000 ??000 00004 01011 101?? 00111 00??? 00001 10000 ?0011 00010 ?1000 0?1?? ???1 01000 00000 00012 10000 21101 00010 10000 00100 00101 00000 10000 00011 ?1100 0010? 10000 01?00 0?010 00100 0020? 101?? ?1?00 00010 01?01 01010 00000 1110? 10010 11111 11101 21101 1001{12} 0???0 000?? 00000 01000 00000 00111 00000 00010 001?0 01010 ????? 1???? 01000 10000 00004 00001 1011? ????? ?0??? 00000 10011 40010 00010 01000 ????? ???? 00010 ?0000 01100 00200 0000? ????? Dolichopareias disjectus 1 51 101 151 201 251 301 0??00 ?0?00 00??0 ????? ????? ????? ????? 0?110 10001 000?0 ????? ????? ????? ????? 10000 ???10 ????? ????? ????? ????? Ecolsonia cutlerensis 1 51 101 151 201 251 301 11000 10000 00120 00011 10101 1??1? 00001 Edops craigi 1 51 101 151 201 251 301 0??10 10000 00100 00000 ?0101 1001? ????1 00000 00101 00101 001?? ????? ????? 00011 00101 00000 ?1000 ????? ????? 00000 00000 10100 0001? ???11 00??? 21100 00010 0?011 01101 ?1101 01000 00000 00010 00000 00010 21101 00000 Eocaecilia micropoda 1 51 101 151 201 251 301 11101 ?0000 00000 00011 ?1??? 11111 00001 00111 ?1000 0000? 01000 00?10 01010 000?? 20110 00000 00001 ??1?? ?0000 00001 ????? ??100 0001? 10000 00?10 1?0?1 00004 40101 11110 01010 0001? ????? 0?010 01011 001?? 01011 00?00 00200 121?? ?1100 000?? 1??1? ??1?? 1???1 ????? ??111 0000? ????? ??0?? ?1??? ?11?? ?1?11 1{12}??? ????1 00110 1???? ???10 0011? 00000 01110 ?0000 11111 ?01?? ?111 Pholiderpeton (= Eogyrinus) attheyi 1 51 101 151 201 251 301 0??01 ?0000 00110 00000 10101 ????? 00001 0011? 00000 020?0 001?? 01010 ????? 111?? ?1000 10000 00002 ????? 10100 ????0 ?0??? 00000 00101 3?100 ???01 ?1000 ?11?? ???? 00010 ?0000 01100 10200 0100? ????? 00000 00100 00001 00??? ????? ????1 00011 00100 00000 ?1100 ????? 000?? 00010 00010 00000 00011 ????? 00000 21100 10110 0?001 01101 ????? 01110 00001 00010 00000 00010 ????? 00010 Early tetrapod relationships revisited 337 Eoherpeton watsoni 1 51 101 151 201 251 301 0??01 ?0000 0020? 00?00 00101 1011? 00001 00110 00000 02030 0???? 01010 ???11 111?? ?1000 0?000 00004 ????? 101?0 1011? 00??? 00000 00101 41??? ??1?1 ?1000 ??1?0 ???1 00010 ?0000 ????? 10200 0000? 110?? 00100 00?00 ?000? ????? ?1??? ????1 00011 00?00 00000 ?1?00 ??010 0000? 00010 00000 ?0000 00111 1?100 00000 21100 10010 0?00? 01101 10101 01100 00000 00?10 ?0000 00010 10001 00000 ?1000 ??000 00004 ????? ?0??? 00111 00110 01000 1000? ??011 ???1? ??000 101?0 0111 01000 ?0100 10010 ???01 0000? 10012 00010 00100 10?11 ??1?? ?110? 01100 00012 00101 00000 ???00 000?1 00100 10000 00000 ?1100 00??? ???0? 00000 21100 10010 10000 00110 0?010 00101 01101 0???? ?111? ?1?1? 010(01)0 00000 11000 ??000 00104 01001 10111 101?1 00110 00000 10001 40011 00010 11000 01120 0111 00010 ?0000 01100 00200 00000 11012 00100 00100 00101 001?? 1110? 01100 00011 00000 21100 00010 00101 00000 00010 00010 00000 10100 0?010 00100 ?1000 00010 01101 00010 00000 11111 11101 21100 000(01)0 00000 01000 00000 ?1000 10010 0000? ????? ????? ?011? ??1?? 00000 100?1 ??010 ???01 ??000 ???20 ???1 00000 ?0000 0110? 10200 00000 1??1? 00000 00012 00100 00000 ?0101 00000 00??? ???00 ?11(01)(01) ???00 000?? 01000 ??1?? 00004 00101 ????? ?0111 011?? 