LUNGFISHES, TETRAPODS, PALEONTOLOGY, AND · ABSTRACT Weconclude that the internal (excurrent) nos- tril of Recent lungfishes is a true choana, as judgedbyits comparisonwith(1) theinternal

LUNGFISHES, TETRAPODS,

PALEONTOLOGY, AND

PLESIOMORPHY

DONN E. ROSENCurator, Department of Ichthyology

American Museum of Natural HistoryAdjunct Professor, City University ofNew York

PETER L. FOREYPrincipal Scientific Officer, Department of Palaeontology

British Museum (Natural History)

BRIAN G. GARDINERReader in Zoology, SirJohn Atkins Laboratories

Queen Elizabeth College, London

COLIN PATTERSONResearch Associate, Department of Ichthyology

American Museum of Natural HistorySenior Principal Scientyifc Officer, Department of Palaeontology

British Museum (Natural History)

BULLETINOF THE

AMERICAN MUSEUM OF NATURAL HISTORYVOLUME 167: ARTICLE 4 NEW YORK: 1981

BULLETIN OF THE AMERICAN MUSEUM OF NATURAL HISTORY

Volume 167, article 4, pages 159-276, figures 1-62, tables 1,2

Issued February 26, 1981

Price: $6.80 a copy

ISSN 0003-0090

Copyright © American Museum of Natural History 1981

CONTENTS

Abstract ........................................ 163Introduction ...................... ........................ 163Historical Survey ...................... ........................ 166Choana, Nostrils, and Snout .............................................. 178

(A) Initial Comparisons and Inferences .......................................... 178(B) Nasal Capsule ............. ................................. 182(C) Choana and Nostril in Dipnoans ............................................ 184(D) Choana and Nostril in Rhipidistians ........................................ 187(E) Choana and Nostril in Tetrapods ........................................... 195

Nasolacrimal Duct, Labial Cavity, Rostral Organ, and Jacobson's Organ .... ......... 196Paired Fins and Girdles ...................... ........................ 202

(A) Paired Fin Skeleton .................... .......................... 202(B) Pectoral Girdle .............. ................................ 216

Dermal Bones of the Skull ....................................................... 221Palate and Jaw Suspension .............................................. 227

(A) Historical ......................................................... 227(B) Dermal Bones of Palate .............................................. 227(C) Palatoquadrate .............. ................................ 232(D) Relation Between Palate and Braincase .................. ................... 233

Hyoid and Gill Arches ...................... ........................ 235Ribs and Vertebrae ............. ................................. 242

(A) Ribs ....................................... 242(B) Interpretation of Gnathostome Vertebrae ................ .................... 246(C) Heterospondyly in Caturus furcatus, and Vertebral Homologies ..... .......... 249(D) Vertebrae of Rhipidistians, Lungfishes, and Tetrapods .......... .............. 249

Dermal Skeleton ...................... ........................ 252(A) Cosmine ......................................................... 252(B) Folded Teeth ............. ................................. 254

Synapomorphy Scheme ...................... ........................ 255Conclusions ...................................... 262Acknowledgments ......................................................... 264Literature Cited ...................... ........................ 264

ABSTRACT

We conclude that the internal (excurrent) nos-tril of Recent lungfishes is a true choana, asjudged by its comparison with (1) the internal nos-tril of a Devonian lungfish species which opensthrough the bony palate internal to an arcade ofmaxillary and premaxillary teeth; (2) the choanaof the Devonian ichthyostegid amphibians, and(3) nostril development in Recent urodeles. Theidea that lungfishes might therefore be the sistergroup of tetrapods is compared with the compet-ing, deeply entrenched theory that rhipidistianfishes and eusthenopterids in particular includethe ancestor of tetrapods. Our own theory, de-rived from study of Recent and fossil material,and an analysis of literature spanning 140 years,is framed in the context of a classification of themain groups of fossil and living gnathostomes:acanthodians, chondrichthyans, cladistians, ac-tinopterygians, rhipidistians, actinistians, dip-noans, and tetrapods. In formulating our proposalwe have reviewed the anatomy of the nasal cap-sule, nostrils and related structures, paired finsand their girdles, dermal bones of the skull, palateand jaw suspension, hyoid and gill arches, ribsand vertebrae, and scale and tooth structure. We

hypothesize, in agreement with most nineteenth-and many twentieth-century biologists, and in dis-agreement with the current paleontological view,that lungfishes are the sister group of tetrapods,and further that actinistians are the sister groupof those two, and that Eusthenopteron is the sistergroup of those three. We also conclude that thecharacters used formerly to link Eusthenopteronwith tetrapods either (1) are primitive for all bonyfishes (including cladistians and actinopterygians)or for living gnathostomes (including chondrich-thyans); (2) are convergent with those of severalgroups of gnathostomes; (3) only justify the inclu-sion of Eusthenopteron in a group with actinis-tians, dipnoans and tetrapods; or (4) are spurious.We attribute the century of confusion about thestructure and position of lungfishes to the tradi-tional paleontological preoccupation with thesearch for ancestors, to the interpretation of Eus-thenopteron in the light of tetrapods and the re-ciprocal interpretation of fossil amphibians in thelight of Eusthenopteron, and to the paleontologi-cal predilection for using plesiomorphous char-acters to formulate schemes of relationships.

"The convictions of our palaeontological colleagues are very real to them,and under the drive of these convictions they have quite honestly contendedfor their theories. The colour-blind man sees the scarlet robe and the greenlawn the same colour, to him they are the same colour, but he is wrong."

Kesteven, 1950, p. 99

INTRODUCTION

Miles's (1977) descriptions and illustra-tions of the head of the Devonian lungfish,Griphognathus, caused the independent re-alization in London (BGG) and New York(DER) that the internal nostrils of lungfishesare true choanae. This conviction meant tous that the only synapomorphy of Miles's(1977, p. 316) Choanata (tetrapods and rhip-idistians; cf. Gaffney, 1979b, p. 93) is alsopresent in lungfishes. It also meant to us thatAllis's (1919, 1932a, 1932b) arguments forrejecting the internal lungfish naris as achoana, based partly on the interpretation ofcertain folds of soft tissue as primary, sec-ondary, and tertiary upper lips, might befaulty. With the idea of reviewing the struc-

ture of the two outermost lip folds discussedby Allis, our attention was drawn to a largeopening in the outermost fold of Neocerat-odus near the corner of the mouth that wasfirst described by Gunther (1871, p. 515):"At the angle of the mouth, and hidden be-low a duplicature of the skin, there is anopening wide enough to admit an ordinaryquill (pl. XXX, fig. 2, a); it leads into a spa-cious cavity (b), irregular in shape, clothedwith a mucous membrane, and containingcoagulated mucous in which an immensenumber of mucous corpuscles are deposited.This cavity is separated from the cavity ofthe mouth by the membrana mucosa only,and there is no direct communication be-

163

BULLETIN AMERICAN MUSEUM OF NATURAL HISTORY

tween them; a branch cavity runs forwardinto the interior of the upper lip."The extent of this rostral structure, which

Gunther called the labial cavity, is from theanterior wall of the orbit forward to the an-terolateral part of the nasal capsule (fig. 18).It is subtriangular, narrow anteriorly near thenasal capsule and wider posteriorly in frontof the orbit where the internal wall of thecavity is thrown into a series of folds. Theopening of the labial cavity to the outside isthrough a short and broad tube that leadsdirectly to the pore in the outermost acces-sory lip (fig. 19). Visual examination of thesame region of the snout in premetamorphicNeoceratodus of about 2 cm. in total length(prior to development of pelvic fins) revealedneither a pore nor a labial cavity, but, inlarger young, serial sections show thesestructures (fig. 20). The presence in largeryoung and adults of an apparently mucous-secreting organ extending between the nasalcapsule and the orbit and its opening to theoutside on the side of the head not far infront of the eye, and its absence in premeta-morphic young, suggested to us the anatom-ical relations and ontogenetic history of thenasolacrimal organ of living amphibians.These two kinds of observations, and their

interpretations, that dipnoans may have achoana and nasolacrimal structure synapo-morphous with such features in tetrapods,caused us to look critically at the dipnoans,and also the rhipidistians, Paleozoic fishesnow universally regarded as the closest rel-atives of tetrapods. In searching for synapo-morphies that would justify the latteropinion, we have also reviewed theinterrelationships of the main groups of fossiland living gnathostomes.

Schaeffer (1965, p. 115), discussing thehistory of ideas about tetrapod origins, wrotethat Cope (1892) "broke through the 'dip-noan barrier."' By "dipnoan barrier,"Schaeffer meant the late nineteenth-centuryview that dipnoans are the closest relatives("nearest allies"-Huxley, 1876, p. 56) oftetrapods. Cope and his successors have suc-ceeded so well that our problem has been tobreak through the "rhipidistian barrier," to-day's unquestioned doctrine that rhipidis-

tians, and osteolepiforms in particular, arethe closest relatives, or direct ancestors, oftetrapods. This doctrine is so pervasive thata researcher seeking the primitive tetrapodcondition of some structure turns automati-cally to the rhipidistians, usually to Eusthe-nopteron, forgetting that Eusthenopteron hasbeen interpreted in the light of tetrapods andthat tetrapods have been interpreted in thereciprocal light of Eusthenopteron. Our as-sault on the "rhipidistian barrier" beginswith an account of the history of opinion onlungfishes, tetrapods, rhipidistians and othercrossopterygian fishes. We believe that thisaccount shows that current beliefs aboutrhipidistians have grown up among a thicketof preconceptions and wishful thinking, es-pecially the belief that processes can be dem-onstrated in the fossil record. To give oneclassic example, Watson, in his Croonianlecture on amphibian origins, wrote (1926, p.189) "It is possible to view the problem asone of purely formal morphology, the estab-lishment of a series of stages which showintermediates between typical fish structuresand those of the homologous organs in Am-phibia. ... But such studies ... are nowsatisfying to few men: the centre of interesthas passed from structure to function, and itis in the attempt to understand the processby which the animal's mechanism was soprofoundly modified ... that the attractionof the problem lies." Watson emphasizedprocess both in morphology ("establishmentof ... intermediates") and in "the attemptto understand the process" through whichmorphology is modified. Our approach issimpler. It rests on the assumption that mor-phological pattern must be recognized beforeone can speculate about process.We have accepted a number of assump-

tions or axioms. First, that tetrapods aremonophyletic. Gaffney (1979a) listed 11 syn-apomorphies which characterize living andfossil tetrapods as a monophyletic group.Four of these also occur in lungfishes (Gaff-ney's numbers 1, 8, 10, and 11: absence ofintracranial joint; pectoral girdle not joinedto skull; pelvic girdle with well-developed is-chiac ramus and pubic symphysis; well-de-veloped, ventrally directed ribs), but we re-

164 VOL. 167

ROSEN ET AL.: LUNGFISHES AND TETRAPODS

gard the remainder, especially the carpus,tarsus, and dactyly as uncontradicted evi-dence of tetrapod monophyly.

Second, we accept Gaffney's (1979a) char-acterization of Ichthyostega as the sistergroup of the remaining tetrapods (Gaffney'sNeotetrapoda). He listed five synapomor-phies which distinguish all neotetrapods1from Ichthyostega, and we accept all thosebut the last (ethmosphenoid and otico-occip-ital portions of braincase not separated bysuture), preferring to wait until the reputeddivision in the braincase of Ichthyostega isdescribed. Accepting Ichthyostega as thesister group of all other tetrapods, we turnto it in seeking primitive tetrapod conditionsin the skull roof, nostrils, etc. But there aremany structures which are still unknown inichthyostegids (e.g., manus, braincase, de-tails of vertebral column and course of sen-sory canals, all soft parts), and we have tolook elsewhere for information on primitivetetrapod conditions. At present, there is noacceptable phylogeny of tetrapods (see L0v-trup, 1977, p. 176; Gaffney, 1979a, 1979b;Carrol and Holmes, 1980). Lack of such aphylogeny leaves us three alternatives: tofollow Gaffney (1979b) in regarding lepo-spondyls as the sister group of other tetra-pods; or to survey Paleozoic "amphibians"and select what appear to be primitive con-ditions; or to select a living tetrapod group.In any case our selection could be justifiedonly by outgroup comparison, i.e., compar-ison with non-tetrapod groups, and wouldtherefore depend on a previous solution tothe problem we are trying to solve. Ofcourse, all previous attempts to solve thatproblem have depended on similar thoughrarely acknowledged circular arguments.Whichever alternative we choose we can be

' We do not, however, endorse the use of Gaffney'sterm Neotetrapoda, or any other named taxon for Re-cent organisms formulated for, or because of, the cla-distic position of a fossil; elsewhere (Patterson and Ro-sen, 1977) reasons were advanced -for believing thatallowing fossils to affect the taxonomic structure of clas-sifications of extant organisms is inevitably self-defeat-ing.

accused of bias, but since most previous dis-cussions of the problem have relied largelyon one or another Paleozoic amphibian, weshall select a living group, the urodeles. Asis well known (Kerr, 1932; Holmgren, 1933,1949a, 1949b; Kesteven, 1950; Lehman,1956; Fox, 1965), urodeles share many fea-tures of soft anatomy with lungfishes. Ourselection of urodeles as primitive tetrapodsmeans that we rate these features as synapo-morphies of choanates, symplesiomorphiesof tetrapods.

Third, rhipidistians are not a monophyleticgroup, so that no unique statements can bemade about their structure. So far as weknow, the same is true of osteolepiforms.This means that osteolepiforms have to betreated as a series of nominal taxa, amongwhich Eusthenopteron foordi Whiteaves ismost completely known, and is selected fordiscussion. The porolepiforms are united byone derived character, dendrodont toothstructure (Schultze, 1970a), known in Poro-lepis, Glyptolepis, Holoptychius, andLaccognathus. On the basis of this one syn-apomorphy, we regard the group asmonophyletic.

Fourth, Dipnoi are monophyletic. Theform of the snout and dentition (radiate tooth-plates or marginal tooth ridges-Miles, 1977,p. 288) are synapomorphies of all those Re-cent and well-preserved fossil fishes currentlyplaced in the Dipnoi. Whether other Paleo-zoic fishes, such as Powichthys (Jessen,1975), are dipnoans we cannot say.

Fifth, osteichthyans are monophyletic,and their sister group is the Chondrichthyes.Sixth, actinopterygians are monophyletic,and include the Chondrostei and Neoptery-gii; seventh, the Cladistia (=Brachiopterygii;Polypterus and Erpetoichthys) are monophy-letic; and, eighth, actinistians (coelacanths)also are monophyletic.

Finally, almost every topic discussed inthis paper is already embedded in depositsof theory and interpretation-vertebral the-ory of the skull, fin-fold theory, gill-arch con-tributions to the skull and forelimb skeleton,bone fusion versus bone loss, constancy ofbone/sensory canal relationships, and so on.In dealing with these topics, we do not claim

1651981


to have eyes unclouded by theory. But wehave tried to be guided by one principle, thatagreed interpretations of soft parts are onlyattainable among living animals, and that the

only changes in anatomical relationships thatare demonstrable are those which occur inthe ontogeny of one or another livingspecies.

HISTORICAL SURVEY

When living lungfishes were first discov-ered, lepidosirenids (Fitzinger, 1837; Natter-er, 1837) and neoceratodontids (Krefft, 1870)were both recognized as amphibians. Thepurpose of this review is to map the pathfrom those first thoughts to recent opinionssuch as that found in today's textbooks, thatlungfishes are more remote from amphibiansthan are coelacanths, or those of von Wah-lert (1968) and Bjerring (1977), that lungfish-es are more remote from amphibians than areactinopterygians, or of Jarvik (1968a), thatthey may be chondrichthyans.

Lungfishes are now universally regardedas fishes (Kesteven, 1950, was the most re-cent exception), but in the middle of thenineteenth century there was a controversyover the class of vertebrates to which theybelonged. The first detailed account of Lep-idosiren was Bischoffs (1840a, 1840b). Heagreed with Fitzinger that the animal was anamphibian, and since several zoologists dis-agreed, judging it to be a fish, Bischoff gavethe reasons for his belief in full (table 1).

L0vtrup (1977, p. 43), in his axiomatiza-tion of phylogenetics, gave as one theorem"If, in a basic [i.e., three-taxon] classifica-tion, one . .. taxon has several characters incommon with both of the other taxa, thenone of the two sets of characters will repre-sent the presence of plesiotypic charactersabsent, and/or the absence of teleotypic [i.e.,apomorphic] characters present, in the thirdtaxon." If the three taxa were Lepidosiren,a salmon and a cow, we see that in table 1the four salmon-like characters (fishlike, nos.1, 2, 4, 5) consist of three primitive charac-ters absent in cows (1, 2, 4) and one derivedcharacter (5) absent in fishes, as L0vtrup'stheorem predicts.

While Bischoffs work was in press, Owen(1839) published a preliminary account of the

African Protopterus, followed by a detailedaccount in 1841. Owen was the first (1839, p.331) to notice the resemblance between Re-cent lungfish toothplates and the MesozoicCeratodus Agassiz (1838). Owen disagreedwith Fitzinger's and Bischoffs opinion thatlungfishes were amphibians and differedfrom Bischoff on two particulars, the auricleand internal nostril, finding both the auricleand nostril to be single. He gave an interest-ing and thorough discussion of the system-atic position of Protopterus, consideringevery system of the body in a search for theessential character that distinguishes fishesfrom tetrapods. After surveying the skeletonand the respiratory, digestive, reproductive,and nervous systems Owen concluded thatnone of these is essentially different in fishesand tetrapods, and wrote (1841, p. 352) "Inthe organ of smell we have, at last, a char-acter which is absolute in reference to thedistinction of Fishes from Reptiles. In everyFish it is a short sac communicating with theexternal surface; in every Reptile it is a canalwith both an external and an internal open-ing. According to this test, the Lepidosirenis a Fish: by its nose it is known not to be aReptile . . . so that at the close of our anal-ysis we arrive at this very unexpected result,that a Reptile is not characterized by itslungs, nor a Fish by its gills, but that the onlyunexceptionable distinction is afforded bythe organ of smell."Having seen Owen's 1839 paper, Bischoff

added a supplement to his own, commentingon the differences between his own obser-vations and Owen's. In particular he reiter-ated his opinion on the divided auricle andthe internal nostril. Milne Edwards (1840)added some remarks to the French transla-tion of Bischoff's paper, pointing out thatdissection of a Paris specimen ofProtopterus

166 VOL. 167


TABLE ICharacters of Lepidosiren paradoxa(From Bischoff, 1840b, pp. 144-151)

Fishlike Amphibian-like

1. Scales 1. Internal nostril2. Sensory canals 2. Large paired lungs, connected with the oesophagus3. No ossified vertebrae, and notochord by a large glottis

penetrating base of cranium 3. Gills much reduced, most blood bypasses them4. Opercular bones 4. Heart with two auricles, one receiving blood5. No external ear from lungs6. Labial and nasal cartilages 5. Conus with two valves which separate the pulmonary

and branchial blood

confirmed Bischoffs observation of a divid-ed auricle and internal nostrils.

In this first exchange in the controversy,we see Bischoff correctly interpreting syn-apomorphies, and Owen basing his opinionpartly on a mistaken observation, and partlyon primitive characters. Owen never (e.g.,1846, p. 256; 1866, p. 474) agreed that theauricle was divided in lungfishes. He later(e.g., 1846, p. 201) allowed that there weretwo nostrils, but argued that neither openedinto the mouth. This first exchange set thepattern for the next 20 years. Those workerswho believed lungfishes to be amphibians(Milne Edwards, 1840; Hogg, 1841; Heckel,1845; Melville, 1848, 1860; Vogt, 1845, 1851;Asmuss, 1856; Gray, 1856) relied on thesame characters as Bischoff, and includedthose who had observed living animals(Gray, Melville). Those who put lungfishesamong fishes (Muller, 1840, 1844; Oken,1841; Fitzinger, 1843-a change of mind;Agassiz, 1843; Hyrtl, 1845; Newman, 1857;and others-see Dumeril, 1870) were, whenthey gave reasons, weighting primitive char-acters.A third viewpoint was possible, that it was

futile to search for an essential distinctionbetween fishes and amphibians. Heckel(1851) took this view, as did M'Donnell(1860, p. 18) in these words: "our presentknowledge . . . leads us to regard it[Protopterus] as a form so transitional as tosupport the view taken by some systematiz-ing zoologists in later days, that no distinctlydefined line of demarcation exists between

Fishes and Reptiles . .. an order uniting fish-es with the amphibian reptiles, seems to bethe true mode of giving it its proper place."By "systematizing zoologists" M'Donnellmight have been referring to Owen (1859),who named a group Haemacrymes for cold-blooded vertebrates (equivalent to his 1866Haematocrya) and wrote that within thisgroup "Archegosaurus conducts the marchof development from the fish-proper to theLabyrinthodont type: the Lepidosiren con-ducts it to the perennibranchiate batrachiantype." Or M'Donnell might have been refer-ring to Darwin and Wallace. In either case,such views as his were readily adapted to theevolutionary doctrine, and in Haeckel's firsttrees (1866) the Dipnoi are placed as the sis-ter group of the tetrapods, with a line of"Dipnoi ignoti" extending back into the De-vonian. In Haeckel's later versions (e.g., fig.1) lungfishes are placed as the direct ances-tors of tetrapods. Thus with acceptance ofevolution, the problem of distinguishing tet-rapods from fishes, i.e., the problem of char-acterizing tetrapods, disappeared, since dif-ficulties in distinguishing the two groupswere agreeable to evolutionists.

In 1870, Krefft announced the discoveryof Neoceratodus, "a gigantic Amphibian."As with the lepidosirenids a third of a cen-tury before, such opinions found no supportin the British Museum, where Gunther wrotethe first detailed account of Neoceratodus(1871). Before reviewing Gunther's conclu-sions, it is worth recalling that he was, likemost of the senior naturalists in the British

1981 167


Semaec

Fulcrati

EfU1CriRhombifei

Plectognathi

Lophobran

StichobranchiiPhysoclisti

Enchelygenes

ThrissogeiiesPhysostomiTeleostel

)pteri

Thrissopides

-'1 Amiadesri

Leptolepides

Sturiones

Cephalaspides

PlacodermiTabuliferlGan4

AnuraPeromela

SalamandrinaclhiiLabyrinthodonta Cryptobranchia

! PerennibranchiaStegocephlia Lissamphibla

Phractamphibia

Amphibia

Dipneumones

Monopneumones

Ctenodipterini

Dipneusta

Crossopterygii (?)

Cycliferi

oides Rajacei

Squalacei Chimaeracei

HolocephaliPlaglostomi

SelachiiPisces

Amphirhina CycloSmono

. I1

istomarhina

Craniota

AcraniaFIG. 1. Phylogeny of lower vertebrates, from Haeckel (1889, p. 613).

168 VOL. 167


Museum at that time, an opponent of evo-lution. Gunther's most recent biographer re-marks that in the memoir on Neoceratodushe "had taken the opportunity to trail hiscoat before the evolutionists" (A. E. Gun-ther, 1975, p. 335).Gunther (1871) found that Neoceratodus

was less amphibian-like than the lepidosiren-ids, since it had ganoid valves in the conus,a single lung, well-developed gills, and so on.Concerning the nostrils of lungfishes, itwould be hard to find a better illustration ofthe idea that all observation is theory-ladenthan the contrast between Gunther's view(1871, p. 517) that both the anterior and pos-terior nostrils open inside the mouth, andOwen's (1866, p. 329) that neither openswithin the mouth. Gunther concluded thatlungfishes were ganoid fishes, and named anew subclass Palaeichthyes to include allfishes except teleosts. Gunther was the firstto realize that Paleozoic fishes such as Dip-terus and Phaneropleuron, included by Hux-ley (1861) in his Crossopterygii, were lung-fishes. Although his discovery carried thedipnoans back into the Devonian, as Haeck-el's trees demanded, Gunther saw this not asevidence of evolution, but as "a proof of theincompleteness of the palaeontological rec-ord" and as "the most remarkable exampleof persistence of organization, not in Fishesonly, but in Vertebrates."Huxley, through whose Crossopterygii

Gunther had trailed his coat, replied in 1876.He distinguished autostylic, hyostylic, andamphistylic skulls, showed that lungfisheshad the same autostyly as amphibians, anddiscussed the paired fins of Neoceratodus asforerunners of tetrapod limbs. On classifi-cation, he argued against Gunther's Palae-ichthyes, and for Muller's original (1844)groups (Dipnoi, Ganoidei, etc.). Huxleymaintained that Dipnoi are the "nearest al-lies of the Amphibia" (p. 56), whereas in1861 (p. 27) he had argued that they are"next of kin" to the crossopterygian gan-oids. This ambiguity was not cleared up inHuxley's next contribution (1880). There hewrote of Neoceratodus (p. 660), "This won-derful creature seems contrived for the illus-tration of the doctrine of Evolution. Equally

good arguments might be adduced for theassertion that it is an amphibian or a fish, orboth, or neither." In this paper, Huxley pro-posed the names Chondrichthyes and Oste-ichthyes, but included in the latter only theganoids and teleosts, placing dipnoans in agroup Herpetichthyes named "for that par-ticular stage of vertebrate evolution of whichboth the typical Fishes and the typical Am-phibia are special modifications." Huxley'sdiagram seems to imply that he regarded thecrossopterygians as descendants of dipnoan-like fishes. Thus, in contrast to Gunther,Huxley sought to blend lungfishes with bothcrossopterygians and tetrapods.

Similar uncertainties about the status, andcontent, of the Crossopterygii are evident inthe publications of other workers during thisperiod (see Schaeffer's 1965 review). For ex-ample, Gill (1872; pp. xxxii, xliii) had thecrossopterygians as the sister group of coela-canths in one of his genealogical trees, andas the sister group of actinopterygians, witha question mark, in the other (our fig. 2);Haeckel's question mark against the cros-sopterygians in figure 1 is also symptom-atic. Cope (1872, 1885, 1887, 1892) vacillatedon the content of the Crossopterygii [Wood-ward (1891, p. xxii), whose Catalogue sta-bilized things, complained of "the somewhatfluctuating classifications of Cope"]. In 1887Cope removed the Rhipidistia (Tristichopter-idae) and Actinistia (Coelacanthidae) fromthe Crossopterygii, leaving in the latter onlythe Holoptychiidae, Osteolepididae, Polyp-teridae, and Phaneropleuron (a dipnoan thathe had formerly excluded). By 1892 the hol-optychiids and osteolepids had also joinedthe rhipidistians and actinistians, leaving thepolypterids as the only crossopterygians.A passage from Cope's 1892 paper is cited

by Schaeffer (1965, p. 115) as the source of"our present understanding of amphibian or-igin." Like Gill, Huxley, Haeckel, and otherevolutionists of this period, Cope had pre-viously believed that tetrapods were de-scended from dipnoans (e.g., Cope, 1884, p.1256, "The Batrachia have originated fromthe sub-class of fishes, the Dipnoi, thoughnot from any known form"). But in 1892 heemphasized the lack of the maxillary arch in

1691981


14 fl--ELEOST SERIES (I.-EI.).p

0 >

-

0 L CYCLOGANOIDZA (XIII.).x Om r R}HOMBOGANOIDEA (XII.).

-? CROSSOPTERYGII (XI.). rAx

RHOLOCEPHALI (VI.).

-RAIAR (V.).

SQUALI (IV.).

W 'HYPXROTB:TA (IL).

-P4

-LEPTOARDII.-cxuaHosTom (I.).

FIG. 2. A genealogical tree of the relations of major groups of vertebrates, from Gill (1872, p. xliii).

Dipnoi, their large dorsal and anal fins, andthe fact that autostyly developed during on-togeny in amphibians, and concluded thatamphibians therefore originated from a hyo-stylic fish with a complete maxillary arch,feeble median fins, and amphibian-likepaired fins. The rhipidistians, and Eusthe-nopteron in particular, were offered as suchfishes.

Pollard's (1891, 1892) papers on Polypt-erus are a forthright attack on dipnoan-tetra-pod relationships. He argued that Polypterusresembles the urodeles in features of the in-ner ear, braincase, palate, musculature,nerves and ribs, and concluded that "the an-cestry of the Urodela must be sought amongthe Crossopterygian forms now representedonly by Polypterus and Calamoichthys"

(1891, p. 341). Pollard also asserted that theposterior nares of dipnoans are not homolo-gous with the choanae of urodeles. Kingsley(1892) argued in the same vein, that sinceautostyly developed during ontogeny in uro-deles, they could not be descended from dip-noans. Baur (1896) reached similar conclu-sions from a study of labyrinthodonts, andwas the first to suggest (p. 670) that the re-semblances between dipnoans and amphibi-ans are merely parallelisms.

Dollo (1896) produced the first detailedphylogeny of dipnoans, and summarized ear-lier work on the interrelationships of dip-noans, crossopterygians, and tetrapods. Heexpressed his conclusions in a diagram (fig.3) which places the dipnoans as the sistergroup of tetrapods, and crossopterygians as

VOL. 167170

1981 ROSEN ET AL.: LUNGFISHES AND TETRAPODS 171

MAMMIFtRES. OISEAUX.PHYSOCLYSTES.

REPTILES.PHYSOSTOXES.

BATRACIENS. DIPNEUSTES.

A mioldes.

Vie terrestre. Vie en eau corrompue,

puis dans la vase. Lepidost6oldes.

Crossopt&ygiens. Acipensboldes.

Continuation de la vie Retour progressif, eten eau douce. de plus en plus

complet, a as mer.HOLOCfrHALIS. SLAC1ENS.

ICHTHYOTOINS.

GANOIDES.

PLEU1OPTi1YGIES.

AcAmoDIsES. ? OSTRACODMES.

Chondropt6rygiens. Ost6opt6rygiew.

En geniral, sejour ininterrompu Adaptation generaledans la mer. a l'eau douce.

Gnathostomes.

Souche marine.

POISSONS.

FIG. 3. Phylogeny of gnathostome vertebrates, from Dollo (1896, p. 113).


the group ancestral to both. This was thegenerally accepted scheme for the next 40years, though with many variants. Thus, bythe end of the nineteenth century, the dip-noan ancestry of tetrapods favored by earlyevolutionists (fig. 1) had been replaced by acrossopterygian ancestry (fig. 3). In cladisticterms, figure 3 implies that dipnoans are thesister group of tetrapods (i.e., those twogroups share synapomorphies), whereas thecrossopterygians are a paraphyletic group,some members of which share synapomor-phies with dipnoans and tetrapods, butwhose relationships are yet to be workedout. Of course, we find nothing to quarrelwith in that statement. It is worth summariz-ing the evidence so far produced to cause theconceptual change from figure 1 to figure 3:the following characters were used by latenineteenth-century biologists to relate tetra-pods, dipnoans, and crossopterygians in theway shown in figure 3:

I. Characters excluding dipnoans from tetrapodancestry1. Dentition too specialized (Boas, 1882, p.

557: cf. Miles, 1977, fig. 157-the toothplates are a synapomorphy of higher dip-noans)

2. No maxillary arch (Cope, 1892: for ourview see p. 180)

3. Autostyly develops during ontogeny inamphibians (Cope, 1892; Kingsley, 1892)

4. Ribs not homologous in dipnoans and tet-rapods (Pollard, 1892; Baur, 1896: for ourview see p. 242)

II. Characters relating crossopterygians to tetra-pods

1. Large maxilla, dentition not modified(Cope, 1892; Pollard, 1892: cf. actinop-terygians-a synapomorphy of osteich-thyans?)

2. Hyostyly (Cope, 1892: cf. sharks-a syn-apomorphy of gnathostomes?)

3. Most of the bones can be homologized inPolypterus, fossil crossopterygians andlabyrinthodonts (Pollard, 1892; Baur,1896; Woodward, 1898, p. 123-for ourview see p. 221, 227)

4. Many sclerotic plates in fossil crossopte-rygians and tetrapods (Woodward, 1898,p. 123-a synapomorphy of sarcopteryg-ians: cf. Miles, 1977, p. 249)

5. Pineal foramen present in fossil crossop-

terygians and labyrinthodonts (Wood-ward, 1898, p. 123-a synapomorphy ofbony vertebrates)

6. Paired fins of Eusthenopteron more liketetrapod limbs than are those of dipnoans(Cope, 1892; for our view see p. 209)

7. Polypterus has dorsal ribs (Pollard, 1892;Baur, 1896: for our view see p. 242)

8. Fossil crossopterygians and labyrintho-donts have an infradentary (splenial) inthe mandible (Woodward, 1898, p. 123-so have all non-actinopterygianosteichthyans)

9. The median fins are reduced in Paleozoiccrossopterygians (Cope, 1892; Baur,18%: no comment necessary)

10. Fossil crossopterygians and labyrintho-donts have folded teeth

As is clear from the above list, the char-acters so far offered in support of the rela-tionship between crossopterygians and tet-rapods were a melange of primitivecharacters, characters of Polypterus alone,and vague assertions about Paleozoic cros-sopterygians. Concerning fin structure,Goodrich's (1901) account of the pelvic finand girdle ends with the statements that Po-lypterus "probably belongs to the actinop-terygian line" and that Eusthenopteron is"very far removed from Polypterus."

Nevertheless, the belief that the Crossop-terygii, containing polypterids and rhipidis-tians, was a natural group, and that this groupgave rise to the tetrapods, continued to gainstrength. Since Polypterus lacks many of thederived features in which dipnoans agreewith tetrapods, it followed that these dip-noan/tetrapod similarities must have arisenin parallel, as Baur (1896) had realized. Thiswas acceptable to some (e.g., Boulenger,1901, p. 30; Bridge, 1904, p. 520), but not toothers (e.g., Semon, 1901, who listed morethan 20 resemblances between dipnoans andamphibians; Jordan, 1905, p. 600; Goodrich,1909, p. 230). Goodrich (1909) believed thatdipnoans "present many striking points ofresemblance to the Amphibia, which cannotall be put down to convergence." In his phy-logenetic diagram (p. 29) Goodrich revertedto an unfashionable view in placing dipnoansas the sister group of teleostome fishes, withthe tetrapods entered, with a query, as a tri-

172 VOL. 167


,COELACANTHS

RHIPIDISTIANS

BDIPNOANS

TETRAPODS

C O E LAC A N T H S

R H I P I D I S T I A N S

DIPNOANS

TETRAPODS

,DIPNOANS

COELACANTHS

D,

RH I PI DI ST IAN S

TETRAPODS

COELACANTHS

DIPNOANS

, RHIPIDISTIANS

' TETRAPODS

FIG. 4. Cladograms to show evolution of opinion on interrelationships of crossopterygians, lung-fishes, and tetrapods. A, Goodrich (1924); B, Watson (1926), Romer (1933); C, Save-Soderbergh (1935),Westoll (1961, etc.), Miles (1975); D, Miles (1977), Wiley (1979).

chotomy with those two groups. In that dia-gram, polypterids are also marked by aquery, since Goodrich's assessment was thatthey are closer to actinopterygians than tocoelacanths and rhipidistians. Goodrich'sopinions were clearer by 1924. In that paper,he still insisted that the resemblances betweenlungfishes and tetrapods "can scarcely bedue to convergence" and wrote "the earliestOsteichthyes diverged into teleostome anddipnoan branches and that the tetrapodsarose from the base of the latter" (cf. fig.4A). Goodrich is often cited as one of thearchitects of the theory of the crossopteryg-ian ancestry of tetrapods (e.g., Romer, 1956,p. 159; Schaeffer, 1965), but in our readinghis work shows nothing of the sort.