00000 00001 00000 00010 00010 21100 00000 ?0000 00010 00010 01?00 00010 10000 11010 {01}101? 01100 01100 00001 11000 0?01? 00000 00000 00200 02??? ?1100 001?? 1??01 01011 ?1000 0001? ?1?11 1001? 1?101 11101 21101 10120 1001{12} ????? ???01 00??? 01110 10000 ???1 Eoscopus lockardi 1 51 101 151 201 251 301 11100 ?0000 01110 00011 ?0?0? ???11 00001 0011? 00001 020?1 101?0 0?0?0 12101 11111 Eryops megacephalus 1 51 101 151 201 251 301 11000 10000 00100 00011 10101 10011 00001 00111 00001 00011 001?0 01010 12101 11111 Eucritta melanolimnetes 1 51 101 151 201 251 301 0??00 ?0000 00111 001?? ????? 1001? ????? 0?11? 00000 000?0 ????? ????? 1111? ????? 00000 000?? ??000 00??? 000?? ????? 21100 10000 0?0?1 ????? ?1?0? ????? 000?0 00?10 ?0000 ????? ?0101 ?0001 ????? Euryodus primus 1 51 101 151 201 251 301 0??01 10100 00000 11000 0???? 11111 ???11 10111 00111 020?1 0?1?? ????? ??1?1 1???? Eusthenopteron foordi 1 51 101 151 201 251 301 0??00 00000 ?0000 00000 00010 00000 00000 00000 ?0010 02000 00000 00100 00000 00000 ?0??? 00000 00?00 00000 0?0?? 00000 000?? 00000 00000 00000 00000 ?0000 ?0000 ?000 00000 00000 00000 00000 00000 0000? 00000 00001 00000 01000 0000? ????0 00000 00000 00000 0?000 00000 000?0 00?00 00000 00000 00000 00?00 00000 00000 00000 0?000 00000 00000 01000 00000 00000 00000 00000 00000 00000 Marcello Ruta, Michael I. Coates and Donald L. J. Quicke 338 Gephyrostegus bohemicus 1 51 101 151 201 251 301 0??01 ?0000 00110 00000 00001 10111 00001 0011? 00000 020?1 001?? 00010 12111 1???? ?1000 0?000 00004 0000? 10111 1011? ?0101 00000 001?1 41010 00?0? 01000 ?0120 0111 00000 ?0000 0???? ??20? 00000 11012 00100 00100 00101 ??1?? 11111 11101 00011 00100 00000 ?1100 00010 00000 10110 00000 11100 00111 11101 00000 21100 10000 1?001 00100 10101 01100 00010 00010 00000 00010 20101 00000 Rhynchonkos (= Goniorhynchus) stovalli 1 51 101 151 201 251 301 0??01 10100 00000 01000 01??? 11111 ???11 10111 01000 00000 00000 001?0 00012 00010 21100 00000 00111 ??1?? ?0000 00010 00010 01100 00010 1000? ??010 0(12)0?1 00004 4?100 01100 01000 00001 01000 0?000 00000 001?1 00101 00000 00200 021?? ?1100 000?1 01101 01010 01100 1???? ?1000 0000? ????? ????? ???01 ?111? ?1101 12101 00111 00120 10012 ?1111 00011 00??? 01110 10000 111?? 011?? ?111 Greererpeton burkemorani 1 51 101 151 201 251 301 0??00 10000 01100 10000 00010 10010 00001 01110 ?1101 10101 10000 00000 00011 00101 21001 10000 1000(01) ??000 10010 10000 00001 10000 00000 00000 00010 00010 00002 20000 01100 00000 00000 00000 0?001 00000 00010 00001 00010 01200 00010 11001 00110 00101 00010 01010 10101 ?0000 10000 0100? 01000 11100 10101 10001 12101 00111 11120 10012 0?110 11000 00000 01000 00000 111?? 00101 01?1 Hapsidopareion lepton 1 51 101 151 201 251 301 0??01 10100 00000 01000 0???? ????? ????? 00111 00101 020?1 0???1 ????? ????? ????? 01000 ??1?? 01014 00101 ????? ????? ????? 00000 00000 00100 00011 00010 21100 00000 ?0000 00100 00000 01000 00010 10000 11010 {01}1100 ?1100 01000 00001 01100 0?01? 00000 00000 00200 02??? ?1100 001?? 1??01 01011 ??000 0000? ????? ????? ????? ????? ????? ????? ????? ???0? ???00 ????? ????? ????? ???? Hyloplesion longicostatum 1 51 101 151 201 251 301 0??01 ?0000 00200 00000 ????? 