After Goodrich's 1909 work, Polypterusplayed a less central role in discussions oftetrapod origins. For example, Watson(1912) used "'crossopterygian" as synony-

mous with "rhipidistian," in a paper includ-ing new information on Megalichthys, andthe observation that the mode of tooth re-placement was identical in rhipidistians andlabyrinthodonts.Gregory (1915) published the first detailed

review of the question of tetrapod origins.On dipnoans, Gregory believed that whiletheir similarity with tetrapods, and urodelesin particular, "might be ascribed to conver-gence" they implied "a similarity in the 'po-tential of evolution,' that is, of structuralpossibilities, in the forerunners of thesegroups" (p. 322). Gregory argued that "Theonly crossopterygians that can claim evenremote relationships with the Amphibia arethe Devonian Rhipidistia, especially the Os-teolepidae and the nearly allied Rhizodonti-dae" (p. 326). In his review he concentratedon seeking homologies in the dermal bonesof the skull and in the forelimb skeleton be-

A

CD

1731981


tween rhipidistians and labyrinthodonts.Concerning the forelimb, Gregory tried tosettle the much-discussed question of the rel-ative status of the fin of Neoceratodus andof rhipidistians. Unlike many of his prede-cessors, Gregory rejected "the traditionalview that the 'archipterygia' of Dipnoi areprimitive structures and [regarded] the im-perfect archipterygium of the Devonian Os-teolepis as more primitive" (p. 350). ThusGregory imagined that tetrapods are de-scended from rhipidistians, and argued for a"common origin of the Dipnoi and Crossop-terygii" (p. 325).Gregory did not describe exactly how he

envisaged the interrelationships of rhipidis-tians and dipnoans, but this was made clearby Watson (1926), the next major reviewerof tetrapod origins. He wrote (p. 195) "themost primitive known Dipnoan is ... Dip-terus valenciennesi . . . this-fish presents somany resemblances in structure to the con-temporaneous Osteolepids as to show thatthe two groups arose from a common ances-tor not much earlier in date. It is from thishypothetical fish that I believe the Amphibiato have arisen." Thus Watson envisages atrichotomy of dipnoans, rhipidistians, andtetrapods (fig. 4B), this arrangement, ratherthan derivation of dipnoans and tetrapodsfrom rhipidistians (fig. 3), being necessitatedby the intracranial joint of rhipidistians (cf.Watson and Gill, 1923, p. 210), which Wat-son regarded as a specialized feature. On theresemblances between dipnoans and uro-deles, Watson wrote (1926, p. 190) "it is cer-tain that the existing resemblances dependon the parallel evolution of not distantly re-lated stocks." Romer (1933) followed Wat-son in adopting a trichotomous diagram ofthe relationships between dipnoans, tetra-pods and rhipidistians (his fig. 51), but Ro-mer preferred to regard the similarities be-tween dipnoans and amphibians as "retentionof characters present in their crossopteryg-ian ancestors." Given the trichotomous pat-tern of relationships inferred by Watson andRomer (fig. 4B), it should be pointed out thatspecial similarities between rhipidistians andtetrapods are open to just the same interpre-tations as those applied by Watson and Ro-

mer to dipnoan/tetrapod similarities: theymay be parallelisms or retained primitivecharacters.

In 1932 Save-Soderbergh announced thediscovery of the Devonian ichthyostegids.These animals have had a profound effect onsubsequent discussions of tetrapod origins,because they are both stratigraphically olderand anatomically more primitive than theCarboniferous amphibians that Watson hadused to demonstrate similarities between os-teolepids and early tetrapods. As Parrington(1967, p. 231) put it, "It was to be expected,therefore, that any Amphibia from the UpperDevonian would be intermediate in theirstructures between the Middle Devonian os-teolepids and the Carboniferous labyrintho-donts, but when discovered the ichthyo-stegids did not conform at all well to thisexpectation. While their skulls showed somevery primitive features which might havebeen expected, the pattern of the dermalbones did not conform to plan." In attempt-ing to explain the differences in the skull roofbetween osteolepids and ichthyostegids,Save-Soderbergh adopted an explanation in-volving differential fusions from an ancestralpattern which, as he said (1934, p. 5), is"astonishingly" dipnoan-like. After the dis-covery of the ichthyostegids, discussions ofthe dermal skull roof shifted away from thesearch for homologies between rhipidistiansand tetrapods, toward discussions of the fu-sion theory and of Westoll's (1938) alterna-tive explanation requiring changes in pro-portion (see Jarvik, 1967; Schmalhausen,1968; and section on Dermal Bones).Of his 1932 paper, Save-Soderbergh wrote

(1935, p. 201) "the investigations of the De-vonian Ichthyostegids ... show that these. . . are more nearly related to the Crossop-terygians [i.e., rhipidistians and coelacanths]than to the Dipnoans." Thus he had adoptedthe scheme of relationships shown in figure4C, though it is not clear to us what char-acters of ichthyostegids or crossopterygiansprompted this conclusion. In any case, thisopinion was short-lived, for in 1933, afterelaborating on the characters common todipnoans, crossopterygians, and tetrapodsand naming the group Choanata for them,

VOL. 167174


Save-S6derbergh suggested (p. 118) that thetetrapods were polyphyletic, urodeles beingrelated to dipnoans, and the labyrinthodontspossibly being polyphyletic. The idea of tet-rapod polyphyly, and of direct dipnoan/uro-dele relationships, was taken up by Holm-gren (1933; see also 1939, 1949a, 1949b), atfirst on the structure of the limbs. He con-cluded (1933, p. 288) "regarding the limbsthere is a greater agreement between a dip-noan and an urodele, than between an uro-dele and an anuran."Save-Soderbergh adopted Holmgren's

views on urodele/dipnoan relationships (1934,1935) and so began the central theme of mod-em discussions of tetrapod origins, the dis-pute between the advocates of polyphyly,and the advocates of monophyly. Holm-gren's original advocacy of urodele/dipnoanrelationships can be seen as an attempt toreconcile the conflicting claims of neontolo-gy (resemblances between Recent dipnoansand amphibians, especially urodeles) and pa-leontology (resemblances between labyrin-thodonts and rhipidistians). However, Holm-gren's views received little support (seereferences cited in Holmgren, 1949a), andthe emphasis soon passed back to paleontol-ogy, with Jarvik's (1942) monumental workon the snout. In that monograph, Jarvik gaveincomparable accounts of the morphology ofthe snout in Eusthenopteron and porolepi-forms. Like Holmgren and Siive-Soder-bergh, he advocated a diphyletic origin oftetrapods, but as urodele ancestors he re-placed the dipnoans by the porolepiforms.Jarvik's principal reason for excluding thedipnoans from consideration was his argu-ment that they have no choanae (see sectionon Nostrils).

Westoll's 1943 review of tetrapod origins,though published after Jarvik's monograph,was written without knowledge of it, becauseof wartime conditions. Westoll's paper is inthe tradition of Watson's (1926) review. LikeWatson, Westoll emphasized process, bothin his discussion of the significance of thechange from fish to tetrapod, and in his mor-phology, where sequences of fossils are se-lected to illustrate trends. Westoll's generalconclusion was that tetrapods are monophy-

letic, and "that all known tetrapods can bederived from a single basic structural pat-tern, which resembles osteolepid crossopte-rygians in so many details that this group isthe only possible source of proto-tetrapods"(p. 79). This conclusion is not surprising, forWestoll's discussion was much influenced byhis (1938) discovery of Elpistostege, "an im-portant proto-tetrapod" (Westoll, 1943, p.78) now interpreted as a panderichthyid fish(Vorobyeva, 1977a), or, in Westoll's classi-fication, an osteolepid. Later, Westoll wrote(1958b, p. 100; 1961, p. 605) of "overwhelm-ing evidence" of the descent of tetrapodsfrom osteolepid rhipidistians. On lungfishes,Westoll (1949; Lehmann and Westoll, 1952)did not accept Jarvik's argument that theposterior nostrils of dipnoans are not choa-nae, and maintained that dipnoans, like tet-rapods, are descended from osteolepids.Westoll's scheme of relationships (e.g.,1961, fig. 1) is like that introduced by Siive-Soderbergh (fig. 4C), with coelacanths closerto tetrapods than are dipnoans, but the coel-acanths are regarded as descendants of po-rolepiforms, and, following Jarvik, the pos-sibility of urodele descent from porolepiformsis allowed. Romer's views (e.g., 1968) andhis trees (e.g., 1956, fig. 19) are similar toWestoll's; like Westoll he did not accept Jar-vik's argument that dipnoans have no choa-nae, but unlike Westoll he never allowed thepossibility that tetrapods are non-monophy-letic.On the relationships of dipnoans, crossop-

terygians, and tetrapods most authoritieshave followed the pattern shown in figure 4C(e.g., Trewavas et al., 1955-"the closer re-lationship of [tetrapods to] Osteolepidoti andCoelacanthini than to the Dipnoi"), and sotoday's textbook view became established.As for the dipnoans, they were free to

wander farther down the family tree, sincecoelacanths, now known (Latimeria) to lackchoanae, and judged to show nothing in com-mon with amphibians, were interposed be-tween lungfishes and tetrapods. White (1965)argued that dipnoans were distinct from oth-er osteichthyans; Schaeffer (1968) that theyformed a trichotomy with actinopts and cros-sopts; Bertmar (1966) that they are related to

1751981


FIG. 5. Phylogeny of gnathostomes, modified from Jarvik (1960, fig. 28).

actinopts; von Wahlert (1968) that they arethe sister group of all osteichthyans; and Jar-vik (1968a, 1968b), citing resemblances be-tween dipnoans and holocephalans, con-cluded (1968b, p. 518) "if we ask which ofthe recent vertebrate groups is the sistergroup of dipnoans (holocephalians? coel-acanthiforms?) no answer can be given."Thus the fossils had taken over. Dipnoans,placed as amphibians or the closest relativesof tetrapods by nineteenth-century neontol-ogists, had been ousted by the rhipidistians,and relegated to a limbo between osteichthy-

ans and chondrichthyans (cf. fig. 5). Despitea few dissenting neontologists (e.g., Fox,1965), it is now accepted that the resem-blances between dipnoans and amphibiansare convergent, and that the resemblancesbetween rhipidistians, especially osteolepi-forms, and tetrapods are evidence of rela-tionship (i.e., synapomorphies). What, then,are these resemblances? Above, those usedby nineteenth-century biologists are listed;the following list contains characters empha-sized by twentieth-century workers to relateosteolepiform rhipidistians and tetrapods,

VOL. 167176


additional to or amplifying those of nine-teenth-century biologists:

1. Detailed correspondence between bones ofskull roof and cheek (e.g., Westoll, 1943; cf.Jarvik, 1967; Parrington, 1967)

2. Intracranial joint is immobilized in some os-teolepiforms (e.g., Westoll, 1943; Vorobyeva,1977a)

3. Dermal bones of palate "practically identi-cal" in osteolepids and primitive tetrapods(Westoll, 1943, p. 84)

4. Choana present, with the same relationshipto the dermal bones in both groups (Jarvik,1942; Westoll, 1943)

5. Mandible "almost identical" in early tetra-pods and osteolepids (Westoll, 1943, p. 85)

6. Tetrapod septomaxilla is found in osteolepi-forms (Panchen, 1967)

7. Opercular and preopercular present in ichthy-ostegids (Jarvik, 1952)

8. Suture in braincase of ichthyostegids repre-sents intracranial joint (Jarvik, 1952)

9. Notochord penetrates base of braincase inichthyostegids (Jarvik, 1952)

10. Cranial cavity and its inferred contents simi-lar in Ectosteorachis and amphibians (Romer,1937)

11. Osteolepiforms have ethmoid, basal, ascend-ing and otic processes on the palatoquadrate(Jarvik, 1954)

12. Hyomandibular double-headed, like tetrapodstapes, in osteolepiforms (Jarvik, 1954)

13. Distribution of dentition, and mode of re-placement of fangs, similar in the two groups(Westoll, 1943; Jarvik, 1954, 1963)

14. Osteolepiforms have folded teeth of the sametype as labyrinthodonts (Schultze, 1970a)

15. Vertebrae of osteolepiforms are "protora-chitomous" (Andrews and Westoll, 1970a,1970b)

16. Eusthenopteron has bicipital dorsal ribs (An-drews and Westoll, 1970a)

17. Scapulocoracoid of Eusthenopteron like thatof early tetrapods, with a screw-shaped glen-oid (Andrews and Westoll, 1970a)

18. Humerus of Eusthenopteron and Strepsoduscomparable with tetrapod humerus (Andrewsand Westoll, 1970a, 1970b)

19. Eusthenopteron may have a sacral attach-ment (Andrews and Westoll, 1970a)

20. Ichthyostega has a fishlike tail (Jarvik, 1952)

This list may seem short to those who re-call the extraordinarily detailed comparisonswith tetrapods in Jarvik's papers (1942, 1952,

1954, 1963, 1964, 1966, 1967, 1972, 1975) onosteolepiforms. But it must be emphasizedthat Jarvik never pretended to cite ch'arac-ters common to osteolepiforms and tetra-pods; instead, he cited characters commonto Eusthenopteron and frogs. Our main taskin this paper is to review the characters listedabove, in osteolepiforms, tetrapods, and dip-noans. Some of them (nos. 7, 9, 20) may beimmediately discarded as primitive for os-teichthyans or gnathostomes. Others may bediscarded because of lack of published evi-dence (nos. 8, 19), or because dipnoans areknown to share the same features (nos. 2,10, 12; cf. Miles, 1977). The remainder arediscussed in subsequent sections.

During the last few years, there has beena changed approach to dipnoan relation-ships, stemming mainly from cladism. Miles(1975) reviewed the question and concludedthat the textbook solution (fig. 4C) was cor-rect, but later (1977) changed his mind, afterstudying the Devonian Gogo lungfishes, andadopted a new scheme (fig. 4D). Meanwhile,others had suggested that coelacanths be-longed elsewhere (L0vtrup, 1977-sistergroup of chondrichthyans; Wiley, 1979-sis-ter group of osteichthyans) on evidence fromthe anatomy, physiology, and biochemistryof Latimeria. In other words, further inves-tigation of Latimeria repeatedly turns upprimitive osteichthyan or chondrichthyan-like characters. Further investigation of Re-cent lungfishes, on the other hand, repeat-edly produces characters which seem to besynapomorphous with tetrapods; e.g., lensproteins and bile salts (L0vtrup, 1977), cil-iation of the larva (Whiting and Bone, 1980)and gill-arch muscles (Wiley, 1979). It is un-fortunate that, so far as we know, no pro-teins of lungfishes or Latimeria have yetbeen sequenced. In any case, the number ofapparent synapomorphies between lungfish-es and tetrapods continues to increase. Pro-ponents of the rhipidistian ancestry of tet-rapods have continued to rate such charactersas convergent, following Baur (1896). Alter-natives are to rate these characters as syn-apomorphies, either of tetrapods and dip-noans, or of some more extensive group.Since most of these characters are in soft

1981 177


anatomy and embryology, they can, whenassumed to have occurred in rhipidistians,be rated as synapomorphies of a group com-prising tetrapods, lungfishes, and rhipidis-tians. As a paraphyletic group, rhipidistiansmight then include members of the stem-groups of tetrapods, of dipnoans, and of tet-rapods plus dipnoans. That statement recallsour comment on Dollo's (1896) diagram (p.170), implying that remarkably little progresshas been made during the last eighty years.

In our view, the reason for that lack ofprogress is plain enough. It is acceptance ofevolution, which has directed attention awayfrom the unsolved problem that occupiedpre-Darwinians-what are the characters oftetrapods, how are they distinguished fromfishes, and from lungfishes in particular-andtoward a search for sequences of fossilswhich will fit the evolutionary doctrine, andtoward an interest in process rather than pat-tern. The search for fossils has produced su-perficially acceptable sequences, as it wasbound to, for few transformations, howeverfantastic, are forbidden by the Darwinian or

Neo-Darwinian picture of the evolutionaryprocess. Yet the sequences consist of noth-ing more than abstractions from paraphyleticgroups such as rhipidistians, osteolepiforms,and labyrinthodonts. As for the pattern, veryfew plausible derived characters have yetbeen presented to link some rhipidistians andsome tetrapods (e.g., folded teeth). For therest, the characters used by twentieth-cen-tury biologists seem to us little different fromthose of nineteenth-century workers-a me-lange of primitive characters, and assertionsof identity or general similarity whose pre-cise content is still to be established.Acceptance of cladistic methodology dur-

ing the last few years has begun to directattention back towards pre-Darwinian prob-lems, problems of pattern. How are choa-nates and tetrapods characterized (e.g.,Miles, 1977; Szarski, 1977; Schultze, 1977a;Gaffney, 1979a, 1979b)? With the realizationthat some generally accepted groups (rhipi-distians, osteolepiforms, labyrinthodonts)are uncharacterized, or uncharacterizable, theway is open for a new attack on the problem.

CHOANA, NOSTRILS, AND SNOUT

(A) INITIAL COMPARISONS ANDINFERENCES

Our interpretation of the dipnoan internalnostril as a choana (fig. 6) has already beenalluded to in the Introduction. This interpre-tation has resulted from the rejection of acomparative anatomical argument by Allis(1919, 1932a, 1932b) concerning living lung-fishes from which he deduced that the trueupper lip and excurrent, external nostril hadmigrated together into the oral cavity. Allis'sreasoning predicts that if marginal upper jawteeth were present in lungfishes, these teeth(which are absent in extant and most fossilspecies) would lie medial to the internal nos-tril. The prediction, however, is inconsistentwith Miles's (1977) account of several moreor less complete skulls of the Devonian lung-fish, Griphognathus whitei Miles, in whichthe internal nostril is clearly marked and in

which marginal jaw teeth are present lateral,rather than medial, to this opening (fig. 7).

In vertebrate systematics identification ofa choana must satisfy not only anatomicalcriteria, but other criteria of homology aswell. To do this, the lungfish choana eithermust be indistinguishable from the choana oftetrapods, or one must be regarded as a mod-ification of the other. Primary comparisonsare, therefore, made between Miles's (1977)excellent specimens of Griphognathuswhitei, as interpreted partly on the basisof our own dissections of Neoceratodus for-steri and ichthyostegid amphibians, whichwe consider to be the sister group of all othertetrapods (see Introduction). Our commentson ichthyostegids are based on Ichthyostegasp. as described by Save-Soderbergh (1932)and Jarvik (1952) and on two casts of well-preserved palates (fig. 10).

178 VOL. 167


TONGU E

ANTERIOR NARIS

CHOANA

.4

FIG. 6. Neoceratodus forsteri (Krefft), approximihead and bases of pectoral appendages.

In Griphognathus the preoral region of thesnout medial to a pair of nasal recesses bearsa pair of tooth ridges fused with the rostrumand covered with enamel. Miles (1977, p.187) refers to these tooth ridges as premax-illae because, when the lowerjaw is in place,the mandibular symphysis closes just behind

ately 120 cm. total length, AMNH 40800, showing

them. Although Miles seemed to have somereservations that these elements are eithertooth plates or premaxillae, for he identifiedthem in his figures only as preoral emi-nences, their interpretation as dentigerouson the one hand and as paired elements onthe other is supported by the presence of a

1981 179


ANTERIOR PALATAL RECESS

PREMAXILLARY TEETH

VOMER

"EXTRA"

PALATIN E

SUBNASAL RIDGE

POSTERIOR LIMIT-MAXILLARY TEETH

CHOANA-

A

C

5mm

FIG. 7. Griphognathus whitei Miles. A, ventral view of skull, from BMNH P. 56054, after Miles(1977, fig. 6); B, camera lucida drawing of left choana and surrounding dermal bones in BMNH P.56054; arrow indicates a crack not a suture; C, camera lucida drawing of left choana in BMNH P. 50996(holotype), an individual in which the bones surrounding the choana are fused.

similar but more distinct pair of toothed ele-ments medial to the nasal recesses in Gan-orhynchus woodwardi (fig. 8). Our conclu-sion that they represent premaxillae that

have fused with the bone of the snout isbased on the observations that they arepaired, are in series with a more posteriormarginal upper jaw dentition (as are premax-

180 VOL. 167


PREMAXILLARYTEETH

ANTERIORNARIS

MAXILLARYTEETH

SUBNASAL,-RIDGE

- U

'1-02iGa1.I _

FIG. 8. Ganorhynchus woodwardi Traquair. Snout of a Devonian "hard-snouted" lungfish in ventralview (holotype, BMNH 44627, whitened with ammonium chloride). A paired row of premaxillary teethon the preoral area (Miles, 1977, p. 179) and a patch of maxillary teeth on the subnasal ridge lateralto the notch leading to the anterior nostril are shown. x 1.5.

illae with maxillae), are the most anteriorteeth in the upper jaw, and are immediatelyin front of an anterior palatal recess that, inosteichthyans, normally follows the concaveoral border of the premaxillae, as it does inGriphognathus (figs. 7, 10, 14). Lateral toeach premaxilla there is a sulcus or nasal re-cess in the consolidated ethmoid block andthe latter is notched along its oral marginwhere the sulcus extends obliquely towardthe palate. Posterolateral to the sulcus is ahigh tooth-bearing knob (the subnasal ridge,figs. 7, 8), also enamel-coated, and this isfollowed in Griphognathus by a long, lowtooth-bearing ridge (the maxilla) that is an-kylosed with the bones of the cheek (fig. 7B,C). We interpret the high, tooth-bearingknob as the anterior part of the maxilla sinceits enamel coating continues posteriorly ontothe low dentigerous surface behind.The maxilla has a medially directed flange

posteriorly which is sutured to the pterygoid(fig. 7). This flange forms the posteriorboundary of an elongate, oval opening intowhat was probably the nasal capsule. Theanterior part of the oval opening is boundedlaterally by the maxilla, medially and ante-

riorly by a series of dermopalatines (fig. 7).The series of dermopalatines and the vomerextend forward around the pterygoid andmeet in the midline. An additional paired,toothed dermal bone ("extra" dermal bone,fig. 7) is present between the dermopalatinesand the inner, hind, edge of the nasal recessand this element may represent a supernu-merary dermopalatine as suggested by Miles(1977) or a vomer (but see p. 230). Betweenthe anterior end of the pterygoids andthe encircling vomers and the ethmoid blockwith its ankylosed premaxillae there is along, shallow depression which, in other sar-copterygian fishes and tetrapods, has beentermed the anterior palatal recess. The po-sition of the nasal capsule between the elon-gate, oval posterior opening in the palate anda region just behind the inner end of the nasalrecess is suggested by the position and ter-mination of the canal for the olfactory tractsand by the course of various canals (fig. 9)that are comparable with those for nerve andblood vessels in extant fishes and amphibi-ans. The posterior oval opening in the palateis therefore interpreted as the fenestra exo-choanalis, and the nasal recess, present in

1981 181

.%.11.


PROFUNDUS CANAL-

VASCULAR CANAL-

FIG. 9. Plans of snout of two Devonian dipnoans to show major canals as interpreted by Miles. A,Griphognathus whitei Miles (after Miles, 1977, fig. 63); B, Chirodipterus australis Miles (after Miles,1977, fig. 66).

some form in all fossil and extant lungfishes,as housing the anterior, external, incurrentnostril. When the lower jaw is in place inGriphognathus the mandibular symphysis isseated in the anterior palatal recess just be-hind the premaxillae, as in all osteichthyans,and the anterior tip of the nasal recess is justvisible beyond the lower jaw margin as inliving lungfishes.The essential features of Griphognathus

for comparison with tetrapods are (1) theventrally opening anterior nostril and choana;(2) the oval outline of the fenestra exochoa-nalis (choana); (3) the interruption of thedentigerous border of the upper jaw (be-tween the premaxilla and maxilla) at the siteof the anterior nostril; (4) the formation ofthe fenestra exochoanalis by the maxilla andthe dermal bones of the palate that encircle

the pterygoids anteriorly, and (5) the junc-tion of the broad pterygoids in the midlineunder the anterior part of the parasphenoid.Each of these features may be seen also inIchthyostega (cf. figs. 10 and 46B).

(B) NASAL CAPSULEThere are basically four major openings in

the wall of the endoskeletal nasal capsule ofgnathostomes that transmit water or air cur-rents, nerves, and blood vessels. One is forthe olfactory tract, and this is situated pos-teromedially. Another, anterolateral openingis for the nostrils extending to the outside.This opening may be subdivided by soft tis-sue, dividing it into an incurrent and excur-rent port. These two ports may each formseparate tubes that open remotely from each

182 VOL. 167


E X T E R N A LNARI S

CHOANA

ANTERIOR PALATAL RECESS

FIG. 10. Casts of ventral surface of snout in specimens of Ichthyostega sp., to show choana. A,AMNH P. A756a; B, BMNH R. 9469.

1981 183


FENESTRA EXONARPO S T E R I O R

angular opening (palatal fenestra, fig. 12) inthe bony palate just below a large fenestra inthe ventrolateral wall of the nasal capsulewhich can be seen in cleared and stained

F E N E S T R A specimens, but not in fresh specimens unlessE X ON ARI N A the covering soft tissues of the oral mucosaANTERIOR are dissected away. The coincidence of the

palatal fenestra in the nasal capsule and der-mal palate involving premaxilla, maxilla, anddermopalatine is similar to the "choana" ofEusthenopteron as described by Jarvik

I NA (1942).

FIG. 11. Snout of Polypterus senegalus Cu-vier, after Jarvik (1942, fig. 84), showing bony lim-its of fenestra exonarina communis.

other, but at their bases they come togetherinto the anterolateral wall of the nasal cap-sule. The latter condition is found in cladis-tians [Polypterus (fig. 11) and Erpetoichthys].A third opening, in the posterior wall of thecapsule, transmits the maxillary branch ofthe fifth nerve and the palatine branch of theseventh. The fourth main opening,4 in theventral or ventrolateral part of the capsule,is for the ramus buccalis lateralis and, often,one or more blood vessels. In some cases,for example, Polypterus, the third and fourthopenings are represented by only a single fe-nestra. In other cases there are numeroussmaller perforations in the capsule for nervesand blood vessels.Of the various principal openings in the

capsule, the last two, for branches of the fifthand seventh nerves, are most interestinghere because of their relations to surround-ing bony structures. In cladistians thesenerves enter the capsule through a single,large, posteroventral fenestra, but the ramusbuccalis and at least one of the main bloodvessels pass together between the capsuleand the anteroventral wall of the orbit dorsalto the bony palate. Where they turn dorsallyto enter the posteroventral capsule wall thereis a triple junction of the premaxilla, maxilla,and dermopalatine (fig. 12). The premaxillais notched posteriorly just below the fenestrain the nasal capsule and the maxilla and der-mopalatine fail to cover completely the pre-maxillary notch. The result is an oval or tri-

(C) CHOANA AND NOSTRIL INDIPNOANS

The internal nostril of living lungfishes wasregarded as a choana by all nineteenth-cen-tury anatomists, and by virtually all in thefirst half of the twentieth (references in Jar-vik, 1942, p. 273). Allis (1932a, 1932b) arguedthat this nostril was homologous with the ex-current nostril of chondrichthyans and actin-opterygians, but not with the choana, be-cause the choana lies medial to the secondaryupper lip, whereas the dipnoan excurrentnostril lies external to that lip (Allis, 1932a,p. 666). Allis here interpreted the anteriortoothplates of lungfishes, usually called vo-mers, as premaxillary, and so identified thesecondary upper lip, which always lies im-mediately external to the premaxillary andmaxillary teeth. Allis had earlier (1919) ar-gued that dipnoans have no secondary upperlip, but a tertiary upper lip, unique to them,lying external to both nostrils.

Jarvik (1942) developed Allis's argumentfurther, bringing in interpretations of thesubnasal cartilage of lungfishes, the courseof the maxillary nerve, and the interruptionof the infraorbital sensory canal opposite thenostrils. Jarvik concluded, like Allis, thatdipnoan nostrils are ordinary fish nostrilswhich have migrated back so that one opensinto the roof of the mouth, and that dipnoansare not choanates. The question has sincebeen reviewed by Thomson (1965), Bertmar(1966), Panchen (1967), and Miles (1977, p.147), each of whom disagreed with Jarvikover the interpretation of one structure oranother, while agreeing with his conclusion.

184 VOL. 167


NFRAORBITALS

FIG. 12. Polypterus ornatipinnis Boulenger, ventral view of roof of mouth ofjuvenile (4 cm. standardlength), still with large external gills, to show fenestra beneath nasal capsule, between premaxilla,maxilla, and dermopalatine. Note that maxilla and infraorbital bones are still recognizably discrete. Eyeand nasal capsule are indicated by broken lines. Cartilage stippled.

Thus Panchen (1967, p. 379) wrote "the evi-dence for the homology of the dipnoan pos-terior nares with those of actinopterygianfishes and their non-homology with tetrapodchoanae seems conclusive." Rather than re-view these interpretations yet again, we referthe reader to pages 376-379 in Panchen's pa-per, and suggest that the conclusion quotedcomes as a surprise if the whole passage isread with an open mind.

Miles's (1977) figures of the Devonianlungfish Griphognathus show an outer dentalarcade (premaxillary-maxillary dentition inother osteichthyans), identifiable by its po-sition outside the lower jaw when the mouthis closed. This tooth row is external to theposterior nostril, which is completely en-closed by dermal bone in some individuals(Miles, 1977, fig. 80; fig. 7C). The tooth row

is interrupted anteriorly by a notch (fig. 7A)which is identifiable as the anterior marginof the anterior nostril, by comparison withRecent lungfishes. The same notch is evidentin other fossil "hard-snouted" dipnoans(e.g., White, 1962, pl. 2-Dipterus, Rhino-dipterus, Ganorhynchus; Miles, 1977, fig.4-Holodipterus, Chirodipterus; fig. 8). Asargued above (pp. 180-181), we identify theteeth anteromedial to this notch (e.g., in Gri-phognathus and Ganorhynchus, figs. 7, 8) aspremaxillary, as did Miles (1977, p. 187),since they bite outside the lower jaw, are themost anterior teeth in the upper jaw, and lieimmediately in front of an anterior palatalrecess similar to that which normally lies be-hind the premaxillae in osteichthyans (figs.7, 10, 14). Posterolateral to the anterior nos-tril, Griphognathus has a tooth row which

1981 185


runs back on a series of bones which Miles(1977, fig. 57) interprets as lateral nasaltoothplates and ectopterygoid. These teethbite outside the lower jaw (Miles, 1977, p.181), and we therefore interpret them asmaxillary, since in no known gnathostomedo palatal teeth bite outside the lower jaw.Maxillary teeth occur in some other Devo-nian lungfishes (Holodipterus, Dipterus;Miles, 1977, p. 186), but in most hard-snout-ed dipnoans they are represented only by thesubnasal ridge (Miles, 1977, p. 185), a den-ticulated or enameled knob lateral to the an-terior nostril (fig. 8).The posterior nostril is, as noted above,

completely bone-enclosed in some speci-mens of Griphognathus (fig. 7). In thesespecimens, the opening is a rostro-caudallyelongate ellipse, with smoothly inturned,toothed margins (fig. 7C). The opening isbordered laterally by Miles's ectopterygoid,our maxilla, and medially by Miles's der-mopalatine 2, our palatine. Miles noted (p.175) that the opening has the position of achoana, but rejected that homology in favorof "chance similarity." This is because he(p. 147) interpreted dipnoan nostrils to bring"dipnoans into line with other fishes." Butsince his overall conclusion (p. 313) was thatdipnoans are the sister group of choanates,it is with them, not other fishes, that weshould expect dipnoans to fall in line.The outer dental arcade of Griphognathus,

which lies external to the posterior nostriland is notched or interrupted by the anteriornostril, meets Allis's (1919, 1932a) criteriafor choanates, and removes his reason fordenying the homology of the lungfish poste-rior nostril and the choana. This interpreta-tion carries the implication that the two nos-trils of choanates (lungfishes and tetrapods)are homologous with those of other gnatho-stomes, so that the choana is the homologueof a fish nostril. We note Bertmar's (1966, p.140) conclusion on the origin of the dipnoancondition-"The posterior nostril migratedventrally simultaneously with a reduction ofthe maxillary, the premaxillary and the an-terior part of the infraorbital sensorycanal"-and Panchen's (1967, p. 379) pro-

viso-"it is very unlikely that there was everin the ontogeny or phylogeny of the Dipnoian actual physical ventral migration of thenares. Rather there was a failure of the dor-sal migration of the nasal placode normal inthe ontogeny of other bony fish." If the dip-noan posterior nostril is a choana, this ex-planation should also apply to tetrapods.Panchen (1967, p. 380) reviews this hy-

pothesis: that the tetrapod choana is thehomologue of a fish nostril. He specifies oneprediction from it, that primitive tetrapodswould have "a single external naris on eachside, situated at or very near the jaw marginand almost confluent with a laterally placedchoana inside that margin. Thus the condi-tion in the amphibian Ichthyostegalia (Siive-Soderbergh, 1932; Jarvik, 1952) would be re-garded as primitive." Later (1967, p. 407)Panchen concluded that this condition wasindeed primitive for tetrapods, a view he stillholds (Panchen, personal commun.). Pan-chen also pointed out the choana/fish nostrilhomology would be refuted if any fish hadtwo external nostrils in addition to a choana,and noted that this was the basis for Wes-toll's (1943) doubt that porolepiforms havetwo external nostrils, since porolepiformsare the only fishes described as having twonostrils and a choana. Panchen went on tosummarize Jarvik's more recent interpreta-tions of porolepiforms as removing Westoll'sdoubts, and refuting the choana/naris ho-mology. But even if two external nostrilswere demonstrated beyond any possibledoubt in porolepiforms, we do not agree thatthe choana/naris hypothesis would be refut-ed, for we would question the existence ofa choana in porolepiforms. To explain why,it is necessary to review the choanae of rhip-idistians. Before doing so, we emphasize thatthe varied interpretations of lungfishes showthat even in living animals, with full onto-genetic information, identification of choa-nae can be a matter for debate. In fossils,where soft anatomy can only be interpretedafter some Recent model, and where vaga-ries of preservation lead to disputed inter-pretation even of hard parts, there is endlessroom for debate.

VOL. 167186


(D) CHOANA AND NOSTRIL INRHIPIDISTIANS

Dollo (1896, p. 109) suggested that if os-teolepid crossopterygians were ancestral tolungfishes and tetrapods, they would have achoana. Such an aperture was first reportedby Watson (1912, p. 9) in Megalichthys, asan opening whose front margin is formed bythe vomer, and then by Watson and Day(1916, p. 11, pl. 1, fig. 5) in Glyptopomus,where it was said to lie entirely within thepalatine. Bryant (1919, p. 18), perhaps influ-enced by Watson and Day, described a sim-ilar internal nostril, enclosed within the pal-atine, in Eusthenopteron. Stensio (1918, p.120) suggested that a choana might be pres-ent in Dictyonosteus, a Devonian fish sinceinterpreted as a coelacanth by Jarvik (1942,p. 579, pl. 17), with the opening in questionlabeled as the fenestra endonarina commu-nis.

In 1921, Stensio referred (p. 136) to choa-nae in the rhipidistians Dictyonosteus, Os-teolepis, Diplopterus (=Gyroptychius), Rhi-zodopsis and Megalichthys, but the last fourare clearly due to misreadings of Dollo (1896)and Watson and Day (1916), who mentionedthese names in discussing nostrils, but re-ferred only to their external nostrils. In 1922Stensio (p. 195) suggested that the coel-acanth Diplocercides had a choana, and thissuggestion seems to be the only source ofWatson's (1926, p. 200) statement "The orig-inal conclusion that the internal nostrils ofOsteolepids imply an air-breathing habit isconfirmed by thefact that whilst the Old RedSandstone Coelacanth Diplocercides pos-sessed similar internal nostrils, the marineTriassic and Cretaceous Coelacanths pos-sess only an external nostril" [italics ours;air-breathing is, of course, not implied by thepossession of internal nostrils-see Broman,1939, on lungfishes, and Szarski, 1977, onteleosts]. In the same paper, Watson (1926,fig. 4) restored a choana in Eusthenopteron,on the basis (p. 249, figs. 34, 35) of two dis-articulated specimens in the Royal ScottishMuseum, and described (p. 246) a choana inMegalichthys, on the basis of disarticulated

specimens in the Hancock Museum. On thisevidence, he wrote (p. 198) "The internalnostril found in all Osteolepids is identical inposition, size and borders with that of certainEmbolomeri."

Stensio (1932, fig. 30) figured two undeter-mined crossopterygian snouts from the De-vonian of Spitsbergen, and described themas showing two external nostrils and an in-ternal nostril, the latter confluent with theposterior external nostril. Romer (1933, fig.53B) illustrated the choana in Eusthenopte-ron, on data from Bryant, Watson, and Sten-si6. Siive-S6derbergh (1933) mentioned in-ternal nostrils in Dictyonosteus andOsteolepis, referring to Stensio (1922) andWatson (1926), and named a group Choana-ta, comprising Dipnoi, Crossopterygii, andtetrapods. Nielsen (1936, p. 32, figs. 14, 15)restored an internal nostril in an undeter-mined coelacanth from the Triassic of EastGreenland. Jarvik (1942, figs. 80, 81) repro-duced Nielsen's figures and reinterpreted the"choana" as the canal for the buccal nerve.Holmgren and Stensio (1936) reproduced

Watson's (1926) figure of Eusthenopteronand Stensio's (1932) figures of the Spitsber-gen crossopterygians, now determined asPorolepis, both showing an internal nostril.In a footnote (p. 348) they referred to Watsonand Day's identification of a choana in themiddle of the palatine of Glyptopomus, andsaid that Save-Soderbergh has found that theopening "ihre normale Lage hat."

Westoll (1937, p. 28) wrote of crossopte-rygians "there is typically a single externalnostril and an internal nostril on each side:in early forms (e.g., Thursius, Diplopterax[=Gyroptychius]) the external and internalnares may be confluent, the former beingmerely notches in the dermal bones of thelip." Also in 1937, Jarvik described the nasalcapsule of two species of Eusthenopteronfrom the Baltic, and showed a large ventralopening identified as the fenestra choanalis.Finally, in 1937 Romer described the nasalcapsule of Ectosteorachis with a ventralopening "to the choana." He named a groupChoanichthyes to include lungfishes andcrossopterygians.

1981 187


So far as we know, this review covers allthe primary literature on crossopterygianchoanae prior to Jarvik's 1942 monograph.Before commenting on that, it is worthsumming up this early work. By 1940,the literature contained references to inter-nal nostrils in 12 genera of Paleozoiccrossopterygians: Megalichthys, Glyptopo-mus, Dictyonosteus, Osteolepis, Gyropty-chius, Rhizodopsis, Diplocercides, Eusthen-opteron, Porolepis, Nielsen's undeterminedcoelacanth, Thursius and Ectosteorachis.Three of these (Dictyonosteus, Diplocer-cides and Nielsen's fish) are coelacanths,and the openings concerned were reinter-preted after the discovery of Latimeria. InOsteolepis and Rhizodopsis no internal nos-tril had been described-the references weremistaken. In Glyptopomus the only illustrat-ed specimen was wrongly interpreted. InGyroptychius and Thursius Jarvik later(1948, p. 36) maintained that Westoll's ob-servations were based on broken specimens.In Porolepis, Ectosteorachis, and Jarvik'sBaltic Eusthenopteron all that had been de-scribed was a ventral opening of the endo-skeletal nasal cavity, not an opening in theroof of the mouth. But all these names surelycontributed to a climate of opinion in whichWatson could write "the internal nostrilsfound in all osteolepids," Save-S6derberghand Romer could name the Choanata andChoanichthyes, and so on. The only unchal-lenged information that had been presentedwas Watson's restoration of Eusthenopteronand his account of Megalichthys. And in themuseums of the world no single specimenhad been reported that might be shown tothe skeptical enquirer anxious to judge thesize and position of this opening for himself.

Jarvik (1942) presented uniquely detailedaccounts of snout structure in Porolepis andEusthenopteron. In Porolepis no palate wasdescribed, so that the opening in the palate,the fenestra exochoanalis (p. 372) could notbe described. In Eusthenopteron he gave anextraordinarily thorough and well-illustrateddescription of the fenestrae endo- and exo-choanalis, and the structures surroundingthem. This account was based primarily ontwo serially-ground specimens, one of which

(series 1) lacked the palate, while the otherwas complete only on one side, and there thepalate was thrust up into the nasal capsule(Jarvik, 1942, pls. 11, 12). Jarvik's restora-tion of the fenestra exochoanalis was basedmainly on this second series, but that open-ing is labeled in one other specimen (pl. 8,fig. 1).We have prepared four BMNH specimens

of Eusthenopteron foordi to show the nasalcapsule and surrounding structures. Themost informative of these is P. 60310 (figs.13, 14), an acid-prepared specimen consist-ing of a snout, both palates, and the maxillaand cheek bones of one side. Our specimensconfirm the accuracy of Jarvik's account ofthese bones, and we will comment only onthe choana.Our restoration of the fenestra exochoa-

nalis and surrounding bones (fig. 14) differsin several ways from Jarvik's (1942, fig. 56-since much reproduced and redrawn, e.g.,Jarvik, 1954, fig. 25; 1966, fig. 17B; 1972, fig.73B; Romer, 1966, fig. 101; Panchen, 1967,fig. 3; Thomson, 1968, fig. 2B; Vorobyeva,1977a, fig. 3C). In his restoration the fenestraexochoanalis is a drop-shaped openingbounded anteromedially and anterolaterallyby the notch between the vomer and pre-maxilla, posterolaterally by the maxilla, andposteromedially by the dermopalatine. Theopening has a maximum width (at the der-mopalatine/vomer junction) of about 12(1942) to 15 (1972) percent of the total widthof the snout at that point, and a maximumlength about equal to the greatest width ofthe vomer (1942) or somewhat more thanthat (1966, 1972). In our restoration, theopening has the same boundaries as in Jar-vik's, but its medial border is irregular, andit is much smaller, almost equal in length toJarvik's, but with a maximum width of onlyabout 5 percent of the width of the snout atthe dermopalatine/vomer junction. This dif-ference is due almost entirely to the form ofthe head of the dermopalatine in the two re-constructions. Jarvik (1942, p. 453, figs. 54A,56) described the head of the dermopalatineas articulating with the vomer by one sur-face, a roughened depression facing antero-ventrally, and overlapping the posterolateral

188 VOL. 167


FIG. 13. Eusthenopteron foordi Whiteaves. Ventral view of an acid-prepared specimen (BMNH P.60310, whitened with ammonium chloride) comprising snout, incomplete palates, and left cheek. Com-pare figure 14.

corner of the vomer. He also argued (1942,p. 457), and has since insisted (1954; alsoBjerring, 1967, 1973), that the joint betweenthe palate and the snout was immobile. In

our specimens (P. 60310, acid-prepared, P.6807, mechanically prepared) the head of thedermopalatine ends in a thumblike processwhich bears three articular surfaces. The

1981 189


ANTERIOR PALATAL RECESS-

PROC ESSUS IN TERMED IUS

/ X<

-N ASALCAPSU LE

ECTOPTERYGOID

FIG. 14. Eusthenopteron foordi Whiteaves. Snout with left palate and maxilla in articulation inventral view, based mainly on camera lucida drawings of BMNH P. 60310 (fig. 13). Body of parasphenoidand ventral margin of postnasal wall (damaged in P. 60310) drawn from other BMNH specimens.Otherwise, the only restoration in figure is by matching two sides of specimen, restoring damaged teeth,and in articulation of palate and maxilla with snout. Inset, drawn on a slightly larger scale, is an earlierattempt at restoring maxilla and palate in articulation with snout, made before left palate and maxillaof specimen were completely free from matrix and each other. Head of dermopalatine is more closelyarticulated with vomer and dermopalatine is turned medially more than in the main figure. Numbers 1,2, and 3 identify three articular facets on the head of the dermopalatine (see text); foramen in postnasalwall is inferred to have been for nerves and vessels (fenestra endonarina posterior of Jarvik); processusintermedius is endoskeletal lining of processus dermintermedius; pit in dermopalatine (labeled on inset)is inferred to have accommodated fang of first coronoid. For other structures, see Jarvik (1942).

most dorsal of these (1, fig. 14) correspondswith the one described by Jarvik. It facesanteroventrally and articulates with a flatsurface on the dorsal side of the posterolater-

al corner of the vomer. The second articularsurface (2, fig. 14) lies ventrolateral to thefirst, faces anteromedially, and articulateswith a facet on the lateral face of the vomer,

VOL. 167190


separated from the dorsal surface by a crest.The third surface (3, fig. 14) is the mostventral, faces anterolaterally, and articulateswith a facet on the posteromedial face of thevomer, above the hind end of the vomerinetooth row. In Jarvik's figures, the externaland internal tooth rows on the dermopalatineextend forward to the junction with the vo-mer, and are continuous with the corre-sponding tooth rows on the vomer so thatthe fenestra exochoanalis has a toothed me-dial margin, as in Ichthyostega (fig. 10; Jar-vik, 1952). In our specimens, the externaltooth row of the dermopalatine is absent onthe anterior part of the bone, and the larger,internal teeth decrease in size and fade outon the upward-sloping part of the bone be-hind the fenestra exochoanalis, and thethumblike articular process of the dermopal-atine is toothless.Because of these differences in the head

of the dermopalatine, the fenestra exochoa-nalis is much smaller in our reconstructionthan in Jarvik's. When the dermopalatine isplaced in articulation with the vomer, itrocks in the dorsoventral plane and in thehorizontal plane, pivoting on the articulationbetween the processus apicalis of the pala-toquadrate and the lateral ethmoid (fig. 14;Jarvik, 1942, figs. 54, 55). The lateral surfaceof the dermopalatine and ectopterygoid wasclosely and apparently immovably bound tothe maxilla and lacrimal (cf. Jarvik, 1942, pl.13, figs. 3-5). If the rostral end of the palatewas mobile, the rostral end of the maxillaand lacrimal must also have been movablyarticulated with the snout. That this was sois indicated by the fact that the two joints(dermopalatine/vomer and lacrimal + max-illa/premaxilla + lateral rostral) are in thesame transverse plane (figs. 13, 14), and bythe form of the joint surfaces on the medialface of the anterior part of the lacrimal andon the lateral face of the lateral rostral andsupraorbito-tectal (Jarvik, 1944a, fig. 7B, od.La + Mx). These two surfaces are a poormatch, and that on the lateral rostral and su-praorbito-tectal shows blurred margins,quite different from the sharply defined over-lap areas of immobile dermal bones such asthe supraorbital. We therefore believe that

the palatoquadrate/cheek unit was movablyarticulated with the snout.Thus our interpretation of the fenestra

exochoanalis differs from Jarvik's in twoparticulars: our opening is smaller, and is tra-versed by a mobile joint-plane. Neverthe-less, the opening does communicate with thenasal capsule in the fossils, and could still bea choana (though we are aware of no choa-nate with a joint through the choana). But ifwe attempt to restore a choanal tube leadingfrom the fenestra exochoanalis to the nasalcapsule, the passage is severely restricted bythe processus intermedius (endoskeletal; fig.14) and the processus dermintermedius ofthe lateral rostral, which occlude the anteriorand widest part of the opening (cf. Jarvik,1942, figs. 47, 48, 52, 53, 56, 58). This meansthat a choanal tube would have had to turnmedially to enter the nasal capsule beneaththe posterior part of the processus interme-dius. But here we meet another problem. Inthe lateral ethmoid (postnasal wall, Jarvik)of Eusthenopteron there is a large foramen(fig. 14). In 1937 Jarvik interpreted this ashaving "most probably transmitted the trun-cus infraorbitalis (the ramus maxillaris tri-gemini and the ramus buccalis of the nervusfacialis) and its accompanying vessels" (Jar-vik, 1937, p. 98, fig. 13). This is a reasonableinterpretation, since it brings Eusthenopte-ron into line with other primitive fishes, inwhich there is a large canal in the postnasalwall (Porolepis, Jarvik, 1942, fig. 36; Glyp-tolepis, BMNH P. 47838), presumably trans-mitting the orbitonasal sinus, in addition tonerves. But in 1942 Jarvik reinterpreted theopening in Eusthenopteron as the fenestraendonarina posterior, or nasolacrimal duct.This interpretation raises certain difficulties(Thomson, 1964; Schmalhausen, 1968), prin-cipally, in our view, because it leaves noroom for the substantial venous drainage ofthe snout which one would expect (cf. Jar-vik, 1942, p. 481; Thomson, 1964, p. 330).Furthermore, the nasolacrimal duct does notpass through the endoskeleton in primitivetetrapods: in living amphibians, the ductpasses through the fenestra endonarina com-munis except in the plethodontids Pletho-don, Hydromantes, and Batrachoseps, and

1911981


in postmetamorphic Breviceps adspersusamong anurans (Jurgens, 1971; Presley, per-sonal commun.); in primitive fossil tetrapodsthe duct penetrates the lacrimal from end toend, like a sensory canal (Watson, 1940, fig.22; Panchen, 1977, fig. 8; Heaton, 1979, fig.15). In Eusthenopteron, Jarvik's (1942, fig.57) reconstruction of the cavities in the snoutshows the canal in the postnasal wall (labeledf. enp.) to be directly in line with a canal ofroughly equal size, the nasobasal canal (c.n.-b.) running into the rostrum from the anteriorwall of the nasal capsule, and breaking upinto three branches. It is the same in ouracid-prepared specimen: looking forwardthrough the opening in the postnasal wall,one looks directly into the posterior openingof the nasobasal canal. If we restore the trun-cus infraorbitalis and vessels running throughthe ventrolateral part of the nasal capsule,from the opening in the postnasal wall to thenasobasal canal, they occlude the only spaceavailable for a choanal tube (cf. Jarvik, 1942,fig. 57).We do not suppose that these remarks will

convince the reader that Eusthenopteronhad no choana, and that is not our intention.We wish only to show that interpretation ofthe opening in the palate as a choana raisescertain difficulties, and is at least debatable.And there are other, non-choanate fisheswith a comparable opening in the palate. InPolypterus the opening (fig. 12, palatal fe-nestra) lies beneath the anterior part of thenasal capsule, between the premaxilla, max-illa, and dermopalatine, and is large in youngfishes. In paleoniscoids such as Mimia (Gar-diner and Bartram, 1977, fig. 2) there is anotch between the premaxilla and vomer,similar to the notch between those bones inosteolepiforms (fig. 15), and when the der-mopalatine and maxilla are in place, there isa drop-shaped opening between the fourbones, as in Eusthenopteron. This openinglies behind the nasal capsule, and does notcommunicate with the latter as it does inEusthenopteron. But our main reasoning forquestioning the existence of a choana in Eus-thenopteron is that the nose does not matchthe condition in other primitive choanates(dipnoans, ichthyostegids, and other primi-

tive tetrapods-temnospondyls, lepospon-dyls, anthracosaurs, hynobiids). In these,the two nostrils are separated only by a nar-row bar of tissue (in ichthyostegids, loxom-matids, and anthracosaurs by the inturnedanterior end of the maxilla; fig. 10, cf. Pan-chen, 1967, p. 407), and the infraorbital sen-sory canal is interrupted. Instead, if Eus-thenopteron is a choanate, it shows a derivedcondition of the snout, with the external nos-tril well above the edge of the mouth, and anuninterrupted infraorbital canal beneath it,as in advanced amphibians (Schmalhausen,1968; see p. 226 on the infraorbital canal ofIchthyostega).

If Eusthenopteron had no choana, thenboth external nostrils must have openedthrough the single fenestra exonarina, asPanchen (1967, p. 402) suggested. We notethat this is also the case in Polypterus (fig.11), Laccognathus (Vorobyeva, 1980), and,according to Jarvik (1966) in Holoptychius(fig. 17D).

Since Jarvik's 1942 account of the fenestraexochoanalis in Eusthenopteron, a similarfenestra has been illustrated (see fig. 15) inthe osteolepiforms Glyptopomus kinnairdi(Jarvik, 1950, pl. 5, fig. 2; 1966, fig. 16C, D),Eusthenodon waengsjoei (Jarvik, 1952, pl.16, fig. 2, text-fig. 29), Panderichthys rhom-bolepis (Vorobyeva, 1975, fig. 2), Thursiusestonicus (Vorobyeva, 1977a, fig. 3) and Gy-roptychius pauli (Vorobyeva, 1977a, figs. 3,29). Among these five genera Glyptopomus,Panderichthys, and Eusthenodon have fangsat the tip of the dentary and differ from Eus-thenopteron, Latvius, and Thursius whichlack them (Jarvik, 1966, p. 57). The dentaryfangs of Glyptopomus, Panderichthys andEusthenodon fitted into the anterior palatalrecess when the mouth was closed (Jarvik,1966, fig. 16C), and this suggested the pos-sibility that the other opening in the palate,the fenestra exochoanalis, might have beenoccupied by the fangs of the anterior coro-noid. We have examined the specimen ofGlyptopomus illustrated by Jarvik (1966, fig.16C), and find that the tip of the first coro-noid fang is preserved beneath the dermo-palatine in about the same position as thedepression in the dermopalatine of Eusthe-

192 VOL. 167


- F PF P e

FIG. 15. Reconstructions of snouts of rhipidistians in ventral view to show variations in fenestra inpalate (fp). A, Porolepis brevis Jarvik; B, Thursius estonicus Vorobyeva; C, Eusthenopteron foordiWhiteaves; D, Megalichthys hibberti Agassiz; E, Gyroptychius pauli Vorobyeva; F, Panderichthysstolbovi Vorobyeva. After Vorobyeva (1977a, fig. 3).

nopteron which we assume housed a coro-noid fang (pit, fig. 14). However, the speci-men is dorsoventrally flattened, and the jawsdisplaced laterally. Janvier (personal com-mun.) informs us that in an undeterminedosteolepiform which he obtained in the Up-per Devonian of Turkey the opening in thefloor of the nasal capsule is in the matchingposition to some fangs in the lower jaw. AndJessen (1966, fig. SB, C) illustrated an un-

determined osteolepid lower jaw in whichthe fangs of the first coronoid are placed lat-erally, just inside the dentary tooth row, andposterolateral to the pit for the vomerinefang. If Jessen's picture is accurate, the cor-onoid fangs would surely fit into the fenestraexochoanalis (fig. 16A). In Panderichthys,where the opening lateral to the dermopala-tine appears very large (Vorobyeva, 1973, pl.36, fig. 4), the lateral articulation between thepalatoquadrate and ethmoid is absent (Vo-robyeva, 1980) and there is no pit in the der-

mopalatine. The coronoid fangs must havefitted into the "choana" (fig. 16B).

In porolepiforms, Jarvik had no informa-tion on the fenestra exochoanalis in 1942 (p.372). In 1966, he restored the fenestra in Po-rolepis brevis (his fig. 2A), adding informa-tion from Glyptolepis. Since the porolepi-form dermopalatine is well known only inGlyptolepis groenlandica (Jarvik, 1972, p.189) we assume that the restoration of Po-rolepis includes the dermopalatine of Glyp-tolepis, so that the shape and size of the fe-nestra exochoanalis are conjectural.

In Glyptolepis, the fenestra exochoanalis(Jarvik, 1972, figs. 30, 31, pl. 22, fig. 1) isrestored as a triangular opening, with asmooth outer border formed by the premax-illa and maxilla, and an angled inner marginformed by the junction of the vomer and der-mopalatine. The opening is very small, witha maximum width of about 3 percent of thebreadth of the snout at that level. This res-

1981 193


FENESTRA EXOCHOANALIS

A~~~~- PIT FOR VOMERINEFANGS

t ~~~~~~~~~A GS5O F FIRS 1T

PIT FOR VOMERINE FANGS PIT FOR DERMOPALATINE FANGSCORONOID FANG

FIG. 16. Possible occlusal relationships between coronoid fangs and fenestra exochoanalis in osteo-lepiforms. A, reconstruction of anterior part of lower jaw, in dorsal view, of an undetermined osteo-lepidid from the Frasnian of Bergisch Gladbach, West Germany, after Jessen (1966, fig. 5B). Notepositional relationships of coronoid fangs and pit for vomerine fangs, compare figure 14; B, Pander-ichthys rhombolepis (Gross). Reconstruction of left half of snout in ventral view (after Vorobyeva, 1975,fig. 2) and of anterior part of right lower jaw in medial view (after Vorobyeva, 1962, fig. 30, reversed).Arrows mark approximate boundaries of pits or openings expected in opposing jaw to housefangs. Endoskeleton stippled in A and B; C, Eusthenopteron foordi Whiteaves, outline of head (afterJarvik, 1944a, fig. 2), with upper and lower fangs (represented by arrows) added (after Jarvik, 1944a,figs. 1 iB, 13). Bracket covering tips of fangs of first coronoid is position of choana in Jarvik's recon-struction (1944a, fig. 13).

toration is based primarily on Jarvik's seri-ally ground specimen. In the published sec-tions, the right fenestra exochoanalis is notpresent in no. 50 (1972, fig. 8B), is shown innos. 58 and 62 (1972, figs. 66C, 8C), and hasdisappeared by no. 67 (1972, fig. 32A). Theleft fenestra exochoanalis is not present inno. 58, is shown in nos. 61, 62 and 67 (1972,figs. 76B, 8C; 1963, fig. 13A) and has disap-peared by no. 72 (1966, fig. IA). This givesthe opening a maximum rostrocaudal extentof less than 1.5 mm. (section interval 100 ,.)in a fish of head length about 75 mm. Wenote that the dermal bones are displaced onboth sides of the sectioned head, and suggestthat one would expect some sort of notch inthe margin of the palate at the junction offour dermal bones (cf. Jarvik, 1972, pl. 22,fig. 1), and that there are no grounds for re-garding such a notch as the choana.The only other supposed porolepiform in

which a choana has been described is Pow-ichthys (Jessen, 1975). The palate of this fishis unknown, and all that has been describedis an opening in the floor of the nasal capsule.

In summary, we regard statements like"the choanate condition of the Rhipidistiahas been regarded as firmly established"(Panchen, 1967, p. 381) as wholly unwar-ranted. No satisfactory evidence of a choanain porolepiforms has been produced. In os-teolepiforms, the only well-known exampleis Eusthenopteron, and we find the fenestraexochoanalis to be much smaller than in pre-vious restorations. Possible interpretationsof that opening are that it was part of a mo-bile joint between the cheek/palate unit andthe snout; and/or that it transmitted nervesand vessels branching from the adjacenttruncus infraorbitalis and its accompanyingvessels; and/or that it accommodated a cor-onoid fang (fig. 16).

VOL. 167194


(E) CHOANA AND NOSTRIL INTETRAPODS

Our theory is that the choana is the homo-logue of a fish nostril. Dismissing that the-ory, Panchen (1967, p. 380) wrote "Argu-ments against this theory follow closelythose used in the rejection of the choanatecondition of the Dipnoi and are reinforced bydescription of the snout in Rhipidistia andthe embryology of tetrapods." Argumentsconcerning the choanate condition of theDipnoi and the snout of Rhipidistia are dis-cussed in the two preceding sections.

In his comment on "the embryology of tet-rapods" Panchen referred to his own sum-mary of the facts (1967, pp. 389-391). Henoted that the amniote and apodan choanaforms at the oral end of a nasobuccal groovesimilar to that formed in the ontogeny of Dip-noi and Chondrichthyes. Urodeles and an-urans differ in that their choanal tube devel-ops in two portions, an anterior choanalprocess, reasonably homologized with thenasobuccal groove, and a gut process thatgrows forward to meet the choanal process.Panchen accepts that this condition is sec-ondary, and cites Medvedeva's and Bert-mar's work on the possible relationship be-tween this mode of development and theinfluence of the infraorbital sensory canal.He concludes that the anuran and urodelemode of development is "probably . . . due... to a precocious development of the [in-fraorbital sensory] canal in the larvae of theancestral forms" (1967, p. 391).The theory of the tetrapod choana that we

advocate is that it is the homologue of thedipnoan choana, and hence of the excurrentnostril of chondrichthyans. The chondrich-thyan excurrent (posterior) nostril lies exter-nal to the primary upper lip which, in turn,borders the external edge of the dental ar-cade of the palatoquadrate. When a second-ary upper lip extends forward from the angleof the mouth to the nasal region, this fold oftissue will abut against the nasobuccalgroove that connects the excurrent (poste-rior) and incurrent (anterior) nostrils in chon-drichthyans. Given such a development ofprimary and secondary upper lips as primi-

FEC

B"s

C DvF E P

FIG. 17. External nostrils in rhipidistians. A,Glyptopomus kinnairdi (Huxley) and B, Megalich-thys hibberti Agassiz, are osteolepiforms showingour interpretation of a fenestra exonarina com-munis (fec, cf. fig. 11); C, Porolepis brevis Jarvik;and D, Holoptychius sp., are porolepiforms, thefirst with separate anterior (fea) and posterior(fep) fenestrae exonarinae, the second with afenestra exonarina communis, as in A and B(A, B after Jarvik, 1966, fig. 14B, C; C, D afterJarvik, 1972, fig. 7A, C).

tive for osteichthyans, the dorsal migrationof the nasal placode during embryogenesisdisplaces both nostrils external to the second-ary upper lip, which includes the dermalbones of the maxillary-premaxillary arch andits dental arcade. This migration from theprimitive ventral position of the placode ex-plains the position of the nostrils on the sideof the snout in front of the eye in cladistians,actinopterygians, rhipidistians and actinis-tians.2 The choanate condition, (dipnoans,

2 We note Panchen's (1967, p. 380) summary of thecontroversy over whether the teleostean nasal placodemigrates up the tip of the snout, so reversing its rostro-caudal direction, the anterior nostril becoming the pos-terior and vice versa, or migrates up the side of thesnout, so that the orientation of the nostrils is un-changed. Since, as Panchen says, it is the placode notthe nostrils that migrates, and the nostrils differentiatesubsequently, we consider it immaterial whether thechoana is homologized with the excurrent or incurrentnostril of other osteichthyans.

1981 195

F EC


tetrapods) arose by a modification of thisprimitive osteichthyan pattern, in which theplacode failed to migrate, and the excurrentnostril came to open in the roof of the mouth,between the primary and secondary upperlips. As one consequence of this modifica-tion, the secondary upper lip and outer den-tal arcade were interrupted in primitivechoanates. This condition is shown by the"hard-snouted" dipnoans (figs. 7, 8), and ishardly modified in larval urodeles (e.g.,Schmalhausen, 1968, fig. 90) and in primitivefossil tetrapods such as Ichthyostega andother genera reviewed by Panchen (1967, pp.402-407), where the choana and externalnostril are separated only by an inturnedprong of the maxilla (fig. lOB). A secondconsequence of this modification was inter-ruption of the infraorbital sensory canal, as indipnoans and primitive tetrapods. One pre-diction of our interpretation is therefore thatthe lateral rostral of Ichthyostega (Jarvik,1952), a tubular ossicle carrying the infraor-bital sensory canal between the nostrils, doesnot exist. In more advanced tetrapods, the

distance between the choana and the externalnostril increased, and in those forms withaquatic larvae, restoration of continuity ofthe infraorbital sensory canal entailed a fur-ther modification of choanal ontogeny, thedevelopment of a gut process.

Other predictions of our hypothesis (seefig. 17) are that the porolepiforms, if cor-rectly restored with two external nostrils,had no choana; and that osteolepiforms, ifcorrectly restored with one external narialopening, either had two narial tubes emerg-ing from that opening (like Polypterus) andno choana, or were choanates exhibiting anadvanced condition of the nostrils, with asecondarily intact infraorbital canal betweenthe choana and external nostril. Since thesecond interpretation demands the indepen-dent acquisition of an uninterrupted infraor-bital canal, we favor the first. Panderichthysis restored as an osteolepiform with two ex-ternal nostrils. If correct, this would corrob-orate our view that other osteolepiforms hada fenestra exonarina communis.

NASOLACRIMAL DUCT, LABIAL CAVITY, ROSTRALORGAN, AND JACOBSON'S ORGAN

As noted above in the Introduction, thenasolacrimal duct of amphibians and the la-bial cavity of Neoceratodus (figs. 18-20)have similar anatomical relations and onto-genetic histories (figs. 18, 19). The only otherstructure in gnathostomes anatomically simi-lar to these dipnoan and tetrapod features isthe rostral organ of actinistians. The rostral or-gan, which is known in some detail in Lati-meria (fig. 22) (Millot and Anthony, 1965) is,like the labial cavity in Neoceratodus, filledwith a cavernous mucosa, partly fills the re-gion of the snout between the orbit and thenasal capsule, and opens by pores to the out-side. The rostral organ, however, is very ex-tensive, the right and left ones are intercon-nected in the midline and, on each side, thereare three pores opening to the outside, oneon the snout and two between the orbit andthe posterior external nostrils.

Jarvik (1942) followed Allis (1932a) in pro-

posing that the nasolacrimal duct of tetra-pods is homologous with the posterior ex-ternal nostril in bony fishes. He saw inEusthenopteron a canal in the endocranialpostnasal wall connecting the nasal capsulewith the orbit approximating the course of thenasolacrimal duct of tetrapods, that Eusthe-nopteron has but a single external narialopening in the dermal bones of the snout andthat Porolepis lacks the canal but has a nor-mally developed posterior external nostril aswell as an anterior one. From this he con-cluded that the posterior external nostril ofEusthenopteron has sunken into theendocranium of the snout to form a primitivenasolacrimal duct, but that in Porolepis theposterior external nostril is essentially un-modified. What, to us, seems basically un-convincing about Jarvik's proposal is thatthe canal in the endocranium of Eusthenop-teron exactly parallels the course of the

196 VOL. 167


NASAL CAPSULE

A BFIG. 18. Dissected snouts, in dorsal views, of A, Neoceratodusforsteri (Krefft) and B; larval Triturus

alpestris (Laurenti), to show position of lungfish labial cavity and tetrapod nasolacrimal duct (A, afterJarvik, 1942, fig. 1 1A; B, after Jarvik, 1942, fig. 19B).

maxillary and buccal nerves and accom-panying vessels as they emerge from theanterior orbital wall and pass forward intothe posterior wall of the nasal capsule in elas-mobranchs, Polypterus, Amia, dipnoans,and urodeles and that there might well havebeen an anterior and posterior nostril sepa-rated only by soft tissue as was probably trueof Laccognathus (Vorobyeva, 1980) andHoloptychius (the nariodal bone shown byJarvik within the external nasal opening hasbeen included in the nasal opening of Hol-optychius by supposition, not by observa-tion). At best, Jarvik's proposal ignores asimpler interpretation of the Eusthenopteroncanal based on frequently observed andwidespread anatomical relations of gnatho-stomes.

Although Schmalhausen (1968), like Jarvik(1942) before him, had accepted Allis's(1932a) theory of the transformation of thefish posterior nostril into the nasolacrimalduct and the development of the choana asa neomorph (hence, a third functional open-ing in the nasal sac of choanates), Schmal-hausen disagreed with Jarvik and Allis byproposing that the nasolacrimal duct incor-porates material from the cephalic lateral linesystem, resulting in an interrupted infraor-bital canal.We have two major objections to Schmal-

hausen's theory, as well as to all others (e.g.,Allis, 1932a; Jarvik, 1942; Bertmar, 1965,1968) postulating a former occurrence of twopairs of posterior nares. One is that a theoryin which the nasolacrimal duct is said to formfrom a seismosensory canal and a nostril isinconsistent with the fact that the nostrils arefully developed in urodeles prior to meta-morphosis, whereas the nasolacrimal ductappears later and, except in salamandrids,becomes confluent with the nasal sac.Schmalhausen's and Allis's theory requiresthe presence of a posterior external nostrilprior to metamorphosis and this is not ob-served in any choanate animal. The secondobjection is that no vertebrate animal isknown to have had three pairs of nostrils.We see our own theory, in which the nasal

placode fails to migrate outward and upwardto the side of the snout as in other fishes, asmore simply accounting for the presence ofthe two nostrils (an anterior one and a pos-terior one, or choana) in tetrapods, and indipnoans. This theory, as outlined above(pp. 195-196), interprets the external nostrilsof actinopterygians, actinistians, and rhipi-distians as a result of ontogenetic migrationof the nasal placode.

Nevertheless, we find much merit inSchmalhausen's ideas, particularly as theypertain to the homologies of the nasolacrimal

1981 197


ANTERIOR 5 \NA RES -.