11110 ???11 0011? 00001 020?1 001?? ????? 12101 111?? ?1101 ??1?? 00004 ??1?? ????? 00111 ?1100 00000 00000 001?0 00010 00011 21111 10010 ?0000 00100 00?00 01100 00010 10000 00010 {01}1?11 0??0? 01101 00001 11111 0?000 00000 ???00 00200 01??? ??100 00??? 0?10? ?1??? ??000 00000 11??? ??01? ???01 ??11? ?1?01 00120 10012 01101 000?? 00??? 0???0 ?0??0 11?1 ?1000 ??000 00000 00000 11000 00111 001?? 00000 10000 00100 00000 ?0000 010?0 ?101 Ichthyostega stensioei 1 51 101 151 201 251 301 0??00 ?0000 01100 00000 00110 10001 00001 01000 00001 00030 0000? 01100 12111 11100 00010 10000 10100 00000 00000 00011 00000 00000 01000 00010 00111 00011 00012 00000 00001 11000 10010 000?? 00001 00000 00000 00010 1?100 00000 11100 00000 0?001 00000 00001 01000 001?0 00000 00000 00010 00101 00000 Early tetrapod relationships revisited 339 Isodectes obtusus 1 51 101 151 201 251 301 10000 10000 00100 00011 ????? ???1? ????? 00111 00000 010?0 00010 ????0 ????? ????? 01001 10000 00001 00001 10111 ????? ????? 00000 00000 10001 00010 ?1000 ????? ???? 00000 00000 01110 00200 1000? ????? 00000 00100 00?1? 12??? ???10 ????? 00010 00100 1000? ??000 00??? ????? 00000 00011 ???0? 00??? ????? ????? 21000 10000 0?000 0?10? ????? ????? 100(01)0 00010 0?000 ?1010 ??0?? ????? ?1010 ??1?? 00004 01011 ????? ?01?? 00110 01001 ?1??? ?0101 00?00 ????? ???2? 01?1 110?0 ??1?1 ?1111 00210 ????? 1???? 00?11 ????? 1???? 12??? ?1??? ???00 01012 ??000 ???1? ??1?? ??0?? 0011? 100?? 001?? ????? 000?? ?1?01 00000 {12}1110 ?1??? ????0 11?10 0?100 0?110 1??01 ??1?? ?111? ?1?1? 0???0 11??0 ?1001 ??1?? 00004 ????? ????? 001?? 001?? 00001 ?00?0 4???? ????? ??000 ???20 ?111 00000 01000 ????? ????? ?0001 1???? 01100 00000 ????? ????? 1100? ????1 00?10 10100 0???? ??100 01??? 000?1 00011 0101? ????? 00??? ?1?0? 11110 21111 10000 0?00? ???0? ?11?? 0???0 10000 00?10 0?000 ?1010 ?0?01 10??0 ?1000 11000 00004 00001 10111 ?01?1 011?? 00000 10001 41100 00001 11000 11120 ?111 00000 ?0100 01100 10200 00000 1001? 00100 00100 01000 011?? 11011 0?101 00010 00101 00001 ?1100 10000 0000? 00010 000?0 00000 00110 11?11 00000 11101 10000 10010 01101 ?1101 01111 10000 00010 00000 00010 21?01 10000 ?1011 ??000 00004 ????? 10111 ????? ?01?? 01000 10??1 40101 ???10 11011 ????? ???1 01000 ?0000 01110 00200 0000? ????? 00010 00100 01010 101?? ???0? ????? 01011 00101 ?1101 ?1??? 00??? ????? 10000 00000 01010 00010 ???0? 00000 21101 10000 0?000 01101 ?111? ????0 ?00?? 00?10 00111 01010 ?1?1? 00??0 ?1101 ??001 10014 ????? ????? ????? 000?? 00000 000?? ????? ????? ????? ????? ?11? 00000 ?00?0 ????? ????? ???0? ????? 00010 00001 ????? ????? ????? ????0 00?1? 10000 ????? ??1?? ????? ????1 ?0??1 0???? ????? 00??? ?1??? 0???? 21001 ??000 ????? ????? ????? 0???0 10?00 ???10 ????? ????? ????? 101?0 Karaurus sharovi 1 51 101 151 201 251 301 11100 ????? 11110 00011 ?1??? ???11 00001 0011? ?1??1 1?0?1 101?? 1???1 1110? 11111 Keraterpeton galvani 1 51 101 151 201 251 301 0??00 ?0000 00200 00?00 ????? 10010 10001 0?110 10001 020?0 0???? ????? 0110? 11111 Kotlassia prima 1 51 101 151 201 251 301 0??01 10000 00200 00000 00101 10011 00011 0011? 00000 01011 001?0 01010 ??111 11111 Leptorophus tener 1 51 101 151 201 251 301 1?100 ?0000 11110 00011 10101 ???1? 000?1 0011? 00011 020?1 101?? 01010 110?? 1???? Lethiscus stocki 1 51 101 151 201 251 301 0??00 ?0000 00110 ????? ????? ????? ???01 ??11? 1000? 02??0 ????? ????? ????? 11111 Marcello Ruta, Michael I. Coates and Donald L. J. Quicke 340 Limnoscelis paludis 1 51 101 151 201 251 301 0??01 01000 00000 00000 0???? 10011 00011 10111 00001 02021 001?1 01010 12111 11111 11000 ??000 00004 00001 10??? 10110 00101 00000 00001 411?? 10101 ?1000 ?1010 0111 00000 ?0100 011?0 00200 00000 1101? 00100 00100 01?0? 01??? 11??? ???01 00011 00100 00001 ?1000 ??010 00001 00111 0001? 0??00 00??? 1?011 00000 21101 10000 1101? 1??01 10101 01111 001?0 00110 00000 01010 21101 10000 00000 ?0000 01100 01200 0000? ????? 10000 00000 00101 00010 ????? ????? 00011 00100 00000 11000 ????? ????? 00000 00001 10100 00111 ????? ????? 21100 00100 0?001 01101 ????? ????? 00000 00010 00000 00010 ????? ????? Megalocephalus pachycephalus 1 51 101 151 201 251 301 0??00 00000 01120 001?? 00111 ????? ????? 00111 00001 00010 ?0000 01110 ????? ????? 01000 ??010 00003 00001 11100 ????? ????? 10010 1001? 30010 00100 ?1000 ????? ???? Micraroter erythrogeios 1 51 101 151 201 251 301 0??01 10110 00000 01000 0???? ????? 00011 10111 00111 1?011 001?1 ????? ???01 11111 01000 ??1?? 01004 00101 ????? 00111 011?? 0000? 0000(01) 001?0 00012 00010 21100 00000 ?0000 10110 00010 00?00 000?? 10000 11010 ?1(01)10 01100 01100 00001 11000 0?001 ?0000 00000 01200 02??? ?1100 001?? 01101 01010 ?1000 00000 11011 101?? 11??? ????? ????? 001?? ????? ???01 00001 00000 01110 10000 ?111 Microbrachis pelikani 1 51 101 151 201 251 301 0??01 10000 00200 10000 00101 11110 00011 0011? 00001 020?1 001?? 01010 11011 11111 ?1101 00000 00000 01100 00010 00011 211?1 00010 ??1?? ?0000 00100 000?0 01100 00010 10000 00010 0000{23} 31110 01100 01101 00001 11100 0?01? 00000 00101 00010 00200 011?? ?1000 00110 01101 01010 10110 11000 00000 11111 101?? ???01 ?01?? ?0101 00111 00120 1001? ???01 00001 00000 01110 001?0 11100 1111 Micromelerpeton credneri 1 51 101 151 201 251 301 11100 ?0000 01110 00011 10??? ???11 ???01 00111 00001 020?1 1000? 01010 1100? 11111 11000 ??000 00001 ????? 10111 ?01?? 00110 01000 01010 00000 00010 100(01)0 21100 00010 100?1 ?0000 00100 00101 00000 10000 00?10 {23}0001 01110 00010 10101 01000 0?000 00111 ???10 00200 111?? ?1000 00010 01101 01010 11000 0000? ??10? 00??? ???0? ?111? ?1?0? ???2? 1???? ???00 001?? 00??? 01000 00000 0111 Neldasaurus wrightae 1 51 101 151 201 251 301 ???00 10000 00100 00011 ????? 10111 00001 0011? 00000 02010 00010 ????? 1210? 1???? ?1000 10000 00000 00000 00012 00000 21000 00000 10000 10000 10000 00100 00101 00010 100?0 00?10 0000{23} ?00?1 11100 00?00 00000 ??000 0?010 00000 00001 00010 00200 00??? ??000 00??? 01101 00010 ????? ?1000 0000? ??00? 010?0 ?110? 11101 21001 001?1 0?1?? ????? ????0 000?? 00000 01000 00000 00110 01?1 Early tetrapod relationships revisited 341 Odonterpeton triangulare 1 51 101 151 201 251 301 0??01 ?0000 00000 00000 ????? 11110 ????1 0011? 00?01 020?1 001?? ????? 110?? 1???? ?1101 ??1?? 