LAB IAL CAVI TY

?b~ 11111!4S 1

MEMBRANOUS iROOFOF4LABIAL CAVITY -

FIG. 19. Neoceratodus forsteri (Krefft), approximately 120 cm. total length, AMNH 40800. Photo-graphs of head to show A, opening of labial cavity; and B, membranous roof of cavity with a probeinserted.

duct. We find interesting, for example, thetopological and dependent developmental re-lations between the duct and the lacrimal andseptomaxillary bones (fig. 23), as well as thefact that extensions of the infraorbital canalcan become confluent with the nasal sac, as

indications that the duct could have had aseismosensory origin.3 If, as discussed

3 Schmalhausen argued that the nasolacrimal ductforms from a detached extension of the infraorbital lat-eral line canal that has become confluent with the nasal

198 VOL. 167


ROOF OF NASALC A P S U L E

PTERYGO IDT0 T H P L AT E

EXT ERN AL/NASAL VEIN-

-BUCCAL NERVE-

INFRAORBITALSENSORY CANALLAMI NA

-ORBITONASALIS-

,LABIAL CAVITY-

-LOWER LIPFOLD-

MAN D tBU LARTOOTH PLATE

_-TON GUE--

.MECKEL'S -CARTILAGE

0.5mm

FIG. 20. Neoceratodus forsteri (Krefft). Transverse sections through deepest part of labial cavityin 27 mm. (A) and 34.5 mm. (B) larvae (series F120, F121, Hubrecht Laboratory, Utrecht; same seriesused by Fox, 1965). Labial cavity at these stages is a deep pit at angle of mouth, whose apex is closeto distal end of lamina orbitonasalis, and is attached to latter by connective tissue. This relationship ismaintained in the adult, where cavity is greatly enlarged by increase of gape.

above, the labial cavity of Neoceratodus andthe rostral organ of Latimeria might behomologues of the tetrapod duct, and if theduct has a seismosensory origin, then oneshould expect to find in the labial cavity, ros-tral organ, and in the primitive tetrapod duct,

sac because: (a) the formation of the lacrimal and sep-tomaxillary bones is dependent on the presence of thenasolacrimal duct (Schmalhausen cites the experimentsof Medvedeva, 1959, 1960, in which the anlage of theduct was removed and the two bones failed to develop);(b) the lacrimal and septomaxillary bones are membranebones that, at metamorphosis, come to surround com-

pletely the nasolacrimal duct but underlie the duct atfirst-thereby closely resembling the development ofseismosensory canal bones in bony fishes (Schmalhau-sen's fig. 103 shows these relations in a larva of theurodele Onychodactylus); and (c) parts of the infraor-bital canal are known to become confluent with the nasalsac in some fishes (Schmalhausen illustrates this, fig.104, in an individual of Amia).

sense organs comparable with neuromastsand an appropriate innervation. So far aspresently known, the rostral organ of Lati-meria, according to Millot and Anthony(1965), has an extensive innervation entirelyfrom the n. ophthalmicus superficialis (V)which is characteristically associated withlateral line innervation (see, e.g., Strong,1895; Norris, 1913). Fibers from at least theventral branches of the n. buccalis would notbe expected in the rostral organ since Lati-meria has a well-developed and uninterrupt-ed infraorbital canal; n. buccalis innervatesneuromasts of the posterolateral snout re-gion between the eye and anterior nostril. Dothe nerve fibers in Latimeria go to someform of seismosensory organ? Are theresuch organs in the labial cavity of Neocerat-odus and are they innervated by n. ophthal-micus superficialis? Do the nasolacrimalducts of the anuran Xenopus and apodans,

1981 199


FIG. 21. Neoceratodus forsteri (Krefft). Snout of 15 mm. larva, AMNH 40801 (alcian blue/alizarincounterstained, cleared preparation), in ventral view.

which are associated with tentacular organsof presumed sensory function, have a similarinnervation and are there seismosensory or-gans in the nasolacrimal ducts which, in the

Salamandridae, do not become confluentwith the nasal sac? Future work will decide.If Schmalhausen's theory is correct, how-ever, the duct should be associated ontoge-

POSTERIOR EXTERNAL OPENINGSOF ROSTRAL ORGAN

LATERAL ROSTRAL B

CANAL FOR BUCCAL NERVE

FIG. 22. Latimeria chalumnae Smith. A, outline of head to show nostrils and openings of rostralorgan; B, diagram of rostral organ and nasal sac of right side in dorsal view; C, snout in lateral view(A, after Millot and Anthony, 1958; B, C, after Millot and Anthony, 1965).

200 VOL. 167


OUTLINE OF CHOANA

FIG. 23. Nasal structure in hynobiid urodeles. A, Ranodon sibiricus Kessler, diagrammatic recon-struction of snout of a 22 mm. larva, ventral view; B, Onychodactylusfischeri (Boulenger), diagrammaticreconstruction of left nasal sac of 85 mm. (postmetamorphic) individual in dorsal view. After Schmal-hausen (1968, figs. 98A, 103).

netically with the infraorbital system and,hence, with n. buccalis, and, if that provesto be the case, it would appear to rule outthe n. ophthalmicus superficialis innervatedrostral organ of Latimeria as a homologue.We find ourselves in disagreement with

one other feature of Schmalhausen's analy-sis, and that concerns the explanation of theinterruption of the infraorbital canal. The in-terruption typically is at the point where, inlungfishes and tetrapods, the canal meets thenostril and does not appear to be associatedwith the ontogenetic history of the choana,whatever that history might be. In bothgroups, the infraorbital canal bends down-ward toward the oral border [in dipnoans,temnospondyls, lepospondyls, and in thoseanthracosaurs with a ventrally situated ex-ternal nostril, and in larval Necturus andRanodon (see, e.g., Stensio, 1947, figs. 2,24A; Schmalhausen, 1968, figs. 100, 101;Miles, 1977, figs. 18, 35, 111, 112)] and thesimilarity in the courses of the supraorbitaland infraorbital lateral lines is particularlystriking between lungfishes and such larvalhynobiid urodeles as Ranodon (Schmalhau-sen, 1968, fig. 90). Differences between lung-fishes and urodeles in the external form ofthe snout appear to be associated primarilywith turning under of the tip of the snout in

lungfishes. But even this difference appearsto be minimal when lungfishes are comparedwith 22 mm. Ranodon larvae in which, ac-cording to Schmalhausen's (1968) figure 98A(fig. 23A), the external nostril opens ventrallyon the margin of the lip as in lungfishes andichthyostegids. It is the snout in modernurodeles, anurans, and amniotes that ap-pears to have turned upward, carrying theexternal nostril into a more dorsal position.One additional similarity between the na-

sal apparatus of lungfishes and tetrapodsconcerns a lateral diverticulum of the nasalsac. In lungfishes it has generally been giventhe neutral term of diverticulum laterale,whereas in tetrapods it is often identified asa vomeronasal organ or Jacobson's organ. Inboth groups the diverticulum is a lateral out-pocketing of the posterior part of the nasalsac (fig. 24) and consists largely of olfactoryepithelium. In the more terrestrial amphibi-ans Jacobson's organ is better developed andis associated with the nasolacrimal duct. Insome forms it includes a narrow strip of res-piratory epithelium along its dorsal part (Jur-gens, 1971, fig. 31). In this dorsal position inat least some Protopterus the epithelium ap-pears to be somewhat differentiated and isless nucleated (Bertmar, 1965, fig. 18D). Acomparison of Bertmar's (1965) figure 18C of

1981 201


FIG. 24. The nasal sac in A, Triturus alpestris(Laurenti); and B, Protopterus annectens Owen.A, diagram of right sac in dorsal view, after Jur-gens (1971, fig. 39A); B, diagram of right sac, indorsal view, of 20 mm. larva, after Bertmar(1965, fig. 18B).

a transverse section of the nasal sac and lat-eral diverticulum of Protopterus with com-parable sections in various urodeles (Jur-gens, 1971, fig. 32; Higgins, 1920, pl. 8, fig.55; Medvedeva, 1968, fig. 1 showing Tritonand Ranodon) makes it clear why the dip-noan structure has been likened to the tet-rapod vomeronasal organ. The similarity isheightened by a comparison of Bertmar's(1965) reconstruction of the whole nasal sacof Protopterus (his fig. 18B) and with Jur-gens's (1971) of the sac in Triturus (his fig.

39A) (fig. 24). Bertmar (1965, 1968) has ar-gued that a large diverticulum laterale is cer-tainly formed in all three dipnoan genera butthat "it develops into an olfactory fold as doother lateral diverticles of the nasal sac, andtherefore it does not seem to be homologouswith the vomeronasal organ . . . in tetra-pods." Since the lateral diverticulum is, infact, an outpocketing of olfactory epitheliumin both dipnoans and tetrapods, Bertmar'sreasoning escapes us. Perhaps he is saying,however, that because dipnoans have otherdiverticula, the large posterolateral one can-not be homologous with the tetrapod diver-ticulum in the same position. We are equallymystified by this argument. The question is:do other gnathostomes besides dipnoans andtetrapods have a lateral diverticulum of thistype and in this position on the nasal sac? Ifthe answer is negative, and Bertmar has notsuggested that such a modification of the na-sal sac occurs elsewhere, the structure in thetwo groups is most simply regarded as syn-apomorphous. Bertmar (1965) does not re-fute Rudebeck's (1945) claim in his study ofthe forebrain of Protopterus that Jacobson'sorgan is present in the genus as a rudimen-tary accessory olfactory bulb and that a ru-dimentary vomeronasal nerve may beformed.

PAIRED FINS AND GIRDLES

(A) PAIRED FIN SKELETONAlthough the paired appendages of lung-

fishes have been cited as precursors of tet-rapod limbs (Braus, 1901 and many earlierworkers; Holmgren, 1933, 1949a, 1949b, as aprecursor of urodele limbs), the osteolepi-form crossopterygian fins are today assumedto be closer to the tetrapod structure (Greg-ory, 1911; Watson, 1913; Westoll, 1943,1958a; Jarvik, 1964, 1965; Thomson, 1968,1972; Andrews and Westoll, 1970a, 1970b;Rackoff, 1976, 1980). The favored compari-son of osteolepiforms is attributable to theirpossession of a pair of distal bones (one shortand stout, the other long and slender) artic-ulating on a single stout basal radial with a

dorsal process that has been likened to a tet-rapod humeral process. The terms humerus,ulna, radius and femur, tibia, and fibula arein general use for the three radial elementsin the osteolepiform pectoral and pelvic fins.Preference for the osteolepiform structuresseemed to be reinforced by the observationthat, at the base of lungfish fins, the principalradials are in a single series with smaller onesradiating outward from each side of the cen-tral axis as do the barbs on the rachis of afeather. Moreover, Neoceratodus fins wereseen to be very similar to those of Paleozoicxenacanth sharks and Gegenbaur (1895) hadrepresented this "archipterygial" fin as theprimitive type for gnathostome fishes. It ap-

202 VOL. 167


parently went unnoticed that, during Gegen-baur's era, Semon (1898), in a study of theembryology of the pectoral fin of Neocerat-odus, showed that the second axial radialelement is ontogenetically double.

In reviewing the skeletal anatomy of thepaired appendages and their girdles ingnathostomes we have concluded that lung-fish fins include synapomorphies with tetra-pod limbs and that there are synapomorphiesof the fins and girdles in all chondrichthyansand osteichthyans.

Osteichthyan paired fins, as pointed out byGoodrich (1930), are of two kinds, those ofsarcopterygians in which pectoral and pelvicfins are constructed according to the samebasic plan, and those of actinopterygians inwhich the pectoral and pelvic differ. Thedifferences between the two groups of bonyfishes may be understood by examining thefin skeletons of other gnathostome fishes. Inacanthodians all that is known about endo-skeletal fin supports is that in Acanthodesbronni the pectoral fin has a few small nod-ules arranged irregularly at the fin base(Miles, 1973). In arthrodires a conditionmore nearly like that of chondrichthyans hasbeen found, especially in Brachyosteus diet-richi Gross, in which the pectoral fin includesa segmented metapterygial axis articulatedwith a posterior process on the scapulocor-acoid and a series of radials arranged alongthe articular surface of the scapulocoracoidanterior to the metapterygium (Stensio,1959, pl. 15, figs. 1-3). The term metapte-rygium as used here refers simply to one ormore basipterygial cartilage or bone seg-ments aligned in series, joined to a posteriorarticulatory process or surface on the scap-ulocoracoid, and forming the posterior ortrailing endoskeletal support of the fin. In thefin of Brachyosteus, as in all primitive sharkfins, the metapterygium appears to be a mainaxis of fin support. An arrangement of finsupports like that of Brachyosteus occursalso in the arthrodire Enseosteus jaekeliGross (Stensio, 1959, pl. 2). In another groupof placoderms related to arthrodires, the pe-talichthyomorphs (Miles and Young, 1977),pectoral radials are also present. In Gemuen-dina sturtzi Traquair, radials are arranged

around the fanlike shoulder girdle, in one rowlaterally and in three or more rows poste-riorly (Stensio, 1959, fig. 14A). In Pseudo-petalichthys problematicus Moy-Thomas, allthat is known of the endoskeletal support isa pair of somewhat triangular basal elementseach of which articulates separately in anotch in the shoulder girdle (Gross, 1962, fig.7). This last fossil also shows that pelvic ra-dials were present in placoderms, in thiscase 10 or 11 rods parallel to each other and,as a group, parallel to a more medial elongatebone. Gross (1962, p. 76 and fig. 7) inter-preted this elongate bone as a basipterygium(=metapterygium). He described it as havinga longitudinal ridge, which makes it unlikethe metapterygial axes of other fishes. Itseems to us that the row of parallel radialsis most simply interpreted as a primitive se-ries of basal radials, and the medial bonelying against the radials on one side, and atthe anterior end of the series of radials onthe other, as a pelvic girdle. A summary ofacanthodian and placoderm fin anatomy in-dicates that the former have nothing com-parable with the rodlike radials of othergnathostomes and that the latter have basalradials but lack the details of endoskeletalanatomy of sharks and bony fishes.The pectoral fin of living chondrichthyans

has often been characterized as tribasal be-cause there are three distinct cartilages artic-ulating with the shoulder girdle that supportthree or more rows of distal radials. In fossilchondrichthyans the fin anatomy may bemuch simpler in lacking a definite propteryg-ium along the fin's leading edge and a centralcartilage, the mesopterygium, between theformer and the metapterygium at the fin'strailing edge (fig. 26A, B). The tribasal finseems to be characteristic only of some eu-selachians.4 Among non-euselachian chon-drichthyans present in the Paleozoic themodern tribasal condition did not exist (Zan-gerl, in press). Instead, the most general con-dition, as exemplified by species of Clado-selache (fig. 25A), Cladodus, Cobelodus (fig.25B), Fadenia, and Xenacanthus is the

4Euselachians include all modern sharks andrays, and some fossil groups (Zangerl, in press).

1981 203


FIG. 25. Left pectoral fin skeleton of A, Cla-doselache brachypterygius Dean; and B, Cobel-odus aculeatus Zangerl and Case (after Zangerl,in press, fig. 34).

presence of a well-defined metapterygial axismade of an elongate basal segment that ar-ticulates with the girdle followed by one to13 cuboidal or rectangular segments. On thepreaxial side of the metapterygium (facingthe fin's leading edge) there are numerousparallel radials arranged at an angle between70 to 90 degrees with the metapterygial axis.Zangerl (in press) regarded those species withunsegmented radials and a long whiplike me-tapterygium as being somewhat specialized.He constructed what he considered to be ageneralized non-euselachian pectoral (his fig.35A) in which there are only four metapter-ygial segments, and proposed that the earli-est transition toward the pectoral of modernsharks involved segmentation of radials (hisfig. 34). We would state the generalized con-dition of the pectoral somewhat differentlyby observing that in Zangerl's cladogram ofthe main groups of sharks (his fig. 52), inwhich euselachians and hybodonts are de-picted as the sister group to all other sharks,segmented radials are present in both sistergroups and that long whiplike metapterygiaoccur only in groups with cladistically apo-morph positions among non-euselachians.Also characteristic of these primitive sharkpectorals is the presence of radials anteriorto the base of the metapterygium as in somearthrodires. As many as 10 of these anteriorradials attach along the posteroventral mar-

gin of the scapulocoracoid in front of themetapterygium. Various morphologicallytransitional conditions in some non-eusela-chians suggest that anterior radials enlargeor fuse to form the propterygium and mesop-terygium of euselachians (cf. figs. 25-27).Correlated with the development of a tribasalfin is the shift in the main point of fin artic-ulation from the metapterygium in non-eu-selachians to the propterygium in euselachi-ans. It is clear, therefore, that cladisticcomparisons involving chondrichthyan finstructure should be with a Cobelodus-like finin which there are no more than four or fivesegments in the metapterygium, seven oreight segmented preaxial radials articulatingon the scapulocoracoid anterior to the me-tapterygial base, and the main axis of fin sup-port on the trailing (metapterygial) ratherthan leading (propterygial) side of the fin.The pelvic fin skeleton of chondrichthyans

has apparently had a history similar to, butless complex than, that of the pectoral. Ametapterygium, also the main axis of fin sup-port, has numerous radials attached along itspreaxial surface. In the pelvic skeleton, how-ever, only a few radials extend anterior tothe metapterygial base, the metapterygiumhas only one to three segments distal to thebasal articular segment, the metapterygiumretains its function as the main axis of sup-port for the fin even in euselachians in whicha small propterygial element develops (fig.26B), and the fin is characteristically eitherunibasal or (in some euselachians) dibasal.In male chondrichthyans the terminal metap-terygial segments (fig. 26B) form internalsupport for a clasper (claspers are not knownto be present in some cladodont sharks).

In the discussion that follows we will in-terpret actinopterygian paired fin structureas a transformation of a primitive (chon-drichthyan) metapterygial fin into a propter-ygial type, paralleling the development of theeuselachian fins, and sarcopterygian finstructure as an alternative transformation ofa primitive metapterygial fin into an exclu-sively metapterygial structure primarily bythe loss of all radial elements anterior to themetapterygial base. We feel justified in pur-suing these comparisons because the basic

204 VOL. 167


POSTAXIAL RADIALS

METAPTERYGIUM,_) Gt

A

FORAMINA FORDIAZONAL NERVES

PROPTERYGIUM -PREAXIAL RADIALS

FIG. 26. Left pectoral (A) and pelvic (B) fin skeleton, in dorsolateral view, of 17 cm. d Squalusacanthias Linnaeus, AMNH 40804.

structural elements of the pectoral and pelvicfins of osteichthyans are present in primitivesharks and placoderms as well (e.g., Brach-yosteus and Pseudopetalichthys). To infer a

transformation series between osteichthyanfin structure and that of acanthodians, aswould be required by Miles's (1973) theoryof relationships, requires the assumption that

1981 205


JUVENILE CLASPER

METAPTERYGIUM

10

FIG. 27. Left pelvic fin skeleton of 20 cm. dHydrolagus colliei (Lay and Bennett) in dorsolat-eral view, AMNH.

complex similarities between osteichthyansand chondrichthyans were attained indepen-dently.The difference between actinopterygians

and sarcopterygians in pectoral and pelvicstructure can be appreciated by contrastingthe paired fin endoskeletons of Acipenserand Eusthenopteron (fig. 28). Even a cursorycomparison shows that the most complete ofthe pectoral and pelvic fin skeletons of Eus-thenopteron illustrated by Andrews andWestoll (1970a, fig. 8J) have branching pat-terns that resemble in detail the arrange-ments of metapterygial elements and theirpreaxial radials in Acipenser. Acipenser finsappear to differ from those of Eusthenopte-ron only in the presence of three sets of ra-dials anterior to the metapterygial axis andof a heavy, fenestrated pectoral propterygialsegment that Jessen (1972) identified as acharacter of actinopterygians. It is also evi-dent that the fins of Acipenser resemble, inconsiderable detail, the paired fins of chon-drichthyans, differing from the primitiveshark fins (fig. 25), but resembling those ofeuselachians (fig. 26), in the development ofa strong propterygial articulation. The fins of

both Acipenser and Eusthenopteron differfrom those of chondrichthyans and resembleone another in the lower number of distalradials anterior to the metapterygial base,and the alignment on the metapterygial axisof one basal preaxial radial per metapterygialsegment. The metapterygial axes in Eusthe-nopteron may be distinguished from those ofAcipenser by having posterodorsal process-es (=entepicondyles of Andrews and Wes-toll, 1970a) in the first and third, second andthird (AMNH P. 3871), or second (AMNHP. 8095) pectoral metapterygial segments(although some specimens have none at all;Holmgren, 1933, fig. 18) (fig. 29) and by hav-ing a greatly reduced pelvic skeleton inwhich there are only two metapterygial seg-ments with three basal radials attached tothem. Hence, Acipenser appears to have re-tained a primitive chondrichthyan arrange-ment of radials but has paralleled euselachi-ans by shifting part of the fin support to thepropterygial side, and Eusthenopteron ap-pears to have retained the primitive chon-drichthyan pattern of four pectoral metap-terygial segments and two pelvic ones and,by losing all radials anterior to the metapter-ygium, to have retained a primitive metap-terygial support for the pectoral and pelvicfins.

With respect to fin structure in general,Jessen (1972) has shown that Acipenser istypical of the members of other plesiomorphgroups of actinopterygians, including variouspaleoniscids (and see also Nielsen, 1942,1949), as Andrews and Westoll (1970b) haveshown that Eusthenopteron is similar to oth-er "rhipidistian" fishes such as Sauripterus(known only from part of a shoulder girdleand fin) and Megalichthys. According toThomson (1972) and Rackoff (1980) Sterrop-terygion brandei, of which both pectoralsand pelvics are known, is, again, much likeEusthenopteron.Although for more than a half-century

Eusthenopteron fin structure has been con-sidered antecedent to tetrapod limbs, and themost recent interpreter of this doctrine haseven identified elbow and ankle joints in thefins of Sterropterygion (Rackoff, 1980), webelieve that Goodrich's (1930, p. 161) cau-

VOL. 167206


/ > ~~~~~~PROPTERYGIUM \/PREAXIAL RADIALS

PELVIC GIR

FIG. 28. Left pectoral (A) and pelvic (B) fin skeleton, in dorsolateral view, of adult Acipenser sturioLinnaeus, BMNH E. 170-1, distal elements in A from Jessen (1972); C, D, left pectoral (C) and pelvic(D) fin skeleton ofEusthenopteronfoordi Whiteaves, in lateral view, after Andrews and Westoll (1970a,figs. 8H, J, 27).

tionary comments are still appropriate. Itwas clear to Goodrich that we should expecta fishlike relative of the tetrapods to possess"6pectoral and pelvic fins alike in structure,with outstanding muscular lobe, extensiveendoskeleton with at least five radials [=me-tapterygial segments], small web, and few ifany dermal rays. But. . . the fin (espe-cially the pelvic fin) is too specialized [inEusthenopteron] ... with short endoskele-ton, large fin-web, and powerful dermalrays. "We specify below seven general condi-

tions of the tetrapod limb that we believe

must be met, including some of the pointsmade by Goodrich:

1. An unpaired basal element (humerus orfemur) supporting entire appendage.

2. Paired, subequal subbasal elements (ulna-radius and fibula-tibia) that are function-ally joined distally.

3. Two primary joints (shoulder or hip andelbow or knee) in each appendage.

4. Distal bony elements (carpals or tarsalsand digits) arising on postaxial side of ap-pendage (see p. 211).

5. Pectoral and pelvic appendages alike in

1981 207


FIG. 29. Eusthenopteron foordi Whiteaves,left pectoral fin endoskeletons, showing variabledevelopment of processes on postaxial (dorsal)margin of metapterygial segments (compare fig.28). Upper left, AMNH P. 8095; upper right,AMNH P. 3871; lower, from Holmgren (1933, fig.18). Note spread radials of upper two fins andcontracted radials of lower fin.

structure (i.e., structurally similar andwith at least four or five primary seg-ments).

6. Extensive muscular lobe that can be ex-tended well below the ventral body wall.

7. Reduction or loss of dermal rays.Of these seven criteria, osteolepiforms sat-

isfy only the first and part of the second, thatis, they have an unpaired basal element (thefirst segment of the metapterygial axis) andthey have a second axial segment with theproximal end of which a large radial articu-lates. The second criterion is only partly metbecause the second axial segment and asso-ciated radial are decidedly unequal in lengthand the two bones diverge distally. In somespecimens of Eusthenopteron, the two bonesare more closely aligned than in others (e.g.,in the specimens illustrated by Andrews andWestoll, 1970a, as compared with two spec-imens seen by us; see figs. 28D, 29) and thissuggests to us that the Eusthenopteron fin,like those of other fishes, could be spreadand contracted, corresponding with a motionof the radial away or toward the second axialsegment. In turn, this means that the distaltips of the two bones are not functionallyjoined. The third criterion is not met be-cause there is no joint of tetrapod typebetween the first and second axial seg-ments (but Rackoff, 1980, disagrees). Thefourth is rejected for osteolepiforms becauseall radials are on the preaxial side of the me-

tapterygial axis, and the fifth is also, becausethe osteolepiform pectoral has three or fouraxial elements and the pelvic only two. Thesixth criterion is rejected because osteolepi-forms have very short muscular lobes asjudged by the posterior limits of the radials(when known), scale cover and position ofthe fin-ray bases, and the seventh criterionis also denied because of the extensive finweb.

In porolepiform "rhipidistians" criteriasix and seven are met with respect to thepectoral fins only. Internal skeletal anatomyof porolepiform paired fins is virtually un-known, however (see Jarvik, 1972).

Dipnoans, on the other hand, meet all ofthe seven criteria (fig. 30A, B), many ofwhich were previously discussed by Holm-gren (1933). There are some minor points ofdisagreement between Holmgren's and ouranalysis which require some comment. Thesecond criterion is not evident in adult lung-fishes, though, because ontogenetically pairedsub-basal elements fuse to form a singlebroad cartilage (fig. 31). But there is no in-dication either in early or late developmentof a significant inequality in length of theseelements and the two sub-basal pelvic fincartilages fuse incompletely and show no in-dications of the divergent angle between thesecond pelvic metapterygial segment and thefirst preaxial radial that occurs in Eusthe-nopteron (figs. 28, 29). In early developmentof the pectoral fin, the radial associated withthe second metapterygial segment actuallycurves down toward that segment at its distalend (fig. 31; and see also Holmgren, 1933,fig. 11). It will be noticed by comparing fig-ures 30A and 30B of the pectoral and pelvicfins of Neoceratodus that the pectoral isstructurally the obverse of the pelvic, al-though the two fins are constructed accord-ing to exactly the same plan. In other words,the larger radials, which stand in a one-to-one relationship with the metapterygial seg-ments, are postaxial (dorsal) in the pectoraland preaxial (ventral) in the pelvic. Likewisethe radial that is functionally and ontogenet-ically associated with the second metapter-ygial segment is postaxial in the pectoral andpreaxial in the pelvic. Semon (1898, p. 68),

208 VOL. 167


/CLEITHRUM

ULNA + RADIUSHUMERUS

PREAXIAL RADIALS IN POSTAXIAL POSITION

ANOCLEITHRUM\

SCAPULOCORACOID

ABALL AND SOCKET JOINTS

POSTAXIAL RADIALS IN PREAXIAL POSITION

TIBIA POSTAXIAL RADIALS

FEMUR /FIBULA PREAXIAL RADIALS

FIG. 30. Neoceratodusforsteri (Krefft), left pectoral (A) and pelvic (B) fin endoskeletons in positionsof rest against body. AMNH 36982, 18 cm. total length.

as mentioned earlier, described the devel-opmental sequence at ontogenetic stage 45of how the pectoral anlage rotates dorsal-ward 90 degrees from the horizontal, and thepelvic anlage ventralward 90 degrees fromthe horizontal, so that the preaxial marginsof the two fins have a 180 degree differentorientation. Part of Holmgren's argument fordrawing a comparison between the twopaired appendages of Neoceratodus and uro-deles is his inference that the carpals and tar-sals and digits of urodeles are preaxial radialsnow in the postaxial position (presumablyalso requiring that in a lungfish-like ancestorof the urodeles the pelvic anlage has rotateddorsalward together with the pectoral). Hemade this inference for two reasons: (1) inNeoceratodus and urodeles the unpairedbasal and paired sub-basal elements of thepaired appendages (the first two metapter-ygial segments and the radial associated withthe second segment) in Neoceratodus andthe humerus (femur) and radius and ulna (tib-ia and fibula) of urodeles all appear in theprimary blastema of the limb endoskeleton

prior to the formation of the prochondral an-lage of the radials in Neoceratodus and car-pals (tarsals) and digits in urodeles; and (2)by identifying the segments of the primitivemetapterygial axis in the developing urodelelimb, one could discover whether the carpals(tarsals) and digits are pre- or postaxial andare therefore homologous with pre- or post-axial lungfish radials. Holmgren's argumentis complex, but reduces to a conclusion thatthe limbs of urodeles include both pre- andpostaxial radials. We believe that Holm-gren's goals are worthy but that his analysisis contrived, and that the problem can be re-duced to a set of simple comparisons.For our primary comparisons with Neo-

ceratodus we select the same urodeles as didHolmgren. We choose the hynobiid salaman-ders because the hynobiids are consideredto be among the most primitive of all liv-ing urodeles (e.g., in their reproductive meth-ods, arterial system, unfused lacrimal, sep-tomaxilla, angular and prearticular bones,size of columella and correspondence in po-sition to the lungfish hyomandibula in ad-

2091981


PREAXIAL SIDE INPOSTAXIAL POSITION

1 VULNA

JOINTS

:LAVICLE

FIG. 31. Neoceratodus forsteri (Krefft), leftpelvic fins are not yet present. AMNH 40801.

vanced larvae; Schmalhausen, 1968, pp.184-187, figs. 110-113). We choose hynobiidsalso because in the earliest stages of limbdevelopment the paired appendages appearas leaf-shaped fins with a strong muscularlobe as in larval Neoceratodus (fig. 32).We agree with Holmgren that the occur-rence in Neoceratodus and Hynobius ofan unpaired basal and paired sub-basal ele-ments in the primary blastema of the ap-pendicular endoskeleton prior to the for-mation of more distal elements is probablysignificant since in primitive gnathostomesthe fin endoskeleton first appears as a pro-cartilaginous plate continuous with the pri-mary girdle and later separates from thegirdle and then subdivides into radials(Goodrich, 1930). But what strikes us as be-ing a more complete, and therefore more in-formative, comparison is that the radial as-sociated with the second metapterygialsegment of Neoceratodus and the ulna ofHynobius and other urodeles are joined tothe basal segment (humerus) in the sameway, viz., mainly along the postaxial (dorsal)surface of the humerus, and that their pairedsub-basal segments (radius and ulna) form adefinite joint surface with the humerus notfound between more distal segments or inany other gnathostome pectoral endoskele-

pectoral fin endoskeleton of 15 mm. larva in which

ton (figs. 31, 33). The existence of this sec-ond joint in lungfishes is easily appreciatedin cleared and stained preparations of smallspecimens which show that the fin is cockedat this juncture and in large and small spec-imens by the presence of a ball and socketjoint surface between the first and secondmetapterygial segments (figs. 30A, 31).These observations therefore satisfy crite-rion number 3, above.The application of the fourth criterion,

whether the distal elements of the fin endo-skeleton are postaxial, depends on discov-ering the position in tetrapods of what re-mains, if anything, of the metapterygial finaxis. In Neoceratodus, the metapterygiumextends distally to the pointed tip of the fin.In developing Hynobius no skeletal elementextehds to the tip of the fin membrane, butinstead the first digit extends ventrally to thepreaxial margin and digits two to four (in thepectoral) or five (in the pelvic) extend dor-sally to the postaxial fin margin. Holmgrenassumed that three parts of the Hynobiuslimb are important for this question, the ra-dius (tibia), the ulna (fibula) and the centralcarpals (tarsals); the lateral carpals (tarsals)and digits he considered to be secondary ele-ments. He reasoned, since there are three

VOL. 167210


primary branches and, since the urodele limbmust have arisen phylogenetically from anarchipterygium with preaxial and postaxialradials on a central metapterygium, that themiddle of the three branches in Hynobiusmust be homologous with the dipnoan me-tapterygium. Holmgren's argument is clearlyloaded by the assumption of what is primaryand secondary for these comparisons and byhis assumption that each of the three ele-ments of a lungfish fin is represented in thetetrapod limb. Our own study of developingHynobius limbs suggests to us that (1) thereare only two morphological units, a ventralunbranched series of segments (radius + ra-diale + prepollex) and a dorsal branched se-ries (ulna + carpals + digits) (fig. 33); (2) theV'entral unbranched series is composed ofmore robust segments; (3) a line drawnthrough the axis of the ventral unbranchedseries intersects at or very near the pointedtip of the larval fin; (4) the base of thefirst element of the dorsal series (the ulna)has a dorsal contact with the humerus asdoes the dorsal radial in the primary fin blas-tema of Neoceratodus; and (5) the base ofthe first element of the ventral series (the ra-dius) has a socket for the ball joint of thehumerus as does the second metapterygialsegment in Neoceratodus. We conclude,as did Romer and Byrne (1934, p. 44),that the ventral unbranched series (radius +radiale + prepollex) represents the primi-tive metapterygium, and that the dorsalbranched series (ulna + carpals + digits) rep-resents preaxial radials that are in the rotatedpostaxial position. The synapomorphy thatwe therefore specify between Neoceratodusand Hynobius is the ontogenetically rotatedappendage with its preaxial radials on thepostaxial side, and this would satisfy crite-rion number 4, as given above. We must alsopoint out, however, if we are wrong andHolmgren is right, then our fourth criterion,restated in Holmgren's terms, would still re-late dipnoans and tetrapods and exclude Eus-thenopteron. We simply feel that our inter-pretation is more closely bound to simplemorphological comparisons and carries asmaller burden of assumptions.

Lungfish fins certainly satisfy the fifth cri-

FIG. 32. Hynobius chinensis Gunther, left pel-vic "fin" of 3 cm. larva. AMNH 30616. Positionsof anlagen of first four of five digits shown bydashed lines.

terion by being of similar form and construct-ed along virtually identical lines (althoughthey are 1800 out of alignment with each oth-er). The high number of axial elements inliving and many fossil dipnoans and the pres-ence of true postaxial radials are probablyderived, however, and are autapomorphousfor these species. In more primitive lungfish-es, such as Griphognathus, the muscular finbases are stout and neither so long nor soslender as in Neoceratodus (Schultze, 1969a,fig. 31).

Criteria 6 and 7 relate to features that arewell known in lungfishes and their signifi-cance is enhanced by the use of the fins, par-ticularly in Neoceratodus, for walking acrossa substrate by their alternate side-to-sidemotions in connection with sinusoidal mo-tions of the trunk and tail as in urodeles.The close similarity in paired appendages

between dipnoans and tetrapods does not im-ply, of course, that Eusthenopteron fins arewithout significance for this problem, butthat, cladistically speaking, their significancelies at a higher level of generality which spec-ifies only that Eusthenopteron is the sistertaxon to a group that includes dipnoans andtetrapods and perhaps also porolepiforms asthe sister taxon to those two.