00004 ????? ????? ????? 11100 00000 ?1??? ?1110 ???00 ??000 ????? 11?1 00000 ??0?0 01100 00200 00000 ????? 00000 000?0 01100 02??? 11??? ???01 00012 0?110 00001 ?1100 ????? 000?? 100?1 00010 01111 0???? ?1?0? 00??? 21110 10000 0?001 0??0? ????? 0???0 00100 11010 00000 ????? ?1?01 00??0 ?1101 ??001 10014 00001 ????? ????? 000?? 10001 00000 4???? 01000 ??000 ????? ?11? 00000 01010 00010 0000? 21001 10000 00001 ????? ??000 00000 10000 00010 ???0? ?1000 000?? ????? 0?00? 00000 00200 01??? ?1?00 00??? 1??01 01010 0(01)000 11??? ????? ?1??? ????? ????? ????? ???00 000?1 1???? 0???0 101?0 01000 ??000 00004 00001 11111 00111 00101 00000 00001 41100 10101 11000 00120 0111 00000 ?0100 01100 10200 00000 10012 Oestocephalus amphiuminum 1 51 101 151 201 251 301 0??01 10000 0?210 0000? 0???? ????? ??001 0011? 10001 00010 0???0 ????? ????? 11111 Paleothyris acadiana 1 51 101 151 201 251 301 0??01 01000 00110 00000 0???? 1???1 00011 00111 00011 00021 001?1 ????0 12101 111?? 00100 00100 01101 01??? 11111 11111 00011 00100 000?? ?1100 10010 00001 10011 00010 ????? 001?? 1?111 00000 21111 10000 11001 1??01 10101 01110 00010 00010 00000 01010 21111 10000 Panderichthys rhombolepis 1 51 101 151 201 251 301 0??00 ?0000 ?0100 00000 00010 ???00 000?0 00000 ?0??? 00000 00000 00000 00010 00?00 00000 10000 ?000(01) ??000 00000 00000 00001 00000 00000 00000 00000 000?0 00?00 00000 00000 00000 00000 0?000 ??0?0 00000 0000? 00000 00?00 00000 00000 0(01)000 00000 00000 00000 00100 01000 ?0000 00000 0000? 00000 00?00 00001 00000 000?? ????? ????0 ???0? ????0 0000? 00000 00??0 ?0000 00000 000?? ?101 Pantylus cordatus 1 51 101 151 201 251 301 0??01 10000 00000 00000 01??? 10111 00011 10111 00101 1?011 001?1 1???0 12101 11111 01101 ??1?? 00004 00101 1111? 00111 11110 00000 ?0000 ?1100 00000 11000 001?0 01?1 00000 00100 10011 00010 21101 10000 10000 00011 01100 00010 11010 00010 01100 00001 0001? ????? 0?001 00000 00200 011?? ?0000 001?? 01101 01010 00010 11011 10000 11001 10101 21101 ???1{12} 01101 00011 00000 01110 001?0 00000 ?0000 ??1?0 00?00 ??000 00120 ?1?1 00000 10110 ?1100 00200 00000 100?? Pelodosotis elongatum 1 51 101 151 201 251 301 0??01 ?0110 00000 01000 ????? 10111 ???11 1011? 00111 1?0?1 0?1?1 ????? 1210? 1???? ?1000 ??1?? 01004 00101 ????? ?0111 ?11?? 00100 00010 01?00 02??? 11111 ???01 00012 00000 000?? ???00 100?1 0000? 00010 0001? ????? 00??? ?1?01 00??? 21100 10000 0?000 ???0? ?0101 01110 00000 11010 ?0000 ?1010 21101 00000 Marcello Ruta, Michael I. Coates and Donald L. J. Quicke 342 Petrolacosaurus kansensis 1 51 101 151 201 251 301 0??01 01000 00110 00000 01??? 11111 00011 00111 00011 02021 001?1 1???0 12101 11111 ?1000 ??000 00004 00001 11111 00111 00101 00000 00001 41110 10101 01000 00120 01?1 00000 ?0100 01100 10200 00000 10012 00100 00000 01101 011?? 11111 11101 00011 00100 01001 ?1100 10010 00001 10011 00010 01110 001?? 11111 00000 21110 00010 10000 00010 11001 00000 01101 01010 10101 21111 0111(01) 10000 00000 ????? 4???? 01??? ??000 ????? ?11? ?01?? ??0?1 ????? ????? 00000 ????? ??110 ????? ????? ?21?? 11??? ???00 00?0? ???00 0???? ?1100 ????? 00001 ?0??? 00010 ????? 000?? ?1??? ????? 21??1 1000? ????? 1??1? ????? 0???0 ?1??? ??010 ????? ??1?? ????? 101?0 ?1000 0?000 00003 00001 10111 ????? ?01?? 00?00 00101 31000 00101 01000 ????? ???1 ?0010 ?0?00 01100 10200 01000 ????? 0010? 00?00 0000? 00000 11?1? ???01 00011 00?00 ?0000 11000 ??000 0000? 00?1? 00001 00000 00111 10??0 00000 ?110? 00000 0?001 01101 ?0?01 01110 ?0001 00010 ?0000 00010 {12}??01 00010 ?1000 ??000 00004 01011 1011? ????? ?0??? 00001 1000? 40011 00010 11000 ????? ???1 01000 ?0000 01100 00200 00000 ????? 01000 00000 0010? 0?1?? 11??? ????0 00011 10000 21100 00010 00101 00000 10000 00111 (01)000? 10?00 0?010 00000 ?1000 00011 01101 01010 ????0 1?111 1?101 2110? 000?? 00000 01000 00000 ?1010 ??000 00004 ????? 1011? ????1 00??? 01000 100?1 400?1 ???10 ?1011 0012? ?1?? 01000 ?0100 01110 00200 00000 ????? 00010 00100 10?11 101?? 11?0? ???00 00012 00101 ?0000 ?1?00 00??? 001?? Phlegethontia linearis 1 51 101 151 201 251 301 0??00 ????? 00110 ??00? ?1??? ????? ??001 00110 ?0011 020?0 0???0 1???1 ????? 11111 ?1100 ??1?? 10014 00001 ????? ????? 000?? Pholiderpeton scutigerum 1 51 101 151 201 251 301 0??01 00000 00110 00000 00101 1111? 00001 00110 00000 00?30 001?0 01010 121?? 1???? Phonerpeton pricei 1 51 101 151 201 251 301 11000 10000 01110 00011 10101 ????? 00001 00??? 00001 02011 001?0 01010 ??1?? 11111 Platyrhinops lyelli 1 51 101 151 201 251 301 11100 ?0000 ?0100 00011 10??? 11111 00001 0011? 00001 020?1 1?1?0 01010 121?? 11111 00000 00000 ?1100 00010 ???0? 00000 20100 10000 0?010 01101 ?110? 01000 00010 00110 01100 0?010 ?1?11 ?0000 Proterogyrinus scheelei 1 51 101 151 201 251 301 0??01 00000 00110 00000 ?0?0? 10011 00001 00110 00000 02030 0???0 01010 11111 11111 ?1000 10000 000?0 00000 00011 00010 21100 00001 00000 00101 ?0000 00100 00100 00011 100?0 00010 0000{12} 41??0 ???0? ?0001 00000 1?000 0?001 00000 00001 00101 10200 00??? ?1000 0011? 0?101 00010 1010? ?1000 01000 11110 00000 11100 10101 20101 10111 01120 11012 11101 00000 00000 01100 00000 00101 0111 Early tetrapod relationships revisited 343 Ptyonius marshii 1 51 101 151 201 251 301 0??00 00000 00110 00010 ????? 11110 00001 00110 00001 020?0 001?? ????? 1110? 11111 ?1000 ??001 00004 00?01 ????? 0011? 00110 00001 000?1 40100 00?10 ??000 ???20 0111 00001 ?0000 ?1100 00200 00000 1???? 01000 10001 01000 01??? 1110? ????1 00011 10100 00001 ??100 010?? 000?1 00111 00010 00000 00??? 11?0? 11111 21001 10000 0?001 ???0? ?11?? 0???0 10001 00?10 00000 ?1010 ?0101 10??0 10001 00000 01100 00200 00000 1??1? 00001 10001 01000 001?? 1110? 0?1?1 01010 10100 000?? ?1100 010?? 00001 00001 00000 ????? 001?? 11?1? 11111 00111 10000 0?00? 1??0? ?11?? 0???0 21001 10000 00010 00000 01010 ?0101 10??0 Sauropleura pectinata/ scalaris 1 51 101 151 201 251 301 0??00 ?0000 00110 00000 0???? 11110 00001 00110 00001 02010 001?? ????1 1110? 11111 ?1(01)0(01) ??001 000?0 00004 4110? 00001 00?00 ????? ?1000 0011? ???20 00110 0111 Saxonerpeton geinitzi 1 51 101 151 201 251 301 0??01 ?0000 00000 00000 ????? 11111 ???11 0011? 00101 020?1 001?? ????? 12101 1???? ?1000 ??1?? 0000? ????? ????? 00111 11110 00000 00000 001?0 00012 ?0010 21100 00000 ?0000 00100 00010 0?000 00010 10000 11?10 {01}111? 01100 0110? 000?? ????? 0?01? 