Since osteolepiforms have a probablyprimitive number of metapterygial segmentsin both fins but fewer preaxial radials thanchondrichthyans, actinopterygians, dip-noans and tetrapods, we infer that osteolep-iforms exhibit a trend in radial reduction.Radial reduction is especially evident in thepectoral fin of Panderichthys (Vorobyeva,1975, fig. 3). Coelacanths also have a singleradial associated with the first two metapter-ygial segments, and these radials are reduced

1981 211


ULNA

GIRDLE

RAD IUS

RAD IUS

FIG. 33. Hynobius chinensis Gunther, left pectoral appendage of 3 cm. larva. AMNH 30616. Upperfigure, diagram of girdle and limb showing four digits clustered into two distinct developmental fields.Lower figure, detail of cartilaginous anlagen of segments of digits within the two fields. Ulna andprepollex are hypothesized to be homologues of fish metapterygium, and radius, carpals and digits,homologues of fish preaxial radials now in postaxial position (compare figs. 30, 31).

to tiny nubbins embedded in connective tis-sue (fig. 34). Such reduction suggests theirmembership in a group including osteolepi-forms. But contrasting with this interpreta-tion is the presence of four metapterygialsegments in the pelvic fin and the mirror-imagecomplete similarity of pelvic and pectoral en-doskeletons (fig. 34), features which are pre-sumably derived and suggest that coelacanthsfit near dipnoans and tetrapods. Also indica-tive of coelacanth-dipnoan-tetrapod relation-ship is the evident rotation of the preaxial sideof the coelacanth pectoral to the postaxial po-sition so that, like dipnoans, the pectoral andpelvic fins are out of phase. Millot and An-thony (1958) illustrated the pectoral fin ofLatimeria wrong way up in their illustrationof the skeleton, but their stated purpose wasto show its similarity to the fin of Eusthe-

nopteron in which the "humeral processes"for muscle attachment have the primitive os-teichthyan orientation (cf. figs. 28, 30, 34).Thus, whereas one character (radial reduc-tion) somewhat ambiguously aligns coela-canths with osteolepiforms, a second (fourpelvic radials) and a third (pectoral rotation)align them as a sister group of dipnoans plustetrapods. Although one might wish to arguethe exact cladistic placement of coelacanthsby using alternative interpretations of pairedfin characters, the placement of these fisheswith the sarcopterygians is unambiguous:coelacanths completely satisfy the first cri-terion, as do osteolepiforms, and, like Eus-thenopteron, they have entirely metapteryg-ial fins, thus rejecting theories of relationshipthat would place them as the sister groupeither of actinopterygians, of all other os-

212 VOL. 167


PREAXIAL SIDE INPOSTAXIAL POSITION

FIG. 34. Latimeria chalumnae Smith, left pectoral (above) and pelvic (below) fin endoskeletons ofunborn juvenile of 34 cm. total length (AMNH 32949) shown in positions of rest against body.

teichthyans (e.g., Wiley, 1979), or of chon-drichthyans (as suggested by some workerswith a penchant for treating shared primitivecharacters as evidence of relationship; cf.Forey, 1980). We also reject the conclusionof Bjerring (1973) that the relationship ofcoelacanths to actinopterygians and sarcop-terygians, or to chondrichthyans and os-teichthyans, is unresolved.

This discussion would be incomplete with-out some comment on the position of cla-distians within the Osteichthyes, since Po-lypterus, as the "type" crossopterygian, hasoften been compared with tetrapods (e.g.,Pollard, 1891, 1892), and Klaatsch (1896)even tried to relate its fin structure to thetetrapod limb. Polypterus, however, has thepropterygial margin of the pectoral finstrongly developed (fig. 35) which, within theOsteichthyes, is a characteristically acti-nopterygian feature.

Pelvic anatomy also contributes to an un-derstanding of cladistian relationships. Sten-sib (1921, 1925) was the first to recognize thatthe pelvic girdle of actinopterygians mayhave a unique ontogenetic history. He sug-gested that the girdle has a compound struc-ture, consisting of the pelvic plate to whichthe metapterygium is fused. We agree withStensio in substance if not in detail. Our ownconclusion that actinopterygians might nothave a primary pelvic girdle derives from acomparison of the endoskeletal fin supportsin gnathostomes generally. In acanthodiansthe only known skeletal support of the pelvicfin is a heavy spine along the fin's leadingedge. All other gnathostomes have a com-plex internal fin skeleton primitively withouta leading spine. In sharks the pelvic is struc-tured much like the pectoral; there are two,or sometimes three, rows of radials joined tothe preaxial side of the metapterygium (fig.

1981 213


FIG. 35. Polypterus ornatipinnis Boulenger,AMNH 40802, 14 cm. standard length.

26B). In primitive actinopterygians such asPolyodon (fig. 36A) and Acipenser (fig. 36B)(and some paleoniscids as well; Lehman,1966), there is a series of radial-like carti-lages that either fuse (in most adult forms) orremain separate (in juveniles and some adultforms). Two rows of radial elements arejoined to the outer surface of these internalcartilages, an inner row of elongate radialsand an outer row of shorter ones, the wholethus closely resembling a metapterygium andits preaxial radials in chondrichthyans andsarcopterygians; no other recognizably sep-arate structure lies anteromedial to these in-ternal cartilages. The various conditions inactinopterygians suggest that there is a trendtoward consolidation of the internal carti-lages both ontogenetically and phylogenet-ically, and a reduction in the length of theseries and number of radials it supports.Among the most primitive living and fossilactinopterygians (see Nielsen, 1942, 1949)varying degrees of reduction are already ap-parent (fig. 37A, B). Since Polypterus has asimple triangular girdle along which tworows of radials are arranged in parallel (an

left shoulder girdle and pectoral fin endoskeleton.

inner row of long radials and an outer row ofvery short ones) comparable with the preax-ial radials of chondrichthyans andosteichthyans, we conclude that the surfacealong which they attach is at least, in part,metapterygial (fig. 38).Goodrich (1930) objected to arguments

like those of Stensio's on the grounds thatthe pelvic girdles of actinopterygians arepierced by foramina for diazonal (=spinal)nerves as are the girdles in other gnatho-stomes, thus indicating a common plan ofdevelopment. These foramina are explainedby the fact that in most gnathostomes thegirdle is primarily outside the myotomes inearly development and is merely an exten-sion inwards from the base of the fin skeletonwith which it initially forms a continuous pro-cartilaginous plate. As the girdle grows inwardit comes to surround the nerves that passfrom the myotomes outwards to supply thefin. But exactly the same developmental his-tory would be expected even if a girdle werereduced, combined with the metapterygiumof the fin endoskeleton, or replaced entirelyby the metapterygium. In the latter case, the

VOL. 167214


B \

FIG. 36. Left pelvic girdle and fin endoskeleton in A, Polyodon spathula (Walbaum), juvenile, 12cm. total length (AMNH 40803); and B, Acipenser oxyrhynchus Mitchill, juvenile, 11 cm. total length(AMNH 37296).

metapterygium, in consolidating and enlarg-ing to form a support for the preaxial radials,would also come to surround one or morediazonal nerves.The alternative explanation, that the par-

allel series of radials lying outside the bodywall is a remnant of the metapterygium (i.e.,that they are not preaxial radials at all, butmetapterygial segments that have assumedthe form of such radials) seems to us con-trived, and is rejected by the observationthat in sturgeons the external metapterygialaxis is continuous with internal segmentedcartilage so that the metapterygial segments

appear to lie on both sides of the body wall(figs. 28B, 36B), and that, when the externalpart of the metapterygium is finally lost al-together in more advanced actinopterygians(see Lehman, 1966), all that is left is a seriesof parallel radials such as those found in Po-lypterus. We conclude, therefore, that thepelvic fin radials of Polypterus are of a de-rived, typically actinopterygian type and thatits girdle, which lies entirely between thebody wall and the myotomes and lacks thedorsal processes that are primitively presentin chondrichthyans (e.g., Cladoselache andchimaeras, fig. 27) and sarcopterygians, but

1981 215


FIG. 37. Pelvic girdle and fin endoskeleton in A, Boreosomus piveteaui Nielsen; find B, Pteronisculusgunnari (Nielsen), both from Nielsen (1942).

absent in chondrosteans, is also of actinop-terygian type.

(B) PECTORAL GIRDLE

Characters of the dermal bones of the pec-toral girdle have been used by several work-ers to support ideas of relationships between

FIG. 38. Polypterus ornatipinnis Boulenger,left pelvic girdle and fin endoskeleton. AMNH40802, 14 cm. standard length.

sarcopterygian groups and also to distinguishone group from another (Jarvik, 1944b, 1948,1972; Andrews and Westoll, 1970a, 1970b;Andrews 1973).

In osteolepiforms (the Rhizodontidae sen-su Andrews and Westoll are very poorlyknown arnd will be dealt with separately),porolepiforms and primitive lungfishes suchas Chirodipterus the girdle consists of fivepaired dermal bones arranged in an overlap-ping series (fig. 39C-G). A median interclav-icle is present ventrally. Primitive actinop-terygians show a series of four paired bonesplus a median interclavicle (fig. 39A, B; Jes-sen, 1972). We may assume therefore thatthe primitive osteichthyan had a girdle whichconsisted of, at least, a post-temporal, su-pracleithrum, cleithrum, clavicle and inter-clavicle and that successive elements in theseries overlapped one another from dorsal toventral. The main lateral line runs throughthe post-temporal and descends through thesupracleithrum before running through thebody scales. Polypterus is unusual in that the

VOL. 167216


4-4- '9 ~~~~~~4-EXTRACLEITHRUM

E FCH-CLAVICL~E-_

FIG. 39. Diagrams of dermal shoulder girdles (interclavicle omitted) of osteichthyans to show de-velopment of anocleithrum (actinopterygian postcleithrum) in sarcopterygians and transformation seriesin the increasing importance of clavicle and dorsal migration of fin insertion (arrow). Stipple distinguishesornamented postcleithrum and anocleithrum. A, Mimia toombsi Gardiner and Bartram (original, BMNHspecimens); B, Polypterus bichir Saint-Hilaire (from Jarvik, 1944b); C, Eusthenopteronfoordi Whiteaves(from Jarvik, 1944b); D, Rhizodus hibberti (Agassiz) (from Andrews and Westoll, 1970b); E, Holop-tychius sp. (from Jarvik, 1972); F, Latimeria chalumnae Smith (from Millot and Anthony, 1958); G,Chirodipterus australis Miles (original, BMNH P. 52560, P. 52586); H, Neoceratodus forsteri (Krefft)(from Wiedersheim, 1892).

sensory canal does not penetrate the supra-cleithrum but runs directly from the post-temporal to the first lateral line scale.The skull-girdle connection is provided by

the post-temporal which abuts against thehind wall of the braincase and is partiallyoverlain by the extrascapular series. In Re-cent actinopterygians and in Polypterusthere is a pronounced anteroventral processon the post-temporal which provides a firmunion with the braincase. This process is ab-sent from the post-temporal of Mimia andMoythomasia and therefore it is consideredas a synapomorphy of all actinopterygiansexcept those most primitive genera.The sarcopterygian girdle differs from that

of the actinopterygians in the incorporationof the anocleithrum which lies between andseparates the cleithrum and supracleithrum(where present) and is overlapped by both ofthese elements. The overlap relationships ofthe anocleithrum suggest that it is not a partof the primitive series in the osteichthyangirdle. The homologous bone in actinopte-rygians is the dorsalmost postcleithrumwhich lies slightly behind the supracleithrumand cleithrum, which are in contact, but isnevertheless overlapped by both of thesebones (fig. 39A, B; Jessen, 1972). Althoughthe postcleithrum may develop a ventral pro-cess in some neopterygians it remains fun-damentally scalelike, a condition well dis-

1981 217


played in the paleoniscid Mimia where atypical scale peg-and-socket is still present.The anocleithrum of Eusthenopteron (Jar-

vik, 1944b; fig. 3) shows a small ornamentedarea which lay superficially but most of thebone is represented as overlap surfaces. Theanocleithrum of porolepiforms is very poorlyknown but there are indications (Jarvik,1972, pl. 18, fig. 2) that in Glyptolepis it islarge and platelike and deeply embedded,bearing no ornament. In fact, it looks verysimilar to the anocleithrum of Chirodipterus(fig. 39G), both showing anterior processes.In Recent lungfishes and in coelacanths theanocleithrum is the most dorsal functionalelement in the girdle. It varies in shape froman ovoid plate in Neoceratodus (fig. 39H), tosplintlike in Latimeria (fig. 39F) and Protop-terus (Wiedersheim, 1892), while it is absentin Lepidosiren (Bridge, 1897).The incorporation of the anocleithrum/

postcleithrum as a functional unit in the gir-dle is a feature common to osteolepiforms,coelacanths, porolepiforms and lungfishes,and this stands in contrast to the scalelikepostcleithrum in actinopterygians. Jarvik(1944b, p. 27) took the view that, since theincorporation of an anocleithrum occurs inrhipidistians, then "it must be assumed tohave undergone a regressive development inActinopterygians and Brachiopterygians."In other words, according to Jarvik the an-ocleithrum condition is primitive for os-teichthyans. We would argue that the con-verse is more plausible; that is, theanocleithrum condition is derived and can beconsidered as a synapomorphy for sarcop-terygians, later lost in tetrapods. Which ofthe two character states is derived cannot bedecided simply on outgroup comparison,since the only other group with a well-de-veloped dermal shoulder girdle is the Placo-dermi. Despite attempts of Stensio (1959) tolist detailed homologies between the bonesof the trunk shield and the bones of theosteichthyan girdle the differences seem toogreat to allow any but the most tenuous com-parison. But even if we allow Stensio's sug-gestion that the placoderm posterior lateralis equivalent to the postcleithrum/anocleith-rum, then the condition in the primitive out-

group, where the posterior lateral lies whollybehind the supracleithrum and cleithrumequivalents, only reinforces the view that theactinopterygian postcleithrum exemplifiesthe primitive character state.

It does, in fact, seem more reasonable toassume that the superficial, scalelike post-cleithrum represents the primitive conditionand that the anocleithrum was developed bya forward migration and sinking inward ofthe postcleithrum. The stages in this hypoth-esis would be represented by Polypterus-Eusthenopteron-coelacanths, lungfishes, andpossibly porolepiforms. A similar hypothesiscan be suggested for another bone of theshoulder girdle, the interclavicle, which hasalso apparently sunk inward to become partof the teleostean urohyal (Patterson, 1977).The coelacanth interclavicle has also sunkinwards to come to lie between the urohyaland the clavicles.

Recent lungfishes have lost both the su-pracleithrum and the post-temporal and theformer element is absent from coelacanths(the anocleithrum is often mislabelled as thesupracleithrum). Furthermore, the anocleith-rum lies deep within the epaxial musculatureand this means that like tetrapods, whichhave lost the anocleithrum as well, there isno bony connection between the skull andthe girdle.5 The absence of such a connectioncannot be regarded as a synapomorphy ofthese three groups since at least Chirodip-terus retains a skull connection by way of apost-temporal and a supracleithrum (fig. 39G).

I Watson (1926) maintained that a skull-girdle con-nection exists in primitive tetrapods such as Pholider-peton and Eogyrinus attheyi; the former was said topossess a post-temporal and the latter a complete seriesof connecting bones. However, Romer (1957, p. 112)reinterpreted the post-temporal as a cleithrum of Phol-iderpeton and suggested that the supposed specimen(DMSW.34) of E. attheyi is in fact a fish. The specimenhas been subsequently reinterpreted as a rhizodont, pos-sibly Strepsodus, by Andrews (1973). At the very best,therefore, the evidence of a skull-girdle connection intetrapods by way of a post-temporal, supracleithrum,and anocleithrum is equivocal. Specialized secondaryconnection between the dorsal tip of the cleithrum andthe "tabular" horn in some nectrideans has been re-ported (Angela Milner, personal commun.).

VOL. 167218


The absence of a skull-girdle connection re-

mains a similarity between coelacanths andtetrapods but in the light of other synapo-

morphies between lungfishes and tetrapodsthis must be assumed to have developed inparallel, in these two groups and in Recentlungfishes. The skull-girdle connection is alsoabsent in some teleosts such as eels.The cleithrum and clavicle, the dominant

elements in the primitive dermal shouldergirdle, always have a complex interlockingsuture between them. In primitive membersof the Actinopterygii, Osteolepiformes, andPorolepiformes the clavicle bears an ascend-ing process which fits along the inner face ofthe cleithrum. In coelacanths and lungfishesthe dorsal part of the clavicle is not clearlydemarcated from the main body of the bonebut nevertheless it rises to wrap around theanterior margin of the cleithrum. It is there-fore misleading to consider that coelacanthslack an ascending process (cf. Jarvik, 1944b,p. 27).

Jarvik (1944) sought to distinguish a par-ticular type of double overlap between theclavicular ascending process and the cleith-rum in porolepiforms from a simple overlapof the clavicle by the cleithrum in osteolep-iforms (Eusthenopteron). But Andrews andWestoll (1970b, p. 407) rightly pointed outthat Osteolepis shows an essentially porolep-iform pattern and intermediate conditionsare seen in Rhizodus.The relative sizes of the cleithrum and

clavicle differ considerably between primi-tive members of osteichthyan groups. Atransformation series of the ratio of exposedsurface area of the cleithrum to exposed sur-face area of the clavicle would read: Polyp-terus, 7; Eusthenopteron, 3.7; Mimia, 3.4;Holoptychius and Latimeria, 1.5 (the lattercombines the area of the extracleithrum withthe cleithrum); Chirodipterus, 1.25. The ratioin a primitive tetrapod such as Archeria isapproximately 0.9, measured from Romer's(1957) restoration. The ratio in Neocerato-dus is 0.93, whereas in adult Protopterus noratio can be calculated since there is onlyone bone present which represents the resultof ontogenetic fusion between an embryolog-ically separate clavicle and cleithrum

(Schmalhausen, 1917). These figures dem-onstrate that there is a progressive increasein the size of the clavicle. On this basis, thetraditional position of Eusthenopteron as themost tetrapod-like fish cannot be supported.The position of the fin insertion follows a

similar sequence, showing a dorsal migrationof the point of fin articulation through theseries primitive actinopterygian-Eusthenop-teron-porolepiforms/lungfishes/coelacanths.It is difficult to quantify the relative positionsof fin insertion in fossil material, particularlyfor round-bodied sarcopterygians in whichpreservation nearly always distorts the trueposition.The rhizodontiforms (Sauripterus, Rhizo-

dus, and Strepsodus) are not our immediateconcern here but since this group is,recog-nized on distinctive features of the shouldergirdle (Andrews and Westoll, 1970b) it is ap-propriate to add a note. The surface area ra-tio between the cleithrum and clavicle is 2.2(Rhizodus); that is a value between Eusthe-nopteron and Holoptychius/Latimeria. Thefin insertion is high in Rhizodus, apparentlycomparable with that in Holoptychius, coel-acanths and lungfishes, but considerablylower in Sauripterus. So, on the basis ofshoulder girdle morphology, rhizodontiformscould be placed on our cladogram immedi-ately above Eusthenopteron.Our interpretation of the dermal shoulder

girdle can be summed up as follows. Theincorporation of the anocleithrum as a func-tional element in the girdle is a synapo-morphy of Eusthenopteron, Latimeria,Holoptychius, and dipnoans. We assume thisto have been lost, with the skull-girdle con-nection, in tetrapods. The ratio of cleithrumto clavicle area follows a similar sequenceexcept that it is not possible to discriminatebetween Latimeria, Holoptychius, and dip-noans. The high fin insertion is a synapo-morphy of Latimeria, Holoptychius, dip-noans, and tetrapods (paralleled by manyacanthopterygians and paracanthopterygiansamong teleost fishes). The anocleithrum ofcoelacanths, Holoptychius and dipnoans hassunk beneath the surface and lost all trace ofornament.The endoskeletal shoulder girdle sheds lit-

2191981


C LE ITHRUM

FIG. 40. "Screw-shaped" glenoid and scapulocoracoid in Eusthenopteron (A) and a Devonian lung-fish (B). Posterior view of scapulocoracoid of A, Eusthenopteron foordi Whiteaves (modified fromAndrews and Westoll, 1970a, fig. 4a, reversed); and B, Chirodipterus australis Miles (BMNH P. 52576).Note especially strong processes, presumably for muscle attachment, developed on scapulocoracoid ofChirodipterus. Andrews and Westoll (1970a) commented on several similarities between scapulocora-coid of Eusthenopteron and that of early tetrapods, including a screw-shaped glenoid. Features of sucha glenoid are that it is pear-shaped, with articular surface twisted, so that narrow lateral end facesdownward, and broad medial end faces upward. In the lungfish, the glenoid fossa is pear-shaped, as inEusthenopteron. It shows no transverse twist, and is much less thoroughly ossified than glenoid in largeindividual of Eusthenopteron figured by Andrews and Westoll (1970a, figs. 4, 6), where a twist is noteasy to detect. In living lungfishes, glenoid is a convex knob (figs. 30, 31), not a hollow as it is intetrapods (cf. Jarvik, 1968a, p. 228). In Gogo lungfishes (the only fossil lungfishes in which the endo-skeletal girdle is accessible) both glenoid and head of humerus are too weakly ossified for us to commenton the form of the joint, though figured specimen suggests that the glenoid was convex.

tle light on the interrelationships of the bonyfishes under discussion. Romer (1924) notedthe basic similarity in the construction of thescapulocoracoid in primitive actinopterygi-ans and primitive tetrapods, with Neocerat-odus differing in the possession of an unfen-estrated scapulocoracoid (the nerves, bloodvessels and muscles pass entirely outside theskeleton). Latimeria shows an unfenestratedscapulocoracoid similar to that of the Recentlungfishes.The scapulocoracoid of Eusthenopteron

has been favored (e.g., Andrews and Wes-toll, 1970a) over that of other bony fishes asbeing of tetrapod type because the glen-oid fossa is said to be screw-shaped as inearly tetrapods (Miner, 1925). The signifi-cance of this resemblance is, however,brought into question in the light of the struc-ture of the scapulocoracoid in the primitive

lungfish, Chirodipterus (fig. 40). Here thescapulocoracoid, which is attached to thecleithrum by three contact areas as inEusthenopteron, bears a large pear-shapedglenoid facet entirely comparable with thatin Eusthenopteron (Andrews and Westoll,1970a, fig. 4a). It seems to us therefore thatthe same tetrapod-like conditions (Romer,1924) seen in Eusthenopteron can bematched in Chirodipterus. The shoulder gir-dle of Chirodipterus is of further interest be-cause the scapulocoracoid is relatively muchlarger than that of Eusthenopteron and be-cause of the development of a broad trans-verse flange, developed along the anterioredge of the cleithrum (fig. 40). The rugosenature of the bone surface suggests that avery large muscle sheet, comparable with thedeltoides scapularis of tetrapods, inserted atthis point.

VOL. 167220


DERMAL BONES OF THE SKULL

Attempts to derive a useful nomenclaturefor the dermal bones of the skull involvesome of the most confused, and confusing,episodes in the history of comparative stud-ies of fishes and tetrapods. Today, consensuson bone homology is still lacking, as we tryto make evident in the following brief re-view.The nomenclature of the skull roof and

cheek bones in osteichthyans has for themost part been derived from that originallyapplied to tetrapods, in particular, man.Thus, the early anatomists confidently ho-mologized the frontals and parietals of mam-mals with similar bones in bony fishes. Thesehomologies were based on positional rela-tionships to other bones and underlying in-ternal structures, as well as embryonic de-velopment. Consequently, over severaldecades the terminology of the dermal bonesof advanced mammals was applied to rep-tiles, amphibians, and fishes (with the excep-tion of the Dipnoi-see later) without anyonedoubting that the bones were homologousthroughout these groups.Then in 1932 Save-Soderbergh proposed

that the dermal skull roof of fossil crossop-terygians and ichthyostegids had been de-rived from a common ancestor with twopairs of frontals and two pairs of parietalsand so concluded (his fig. 5) that the tetrapodparietal and post-parietal should be regardedas frontoparietal and parieto-extrascapular,respectively. He developed this terminologyfurther in his 1935 paper on labyrinthodontswhere he introduced a set of compoundnames for additional, inferred bone fusions,viz., lacrimo-maxillary, supratemporo-inter-temporal, supraorbito-dermosphenotic, etc.By the assumption of different fusion pat-terns, Save-Sbderbergh was then able to ex-plain the varying conditions of the skull roofseen in osteolepid fishes, ichthyostegids, andlabyrinthodonts.A year later Westoll (1936, p. 166) working

on Osteolepis suggested that the frontals andnasals of this fossil fish included homologuesof the tetrapod nasals, frontals, and pari-etals. Then in 1938, with the discovery of the

skull table of the Devonian fossil Elpistos-tege (now thought to be a panderichthyidfish), Westoll decided that the crossopteryg-ian frontal was homologous with the tetra-pod parietal and that both the classical ho-mologies and Save-Soderbergh's (1935)revision were misleading. He concluded(1938, p. 127) that the tetrapod names ofbones have largely been misapplied in fish-es, and suggested that the parietal of cros-sopterygians was homologous with the post-parietal of tetrapods and that the latter bonesdiminished in size passing from crossopte-rygian fishes to amphibians and to reptiles.He therefore imagined that there was a cor-responding backward movement of the pa-rietals.Romer (1941, 1945) subsequently champi-

oned Westoll's theory and today many Brit-ish and North American texts (Jollie, 1962;Moy-Thomas and Miles, 1971) and authors(Thomson, 1965; Parrington, 1967;. Andrews,1973) have adopted Westoll's terminologyand consider the bone previously named asfrontal in crossopterygians to be the parietaland the parietals to be post-parietals.The underlying reason for Westoll's radi-

cal change in terminology was his belief(1938, p. 128) that "the pineal foramen is ab-solutely homologous in position in allforms," a view with which Romer (1941, fig.1) agreed, stating that in the transition fromfish to amphibian to reptile there "has beenno shift in the pineal eye, but a gradual shiftin bone positions." In order to substantiatehis theory Westoll (1943) pointed out that theosteolepid frontal not only included the pi-neal foramen but also lay above the opticforamen, pituitary fossa, and basipterygeroidprocess, all of which he claimed were cov-ered by the parietal in early tetrapods. More-over, the parietals of osteolepids covered theotic capsule, hyomandibular facet, and theexits of the trigeminal and facial nerves, allof which in early tetrapods were covered bythe post-parietals.The pineal foramen may occupy different

positions in vertebrates (Jarvik, 1967, fig. 3)and even within the osteolepids it may be

1981 221


situated between the eyes (Osteolepis; Jar-vik, 1967, fig. 4a) or behind them (Eusthen-odon; Jarvik, 1967, fig. 4c). The foramenmay lie between the parietals, or between theparietals and frontals, or between the fron-tals, or even anterior to the latter (Powich-thys; Jessen, 1975, fig. 2). In Captorhinus(Parrington, 1967, fig. 3c) the frontals lieabove the basipterygoid process and the op-tic foramen, whereas in the urodele Molge(Reynolds, 1913) and in Sphenodon (Siive-S6derbergh, 1946) both this foramen and thehypophysis lie beneath the junction of fron-tals and parietals.The frontal/parietal controversy is associ-

ated with the problem of whether or nothomologues of the tetrapod tabulars andpost-parietals may be recognized in any oth-er group of osteichthyans. There are twopoints of view. The first owes its inceptionto Watson and Day's 1916 paper in which theauthors homologized the crossopterygianmedian extrascapular with the post-parietaland the lateral extrascapular with the tabular(their figs. 1, 5). These homologies were ac-cepted by Goodrich (1919, 1930), Schmal-hausen (1950, 1968) and most Russian,French, and Swedish workers but with Save-Soderbergh (1936, fig. 14) and Jarvik (1967,fig. 13B) later maintaining that the post-pa-rietal in temnospondyls represented a com-posite parieto-extrascapular and the tabulara supratemporo-extrascapular. The secondhypothesis (Westoll, 1938) is that the cros-sopterygian parietal is a post-parietal and theposterior supratemporal a tabular.Schmalhausen (1968, p. 82), in accepting

the theory of Watson and Day (1916), point-ed out that the lateral line furrows in fossilamphibians pass unchanged over the samebones as in fishes, whereas Westoll's (1938)hypothesis presumed that the fish extrascap-ular series had vanished. Schmalhausen(1968, p. 83) cited in support the works ofWatson (1926) and Bystrov (1935), in whichhe believed it had been demonstrated thatthe cephalic division of the main lateral line(infraorbital canal) joined the transverse su-pratemporal canal in the tabular. In none ofthe examples chosen by Schmalhausen,however (viz., Benthosaurus; Bystrov, 1935,

fig. 12: Trematosaurus; Bystrov, 1935, fig.25: Dolichopareias; Watson, 1926, fig. 24b)does the lateral line groove actually connectwith that of the supratemporal canal. More-over, in the majority of stegocephalians thegroove for the cephalic division of the lateralline terminates posteriorly in the supratem-poral and only in such forms as Metopo-saurus (Romer, 1937), Trematosaurus,Lyrocephalus, Lonchorhynchus, andDolichopareias (Bystrov, 1935) does the ca-nal leave a furrow on the anterior limits ofthe tabular. Further, in many temnospondylsthe supratemporal canal is confined to thetabular (Aphaneramma, Lyrocephalus; Siive-Soderbergh, 1935, figs. 31, 60, 61), whereasin others, where the furrow is continued ontothe post-parietal, it rarely if ever meets itscounterpart medially (Trematosaurus, Ben-thosaurus; Save-Soderbergh, 1937, figs. 4, 9).

In his criticism of Westoll's (1936) theory,Jarvik (1967, p. 186) found it difficult to imag-ine how the bones of the skull roof couldmove backward, independently of the un-derlying neural endocranium, yet in accept-ing in its place the theory of Watson and Day(1916) or its modification proposed by Save-<Soderbergh (1936) he demanded the re-verse-that a transverse series of bonesmoved forward onto the endocranium.

In osteichthyans many of the dermal bonesof the skull roof are tightly attached to theunderlying neural endocranium by ventraloutgrowths ofmembrane bone from their cen-ters of growth (descending laminae). The ex-trascapulars are clearly recognizable as atransverse series of scalelike bones variablein number and arrangement (4 in Polypterus,7 in Diplurus, 3 in Osteolepis and Dipterus)and never intimately connected with the neu-rocranium. In many fossil amphibians thetabulars and post-parietals show extensivedescending laminae (Benthosaurus; Bystrovand Efremov, 1940, fig. 23). And from thiswe must conclude that there is no unambig-uous way to recognize the homologues of thetetrapod post-parietal and tabular in the skullroof of fishes.Regarding lungfishes, Save-Soderbergh

(1932, pp. 97-98) commented that, in the

222 VOL. 167


FIG. 41. Skull roofs. A-C, Actinopterygians: A, Kentuckia deani (Eastman) (from Rayner, 1951);B, Moythomasia nitidus Gross (from Moy-Thomas and Miles, 1971); C, Pteronisculus magna (Nielsen)(from Nielsen, 1942). Cladistian: D, Polypterus sp. (from Schmalhausen, 1968). E-I, Rhipidistians: E,Porolepis brevis Jarvik (from Jarvik, 1972); F, Eusthenopteron foordi Whiteaves (from Jarvik, 1967);G, Eusthenopteron, composite, showing variations in bone patterns (from Jarvik, 1944a); H, composite,showing maximum number of separate bones observed in Osteolepis, Thursius, and Gyroptychius (fromJarvik, 1948); I, Holoptychius sp. (from Jarvik, 1972). Arrows show position of eye.

morphology of the skull roof, Dipterus ap-peared more nearly related to ichthyostegidsthan were crossopterygians. But most Brit-ish and American authors (White, 1965; Den-ison, 1968a, 1968b; Thomson and Campbell,

1971; Miles, 1975, 1977) have given up at-tempts at dermal bone homology and use in-stead a system of letters and numbers. Thiscustom stems from Romer (1936) who, find-ing it impossible to reconcile the two mu-

1981 223


A B C

F GFIG. 42. Skull roofs. A, tetrapod; B, D, E, F, G, dipnoans (all but E from Lehman, 1959); C,

actinistian. A, Ichthyostega (from Romer, 1966); B, Soederberghia groenlandica Lehman; C, Diplo-cercides kayseri (von Koenen), after Stensi6, showing condition intermediate between A, B, D-G(this figure) and fig. 41; D, Soederberghia sp.; E, Scaumenacia (various sources); F, Nielsenianordica Lehman; G, Oervigia nordica Lehman. Arrows show position of eye.

tually contradictory systems of Watson andDay (1916) and Goodrich (1925, 1930), aban-doned the use of names. He was subscribingto a view put forward by Gregory (1915, p.327) that the primitive dipnoan had a skullconsisting of a mosaic of small and variableplates. Forster-Cooper (1937) formulated asimilar nomenclature of letters and numberswhich was used by Westoll (1949), White(1965), and Thomson and Campbell (1971).

Disagreements and uncertainties in the useof bone names have several different causes.The most obvious source of disagreement isapplication of different criteria of homology

(e.g., bone fusion, relation of bones to pinealposition or to a part of the cephalic lateral-line canal system). Seemingly irresolvabledifferences have arisen when anatomistshave applied the same criterion (the lateralline) and yet obtained different results. A fur-ther difficulty has been that the evidence forchoosing among the different criteria, or forapplying any one criterion in specific cases,has been of a very indirect, circumstantialkind. The last, perhaps the major source ofdifficulty, is that the exposition of patternand process have been inextricably boundtogether so that assumptions about certain

VOL. 167224


QUADRATOJUGAL

FIG. 43. Cheek in A, Eusthenopteron foordi Whiteaves (after Jarvik, 1944a, fig. 18); B, Porolepisbrevis Jarvik (after Jarvik, 1972, fig. 43); C, Ichthyostega sp. (after Jarvik, 1952, fig. 35); D, Griphognathuswhitei Miles (after Miles, 1977, fig. 112). In A and C there are seven bones in the cheek plate, and thesame bone names can be applied. This pattern is usually treated as synapomorphous, uniting osteo-lepiforms with tetrapods. In B there are 11 bones in the cheek, and in D 14. Using A as a model, bonenames can be applied in B and D only by giving the same name to more than one bone (e.g., squamosals1-3 in B; preoperculars 1-6 in D). The different patterns in B and D must be autapomorphous orprimitive, given that the pattern in A and C is synapomorphous. That synapomorphy may be tested byconsidering other patterns-see figure 44.

processes have had the effect of denying pat-terns that were readily apparent to the ear-liest investigators of the problem. For ex-ample, if we assert that a pattern exists withrespect to the position of the eye in relationto the junction between two pairs of largebones on the skull roof, actinopterygians,cladistians, and "rhipidistians" form onegroup and primitive dipnoans, and tetra-pods form another (figs. 41, 42). But ifwe also assert that the bone pattern hasbeen more or less stable and that in earlydevelopment the eye can migrate anteriorlyor posteriorly, we imply that accepting thepattern is dependent upon accepting this de-velopmental explanation. Our confidence inthe significance of the pattern then comesinto question because we lack information onwhether one process (shifting of the eye inearly development) explains the pattern bet-ter than another process, say elongation ofthe snout, which carries the roofing bones

with it. The issue then degenerates into anirresolvable argument over process and thereality of the observed pattern becomes sec-ondary. Thus, Save-Soderbergh suggestedthat a pattern exists to unite osteolepids andtetrapods if one accepts a specific process ofbone fusion. Or, in the case of Westoll's as-sertion of pattern based on pineal position,one must accept a process of bones shiftingbackward, or, in the case of Jarvik's theory,their shifting forward. Jarvik's hypothesis ofpattern also involves the assumption of aprocess of interaction of the placodes of thecephalic lateral-line system and dermal boneformation such that the latter is invariablydependent on the former. There is, ofcourse, evidence that, in some instances,eyes can assume different relative positionsin the head (flatfishes), that bones can fusein ontogeny or fragment in phylogeny (in-fraorbital bones of characoid fishes), that thepineal maintains a stable relationship to cer-

1981 225


DERMOSPHENOTIC

CS PIRACU LARS

POSTORB ITALT

LACRIMAL

-ERUL A

OPERCULAR )

INF RAORB ITO -

MAX ILLAQUADRATOJUGAL

SUPRAMAXILLA SUBOPERCULAR

B

Sk B0SUBOPERCULAR

QUADRATOJUGAL?PR EO P EERC U LAR

FIG. 44. Cheek bones and operculum in coelacanth (A), an onychodont (B), a cladistian (C), and aDevonian actinopterygian (D). A, Rhabdoderma elegans (Newberry) (after Forey, in press); B, Struniuswalteri Jessen (after Jessen, 1966, fig. 6); C, Polypterus bichir Saint-Hilaire (after Daget, 1950, fig. 26);D, Cheirolepis trailli Agassiz (after Pearson and Westoll, 1979, fig. 20A). Bone names are those used byauthors cited. The pattern in Eusthenopteron and Ichthyostega (fig. 43A, C) is best matched in B(quadratojugal missing) and D (squamosal missing). The quadratojugal (present in C and D) is apparentlya primitive osteichthyan feature, lost or not found in onychodonts (Andrews, 1973, p. 146); and thesquamosal is a primitive sarcopterygian feature (present in A, B and fig. 43). The features commonto the cheek plate of Eusthenopteron and Ichthyostega (fig. 43A, C) seem therefore to be primitivefor sarcopterygians; the extra bones in porolepiforms (fig. 43B) and dipnoans (fig. 43D) are autapo-morphous, as is the absence of maxilla and quadratojugal in coelacanths (A).

tain bones (some tetrapods), and that theneuromasts can influence dermal bone on-togeny or fail to do so. But the occasions onwhich we should call forth these various"6explanations" are unknown, especially inDevonian fishes and tetrapods where exper-imental investigation is not possible. As wesee the problem, the patterns suggested bySave-Soderbergh, Westoll, and Jarvik arenot actually observable and can be said toexist only if one first assumes the reality andgeneral applicability of a process of ques-tionable relevance to the problem at hand.We therefore eschew such lines of argumentand suggest, instead, that there are three ob-servable patterns that can be recognized re-

gardless of what causal agents might be in-voked to explain them. Two of these patternsconcern dermal roofing bones, and, one, thecephalic lateral-line system. The first is theposition of the eye in retation to the largepaired bones on the skull roof which, as men-tioned above, groups dipnoans and tetrapods.The second is a cluster of five bones in theoccipital region of the skull of dipnoans andIchthyostega. And the third is the interrup-tion of the infraorbital canal near the incur-rent (external) nostril in dipnoans and tetra-pods (see p. 196). An interpretation of thedermal cheek bones is suggested in figs. 43,44.