00000 ???00 00200 01??? ??100 00??? ???0? ?1??? ??000 00000 11111 1001? ?1?01 ?011? ?1101 001(12)0 10012 01101 000?? 00??? 0???0 ?0??0 01?1 Schoenfelderpeton prescheri 1 51 101 151 201 251 301 1?100 ?0000 11110 00011 ????? ???1? ????1 0011? 00001 020?1 101?? ????0 1101? 1???? ?1011 ??000 00004 ????? ???11 ?01?? ?0110 01000 ?0??1 40101 ???10 11?00 ????? 0??1 01000 ?0000 01110 00200 0000? ????? 00?11 00100 11010 12??? ???0? ????? 01010 00101 ?1101 ?1?00 10??? ????? 00000 00010 01010 000?? ???0? 00??? 21101 10000 0?000 01101 ?111? ????? 10010 00?10 00111 01010 ?1?0? ????0 ?1001 ??1?? 00004 10001 ????? 00111 001?? 00001 ?00?0 401?? 00?00 ??000 00120 ?111 00000 00?00 ?11?0 00200 ?0000 1??1? 01100 00011 01111 01??? 1100? ????1 00110 00100 0001? ????? 000?1 000?1 00010 01010 ????? ????? 11?11 11110 21101 10010 0?000 ????? ?0??? 01110 ?1??? 00010 00000 ????? ?1101 101?0 00000 ?0100 01100 00200 00000 11012 00100 00100 00101 011?? 11011 01100 00011 00100 00001 ?1100 ?0110 00000 00010 00010 1110? 00111 11111 00000 21100 10010 10011 01101 10101 01111 000(01)0 00110 00000 01010 21101 10000 Scincosaurus crassus 1 51 101 151 201 251 301 0??00 ????? 00200 00000 ????? 11110 00001 00110 ?0101 020?1 001?? ????? 1210? 11111 Seymouria baylorensis /sanjuanensis 1 51 101 151 201 251 301 0??01 10000 00200 00000 00101 10011 00011 00111 00000 02011 001?0 01010 12111 11111 11000 11000 00004 00001 10111 10110 00101 00000 10001 41010 00001 11000 ?1020 0111 Marcello Ruta, Michael I. Coates and Donald L. J. Quicke 344 Solenodonsaurus janenschi 1 51 101 151 201 251 301 ????1 ?0000 0?100 ????? ????? 10?11 00001 ??11? 00001 020?1 ????? ????? 120?? 1???? ?1000 ??000 00004 ????? ????? ????? 001?? ??000 10??? ????? ????? ??000 ????? ?1?1 00000 ?0000 ????? ????? ?0000 ????? 00100 00100 ????? ?0??? 11?11 ???01 00?11 00100 ????? ????? 100?? 000?? 00111 0?01? ????0 ?0??? ????1 00000 21101 10000 ????? ???0? ?0101 01110 100?? 00?1? ????? ??010 21?01 100?0 01000 ??1?? 00004 00101 ????? ?0111 111?? 00000 ?0000 ?1100 00000 ?1000 10110 ?1?1 00000 ?0010 01100 00200 0001? ????? 00000 00010 00001 02??? ???1? ????? 10012 01?00 0001? ??000 1000? ????? 00010 00010 ????? 00??? 11??? 00000 21100 11000 0?001 ???0? ?1??? 01110 00000 00010 ?0000 ?1010 ?1?0? 001?0 01000 00100 ?0?01 00??? 11??? 011?0 00011 00101 000?? ?1100 ??010 000?? 10000 00000 ????? 00??? 1111? 00000 21100 10000 0?010 ???0? 11101 01000 00010 00111 ?0100 ?1010 21100 00000 Stegotretus agyrus 1 51 101 151 201 251 301 0??01 10100 00000 00000 ????? ????1 00011 10111 00101 010?1 001?1 ????? 1211? 1???? Acheloma cumminsi (= Trematops milleri) 1 51 101 151 201 251 301 11000 10000 01110 00011 ????? 0011? 00001 0011? 00001 02011 00000 ????? ??101 1???? ?1000 ??000 0000? 01011 ????? 00111 001?? 00001 ?0001 ????1 00010 ?1000 001?0 ???1 01000 ?0000 ???0? 00200 00000 1??12 ???0? ?1??? ????? 00?00 ????? ?0?20 ?111 ?1??? ???1? 01??? 11??? ?011? ?1??? ??1?1 ????? ???01 001?? ????? ???1? ???1? ?1011 ??11? ????? 0?000 10110 00210 10??? ????? 10??? 1??1? ??1?? ????0 11??? ??1?? ?1?01 ?111? ?1?11 1001{12} 0??00 0011? 00000 0???0 10??0 Triadobatrachus massinoti 1 51 101 151 201 251 301 ????? ????? ??11? 00011 ?1??? 11111 00001 ????? ?1??1 0?0?1 1???? 1???1 1110? 11111 ?1??? ??1?? 0000? 01011 ????? ?1110 001?? Trimerorhachis insignis 1 51 101 151 201 251 301 10000 10000 00110 00011 101?? 10111 00001 00111 00000 00010 00010 01000 11101 0???? ?1000 00000 00000 00000 00010 00000 21000 00000 00000 1000(01) 00000 00001 00001 00000 10000 00010 0000{23} ??001 11100 00000 00001 00000 0?000 00000 00001 00010 00200 011?? ?1000 00011 01101 00010 10010 01000 0000? ??10? 011?0 ?1?0? ?1101 21001 0011? ???10 100?? ???01 000?? 00000 010?0 00000 00110 0111 Tuditanus punctulatus 1 51 101 151 201 251 301 ????? ?0??? ?0200 ?0000 ????? 11111 ???11 ??11? ?1000 0000? 00001 001?0 001?1 ??1?? ?0??0 001?0 00010 0?0?1 0000? ????? ????? ????? 0???? 00101 00?00 00200 00??? ????? ????? ??000 00000 11111 1{12}?11 00111 001?0 1001{12} 11??? ?1110 0111 00?11 00010 21100 00??? 01000 0001? 10000 00?10 0???? ????? 0?00? ??000 ???00 00??? ???0? ?1010 10??? ???0? ?01?? ?1001 {01}1?11 000?? 00??? 0???0 10??0 Early tetrapod relationships revisited 345 Tulerpeton curtum 1 51 101 151 201 251 301 ????? ????? ????? ????? ????? 1000? ????? ????? ????? ????? ??00? ????? 111?? ????? ????? ????? ????? ????? ????? ????1 ??100 ????? ????? 3?01? ????? ??000 11021 0??? ????? ????? ?01?0 ????? ?0000 10112 ????? ????? ????? ????? ?101? 111?? ????? ????? ????? ?1000 00010 ????? ????? ????? ????? 0???? 10100 ????? ????? ????? ????? ???00 00101 0100? ????? ????? ????? ?0??? 10001 ?0000 ????? ????? 00?0? ????? ????? 0?111 00110 ????? ?00?0 ????? ????? ??000 01120 0111 ??001 00??0 ????? ????? ?0000 1??1? 01000 100?1 ?1?00 ????? 1110? ????1 00??? 10?00 ?00?? ?1100 010?? 000?1 ????? 00000 ????? 00??? ?1?0? 11111 ????? 10?00 ??0?? ???0? ?01?? 0???0 ????? 00?1? ?0?00 ?1010 ?0101 101?0 ?1001 ??1?? 00004 01011 ????? ?0111 00110 01001 ?1??? ?0101 00?00 ??1?? 0?120 0111 111?? ????1 11111 00200 0000? 100?? ???11 ????? 0???? 121?? ?1??? ???00 01010 ??000 ???1? ?1100 ??011 001?1 100?? 001?? ????? 00??? 11?01 00010 011?0 ????1 0?100 1??1? ?111? 0???0 ?1??? ???10 1?010 ??1?? ?1?11 11??0 ????? ????? ????0 ????? 11000 ????? ????? ????? ????? ????? ????? ?0000 ????? ???? ??010 ????? ????? ????? 0000? ????? 00000 ????? ????? ??001 ??0?? ????? 00??? ????? ????? 11010 ????? ????? ????? ????? ????? 00010 ????? ????? ????? ????? ????? 00000 ????? ????? ????? ????? ????? 00000 ????? ????? 01100 ??000 0000? ????? ????? 00111 011?? ?000? 00??1 ????? ???00 ??000 10121 ?111 000?? ?0100 ????? 10200 0000? 10012 0?1?? 00100 ??101 01??? ?1??? 111?1 00?11 01?00 ?0001 ??100 ??0?? 000?? 00011 000?? 1?100 00??? ??111 00000 21100 10000 0?00? ???0? 10101 01111 00000 00010 ?0000 ?1010 21?01 10000 Urocordylus wandesfordii 1 51 101 151 201 251 301 0??00 ????? ????0 ??000 ????? 10010 00001 0?1?? ????? ????0 0???? ????? 1110? 11111 Valdotriton gracilis 1 51 101 151 201 251 301 11100 ????? 10110 00011 ?1??? 11111 00001 0011? ?1??1 1?0?1 1?1?0 1???1 1100? 11111 Ventastega curonica 1 51 101 151 201 251 301 0???? ????? ????0 ????? 00010 ????? ????? ????? ????? ????? ????? 00100 ????? ????? Westlothiana lizziae 1 51 101 151 201 251 301 0??01 ?0000 002?0 00000 ????? 11110 00011 00111 000?1 020?1 0???1 ????? 12111 11111