226 VOL. 167


PALATE AND JAW SUSPENSION

(A) HISTORICAL

The earliest comparisons of Recent lung-fishes with Recent amphibians, particularlyurodeles, noted an obvious similarity in thepalates and braincases (Huxley, 1876; Cope,1885). But the late nineteenth century searchfor amphibian ancestors meant that the manyderived features of dipnoans excluded thesefishes from immediate relationship with tet-rapods (see Historical Survey) and placed,in our view, unwarranted emphasis on thefossils. It was reasoned that lungfishes couldnot be ancestral to Recent Amphibia be-cause, among other features, they lack amaxillary arch (Cope, 1885) and the dentitionis too highly specialized (Baur, 1896; Greg-ory, 1915). Pollard (1891, p. 343) argued that". .. Presuming the Dipnoi to be in the an-cestral line of the Urodela such [autostylysensu Huxley, 1876] should be the conditionin young and primitive Urodeles. Such is notthe case." He believed, like Jarvik (1968a),that lungfishes were close relatives of holo-cephalans. Kingsley (1892, p. 679) main-tained that "the. autostylic condition [inurodeles] arises comparatively late in devel-opment, and never attains the completenesswhich a Dipnoan ancestry would require."The embryology of lungfishes was not knownto either Kingsley or Pollard so this view ex-presses the same opinion as that given byDollo (1896), that the dipnoans are more au-tostylic than urodeles and cannot thereforebe considered ancestral to Recent amphibi-ans (see also Schmalhausen, 1968, and pp.233-234 below).

Pollard (1892) went on to suggest a schemeof homology between the palatal bones ofPolypterus (then considered as a crossopte-rygian) and those of a frog.The paleontological arguments for dis-

missing lungfish as amphibian ancestorswere taken up by Watson from Baur (1896)and Cope (1892). Watson (1912, p. 10) com-pared the palate of Megalichthys with that ofthe temnospondylous amphibians Loxommaand Eogyrinus. He concluded that not only

was the arrangement of the bones of the pal-ate very similar but also that "the ordinaryidea of the autostylism of the Tetrapoda isincorrect in postulating a connection be-tween the palato-quadrate cartilage and theotic region. It is, I think, quite certain thatthere never was such a connection in primi-tive forms, except through the dermal bonesof the temporal region."That is still the current view (Kesteven,

1950) and may be summed up as follows:crossopterygians (with or without Polypter-us), rather than lungfishes, are related to am-phibians because of similarities in palataldentition, a closer correspondence with fos-sil amphibians6 in jaw suspension and thepersistence of palatal kineticism, and be-cause it is easy to draw homologies betweenthe dermal bones of the palate of a crossop-terygian and those of both Recent and fossilamphibians.

Yet, Dollo (1896), Kesteven (1950), andThomson and Campbell (1971) recognizedthat the tritoral dentition and the absence ofmaxillae, dermopalatines, and ectopterygoidsin Recent lungfishes are simply unique dip-noan specializations and are therefore irrele-vant for establishing their relationships.

(B) DERMAL BONES OF PALATE

We are concerned here with the presenceor absence and the pattern formed by thepaired pterygoid (entopterygoid or endopter-ygoid), ectopterygoid (transpalatine or trans-verse of tetrapods), dermometapterygoid(dermal metapterygoid), and the dermopala-tine (palatine of tetrapods) which are asso-ciated with the palatoquadrate, and the vo-mer (prevomer) and the median parasphenoidwhich are associated with the neurocranium(figs. 45, 46). A variety of names has beenused for the bones of the dermal palate; these

6 For our purposes in this section we take fossil am-phibians to include ichthyostegids, lepospondyls, tem-nospondyls, and anthracosaurs. It is unlikely that theyform a monophyletic group (Gaffney, 1979a).

1981 227


ENDOPTERYGOID

A

6 ( QUADRATE D

FIG. 45. Palates in ventral view and suggested homology of dermal bones and their relation topalatoquadrate. Actinopterygian: A, Mimia toombsi Gardiner and Bartram (BMNH specimens). Cla-distian: B, Polypterus bichir Saint-Hilaire (BMNH specimens). Rhipidistians: C, Eusthenopteron foordiWhiteaves (from Jarvik, 1954); D, Glyptolepis groenlandica Jarvik (from Jarvik, 1972). Compare fig-ure 46. Code 2 refers to dermopalatine in all figures.

are listed in table 2. The dermal palatal bonesare thought to have been derived from con-densations within a uniform lining of denti-cles within the buccal cavity (Nelson, 1970).We regard the presence of such tooth-bear-ing bones (except the parasphenoid, presentin placoderms) as a synapomorphy of os-teichthyans.Many authors have stressed the close sim-

ilarity in number and arrangement of thebones between the palate of Eusthenopteronand primitive amphibians. In both (figs. 45C,46B, C), the parasphenoid is flanked by alarge pterygoid which extends from the pos-terolateral contact with the quadrate to thevomer. The dermopalatine and ectoptery-goid are sutured to the lateral edge of thepterygoid but a considerable portion of the

228 VOL. 167


2--

C

D (PE

FIG. 46. Palates in ventral view and suggested homology of dermal bones and their relation topalatoquadrate. Dipnoan: A, Griphognathus whitei Miles (from Miles, 1977). Tetrapods: B,Ichthyostega(from Romer, 1966); C, Eogyrinus attheyi Watson (from Panchen, 1972a); D, Benthosuchus sushkiniEfremov (from Bystrov and Efremov, 1940); E, Tylototriton verrucosus Riese (from Noble, 1931). Com-pare figure 45 for bone names.

latter extends posteriorly beyond the ecto-pterygoid. The dermopalatine and ectopte-rygoid lie in series with the vomer and aresecurely attached to the maxilla laterally.In fossil amphibians there is no possibil-ity of movement between the dermopala-tine, ectopterygoid and maxilla. The choana,or its presumed homologue in Eusthenop-teron, is bordered medially by the vomerand dermopalatine. Achoanate Latimeriais very similar except that there are threebones in series in place of the dermopalatine

and ectopterygoid of Eusthenopteron andprimitive amphibians. This difference is ofunknown significance since actinopterygianshave a variable number of bones in thisposition (below). Also, fossil amphibianshave been described with two ectopterygoids(Megalocephalus, Watson, 1926) and somehave a single bone occupying the territoryof the vomer and dermopalatine [Tremato-saurus brauni (Owen)-Watson, 1919; Ba-trachosuchus-Sive-S6derbergh, 1935]. Thepalate of anurans and urodeles (fig. 46E)

1981 229

1I


differs from that described above chieflyin the development of large palatal vacui-ties and absence of the ectopterygoid. Thepalatal vacuities are present in more ad-vanced temnospondyls; they are developedbetween the pterygoid-dermopalatine andthe parasphenoid, in contrast to those foundin amniotes between the pterygoids and themaxilla-jugal. Palatal vacuities and absenceof the ectopterygoid are both derived fea-tures within certain tetrapod subgroups.There are two important differences be-

tween the palatal structure of crossopteryg-ians (including coelacanths) and tetrapods.In crossopterygians the pterygoids do notmeet and suture with one another in the mid-line and the parasphenoid ends posteriorly atthe ventral fissure, as in primitive actinopte-rygians (fig. 45). This contrasts with mediallyunited pterygoids and a long parasphenoid intetrapods (we are uncertain about the con-dition of the parasphenoid in Ichthyostega;see Jarvik, 1952).

In all dipnoans, like tetrapods, the ptery-goids and vomers meet in the midline andthe parasphenoid is long.7 Kesteven (1950)noted the close similarity in shape and ar-rangement of the vomers and pterygoids be-tween Dipterus and Ichthyostega and wenton to say (p. 125) ". . . The essential simi-

7 Three Lower Devonian lungfishes have been de-scribed as having short ("stalkless") parasphenoids andthis is presumed primitive for dipnoans (Miles, 1977).The precise condition of the parasphenoid in each ofthese forms is, however, uncertain. For Dipnorhynchuslehmanni Lehman and Westoll (1952, p. 410) stated that"The details, however, are not well shown, and a very

small posterior expansion may be present." In Dipno-rhynchus sussmilchi there is no separate parasphenoid(Thomson and Campbell, 1971) although it is presumedto have fused with the pterygoids (Miles, 1977). Theparasphenoid of Uranolophus has a large toothed area,

behind which is a smooth flange, broken posteriorly(Denison, 1968b, figs. 8, 9). Here too, there is doubtabout the posterior limit of the parasphenoid. The cru-

cial fact is, not so much whether there is a "stalk" or

not, but whether the parasphenoid has grown back suf-ficiently to occlude the ventral cranial fissue, as it hasin all other dipnoans. The relative position of the ventralfissue and parasphenoid is unknown in both D. lehmanniand Uranolophus.

larity of these bones is only partly disguisedby the peculiar teeth on those of Dipnoans,and this difference in the teeth should not bedeemed of phylogenetic importance becausewe observe marked differences in the teethof relatively closely related Elasmobranchs."Since Kesteven wrote that sentence we havediscovered a great deal about the primitivelungfish palate (Denison, 1968a, 1968b, 1974;Thomson and Campbell, 1971; Miles, 1977).Our current knowledge strengthens the cor-rectness of Kesteven's view. The specializeddentition of the Recent lungfishes has clearlyarisen within the group (Denison, 1974) andthere are several genera (Griphognathus,Uranolophus, Holodipterus, Fleurantia andConchopoma) in which the palate is coveredby a shagreen of tiny teeth. The pterygoidsare also primitively long (Miles, 1977, p. 173)as in crossopterygians and primitive tetra-pods. Griphognathus whitei Miles showslong pterygoids which, in some specimens(e.g., BMNH P. 56054) are flanked anteriorlyby two paired bones identified here as vomerand dermopalatine (figs. 7A, 46A). They areequivalent to dermopalatine 1 and 2 of Miles.The vomers meet in the midline anteriorly andthe dermopalatine forms the medial margin ofthe choana. There is a third paired palatal bonewhich lies anterolateral to the vomer (der-mopalatine 3 of Miles; "extra" dermal bone,fig. 7). This could be interpreted as a der-mopalatine, in which case the bone we havelabeled by that name would be interpreted asthe ectopterygoid. Be that as it may, there aremarginal palatal bones and at least the poste-rior member of this series is firmly sutured toa maxilla (see p. 181). It is possible that Hol-odipterus gogoensis also shows a separatepalatine and ectopterygoid since there areseveral small tooth plates in this area (Miles,1977, fig. 72). We do, however, agree withMiles that it is very difficult to restore theseplates to their mutual position and theyequally well might represent fragments of themaxilla or premaxilla as seen in Griphogna-thus (fig. 7B).

In Griphognathus therefore, we can seethe same so-called special similarities be-tween the crossopterygian palate and the pal-ate of a primitive tetrapod. The lungfishes

230 VOL. 167


TABLE 2Synonymy of Bone Names for Dermal Palate(Those in brackets are less commonly used.)

Actinopterygians CrossopterygiansThis paper + and

(figs. 45, 46) Polypterus Lungfishes Tetrapods

Dermopalatine (2) dermopalatine dermopalatine palatineEctopterygoid (3) posterior dermopalatine ectopterygoid transpalatine (transverse,

ectopterygoid)Pterygoid (4) ectopterygoid entopterygoid pterygoid (entopterygoid)

(endopterygoid)Dermometapterygoid (8) dermometapterygoid -

(dermal metapterygoid)Endopterygoid (7) endopterygoidVomer (1) vomer (prevomer) vomer (prevomer) prevomer (vomer)

show an additional feature of tetrapods,namely the medially united pterygoids.

In porolepiforms (fig. 45D) the palate, asrestored by Jarvik (1972), is similar to thatof osteolepiforms. The vomers do not appar-ently meet in the midline, in at least Poro-lepis or Glyptolepis, and this is a reflectionof the broad snout of porolepiforms. Unlikedipnoans and tetrapods the pterygoids do notmeet in the midline.The vertebrates dealt with so far show a

similar number and arrangement of palataldermal bones to which a consistent systemof names can be applied. The varied nameswhich have been used resulted from the ap-plication of names coined for actinopterygi-ans or, in the case of the prevomer, from apresumed difference in mammals.8 We em-

8 In non-mammalian vertebrates the bones which areclearly homologues of fish vomers are often called pre-vomers (Watson, 1926; de Beer, 1937) following a rec-ommendation by Broom (1895). The history of the prob-lem has been outlined by de Beer (1937, pp. 433-435)and Parrington and Westoll (1940). Broom noted thatthe premaxilla of several primitive mammals developedas two parts; a marginal tooth bearing portion and apalatine process. The latter remains separate in Or-nithorhynchus (the dumbbell bone) and apparently fusesvery late in ontogeny in several others (e.g., Erinaceus,Tatusia). Since this separate process lay anterior to themammalian vomer Broom called this the prevomer.Broom considered this palatine process to be the topo-graphic homologue of the vomer of non-mammalian ver-tebrates. However, Presley and Steel (1978) have re-

ploy terminology commonly used for fishes(table 2).

Actinopterygians and Polypterus show arather different arrangement of dermal boneson the palate: some homologies can bedrawn between these and the palatal bonesdiscussed above but this group has additionalelements. The Gogo paleoniscid Mimiashows what we believe to be the primitivecondition for the group. There is a pairedvomer9 followed by two series (fig. 45A); anouter row of dermopalatines and an innerrow of endopterygoid followed by one ormore dermometapterygoids. The dermomet-

viewed this problem with particular reference toOrnithorhynchus and showed that the palatine process(prevomer) is ontogenetically part of the premaxilla.They concluded (p. 104) "There is therefore in no mam-mal an additional pair of elements between the vomerand the premaxilla, and thus the vomer(s) in all mam-mals occupy an homologous position to those of rep-tiles" [italics theirs].

9 A special problem surrounds the identification ofthe vomer of Polypterus. Several authors identify thevomer as that bone here called dermopalatine (2, fig.45B). It resembles the vomer of Griphognathus (cf.Miles, 1977, fig. 81; Allis, 1922, fig. 25) and lies closebehind and parallel to the premaxilla. It is here identifiedas a dermopalatine because it is associated with the pal-ate in disarticulated skulls (a true vomer is alwaysclosely attached to the floor of the nasal capsule) andbecause in early ontogenetic stages of Polypterus anidentifiable vomer is recorded (Holmgren and Stensi6,1936; Pehrson, 1947).

1981 231


apterygoid is a primitive characteristic ofthe group Polypterus + actinopterygians.One or several are found in paleoniscids(Watson, 1925; Nielsen, 1942, 1949). Polyp-terus is the only living fish with a (single)dermometapterygoid.The endopterygoid lies anterior to the der-

mometapterygoid and traditionally is thebone considered homologous with the pter-ygoid (sometimes called entopterygoid) ofrhipidistians and coelacanths, lungfishes,and tetrapods. However, there is reason tobelieve that the endopterygoid is unique toPolypterus and actinopterygians and that thehomologue of the pterygoid is the bonecalled ectopterygoid in actinopterygians andPolypterus.The ectopterygoid of actinopterygians and

Polypterus (bone 4; fig. 45B) is the posteriormember of a series with the dermopalatines towhich it is related embryologically (Pehrson,1922, 1947). It reaches back to the quadrate,borders the adductor fossa and primitivelycontacts the maxilla anterolaterally. These areprecisely the relationships of the pterygoidof Eusthenopteron (Jarvik, 1954), porolepi-forms (Jarvik, 1972), lungfishes (Miles,1977), urodeles (Goodrich, 1930), fossil am-phibians (e.g., Panchen, 1964; Beaumont,1977; Save-Soderbergh, 1935, 1936), andLatimeria (Millot and Anthony, 1958) al-though in the last a maxilla is absent. In thepterygoid of all these forms and in the actin-opterygian ectopterygoid, the ossificationcenter lies near the ventral (lateral) margin. Incontrast, the ossification center of the endop-terygoid lies near the dorsal or medial margin.The pterygoid of sarcopterygians is very

large and reaches along the medial edge ofthe more anterior members of the serieswhere it may overlap the ectopterygoid(coelacanths) or the ectopterygoid and thedermopalatine (Eusthenopteron, porolepi-forms, Ichthyostega and many temnospon-dyls, and anthracosaurs; similar overlaps areshown by the ectopterygoid of Polypterusand actinopterygians such as Lepisosteus,Amia, and Saurichthys).

It is difficult to decide if the endopterygoidand\ dermometapterygoid are synapomor-

phies of Polypterus + actinopterygians fortwo reasons. First, a dermometapterygoidhas been figured in a specimen of Eusthen-opteron (BMNH P. 6807, Woodward, 1922).We have examined this specimen and remainundecided whether or not there is a separatebone. Second, if we assume that the groupPolypterus + actinopterygians are the mostprimitive osteichthyans and if we also as-sume that the dermal skeleton has taken anassimilative route (Nelson, 1970) then theextra bones in primitive actinopterygianscould be the primitive osteichthyan condi-tion. If so, the absence of an endopterygoidand dermometapterygoid are sarcopterygiansynapomorphies.

(C) PALATOQUADRATEThe gnathostome palatoquadrate cartilage

primitively ossifies from three centers; ananterior autopalatine; a posteroventral quad-rate at the lower jaw articulation; and a dor-sal metapterygoid (epipterygoid of tetrapods)at the basal articulation. This primitive con-dition may be seen in diverse fishes: Acan-thodes (Miles, 1973) where these ossifica-tions are perichondral only; juvenile Mimia,Moythomasia, some specimens of Pteronis-culus (Nielsen, 1942), and Acropholis (Aldin-ger, 1937), many teleosts, and coelacanths.In osteolepiforms and porolepiforms, as, inlarge specimens of the actinopterygiansMimia, Moythomasia, and Pteronisculus, thepalatoquadrate is ossified throughout as a sin-gle bone. Our knowledge of the growth stagesof the actinopterygians mentioned and thecondition in coelacanths suggests that thepalatoquadrate of rhipidistians also ossifiedfrom three centers representing sites of artic-ulation between the palate and the braincaseand lower jaw.

Reduction or loss of ossification centers istherefore regarded as a derived condition.Thus, among actinopterygians, Lepisosteusand some teleosts have lost the autopalatine(Patterson, 1973) and only the quadrateseems to have survived in Birgeria (Nielsen,1949). Polypterus has lost the metapterygoid(Allis, 1922).

VOL. 167232


The palatoquadrate of Neoceratodus con-tains only a quadrate ossification, as in frogsand urodeles, and even this is absent in Lep-idosiren and Polypterus. Fossil lungfishesappear to show more extensive ossificationof the palatoquadrate in which it is not pos-sible to recognize centers of ossification(Miles, 1977, p. 24) but, since an ascendingprocess is well ossified in many of these lung-fishes, we assume that at least a metaptery-goid center was present. Temnospondyls(Bystrov and Efremov, 1940), anthracosaurs(Panchen, 1964, 1970), turtles, lizards, andSphenodon have both a quadrate and a met-apterygoid (epipterygoid).The picture which emerges is that choa-

nates (lungfishes and tetrapods) have the pal-atoquadrate foreshortened anteriorly so thatit no longer contacts the ethmoid region.This is associated with the loss of an auto-palatine. Loss of an autopalatine is also seenin some actinopterygians but it is never as-sociated with a foreshortening of the pala-toquadrate.

(D) RELATION BETWEEN PALATEAND BRAINCASE

The similarity between lungfishes and am-phibians in the relation between the palateand the braincase was pointed out long agoby Huxley (1876) and Cope (1885), and morerecently by Kesteven (1950) and Fox (1965).This type of jaw suspension was designatedautostylic by Huxley (1876, p. 40) who notedthat autostyly also occurred in holocepha-lans and Recent agnathans. These authorsregarded the union of palate with braincaseas evidence of relationship between lung-fishes and amphibians. Other authors haveregarded autostyly as evidence of a lung-fish-holocephalan relationship (Jarvik, 1968a;Edgeworth, 1935). Jarvik's view can befalsified by several other characters whichsuggest that elasmobranchs and holocepha-lans are sister groups (Schaeffer and Wil-liams, 1977). The "autostyly" of holoceph-alans differs considerably from that oflungfishes which, in turn, is very similar tothat of amphibians.The contact between the palate and the

braincase of dipnoans and lissamphibians is

effected by three processes: basal,10 ascend-ing and otic (defined by de Beer, 1937) andthe anterior projection of the palatoquadratebeneath the neurocranium (pterygoid pro-cess) is reduced or absent (Fox, 1965; Bert-mar, 1968). The relationships of the nervesand blood vessels to these processes and tothe extracranial cavities which they enclose(cavum epiptericum and cranioquadrate pas-sage) in the two groups and the form and sizeof these processes are very similar (Good-rich, 1930; de Beer, 1937; Fox, 1965, figs. 21,34, 38). The otic and ascending processes arefused to the neurocranium; the basal processis fused in dipnoans and some urodeles, butnot in anurans and apodans. The hyoman-dibular is reduced in comparison with thefirst epibranchial and takes no part in jawsuspension.The autostyly of holocephalans is differ-

ent, a fact which led Gregory (1904) to pro-pose the term holostyly for these fishes. Thepalatoquadrate is attached throughout itslength from the nasal capsule to the occiput,there is no ascending process and the iden-tification of the otic process is questionable(Goodrich, 1930, fig. 444). The hyomandib-ular takes no part in jaw suspension but isunreduced. With Gregory we agree there islittle ground for comparison between the au-tostyly of holocephalans and dipnoans + tet-rapods and accept that fusion of the palatewith the braincase has occurred convergent-ly.

Consideration of the palate/braincase re-lationship in primitive amphibians and incrossopterygians has profoundly favoredcrossopterygians over lungfishes as tetrapodrelatives since it is believed that the primitiveamphibian palate is kinetic and that thiscould only have been inherited from a cros-sopterygian ancestor. It is assumed that thepalate of primitive amphibians is movable on

10 There has been considerable argument (see deBeer, 1937, pp. 207, 214; Jarvik, 1954) as to whetherthere is a true basal process in anurans (except Asca-phus), since the relation of the palatine nerve to theprocess is rather different to that in dipnoans and uro-deles. This is of little significance to the arguments heresince, if one wishes to argue for non-homology thiswould simply be an autapomorphy of anurans.

1981 233

BULL,ETIN AMERICAN MUSEUM OF NATURAL HISTORY

the braincase (e.g., Watson, 1912; Panchen,1970; Beaumont, 1977). The palate is cer-tainly movable in Latimeria (Thomson, 1966)and this was probably true of Eusthenopte-ron and porolepiforms. In fact, in Eusthe-nopteron, there is a very complex and, in ourview, movable joint between the dermopal-atine and the vomer (fig. 14). The palate andectopterygoid are keyed to the maxilla andmust have moved with it. The condition ofour Eusthenopteron specimen suggests thatthe maxilla, palate, and the circumorbitalbones moved as a unit against the snout:there are distinct sliding joints between thelacrimal and the supraorbito-tectal/lateralrostral.A kinetic palate in primitive amphibians is

more difficult to believe. Anteriorly the skullis very solid, with dermopalatine, ectopter-ygoid, vomers, and pterygoids firmly keyedto one another and to the cheek bones, andthe cheek bones are keyed to the skull roof-ing bones, at least to those lying anterior tothe pineal. There is no evidence of slidingjoints. So, with the palate firmly fixed ante-riorly the kineticism must be located poste-riorly, at the basal articulation, or at thepoint where the ascending process contactsthe braincase. Ichthyostega may or may notretain the intracranial joint but kinesis at thispoint is apparently impossible because of''a remarkably precocious consolidation ofthe skull roof' (Panchen, 1972b, p. 78). Thecomments of two amphibian workers arepertinent at this point. Panchen (1970), writ-ing about anthracosaurs, denies the exis-tence of kineticism in temnospondyls (prim-itive relatives of lissamphibians, Gaffney,1979b). Certainly the members of this groupgenerally accepted as advanced, the stereo-spondyls, have the palate firmly fixed at thebasal articulation (Siive-Soderbergh, 1935,1936). But according to Beaumont (1977) themore primitive loxommatids have palatal ki-neticism which to us requires bone-bending.The bending would take place in the trans-verse plane. In anthracosaurs (Panchen,1970) a gap between the squamosal and thesupratemporal is thought to enable the palateto move up and down but since the palate isfixed anteriorly this may mean no more than

bone-bending in the longitudinal plane.These remarks are not intended to be face-tious but there does seem to be a case forquestioning the existence of palatal kineti-cism in primitive amphibians. Perhaps therewas some movement but the alleged remnantof kineticism was not of the crossopterygiankind.To sum up this section on the palate and

jaw suspension we propose the followingoutline of character phylogeny. The primi-tive gnathostome condition is an amphistylicjaw suspension with a fully developed sus-pensory hyomandibula and a palate whicharticulates with the braincase at the basalarticulation and rises dorsally from this pointas a cleaver-shaped process (Schaeffer,1975). The palatoquadrate of either sidecurves inward anteriorly and meets its anti-mere beneath the nasal capsules. The pala-toquadrate ossifies from three centers and iscovered with a shagreen of teeth; a vomer isabsent. These conditions are met in Acan-thodes and in primitive sharks, except thatsharks have an unossified palatoquadrate(derived sharks show hyostyly and rays eu-hyostyly; holocephalans are holostylic).Synapomorphies of osteichthyans are: pala-toquadrate of either side not meeting in themidline anteriorly (chondrosteans parallelmore plesiomorph groups in this respect) butarticulating with the postnasal wall; pairedvomer, dermopalatine, ectopterygoid andpterygoid tooth plates present (actinopte-rygians + Polypterus show additional toothplates and methyostyly is developed withinthe group). Sarcopterygians show an articu-lation between the metapterygoid (epiptery-goid) and the dorsal portion of the orbitotem-poral region of the braincase, hyomandibularstraddling the jugular canal, pterygoid en-larged and forming the chief bone of the pal-ate. Choanates show a reduced hyomandib-ula playing no direct part in jaw suspension,absence of an autopalatine associated withan anterior reduction of the palatoquadrate,an immobile dermopalatine-snout contact,pterygoids meeting in the midline, and au-tostyly (but not necessarily involving fusionof bone as opposed to cartilage fusion).

VOL. 167234


HYOID AND GILL ARCHESIn comparing the gill arches of the main

groups of gnathostomes we have accepted,in part, Nelson's (1968), rather than Miles's(1964, 1965) and Moy-Thomas and Miles's(1971), interpretation of the branchial skele-ton ofAcanthodes, and we have accepted Jar-vik's (1954) reconstruction of the parts of theEusthenopteron skeleton preserved in hisspecimens (viz., the first three arches andpart of the fourth). Nelson's and Miles's in-terpretations of the Acanthodes visceralskeleton are out of phase by one element,namely whether the gill arches are supportedventrally on a copula with two ossified ba-sibranchials between arches one and two andtwo and three (Nelson's inference, fig. 47A)or whether the basibranchials identified byNelson are really hypobranchials of the sec-ond and third arches, and the hypobranchialsand ceratobranchials of Nelson are really thetwo parts of a segmented ceratobranchial(Moy-Thomas and Miles's inference). We fa-vor Nelson's view because (1) there is noparticular reason to postulate an unusualcondition of a bipartite ceratobranchial (see,below, comments on Miles's postulate of abipartite epibranchial); (2) the structuresidentified as basibranchials by Nelson aresmall elements of about equal length orientedat an angle to the gill arch as one might ex-pect them to be ifthey were median elements;and (3) in a photograph published by Miles(1973, plate 7), the plainly visible gill rakerson the inner side of the arches are present allalong what Nelson identified as hypobran-chials and ceratobranchials but are absent onthe small ventral elements, as would be trueof ceratobranchials, hypobranchials, and ba-sibranchials in other fishes. Nelson (1968)commented on the dorsal gill arches ofAcan-thodes: "Watson and Miles apparently erredin showing the epibranchial divided in twopieces, and in terming one of these either apharyngobranchial (Watson) or an infrapha-ryngobranchial (Miles) . . . the suprapha-ryngobranchial of Miles appears to be part... of the same pharyngobranchial figuredby Reis and possibly also by Dean." In addi-tion, Miles (1964, 1965) figured the pharyn-

gobranchials extending forward as in bonyfishes, whereas Nelson (1968), by studyingthe details of paired articular surfaces on theepibranchials in relation to correspondingpoints on the pharyngobranchials, concludedthat the pharyngobranchials are directedposteriorly as in chondrichthyans. Fromthese, and other observations of the hyoidarch, Nelson (1969) concluded that "the vis-ceral apparatus of acanthodians appears tobe more primitively organized than that ofany other gnathostome group."

If Nelson's interpretation of acanthodiangill arch structure is correct, then Acan-thodes resembles chondrichthyans in threeimportant respects: the ventral articulationof the first two gill arches fore and aft of asmall basibranchial element, the absence ofa suprapharyngobranchial element articulat-ing with the basicranium, and the posteriororientation of the pharyngobranchials (fig.47). Further, if these are primitive conditionsof gnathostome gill arches, then the Osteich-thyes are defined by three derived condi-tions: the ventral articulation of at least thefirst two gill arches on a single basibranchial,the presence of a suprapharyngobranchialarticulation with the basicranium, and theforward orientation of the infrapharyngo-branchials. These three conditions areillustrated (fig. 48) in the sarcopterygianEusthenopteron (in which the firstsuprapharyngobranchial apparently arisesfrom the proximal end of the first infrapha-ryngobranchial), the paleoniscid Pteroniscu-lus (in which there is a suprapharyngobran-chial on the first and second epibranchials,as in Recent chondrosteans), and the Recentcladistian Polypterus (in which the only su-prapharyngobranchial present appears tohave fused with the first epibranchial). Nel-son (1969) suggested that a condition not un-like that of Eusthenopteron may be presenton the first arch of Latimeria in which thepharyngobranchial is a slender, dorsally di-rected rod (fig. 49A); it suggests the inter-pretation that a compound supra-infrapha-ryngobranchial has been reduced distally,leaving only the erect dorsal process proxi-

1981 235


FIG. 47. Gill arches, dorsal view. A, Acanthodes bronni Agassiz (from Nelson, 1968; compilation offigs. 3 and 4a; fifth arch not shown). B, Squalus acanthias Linnaeus (AMNH 40804); C, Hydrolaguscolliei (Lay and Bennett) (AMNH 40805). Dorsal gill arches unshaded.

mally. The suprapharyngobranchial of thesecond arch in Latimeria, like that in Eus-thenopteron and Pteronisculus, is articulat-ed, not fused, with the epibranchial.Eusthenopteron and Pteronisculus and

other actinopterygians share with acantho-dians and chondrichthyans the apparentlyprimitive condition of having the right andleft gill arches coming together in the ventralmidline on a long, narrow copula, the arch

elements of one side extending forward anddownward more or less parallel with eachother. In Polypterus the arches also run par-allel and articulate against a copula, but thecopula, as in Mimia, is undivided into sepa-rate basibranchials (Sewertzoff, 1927, figuresa specimen of P. delhezi as having two ba-sibranchials, but our material of this species,AMNH 31013, shows only a single basi-branchial); the single basibranchial is also

VOL. 167236


FIG. 48. Gill arches, dorsal view. A, Eusthenopteronfoordi Whiteaves (modified from Jarvik, 1954); B,Polypterus ornatipinnis Boulenger (AMNH 40802); C, Pteronisculus stensioei (Nielsen) (modified fromNielsen, 1942). Dorsal gill arches unshaded.

somewhat broader than in actinopterygiansbut is perhaps similar in outline to what isknown of the basibranchials of Eusthenop-teron (fig. 48A). In addition, the copula ofPolypterus may be secondarily foreshor-tened in connection with the loss of a gillarch (Nelson, 1969, has argued that an inter-mediate arch has been lost and his proposalis supported by the similarity in structureand associated dorsal arch elements on thefirst arch and the absence of dorsal elementson the fourth and last arch with the first andfifth arches of other fishes).

In Latimeria and Neoceratodus (fig. 49A,B) five arches are present, but the copula isgreatly foreshortened (Latimeria) or reduced(Neoceratodus) and hypobranchials are ab-sent (Latimeria) or absent anteriorly andgreatly reduced posteriorly (Neoceratodus).The supposition that Neoceratodus has smallhypobranchials present on arches three andfour is suggested by the presence of small,forwardly directed proximal processes whichclosely resemble similar elements in the hy-pobranchial position of larval urodeles (see,for example, Hynobius, fig. 49C) and by the

1981 237


D

FIG. 49. Gill arches, dorsal view. A, Latimeria chalumnae Smith (AMNH 32949); B, Neoceratodusforsteri (Krefft) (AMNH 36982); C, Hynobius leechi Boulenger (AMNH 58792); D, Glyptolepis groen-landica Jarvik (modified from Jarvik, 1972). Dorsal gill arches unshaded.

fact that one of these processes is free fromthe ceratobranchial on one side of one spec-imen. Latimeria, Neoceratodus, and Hyno-bius have the copular region reduced to ashort triangular area with the apex of the tri-angle posterior, so that the anterior gill archbases are noticeably farther apart than thoseof the posterior arches, and consistentlyhave the posteriormost arch articulating di-rectly with the base of the preceding archrather than lying free or articulating with the

copular region.11 Latimeria, Neoceratodusand urodeles also show an apparent transi-

11 The gill arch bases of Griphognathus, as restoredby Miles (1977, fig. 137) are not farther apart anteriorlythan posteriorly, but Miles has restored the bases of thefirst two arches on the posterior end of a specializedbasihyal. If the known arches were restored entirely onthe triangular basibranchial (on which they would allfit), the arrangement of gill arch bases would be similarto those of Neoceratodus. We have no good clue as towhich is the preferable solution.

VOL. 167238


tion series in reduction of gill arch elements.In Latimeria the anterior epibranchials arereduced in length relative to posterior ones,the first pharyngobranchial is represented bya slender rod (cf. the first arch in Eusthe-nopteron), the second pharyngobranchial is asimple, round nodule loosely associated withthe epibranchial, and the third pharyngo-branchial is either absent or, if present, isrepresented only by one or two tiny, irreg-ular bits of cartilage. In Neoceratodus theanterior epibranchials are reduced relative tothe posterior ones and there are no pharyn-gobranchials. In urodeles there are neitherepibranchials nor pharyngobranchials. Eus-thenopteron foordi, if correctly interpretedby Jarvik (1954), might be considered to bepart of this transformation series in lackinga posterior pharyngobranchial. This loss maybe a unique reductional feature of the speciesbecause it is associated with a very smallthird and no fourth epibranchial in combi-nation with very well-developed epibranchi-als and pharyngobranchials similar to thoseof actinopterygians on the first and secondarches. Or the loss may be apparent and re-flect only poor preservation (see Jarvik,1954, pp. 22-23). A fifth arch was postulatedby Jarvik, but is not actually known. Jarvik(1972) has also reconstructed the ventral partof the gill arches of the porolepiform Glyp-tolepis groenlandica, in which, again, ele-ments of only four arches have been found(fig. 49D). In the porolepiform Laccogna-thus, however, there are five (Vorobyeva,1980). Jarvik shows the copula as suboval,greatly foreshortened, and supporting all butthe last arch; this reconstruction strongly re-sembles the condition in Latimeria, exceptthat Jarvik has identified hypobranchials inGlyptolepis. A further resemblance betweenJarvik's reconstruction and the ventral gillarch structure in Latimeria, Neoceratodusand urodeles is the articulation of the pos-terior arch base with the base of the preced-ing arch rather than with the copular region,but unlike those three groups and other os-teichthyans, the posterior arch has a hypo-branchial. In Laccognathus the bases ofeach of the last two arches (fourth and fifth)articulate with the preceding arch (Vorobye-

FIG. 50. Hyoid arch in relation to palatoquad-rate. Above, Acanthodes bronni Agassiz (varioussources). Below, Chlamydoselachus anguineusGarman (from Allis, 1923).

va, 1980). The Devonian dipnoan Gripho-gnathus is also described (Miles, 1977) as hav-ing only four gill arches, but, as in modernlungfishes, the last arch articulates with thebase of the preceding arch and there are nohypobranchials. In general, it does seem tous that there exists reason here, as in a num-ber of other features discussed in other sec-tions, to place porolepiforms in a group in-cluding coelacanths, lungfishes, andtetrapods. On the basis of its primitive vis-ceral arch anatomy, Eusthenopteron is ex-cluded from the above group and is not evenunambiguously assignable to the Sarcopter-ygii. Polypterus likewise cannot be placedwithin a subgroup of the Osteichthyes on thebasis of its visceral arch structure. Livingand fossil chondrosteans appear to have themost primitive osteichthyan gill arches inshowing the greatest general similarity tochondrichthyans and Acanthodes.

Reorientation and reduction of dorsal ele-ments is characteristic also of the hyoid archof coelacanths, dipnoans, and tetrapods butnot of Eusthenopteron, actinopterygians,

1981 239


FIG. 52. Hyoid arch in relation to palatoquad-FIG. 51. Hyoid arch in relation to palatoquadrate. Above, Latimeria chalumnae Smith (modi-

rate. Above, Pteronisculus stensioei (Nielsen) fied from Millot and Anthony, 1958). Below, Neo-(from Nielsen, 1942). Below, Eusthenopteron ceratodus forsteri (Krefft) (various sources;foordi Whiteaves (from Jarvik, 1972). showing fusion of palatoquadrate with neurocra-

nium).

chondrichthyans, or acanthodians. In thelast four groups (figs. 50, 51) the hyo-mandibula is long and slender, closely asso-ciated with both the ceratohyal and palato-quadrate, and conforms in orientation withthe sharp forward slope of the posterodorsalmargin of the latter. In coelacanths, dip-noans, and tetrapods the hyomandibula andthe posterior part of the palatoquadrate havea more nearly vertical or backward orienta-tion, and the hyomandibula shows a pro-gressive reduction in size in the series, andevidence of increasing functional decouplingfrom the hyoid bar and palatoquadrate. InLatimeria (fig. 52) the hyomandibula is a rela-tively small, flat, subtriangular element thatis displaced posteriorly away from the palato-quadrate. Ventrally the hyomandibula articu-

lates with a rodlike interhyal (stylohyal of au-thors). The interhyal, which in Eusthenopte-ron and actinopterygians is in series with anddirectly articulates between the hyomandibulaand hyoid bar, is in Latimeria laterally offsetfrom the distal end of the hyoid bar and at-tached to it only by a bed of ligaments. InNeoceratodus (fig. 52) the hyomandibula isalso a small, flat, triangular plate, butthe interhyal is absent and a connection be-tween the ventral part of the hyomandibulaand the hyoid bar is made only by an obliqueligament. But in dipnoans (figs. 52, 53)the hyomandibula is completely reoriented,its ventral end directed anteriorly toward thequadrate and, at least in Neoceratodus, itsdorsomedial process fused with the capsularwall of the otic region of the cranium (de

240 VOL. 167


FIG. 53. Hyoid arch in relation to palatoquadrate and neurocranium. Above, Griphognathus whiteiMiles (from Miles, 1977). Below, hynobiid urodele (modified from Schmalhausen, 1968; Hynobius leechiBoulenger, AMNH 58792, and H. chinensis Gunther, AMNH 30616).

Beer, 1937, p. 412). The hyoid bar is ex-panded posteriorly and turned sharply up-ward toward the otic region, a structuralmodification also present to some degree inLatimeria. In hynobiids (fig. 53), the hyo-mandibula has exactly the same orientationas in dipnoans except that here the hyoman-dibula is fused anteroventrally with the quad-rate and its dorsomedial process is greatlyexpanded and occludes an unossified region,the fenestra ovalis, in the capsular wall.Again, as in dipnoans, there is a hyoquadrateligament between the sharply upturned distalend of the hyoid bar and the ventral part ofthe hyomandibula that fuses with the quad-rate (this ligament is all that remains of theprimitive osteichthyan interhyal joint).

Thus, the anatomy of both the hyoideanand branchial arches supports a sister grouprelationship between dipnoans and tetrapods

with coelacanths as the sister group of thosetwo. Eusthenopteron appears to have a prim-itive hyoidean and gill arch anatomy similarto that of actinopterygians. Porolepiformsappear to have retained the primitive hyo-mandibular orientation, as judged by the ar-rangement of dermal cheek bones, and thuslack one of the two classes of visceral archsynapomorphies that unite coelacanths withdipnoans and tetrapods. Polypterus shows ajaw suspension of intermediate type: the hyo-mandibula is only slightly expanded dorsal-ly, is strongly oblique dorsally and nearlyvertical ventrally where it lies against thevertical posterior face of the palatoquadrate,but remains intimately associated with boththe hyoid bar and palatoquadrate as in chon-drosteans. The cladistic position of Polyp-terus is therefore problematical not only withrespect to the branchial apparatus but with

2411981


the hyoid arch as well. Hence, the combinedinformation derived from hyoid and gill archanatomy specifies that porolepiforms are thesister group of coelacanths and dipnoans/tet-rapods, that actinopterygians, cladistiansand Eusthenopteron form an unresolved tet-

rachotomy with those four, that chondrich-thyans are the sister group of bony fishes andthat acanthodians, using Nelson's (1969) ar-gument, are the sister group of other gnatho-stomes.

RIBS AND VERTEBRAE

(A) RIBS

One of the earliest reasons given for ex-cluding lungfishes from close relationshipwith tetrapods was that lungfishes have pleu-ral ribs, whereas tetrapods have dorsal ribs(Pollard, 1892; Goodrich, 1909). And theidentification of dorsal ribs, possibly bicipi-tal, in Eusthenopteron (Jarvik, 1952; An-drews and Westoll, 1970a) seemed an impor-tant confirmation of the relationship betweenosteolepiforms and tetrapods.

Devillers (1954) gave an excellent accountof the interpretation of ribs, on which thefollowing summary is based. The presenceof two kinds of ribs on each vertebra in themiddle part of the trunk of Polypterus wasthe foundation for the recognition of two dis-tinct categories of ribs, dorsal ribs in the hor-izontal septum, and ventral (pleural) ribs inthe wall of the coelom. Many teleosts alsohave these two types of ribs, and in addition,epineural ribs in the epaxial musculature. Inother groups of gnathostomes, with only oneseries of ribs, the problem is to find decisivecriteria for allocating the ribs to one seriesor another. Topographic criteria fail, be-cause in actinopterygians and chondrichthy-ans the ribs vary in position, from species tospecies, and from one part of the vertebralcolumn to another in the same fish, "wan-dering" within the transverse septum. In tet-rapods the problem is aggravated by the factthat the horizontal septum is absent in thetrunk of all except urodeles.

Seeking a more reliable criterion, Emeli-anov (1935, 1936) found that the ventral ribsof Polypterus and teleosts develop centrifu-gally, from cartilaginous anlagen close to thevertebra, whereas the dorsal ribs of thosefishes develop centripetally, from cartilagi-

nous anlagen beneath the lateral line, at theouter junction of the horizontal and trans-verse septa. According to this criterion, theribs of elasmobranchs, previously thought tobe dorsal, are ventral ribs, as are those ofdipnoans and non-teleostean actinopterygi-ans; the ribs of urodeles, previously taken tobe dorsal, are indeed dorsal; and those ofapodans and amniotes, previously thought tobe dorsal, are ventral. However, in urodelesthe characteristic centripetal development ofdorsal ribs is seen only in the anterior ver-tebrae. Passing back down the trunk, the lat-er the ribs appear, the closer their anlagenare to the vertebral column, so that the mostposterior ribs develop as ventral ribs. Be-cause of this, Devillers doubted that Emeli-anov's criterion was as trenchant as its au-thor supposed and noted that the anteriorribs of urodeles are the only exception to theuniversal presence of pleural ribs.From this summary, it is clear that iden-

tification of ribs as dorsal or ventral is nosimple matter. Topographic criteria (positionof the articulation with the vertebra; orien-tation of the rib; relationship to the muscu-lature) do not suffice. In fossils, unless morethan one series of ribs is present, the ribs cantherefore be identified as dorsal or ventralonly by a comparative argument (showingthat the fossil is a member of a group char-acterized by one type of rib), or by evidenceof mode of growth (centripetal or centrifu-gal).The original reason for recognizing two

categories of ribs was the presence of bothtypes in cladistians and teleosts. In Cladistiathe dorsal ribs are peculiar, borne on hyper-trophied parapophyses and firmly tied dis-tally to the lateral line scales, an arrangement

VOL. 167242


IOmm

CENTRUM 2 (AXIS)

FIG. 54. A, Griophognathus whitei Miles. Sketch of three vertebrae (approx. nos. 9-11) in left lateralview, based on BMNH P. 60420. B, Gyrinophilus porphyriticus (Green), vertebrae. Ribs, black. Unos-sified areas hatched.

interpreted by Pearson (in press) as a spe-cialization enhancing the flexibility of thetrunk. In teleosts, dorsal (epipleural) ribs arenow thought to be a derived feature charac-terizing higher (elopocephalan) teleosts (Pat-terson and Rosen, 1977, p. 126) and certainosteoglossomorphs (Heterotis and notopter-ids; Taverne, 1979, p. 61). The alternative,to regard dorsal ribs as a primitive attributeof actinopterygians, would be unparsimo-nious, requiring their independent loss inmany lineages. By the same argument, dor-sal ribs are not a primitive attribute of os-teichthyans. Instead, we should expect prim-

itive osteichthyans to have ribs ofchondrichthyan type-short, oriented likedorsal ribs, but developing like ventral ribs.In our view, this is the simplest interpreta-tion of the ribs of Eusthenopteron (figs. 55,56). The thickening or anterior process onthe ribs in the middle part of the trunk ofEusthenopteron, identified by Andrews andWestoll as a tuberculum, is matched in var-ious elasmobranchs (e.g., Notorynchus,Daniel, 1928, fig. 51; Somniosus, Dalatias,Helbing, 1904, figs. 27, 28), and the absenceof ribs on the first three to seven vertebrae(Andrews and Westoll, 1970a, fig. 22; Jarvik,

1981 243


SPACE FORLONGITUDINAL

:,,° LIGAMENT

A B C

LONGITUDINAL LIGAMENI

HORIZONTAL SEPTUM

RIB.......LATERNERVE

~~~~RIB

EFIG. 55. Form of ribs in Eusthenopteron and elasmobranchs. A-C, restorations of three vertebrae

of Eusthenopteron foordi Whiteaves. A, fifth in posterior view, after Jarvik (1975, fig. 10); B, abouteighteenth in anterior view, after Andrews and Westoll (1970a, fig. 21C); C, about twenty-sixth inanterior view, after Jarvik (1952, fig. 13); D, E, reconstructed thick sections of a mid-trunk vertebra ofD, Mustelus laevis Risso, 100 mm.; and E, Torpedo sp., 30 mm., after Emelianov (1935, figs. 49, 52).In D and E muscle is stippled, nerve cord hatched.

1975, fig. 10) also agrees with sharks (fig.56A, B).An assessment of the primitive condition

of the ribs in tetrapods (dorsal ribs, pleuralribs, or both) requires a reasoned phylogenyof tetrapods, and, as Devillers (1954) sug-gested, investigation of a more extensivesuite of urodeles than the two genera studiedby Emelianov. But one synapomorphy of tet-rapods is the bicipital rib, articulating witha diapophysis on the base of the neural arch,and a parapophysis on the centrum. The dia-pophysis of tetrapods recalls the articulationof the epineurals in teleosts. The separateepineurals of teleosts are a derived featurewithin actinopterygians, where the primitivecondition seems to be an outgrowth or pro-cess of the neural arch in the diapophysialposition (Patterson, 1973, p. 237; these out-growths are also present on the anterior

neural arches of the Gogo paleoniscoids).Such outgrowths are also present in the dip-noan Griphognathus (fig. 54A), and appearto be independently ossified. They are absentin Eusthenopteron (figs. 55A-C, 56B), as areany identifiable diapophyses on the rib-bear-ing vertebrae (Andrews and Westoll, 1970a).These diapophysial outgrowths in Griphog-nathus and primitive actinopterygians aremost completely developed on the foremostvertebrae; that is, they show a morphoge-netic gradient which decreases with distancefrom the occiput. The ribs of Eusthenopte-ron, sharks and primitive actinopterygians(fig. 56C) show a different gradient, whichdecreases rostrally and caudally from themiddle of the trunk. The ribs of primitive tet-rapods (fig. 54B) and Recent lungfishes (fig.57) show the first type of gradient, decreas-ing from the occiput. The expanded head of

244 VOL. 167

T


DORSAL ROOTSUPRANEURAL NEURAL ARCH VENTRAL ROOT

VENTRAL ARCH

A

'TERY

I NT ER VENTRAL

PARASPHENOID\\\I\FIG. 56. Occiput and anterior vertebrae in left lateral view of A, Notorynchus maculatus Ayres,

after Daniel (1928, figs. 47, 51); B, Eusthenopteron foordi Whiteaves, after Jarvik (1975, fig. 10); C,Acipenser oxyrhynchus Mitchill, 150 mm. total length, from AMNH 37296. Cartilage stippled in A andC, bone stippled in B.

1981 245


ARC H

SPINE

FIG. 57. Neoceratodus forsteri (Krefft). A, anterior vertebrae and ribs, left lateral view; B, skull-vertebral articulation, dorsal view. Small bone in front of first neural arch, in the position of the tetrapod"preatlas," is present on only left side of this specimen, AMNH 36982.

the first rib of young Neoceratodus (fig. 57;this rib becomes the cranial rib in full-grownanimals) is evidence of this gradient in lung-fishes. In urodeles (fig. 54B) modification ofthe heads of the foremost ribs is more pro-nounced, with the dorsal struts to the dia-pophysis decreasing in size away from theocciput. We suggest that the anterior verte-brae and ribs of Neoceratodus fit the tetra-pod morphocline far better than those ofEusthenopteron, which seem to be primitivein all respects.

(B) INTERPRETATION OFGNATHOSTOME VERTEBRAE

Four pairs of arcualia are present in thedevelopment of many elasmobranchs, holo-cephalans, actinopterygians, unborn juvenileLatimeria, and Neoceratodus. These arcu-alia are also retained in some adult elasmo-branchs (Chlamydoselachus, fig. 58A) and

holocephalans (Callorhynchus), and, withinthe -Osteichthyes, they are represented byossified or cartilaginous elements in at leasta part of the vertebral column: e.g., in actin-opterygians (Acipenser, fig. 56C; Polyodon,fig. 58B; Pteronisculus, Caturus, fig. 59;Pholidophorus, Patterson, 1968), in actinis-tians (Latimeria, fig. 58C), in dipnoans (Neo-ceratodus, fig. 58D, E) and in tetrapods (Ar-chegosaurus, Chelydosaurus and "Spheno-saurus" sternbergii, Fritsch, 1883-1895,pp. 25, 28). In many bony fishes and primi-tive tetrapods (loxommatids, temnospon-dyls) however, the ventral, caudal pair ofarcualia (interventrals) are wanting, theyhave either failed to develop or have fusedwith the more anterior basiventrals. Thus, inthe trunk and abdominal regions of Caturus(fig. 59A, B), Pholidophorus, Latimeria, andthe temnospondyls Archegosaurus, Chely-dosaurus, and "Sphenosaurus" there is onlyone pair of ventral elements, the ventral ver-

VOL. 167246


FIG. 58. Vertebrae. A, Chlamydoselachus anguineus Garman, trunk region (from Goodey, 1910);B, Polyodon folium Lacepede, region of trunk-caudal transition (from Schauinsland, 1906); C, Latimeriachalumnae Smith, caudal region (from Andrews, 1977); D, Neoceratodusforsteri (Krefft), anterior trunkregion (from Fiirbringer, 1904); E, N. forsteri, posterior trunk region (from Andrews, 1977).

tebral arch, in each segment. In Neocerato-dus (fig. 58D), on the other hand, interven-trals have only been described in one smallportion of the thoracic region (Fiirbringer,1904), whereas in Glyptolepis (Andrews andWestoll, 1970b, fig. 23), Eusthenopteron(Andrews and Westoll, 1970a, fig. 23) andOsteolepis (Andrews and Westoll, 1970b,fig. 5) ossified interventrals are invariablyabsent.

Previously (Jarvik, 1952, 1975; Andrewsand Westoll, 1970a), the vertebral column ofEusthenopteron was considered very similarto that of Ichthyostega. For Eusthenopteronthere are four very different reconstructionsof the vertebral column. The earliest is thatof Gregory, Rockwell and Evans (1939); Jar-vik (1952, fig. 13C) figured vertebrae fromthe anterior part of the tail and from the mostanterior trunk region (Jarvik, 1975, fig. 1OA);Andrews and Westoll (1970a, figs. 19, 20) fig-

ured the first 26 vertebrae. In all of thesereconstructions the interdorsal is interpretedas being smaller than that of Ichthyostega(Jarvik, 1952, fig. 14A, B, C). In addition,the occurrence of a notch in the interdorsalfor the presumed passage of the ventralnerve root in Eusthenopteron (figs. 56B,60B) and Ichthyostega (Jarvik, 1952, fig.14C) is matched by a similar notch for thepassage of the ventral root in the interdorsalof Neoceratodus (fig. 58D).The pre- and postzygapophyses of Eus-

thenopteron (Jarvik, 1952, fig. 13, pr.a,pr.p.) are here presumed to be no more thanexpansions of the neural spine associatedwith the passage of the dorsal ligament andcorrespond with similar projections in Gri-phognathus (fig. 54A), Latimeria (Andrews,1977, fig. 1) and many actinopterygians (Ca-turus, fig. 59). Zygapophyses, in the sense ofprocesses with articular surfaces that join

2471981


A B

D 5mm

FIG. 59. Vertebrae of Caturusfurcatus (Agassiz), as preserved in BMNH P. 20578, anterior to right.A, anterior trunk; B, trunk; C, posterior trunk; D, anterior caudal.

successive neural arches, are found only intetrapods.Complete centra occur within the elas-

mobranchs (Chondrenchelys, euselachians),holocephalans, actinopterygians, osteolepi-forms (Strepsodus, Rhizodus, Andrews andWestoll, 1970b, fig. 7), dipnoans (Gripho-gnathus, fig. 54, Protopterus, Mookejee,Ganguly, and Brahma, 1954), and tetrapods.Within the actinopterygians true (complete)centra have arisen on at least four separate

occasions (Polypterus, Lepisosteus, Amia,and teleosts), whereas within the tetrapodswe conclude that true centra have formed onat least two occasions (Recent amphibiansand amniotes). This latter conclusion isbased on the fact that the amniote vertebralcolumn, unlike that of anurans, apodans, andurodeles, is often composed of more thanone part per segment (viz., intercentrum andcentrum) and is ossified not only perichon-drally, but endochondrally as well.

VOL. 167248


(C) HETEROSPONDYLY INCATURUS FURCATUS, ANDVERTEBRAL HOMOLOGIES

An almost complete specimen of Caturusfurcatus (BMNH P. 20578) was acid-pre-pared by one of our colleagues, Alan Bar-tram. This specimen has a complete anduncrushed vertebral column (fig. 59).Throughout the column individual ossifica-tion centers comprising perichondral shellssurrounding endochondral bone are clearlyrecognizable and there are four such pairsper segment from the thirty-ninth vertebraposteriorly. Furthermore, these ossificationsmay be directly compared with the four pairsof cartilages known to be present in Chlam-ydoselachus and in the developing vertebralcolumn of Amia. We shall therefore callthese ossifications basidorsals, basiven-trals, interdorsals, and interventrals in ac-cordance with Gadow's (1933) terminology.Throughout the column the basidorsals are

continuous with the bases of the neuralspines and sit between the interdorsals. Thepairs of interdorsals are joined to oneanother by half-rings of calcification of thenotochordal sheath as in the caudal region ofPholidophorus (Patterson, 1968, fig. 3). Sim-ilar calcifications of the notochordal sheathare seen in the development of Clupea (Ra-manujam, 1929, p. 377) and Salmo (Fran-gois, 1966, p. 292). Elsewhere calcificationsof the notochordal sheath are seen in someholocephalans, Chondrenchelys, variouselasmobranchs, and Australosomus.

In the trunk region of Caturus (anterior tovertebra 39) the basiventrals and interven-trals of each side are presumed to have fusedtogether, their point of junction beingmarked by the foramen for the intersegmen-tal artery. A similar fusion of basiventralwith interventral is deduced to have also oc-

curred in the ventral element of Eusthenop-teron (Jarvik, 1952, fig. 13E) where a dor-soventral groove marks the passage of theintersegmental artery.

In the caudal region of Caturus these twoelements remain separate as in the tail ofPholidophorus (Patterson, 1968, fig. 1). Inthe anterior trunk region partial chordacen-

tra form in association with the fused basi-ventrals and interventrals, but the ossifica-tions of either side remain separate as inGlyptolepis, Osteolepis (Andrews and Wes-toll, 1970b, figs. 4, 20), and Eusthenopteron(Jarvik, 1952, fig. 13D). Farther back alongthe trunk the fused ossifications of either side(basiventral + interventral) are joined into ahemicentrum by the addition of a medianventral crescentic calcification in the sheathof the notochord (chordacentrum). Thus,these ventral ossifications resemble the con-dition seen in the trunks of many holosteansand Pholidophorus. Fusion of the ventralelements in the midline has also occurred inthe tail region of Eusthenopteron (Jarvik,1952, fig. 13E) and anterior trunk (Jarvik,1975, fig. 10) and throughout almost the en-tire column in temnospondylous amphibians.More posteriorly in the tail region of Ca-

turus (vertebra 39 onward) two ventral hemi-centra are formed in each segment. An an-terior chordacentrum links the basiventralsof either side while a posterior chordacen-trum joins the two interventrals. Thus, themiddle part of the caudal region of Caturusshows diplospondylous hemicentra as inPholidophorus, Australosomus (Stensib,1932, pls. 35-37) and many holosteans andparallels the condition seen in the entire ver-tebral column of primitive amniotes. Fromthis study we conclude that Schaeffer (1967)and Thomson and Vaughn (1968) were prob-ably correct in homologizing the pleurocen-tra of crossopterygians with the dorsal inter-calaries of Amia and furthermore that thesedorsal intercalaries may be directly homolo-gized with temnospondylous pleurocentra.

(D) VERTEBRAE OF RHIPIDISTIANS,LUNGFISHES, AND TETRAPODS

As was the case with the interpretation ofdipnoan ribs, dipnoan vertebrae were alsotaken to be of an extremely primitive typethat has no bearing on the question of tet-rapod affinities. This was partly based on thesupposition that dipnoans lack ossified cen-tra and that their neural and hemal arch com-ponents are extremely simple and primitive

2491981


for gnathostomes generally. Schmalhausen(1968) has written, for example, that the"primitiveness of the Dipnoi is displayed inthe retention of a permanent notochord evenin the modern forms" and the "vertebraeconsisted only of cartilaginous elements, thebases of the upper and lower arches." Thiswidespread opinion is due, in large part, tothe repetition of an old illustration in Good-rich (1909) showing several poorly differen-tiated vertebral segments from the posteriorpart of the axis in Neoceratodus, and, inpart, from failure to realize that the vertebralcentra, ossified in some Devonian formssuch as Griphognathus, are represented inmodern forms by a few barely ossified ringcentra or hemicentra in the posterior part ofthe skeleton. Nevertheless, a study of thefirst few vertebrae of young Neoceratodusshows that the column articulates with theskull on two exoccipital processes (fig. 57B)that closely resemble the exoccipital con-dyles of the bicipital first vertebral articula-tion of modern and fossil amphibians (seeSchmalhausen, 1968, figs. 136, 137). In con-trast, actinistian and rhipidistian fishes, sofar as known, have only the primitive, single,basioccipital contact of the first vertebra thatalso characterizes chondrostean actinopte-rygians and chondrichthyans.

Until now, however, it has been assumedthat the vertebrae of Eusthenopteron andtetrapods (as represented by Ichthyostega,fig. 60A) are significantly alike and that thesimilarity signifies close relationship. The as-sumption goes back to several statementsmade by Jarvik (1952) in his review of theichthyostegids: "each vertebra, both in thetrunk and in the tail, is composed of threeelements, two dorsal and one ventral verysuggestive of those in Eusthenopteron" (p.36), and "neural spines [in the sacral region]gradually become narrower and more in-clined backwards and are converted into thecomparatively narrow spines, round in sec-tion, which in the tail articulate with the ra-dials. The latter strongly resemble the neuralspines in Eusthenopteron" (p. 38). Jarvikalso compared the interdorsal vertebral seg-ments in Eusthenopteron and ichthyostegidsand concluded that their segments are simi-

larly notched for the dorsal and ventral rootsof the spinal nerves. But Jarvik also noted(p. 36) that "The vertebral column in the ich-thyostegids . . . is on the whole imperfectlypreserved in the material available." Whenone compares the published photographs ofthe columns of ichthyostegids and Eusthe-nopteron there is, in.our opinion, little spe-cific resemblance between the trunk and tailvertebrae of Eusthenopteron (Andrews andWestoll, 1970a, pls. 4, 5) and Ichthyostega(Jarvik, 1952, figs. 2, 3, 14, 15) except insofaras they are all unconsolidated vertebrae withossified neural and hemal arches, bits ofhemicentra, and structures of questionableidentity called interdorsals: "The interdorsal[of Ichthyostega] is well shown [in twospecimens] whereas in other specimensshowing the neural arches it generally is im-perfectly preserved or missing [Jarvik, 1952,p. 36]." Only one of these structures is iden-tified in one specimen and perhaps three orfour can be made out in the second; they areof varying sizes and shapes, although some-what hemispherical and, to us, their relation-ship to other structures is unclear. These fea-tures in the two specimens of Ichthyostegacontrast with the positionally similar struc-tures of Eusthenopteron which are smaller,more dorsal in position (behind the neuralspine base), and round, nodular or hemi-spherical (but with a different orientationfrom those of Ichthyostega). Hence, bothtaxa seem to be similar at most in havingpresent in the column three of the four pos-sible arcual elements, basidorsals and basi-ventrals (neural and hemal arches) and inter-dorsals, and incomplete ring (Ichthyostega)or hemicentra (Eusthenopteron) that areprimitive for chondrichthyans and osteich-thyans. The only derived character theymight share, therefore, is the absence ofinterventrals (see, for example, fig. 56C, ofAcipenser).Panchen (1977), however, rejected Jar-

vik's comparisons because he believed thatthe interdorsal of Osteolepis is a bettermatch with those of ichthyostegids than thatof Eusthenopteron in being larger. But inOsteolepis the interdorsals, in being large tri-angular bones, are not only larger and more

VOL. 167250


A B

VENTRAL ELEMENT

FIG. 60. Trunk vertebrae. A. Ichthyostega sp. (from Jarvik, 1952, fig. 14B, anterior to right); B-D,Eusthenopteron foordi Whiteaves, anterior to left; B, anterior trunk vertebrae, from Jarvik, 1975; C,from Andrews and Westoll, 1970a; D, from Jarvik, 1952.

heavily ossified than those of Eusthenopte-ron, they are also larger, better ossified, andbetter differentiated than those of Ichthyo-stega. In other words, Osteolepis has a con-dition of the interdorsals not present in eitherof the other taxa. Interestingly, Panchen'sargument, unlike Jarvik's apparent use ofsymplesiomorphous features for formulatingrelationships, is also an appeal to stratigraph-ic position (Osteolepis is older than Eusthe-nopteron and is therefore a better ancestor)and convergence (Osteolepis resemblesmore advanced tetrapods with large, trian-gular interdorsals). He writes that Osteolepis

"is generally regarded as [a] primitive[osteolepiform], characterized by heavy os-sification" and is therefore a better ancestor,and that the convergent development of ringcentra in some more advanced members ofthe Osteolepididae and Tetrapoda indicatesa relationship between them. Perhaps Pan-chen's line of argument might also be used topropose a relationship between those twoassemblages and the halecomorph acti-nopterygian Caturus furcatus (Agassiz),since different parts of the column of Catu-rus show most of the main features of thecolumn of osteolepids and primitive tetra-

1981 251


pods and amniotes (fig. 59; Baur, 1896; An-drews, 1977; and p. 249 above). Baur (p. 663)wrote: "In the Crossopterygii we find unos-sified and well ossified vertebrae . . . it isvery probable that some of the crossopteryg-ians showed a condition like Archegosaurusand Caturus." And Andrews (p. 284) ob-served: "Fossil chondrosteans and especial-ly holosteans show very diverse conditionsof the vertebrae, but in some forms at least(Caturus, Eurycormus) the centra appear tobe composed, in a similar manner to Osteo-lepis, from 'alternating dorsal and ventralcrescentic hemicentra' " (Schaeffer, 1967, p.192).Each of the proposals, by Jarvik and

Panchen, reminds us of our methodologicalgoal in undertaking this study: to avoid theloose style of argument that has often char-acterized investigations of gnathostome re-lationships (Nelson's 1969 work is one ex-ception), and to approach the problem in anexplicitly cladistic fashion. To accomplishthis in dealing with the vertebrae we mustfirst specify the condition of the vertebraeprimitive for each group. In attempting this,we observe that all but one of the recogniz-able groups, chondrichthyans, actinopteryg-

ians, actinistians, dipnoans, and tetrapods(as represented by Ichthyostega) are primi-tively with unconsolidated cartilaginous orossified arcualia and unossified or poorly os-sified centra. We note that the same is trueof Acanthodes, Osteolepis, and Eusthenop-teron. The Cladistia are the single exception.Features that have been used in the past tounite groups representing patterns of fusionof arcual elements and their co-ossificationwith centra are therefore secondarily derivedfor each group (i.e., convergent). We are leftin doubt about the structure of the vertebraein Ichthyostega, and we reject Panchen'sreasoning which uses every undesirable fea-ture of past phylogenetic argumentation inpaleontology and neontology-symplesio-morphy, stratigraphic position, and conver-gence-as evidence of relationship. We do notrule out the possibility that some unique de-rived characters will be discovered in bettermaterial to unite Ichthyostega with one oranother "rhipidistian," but our own studyhas revealed only a single derived vertebralcharacter, a bicipital type of exoccipital jointwith the first vertebra (fig. 57B), that unitesonly dipnoans and tetrapods.

DERMAL SKELETON(A) COSMINE

Cosmine is a superficial hard tissue con-taining a pore-canal system. It is found onthe dermal bones of some cephalaspido-morphs, rhipidistians and Paleozoic dip-noans. No living fish has cosmine or a pore-canal system, and the function is unknown(Thomson, 1977). The cosmine of cephalas-pids differs from that of osteichthyans in somany ways (0rvig, 1969a; Schultze, 1969a,p. 56) that we agree with 0rvig (1969a) thatcosmine is not a primitive feature of a groupcontaining cephalaspids and osteichthyans;in other words, we do not regard cephalaspidand osteichthyan cosmine as homologous.Schultze (1969a) and Thomson (1975, 1977)accept that the pore-canal system (not nec-essarily enclosed in cosmine) is a primitivefeature of vertebrates. It is true that a pore-canal system also occurs in some acantho-

dians (Poracanthodes), but the architectureof the system is so different in cephalaspids,acanthodians, and osteichthyans (Gross,1956) that it is difficult to recognize any ho-mologies. We therefore regard the cosmineand pore-canal system of dipnoans and rhip-idistians as a synapomorphy (Schultze,1977b, fig. 1, seems now to agree, indicatingthe pore-canal system as a synapomorphy inhis cladogram). The problem is to decidewhat group is characterized by this synapo-morphy.

In rhipidistians, cosmine occurs in poro-lepiforms (Porolepididae-Porolepis, Heim-enia-and Powichthys, if Powichthys is aporolepiform) and in osteolepiforms(Osteolepididae sensu Vorobyeva, 1977a;Rhizodopsis, Andrews and Westoll, 1970b,p. 403). In dipnoans cosmine occurs in Ura-nolophus, Dipnorhynchus and dipterids (sen-

252 VOL. 167


su lato). In onychodonts the only report ofcosmine is a small patch on the premaxilla(Jessen, 1966, pl. 16, fig. 4) and parietal (Jes-sen, 1967, pl. 2B) of Strunius. This tissuehas not been studied histologically, andmight be only superficially cosmine-like, asare patches of porous enameloid on the headof the dipnoan Holodipterus (Smith, 1977, p.48, fig. 66). Cosmine has not been reportedin coelacanths, though Westoll (e.g., 1979,p. 346) has referred to "modified cosmine"in undescribed Devonian specimens fromScaumenac Bay. We have examined thosespecimens, and did not see anything recog-nizable as cosmine.The pore-canal system of cosmine-bearing

dipnoans, osteolepiforms, and porolepiformshas the same layout (Schultze, 1969a, fig.42). In some osteolepiforms the mesh canalslinking the flask-shaped pore-canals are di-vided by a horizontal partition into upper andlower divisions (Gross, 1956). Schultze(1969a) and Thomson (1975, 1977) have tak-en that condition to be primitive, since it re-calls the perforated "sieve-plate" dividingthe mesh-canal in cephalaspids. We regardthis as a far-fetched homology, since it de-pends on the assumption that the pore-canalsystem is primitive for vertebrates. Horizon-tal division of the mesh-canals, which hasnot been found in any dipnoan or porolepi-form, might be a synapomorphy of osteolep-iforms, or its absence might be a synapo-morphy of dipnoans and porolepiforms. Insome porolepids the enamel extends inwardto line the pore-canal (Qrvig, 1969b, p. 294).This condition is also found in some dip-noans (Gross, 1956, fig. 69) but not in others(Smith, 1977, fig. 18).

"Westoll-lines" are a prominent feature ofthe cosmine of dipnoans (Gross, 1956; 0rvig,1969a). Such naked strips are not recordedin porolepiform cosmine, or, according toGross (1956) and Jarvik (1968a) in any osteo-lepiform. However, Thomson (1975) be-lieves that Westoll-lines occur in some os-teolepiforms, and illustrates skull bones ofMegalichthys laticeps. Jarvik (1966, pl. 4,fig. 3) has also illustrated these lines on amaxilla of Megalichthys, and said (1968a, p.227) that their "nature ... is uncertain."Thomson suggests that the difference be-

tween his interpretation and Jarvik's is amatter of how Westoll-lines are defined, anddepends on some theory of the ontogeny ofcosmine. In any case, such lines have beenreported, and observed by us, only on skullbones of Megalichthys. They are unknownon osteolepiform scales. Evidence of re-sorption of cosmine (as distinct from West-oll-lines) is found in dipnoans and osteo-lepiforms, but not in cosmine-bearingporolepiforms (Qrvig, 1969b).

According to Miles's (1977) phylogeny ofdipnoans, cosmine is a primitive feature ofthe group, independently lost at least threetimes. In any alternative dipnoan phylogeny,it seems that cosmine must either be lost ordeveloped more than once, or dipnoans can-not be treated as monophyletic. In osteolep-iforms, Vorobyeva' s recent phylogenies(1977a, 1977b) require that cosmine be in-dependently lost five or six times, or that itdevelop more than once within the group. Inporolepiforms, it is generally assumed (large-ly on stratigraphic grounds) that cosmine isprimitive, and is lost in holoptychiids.

In dipnoans, Smith (1977) has made thor-ough comparisons of the histology of cos-mine-bearing (Chirodipterus) and cosmine-free (Griphognathus, Holodipterus) dermalbones. Of particular interest is her report onsuperposed generations of tubercles in Gri-phognathus and Holodipterus. Unless cos-mine developed independently on severaloccasions, these genera have to be regardedas having secondarily lost cosmine. Yet theyshow a pattern of superposed tubercles (gen-erations of odontodes; 0rvig, 1969a) resem-bling the tubercles of Latimeria (Smith, 1977,p. 41) and like that in other primitive fishes(Schultze, 1977b). We therefore lack any an-atomical criterion which will distinguish"pre-cosmoid" (cosmine assumed never tohave been present) and "post-cosmoid"(cosmine lost) dermal bones. Discriminationbetween those two conditions depends onthe position of the animal in question in aphylogeny, and that phylogeny must be es-tablished on grounds other than cosmine. Wealso therefore reject arguments that Lati-meria is the primitive sister group of otherosteichthyans because it possesses scaleodontodes.

1981 253


To sum up, we believe that cosmine is asynapomorphy of some taxon, but cosminealone cannot tell us whether animals thatlack it (such as coelacanths and tetrapods)belong in that taxon or not, because we can-not distinguish pre-cosmoid and post-cos-moid dermal bones. Cosmine is therefore ir-relevant to the monophyly of Sarcopterygii,or any other taxon. Resorption of cosmine,found only in cosmine-bearing dipnoans andosteolepiforms, could be a synapomorphyrelating those taxa. But if the primitive dip-noan pattern is shown by Uranolophus(scales and skull bones with Westoll-lines,superposed generations of tubercles in thescales, mesh-canals without a horizontal par-tition, Denison, 1968a, 1968b), it does notmatch the pattern generally assumed to beprimitive for cosmine-bearing osteolepiforms(no Westoll-lines, no buried tubercles, mesh-canals with horizontal partition). We have toconclude that, as yet, cosmine provides nophylogenetically useful information.

(B) FOLDED TEETHSchultze (1969b, 1970a) made a very thor-

ough survey of the histology of folded teethin fishes and tetrapods. He found that trueplicidentine occurs only in porolepiforms,osteolepiforms, labyrinthodont amphibians,a few reptiles, and in lepisosteids. Heshowed that these teeth could be separatedinto several categories according to the modeand complexity of folding. All porolepiforms(except Powichthys, Vorobyeva, 1977c)have dendrodont teeth, with very complexfolding and the pulp cavity filled with osteo-dentine. The most primitive amphibians (ich-thyostegids, loxommatids) have polyploco-dont teeth, with simple branching of the foldsin the dentine. Among fishes, polyplocodontteeth are found only in osteolepiforms, Pow-ichthys and lepisosteids. Within osteolepi-forms, there are genera with simple, unfold-ed teeth (Osteolepis, Thursius, Latvius),with polyplocodont teeth (Gyroptychius,Megalichthys, Tristichopterus, Eusthenop-teron, Panderichthys, Rhizodopsis, Rhizo-dus, Strepsodus), and with the more compli-cated, eusthenodont teeth (Eusthenodon,Platycephalichthys, Litoptychus, Sauripte-

rus). Among osteolepiforms with polyplo-codont teeth, only Panderichthys has thesame type as primitive tetrapods (Vorobye-va, 1977c, table 1).

Schultze (1969b, p. 128) summarized thesignificance of these facts neatly: either onemust accept the detailed correspondence intooth structure between Panderichthys andIchthyostega as evidence of close relation-ship (or direct ancestry), or one must pushtetrapod ancestry back to animals with sim-ple, unfolded teeth, and interpret the foldedteeth of tetrapods and osteolepiforms as con-vergent. Schultze preferred the first of thesealternatives, as does Westoll (1979), who en-visaged a direct stratigraphic lineage fromPanderichthys, through Elpistostege, toIchthyostega, and writes (p. 347) "Late De-vonian intermediates are confidently await-ed."When Schultze discussed the teeth of Pan-

derichthys, the animal was known only byisolated fragments-lower jaws, scales andbones of the skull roof and pectoral girdle.Since then, more or less intact individualshave been found (Vorobyeva, 1973, 1975,1980). These serve as a test of Schultze's andWestoll's hypothesis, which predicts thatPanderichthys will be more tetrapod-likethan other osteolepiforms. In some respects,this is true: the head is ichthyostegid-like indorsal view (Vorobyeva, 1980, fig. 1), with"brow-ridges" and a long snout, and theskull roof pattern is somewhat ichthyostegid-like (Vorobyeva, 1973, fig. 1). But in othercharacters Panderichthys does not fit. It hastwo external nostrils (Vorobyeva, 1973), avery short pectoral fin skeleton, of only threesegments (Vorobyeva, 1975, fig. 3), pelvicfins which are "very small, and set far backclose to the caudal fin" (Vorobyeva, 1980,p. 194), and ring-vertebrae in the thoracic andcaudal regions. Vorobyeva's conclusionsfrom her work on Panderichthys, and onfolded teeth (1977c), are that "at presentPanderichthys and Ichthyostega are practi-cally proved to belong to divergent evolu-tionary lines" (1977c, p. 17), and that theresemblances between the two are conver-gent (1974).

Thus, after studying more complete spec-

254 VOL. 167


imens of Panderichthys, Vorobyeva has dis-carded her earlier opinion (1962, fig. 34) thatPanderichthys is the closest relative of tet-rapods among fishes, and now concludes thattetrapods are the sister group of Osteolepi-didae (1977a, fig. 22). The same conclusion,that tetrapods are best compared with osteo-lepidids rather than more specialized osteo-lepiforms, has been reached by Panchen(1977) on vertebral structure, Rackoff (1980)on fin structure, and by Andrews and Thom-son (personal commun.).The polyplocodont teeth of Panderichthys

and tetrapods do not, therefore, appear to besynapomorphous. Nor are polyplocodontteeth synapomorphous within osteolepi-forms, since their distribution (Vorobyeva,1977c, table 1: Gyroptychiinae, Megistole-pididae, Eusthenopteridae, Rhizodopsidae,Rhizodontidae) cuts across every classifica-tion (e.g., Andrews, 1973; Vorobyeva,1977a). Occurrence of such teeth in lepisos-teids and other actinopterygians such as An-arhichas (personal observ.) only emphasizesthis.

SYNAPOMORPHY SCHEME

Comparisons are among cladistically ple-siomorphous or the only well-known mem-bers of each group: Acanthodes bronni (foracanthodians), Eusthenopteron foordi (forosteolepiforms), Latimeria chalumnae (foractinistians), Polypterus (for cladistians),chondrosteans (for actinopterygians), Hy-nobiidae and Ichthyostegidae (for tetrapods),Neoceratodus forsteri (for living lungfishes)and Griphognathus whitei (for fossil lung-fishes). Chondrichthyan characters are se-lected from both sharks and chimaeras, ex-cluding from the comparisons apparentlyderived characters of each (pelvic bridge andposteriorly directed hypobranchials in sharks;fin endoskeleton in chimaeras; etc.). Placo-derms are not included; at least one, Pholi-dosteus, has a complex pectoral fin endo-skeleton without a definite metapterygialaxis (part of character B.7.), and another,Coccosteus, has a sclerotic ring of fourplates (character C-I. 13.) (Moy-Thomas andMiles, 1971).

A. Acanthodes shares with other gnatho-stomes:1. The presence of a lower jaw sup-

ported by a palatoquadrate andhyomandibula.

2. A hyoid bar connecting the bran-chial apparatus with the hyoman-dibula.

3. Anterior branchial arches consist-ing of basibranchial, hypobranchi-

al, ceratobranchial, epibranchialand pharyngobranchial elements.

4. A cephalic lateral-line system thatincludes the following canal sec-tions: a supraorbital, infraorbital,supratemporal, mandibular, su-pramaxillary (jugal) that joinsthe infraorbital (absent in cladis-tians and actinopterygians), andpreopercular that rises toward thesupratemporal canal (connectionabsent in sarcopterygians).

5. Paired pectoral and pelvic append-ages with internal supporting gir-dles.

6. Three semicircular canals in theotic capsule.

B. Chondrichthyans also have the follow-ing derived feature which they sharewith osteichthyans:7. Complex endoskeletal support of

the paired appendages consistingof a metapterygial axis and radialsarticulating with the metapteryg-ium and girdle.

C-I. Actinopterygians have the foregoingcharacters and share with other os-teichthyans:8. Dermal skull bones with internal

processes (descending laminae)attaching to endocranium, and vari-ous toothed dermal bones on thepalate (see pp. 222, 227ff).

9. Endochondral bone.

1981 255


10. The presence of lepidotrichia inaddition to ceratotrichia in thefins.

11. Radials of the fins never extendingto the fin margin.

12. With two or fewer preaxial radialsassociated with the first two meta-pterygial segments of the pectoralfin endoskeleton.

13. Dermal sclerotic ring (primitivelyof four plates (see below, sectionD).

14. The presence of marginal upperand lower jaw teeth (dental ar-cades) on dermal bones lateral topalatoquadrate (as contrasted withteeth fused with, or attached to,the palatoquadrate and Meckel'scartilage in A and B).

15. Interhyal (=stylohyal) present.16. A forward, rather than backward,

orientation of the infrapharyngo-branchials.

17. Suprapharyngobranchials presenton first two gill arches.

18. Gill arches 1 and 2 articulating onthe same basibranchial.

19. A pneumatic or buoyancy organ asan outpocketing of the anterior gut(swimbladder, lung).

C-II. Cladistians (=Brachiopterygii) alsouniquely share with primitive actin-opterygians:20. The absence of the supramaxillary

(jugal) canal joining the infraorbitalcanal.

21. A differentiated propterygium inthe pectoral fin.

22. Modification of the pelvic fin me-tapterygium to form part or all ofthe pelvic girdle, and the endo-skeletal support of the pelvic finexclusively by preaxial pelvic ra-dials.

23. Acrodin caps on all teeth (0rvig,1976).

D. Eusthenopteron has the characters ofA, B and C-I (but character 19 notknown) and shares the following de-rived features with actinistians, poro-lepiforms, dipnoans and tetrapods:

24. An exclusively metapterygial pec-toral and pelvic fin (loss of all ra-dials anterior to base of metapter-ygium) supported by a single basalelement, and an anocleithrum inthe shoulder girdle suspension(see pp. 206, 207).

25. Absence of the connection be-tween the preopercular and supra-temporal lateral-line canal.

26. Enamel on teeth (Smith, 1979).27. Sclerotic ring composed of more

than four plates (see Miles, 1977,pp. 249-250).

E. Actinistians have the characters of A,B, C-I, and D and share with dipnoansand tetrapods:28. Anocleithrum of dermal shoulder

girdle sunken beneath dermis andwith no trace of surface ornamen-tation.

29. Clavicle large relative to cleithrum,and pectoral appendage insertionhigh as compared with these con-ditions in C-I, C-TI, and D.

30. Pectoral and pelvic appendageseach with long, muscular lobesthat extend well below ventralbody wall and with structurallysimilar endoskeletal supports.

31. Preaxial side of pectoral fin endo-skeleton rotated to postaxial po-sition (Braus, 1901; Romer andByrne, 1931).

32. A series of canal-bearing bones(for the supraorbital canal) lateralto large, paired, roofing bones be-tween orbits.

33. Absence of hyostyly in the jawsuspension (the reduced, subtrian-gular hyomandibula and interhyaldecoupled from, and not in directsequence with, the hyoid bar, andonly loosely connected to thesymplectic).

34. Posterior margin of palatoquad-rate erect or sloping backwardrather than sharply forward fromthe jaw articulation.

35. Presence of a rostral organ (a pos-sible synapomorphy with the la-

256 VOL. 167


bial cavity of dipnoans and naso-lacrimal duct of tetrapods, asdiscussed above, but recognizedhere as an ambiguous character).

36. With only a single, broad, trian-gular basibranchial (cladistianshave a similar basibranchial, per-haps associated with the loss ofone gill arch; a single elongate ba-sibranchial is present also in prim-itive paleoniscids; Gardiner, 1973).

37. Last gill arch articulates with baseof preceding arch rather than withbasibranchial.

38. Reduction or loss of hypobranchi-als.

39. Reduction or loss of pharyngo-branchials.

40. An inferior vena cava (Lauder,1979).

41. A pulmonary vein.

Porolepiforms (Porolepis, Glyptolepis, Lac-cognathus, and Holoptychius are the bestknown of these fishes) are very incompletelyknown but have characters reported in theliterature some of which would place thesefishes in a position plesiomorphous, and oth-ers, apomorphous, to actinistians. Nothingis known of their dorsal gill arches and al-most nothing is known of their endoskeletalfin supports. They might be regarded as ple-siomorphous to actinistians because of eyeposition in relation to paired dermal roofingbones and because the pelvic appendage isvery much shorter than the pectoral, al-though the scale-covered (and presumablyfleshy) fin bases are longer than those of os-teolepiforms and extend well below the ven-tral body wall. They agree with actinistians,dipnoans and primitive tetrapods (e.g., ich-thyostegids) in having a relatively large clav-icle and an embedded unornamented ano-cleithrum, a high pectoral appendageinsertion, an expanded parasphenoid, only asingle, compact basibranchial, and the lastgill arch (or the last two in Laccognathus;Vorobyeva, 1980) articulating with the baseof the preceding arch, not with the basi-branchial where these structures are known.

They agree only with dipnoans in having along, leaf-shaped pectoral, and in details ofthe cosmine pore-canal system (in a compar-ison of Porolepis and Dipterus; Schultze,1969a, fig. 42). At most, one can say thatthey form an unresolved trichotomy with ac-tinistians and dipnoans + tetrapods.

F. Dipnoans have all the foregoing char-acters excluding those of C-TI, anduniquely share with primitive tetra-pods:42. A choana.43. A labial cavity (a possible synap-

omorphy of the nasolacrimal duct,as discussed above, but regardedhere as an ambiguous character).

44. Second metapterygial segment ofpaired appendages composed ofpaired, subequal elements that arefunctionally joined distally.

45. Two primary joints in each pairedappendage, between endoskeletalgirdle and unpaired basal elementand between basal element andpaired elements of second seg-ment. In pectoral appendage,preaxial member of paired ele-ments with a ball and socket jointwith basal element and postaxialmember articulating on dorsal(postaxial) margin of basal ele-ment.

46. Reduction in ratio of dermal finrays to endoskeletal supports inpaired appendages.

47. Tetrapodous locomotion in livingrepresentatives (Dean, 1903;Lindsey, 1978; personal observ.).

48. Muscles in paired fins segmented(Braus, 1901, pl. 21; in contrast tothe non-segmented muscles in ac-tinistians, actinopterygians, cladis-tians and chondrichthyans).

49. Fusion of right and left pelvic gir-dles to form pubic and ischial pro-cesses (fig. 61).

50. Hyomandibula plays no part injaw suspension. De Beer (1937)described the condition of Neo-ceratodus, as follows: "The orbi-

1981 257


PUBIC CARTILAGE

A

FIG. 61. Pelvic girdle of A, Neoceratodusforsteri (Krefft), 18 cm. total length (AMNH 36982); andB, Necturus maculosus (Rafinesque), (from Gilbert, 1973). A is entirely cartilaginous.

totemporal region is different fromthat of any other (non-Dipnoan)fish owing to the autostylic attach-ment of the quadrate by means ofbasal, otic, and ascending pro-cesses. The basal connexion isventral to the head vein and dorsaland anterior to the palatine branchof the facial nerve. The otic con-nexion is lateral to the head veinand orbital artery, posterior to thetrigeminal nerve (and ophthalmicand buccal branches of the facialnerves), and anterior to the hyo-mandibular branch of the facialnerve. The ascending connexionis lateral to a branch of the headvein, postero-lateral to the abdu-cens nerve and the ophthalmic pro-fundus nerve (V1), but antero-me-dial to the maxillary andmandibular branches of the tri-

geminal nerve (V. and V3), and tothe orbital artery. All these rela-tions are typical and practicallyconstant throughout Tetrapods."

51. Loss of interhyal (=stylohyal).52. Hyomandibula reduced and, at

least in extant forms, joined dor-sally by ligament and closely as-sociated with an otic recess in thebraincase, and ventrally withquadrate.

53. Loss of dorsal gill arch elements(pharyngobranchials).

54. Opercular bone, in extant forms,reduced and joined posteriorly bya division of the epaxial muscula-ture to the upper half of the pri-mary shoulder girdle.

55. Right and left pterygoids joined inmidline, excluding the parasphe-noid anteriorly from the roof ofthe mouth.

258 VOL. 167


56. Loss of autopalatine.57. Elongation of the snout region.58. Eye somewhat posterior in skull,

at level of junction between twoprincipal bones in skull roof.

59. Pattern of five dermal bones cov-ering otic and occipital regions ofbraincase.

60. Resorption of calcified cartilageand subsequent deposition ofbony tissue in the vertebrae of theDevonian Griphognathus is iden-tical with the formation of bonytissue in uncalcified cartilage inTetrapoda, according to Schultze(1970b, p. 329).

61. There are, of course, numerousfeatures of soft anatomy, devel-opment, and physiology in whichdipnoans uniquely resemble uro-deles and other tetrapods. Someof the more striking of these fea-tures are the internal structure ofthe lung (Johansen, 1970), thepresence of a pulmonary circula-tion, the development of a two-chambered auricle and the ventralaorta as a truncus, the telolecithaldevelopment of the jelly-coatedbipolar egg, ciliation of the larva(Whiting and Bone, 1980), pres-ence of a glottis and epiglottis,flask glands, pituitary structureand a tetrapod neurohypophysialhormone (Perks, 1969; Archer,Chauvet, and Chauvet 1970), lensproteins and bile salts (L0vtrup,1977), and gill arch muscles (Wi-ley, 1979).

Within the framework of this theory of re-lationships (fig. 62), three features of sarcop-terygians deserve some comment: the intra-cranial joint, ventral gill arch muscles, andthe rhipidistian "choana."When present, the intracranial joint sepa-

rates the anterior part of the neurocranium,including the orbit, from the posterior part,including the brain minus the olfactory lobes,and it corresponds, at least ventrally, to anembryonic division of the neurocranium into

trabecular (anterior) and parachordal (pos-terior) cartilaginous primordia. In "rhipidis-tian" crossopterygians, the joint may bekinetic (as in Ectosteorhachis, where a peg-and-socket are present in the skull roof), aki-netic (as in Eusthenopteron, where no hingemechanism has been found), or perhaps evenabsent (as in Panderichthys, Elpistostegeand Powichthys, where there is no externalindication that a joint exists; Jessen, 1975).In actinistians there is a well-developed ki-netic joint that moves on a ventrally situatedgroove and track, but dorsally the joint sep-arates the cranium more posteriorly than itdoes in rhipidistians (Bjerring, 1973), and theclaim has been made that the actinistian jointarose independently of that in "rhipidis-tians." Gardiner and Bartram (1977; alsoGardiner 1973), on the other hand, in a dis-cussion of the primitive subdivision of theactinopterygian cranium along a cartilage-filled, staggered groove, claimed that theventral part of this groove (the ventral oticfissure) is homologous with the ventral partof the intracranial joint. Both a ventral oticfissure and a lateral occipital fissure are alsopresent in Acanthodes so that the actinop-terygian condition might be primitive forgnathostomes. Bjerring's claim that thejoints are fundamentally different in actinis-tians and rhipidistians is unconvincing be-cause in both groups the dorsal part of thejoint is just anterior to the emergence of thefifth and seventh nerves from the endocra-nium and because, ventrally, thejoint is eitherat or slightly in front of the end of the no-tochord. Differences between the two areperhaps most easily related to the probablydifferent degrees of, and certainly radicallydifferent mechanisms for, kineticism. Thetwo kinds of joints are either synapomor-phous, thus specifying a sister-group rela-tionship between "rhipidistians" and acti-nistians, or some form of a joint is primitivefor sarcopterygians and has been lost in dip-noans and tetrapods (the claim that a rem-nant of the joint is present in Ichthyostegais not yet supported by published evidence).We cannot ourselves resolve these questionson the basis of a study of the anatomical fea-tures of the joint in different animals, but weobserve that if there are different degrees of

1981 259


GNATHOSTOMATA

0 STE IC HT HYES 1

ACTINOPTERYGII SARCOPTERYGI

ACT IN OPTERI

8 -41

.24 - 27

I

7 1 = t Acanthodes bronni

2 = t Eusthenopteron foordi

3 = t Porolepiformes

FIG. 62. Character-state tree of major groups of gnathostomes. Numbered characters refer to syn-apomorphy scheme in text. Unresolved cladistic position of porolepiforms is discussed in synapomorphyscheme.

joint development in rhipidistians, and rhip-idistians are not a monophyletic group as wehave every reason to suspect, then actinis-tians would simply represent the last of a

perhaps large series of sister groups whosepositions relative to each other in the hier-archy would be specified by the degree ofkineticism. A parsimony argument would,therefore, require that the joint be consid-ered either an autapomorphy of actinistians,as suggested by Bjerring (1973), or a primitivefeature of sarcopterygians (Miles, 1977). Wi-ley (1979) also reviewed the problem andconcluded (p. 171) "that both Bjerring andMiles are partly correct. Bjerring (1973) is cor-rect in his assertion that the dorsal part of theintracranial joint is non-homologous [in rhipi-distians and actinistians]. Miles (1977) iscorrect in accepting the ventral part of theintracranial joint as homologous. Bjerring isincorrect in asserting that the entire intra-

cranial joint is non-homologous and, as amatter of parsimony, Miles is incorrect inasserting that the entire intracranial jointis homologous."

Wiley (1979) also proposed in a study ofventral gill arch muscles that dipnoans andtetrapods are sister groups but that actinis-tians are excluded from membership in eitherthe Sarcopterygii or Actinopterygii. He ad-vocated that the Actinistia are instead thesister group of all other osteichthyans. Thelatter conclusion was based on his observa-tion that three conditions of the musculatureare present as a synapomorphy of actinop-terygians plus sarcopterygians but areabsent in actinistians. These three conditionsare (1) an obliquus ventralis 1; (2) a trans-versus ventralis 4; and (3) the pharyngocla-viculares. Although we have not ourselvesstudied this musculature, we admit to somesurprise at the absence of three myological

*1-6

260 VOL. 167

I I


features of Latimeria that occur in dipnoansand urodeles, primarily because four of the14 synapomorphies we propose for unitingactinistians with dipnoans and tetrapods arein gill arch characters, three of which are fea-tures of the ventral gill arch skeleton. Forthe present, it is parsimony rather than astudy of Wiley's "absence" characters thatforces us to interpret the absence of thesemuscles in Latimeria as autapomorphic. Wi-ley's evidence that "refutes" a link betweenLatimeria and our Choanata includes notonly the three "absence" characters, butalso two muscles uniquely shared by dip-noans and urodeles: a subarcualis rectus andmultiple pharyngoclaviculares. For us, how-ever, these last two conditions of the mus-culature only further corroborate our Choa-nata and say nothing about the position ofLatimeria.

Finally, a word about the "choana" ofEusthenopteron is in order. We have tried toshow that study of new material of the Eu-sthenopteron palate allows other interpreta-tions of the palatal fenestra which is notablysmaller in our specimens than in Jarvik's:(1) as a space for the passage of nerves andvessels as seen in some Recent fishes; (2) asa space allowing movement along a jointsurface; and (3) as a space for receivingthe tip of a coronoid fang when the mouthis closed. One might even suppose, sincethe palatines and pterygoids were dis-placed anteriorly under the vomer in Jarvik'sspecimen, probably a result of dorsoventralcrushing, that the much larger palatal fenes-tra observed by Jarvik was caused mechan-ically by an abnormally deep penetration ofa coronoid fang into the palate. Whatever thereality of the Eusthenopteron palate mighthave been, there is another view of the prob-lem which it is helpful to remember: to ac-cept the position of Eusthenopteron as a sis-ter group of tetrapods based on this featureand even the alleged "septomaxilla" and"nasolacrimal duct," it would be necessaryto reject at least 33 characters (nos. 28through 61 in our synapomorphy scheme)which unite actinistians, dipnoans and tet-rapods. Some of these 33 characters wouldalso dichotomously resolve the position of

porolepiforms in this sequence if more couldbe learned about details of their anatomy-and that porolepiforms belong somewhere inthis sequence seems evident to us. What thishas meant to us, since actinistians and someporolepiforms have two external narial open-ings in the snout, is that the single narialopening in the snout of osteolepiforms, andEusthenopteron in particular, is reasonablyinterpreted as having housed a commonchamber for superficial membranous canalsof two external nostrils, as in Recent Polyp-terus (compare figs. 11 and 17A, B). Ac-cepting this interpretation, there is no longera reason to search for a missing fenestra exo-narina posterior in the palate of Eusthe-nopteron and other osteolepiforms. Appli-cation of parsimony in constructing theoriesof relationships among tetrapods, dipnoans,actinistians and "rhipidistians" could longago have resolved the question of the rhipi-distian "choana" with the ample data al-ready in the literature.Our conclusions, based on the foregoing

synapomorphy scheme are depicted in figure62, and summarized in the classification be-low; fossils are incorporated according to themethod described by Patterson and Rosen(1977).

SUPERCLASS Gnathostomata Gegenbaur, 1874plesion tAcanthodes bronni Agassiz, 1832CLASS Chondrichthyes Huxley, 1880SUBCLASS Selachii Latreille, 1825SUBCLASS Holocephali Bonaparte, 1832

CLASS Osteichthyes Huxley, 1880SUBCLASS Actinopterygii Cope, 1891INFRACLASS Cladistia Cope, 1871INFRACLASS Actinopteri Cope, 1871SERIES Chondrostei Muller, 1844SERIES Neopterygii Regan, 1923DIVISION Ginglymodi Cope, 1871DIVISION Halecostomi Regan, 1923

SUBCLASS Sarcopterygii Romer, 1955Sarcopterygii incertae sedis t Porolepiformes

Jarvik, 1942plesion tEusthenopteron foordi Whiteaves,

1881INFRACLASS Actinistia Cope, 1871INFRACLASS Choanata Save-Soder-

bergh, 1934SERIES Dipnoi Muller, 1844SERIES Tetrapoda

2611981


In the classification above we use two verysimilar names intentionally: Actinopterygii,Actinopteri. Actinopteri is Cope's originalname for actinopterygians sensu stricto, ex-cluding Cladistia. Actinopterygii is todaycommonly used sensu Goodrich (1930), to

include Cladistia. The distinction we suggestwill allow continued use of the vernacular"actinopt," and it will only be necessary tospecify Actinopterygii or Actinopteri whenthe inclusive (+Cladistia) and exclusive(-Cladistia) groups are to be distinguished.

CONCLUSIONSGoodrich (1909, p. 230), writing of the

"many striking points of resemblance" be-tween lungfishes and amphibians, comment-ed that they "cannot all be put down to con-vergence." Goodrich was careful todistinguish primitive and derived characters,and the "many striking points of resem-blance" were, in his view, derived. SinceGoodrich's time, the number of apparentsynapomorphies between lungfishes and tet-rapods has increased rather than decreased,and the question still remains, can they allbe put down to convergence? The consensustoday is that these resemblances are conver-gent or adaptive, since they are outweighedby detailed correspondence between Paleo-zoic amphibians and osteolepiform fishes.Our purpose in this paper has been to opposethat view, by reemphasizing the shared de-rived characters of lungfishes and tetrapods,and by proposing alternative explanations ofthe similarities between osteolepiforms andtetrapods-that they are convergent or prim-itive.Eusthenopteron is, after all, a very primi-

tive fish, as Jarvik (e.g., 1968b, p. 506) hasoften said. Even so, it seems not to be prim-itive enough to serve as a morphotype of theancestral tetrapod, since others who havesurveyed a wide range of osteolepiforms(Vorobyeva, Andrews, Thomson, Panchen)have concluded that it is osteolepiforms asa whole, or the earliest osteolepidids, thatare closest to tetrapods, rather than any de-rived osteolepiform subgroup. These theo-ries of tetrapod origins follow the traditionalpaleontological method of the search forancestors-reliance on paraphyletic groups.In the approach we have followed, thesearch for synapomorphies specifying sister-

group relationships, there is nothing to sayabout osteolepiforms as a whole, or osteo-lepidids, since they seem to lack synapo-morphies, and therefore present the prob-lems peculiar to extinct paraphyletic groups.Previous attempts to specify the relationshipbetween osteolepiforms and tetrapods interms of synapomorphies (Szarski, 1977;Schultze, 1977a; Gaffney, 1979a, 1979b)have avoided the problem of characterizingosteolepiforms, and have relied almost ex-clusively on the choana, a character whichwe have found to be open to other interpre-tations. If we select Eusthenopteron foordias the best known osteolepiform, we cantake that paleospecies, lungfishes, and tet-rapods as three monophyletic groups, andlist the apparent synapomorphies betweeneach pair. Eusthenopteron and lungfishesshare no plausible synapomorphies apartfrom the elongate basihyal (sublingual rod,also found in Griphognathus). Among theapparent synapomorphies between Eusthe-nopteron and tetrapods (p. 177), thosethat have withstood our criticism best are thepattern of dermal bones on the cheek (fig.43), and the structure of the humerus. Lung-fishes and tetrapods share derived featuresof the choana, dermal roofing bones of theskull, palate and jaw suspension, hyoid arch,gill arches, skull-vertebra articulation, ribs,dermal shoulder girdle, and paired append-ages, in addition to numerous features of softanatomy, physiology, embryology, and be-havior which are not checkable in Eusthe-nopteron. The principle ofparsimony leads usto accept these numerous derived features assynapomorphies. It follows that the charac-ters shared by Eusthenopteron and tetrapodsare primitive for sarcopterygians, conver-

262 VOL. 167


gent or spurious. Abstracting from the list ofthose characters which were claimed to beexclusive to Eusthenopteron and tetrapods(p. 177), our assessment is as follows: primi-tive-cheek pattern, "protorhachitomous"vertebrae; convergent-polyplocodont teeth,structure of humerus; spurious-fenestraexochoanalis, septomaxilla.

In those assessments, we mean to arguean alternative view, not to insist that we areright. Those who are still impressed by re-semblances between Eusthenopteron, orother osteolepiforms, and tetrapods, mayamplify and enumerate those resemblances,and seek to show that they outnumber oroutweigh the lungfish/tetrapod synapomor-phies that we cite. We hope they will do so.In the course of this paper we have criticizedmany people and many ideas. Our criticismsstem from our viewpoint, that of phyloge-netic systematics. Some of those we criticizehave already published their opinion of thatviewpoint (e.g., Panchen, 1979; Westoll,1979, p. 346). We suggest that little will begained by denigrating logic (Westoll), or de-nying the principle of parsimony (Panchen).Nor will progress come from simply reas-serting general similarities between particu-lar osteolepiforms and particular tetrapods,or between an osteolepiform morphotypeand a tetrapod morphotype, unless thosesimilarities are justified as synapomorphiesby means of an explicit scheme of relation-ships, covering tetrapod subgroups and otherrelevant groups such as lungfishes and coel-acanths.

In the 140 years since living lungfisheswere first discovered, there has been evi-dence of a consistent pattern of derived char-acters shared by lungfishes and tetrapods.Nevertheless, during this century, that pat-tern has been treated as irrelevant or hasbeen explained away, since the problem oftetrapod relationships has become one oforigins, the property of paleontology, solvedby the process approach: following backstratigraphic series of tetrapods and findingthat they converge on a morphotype plausi-bly attributed to osteolepiforms. But whenlungfishes are treated by the same method,they also converge on osteolepiforms (Den-

ison, 1968a); so do actinopterygians (Wes-toll, 1979, p. 345). Ancestors are slipperythings, but ancestral groups are very accom-modating, for their only attributes are plesio-morphous.

It seems fitting, therefore, to end our as-sault on the "rhipidistian barrier" by record-ing some personal observations, made nearthe completion of this study, of five partic-ular locomotor and feeding behaviors of twolarge lungfishes (a Protopterus annectensand Neoceratodus forsteri, each over a me-ter long)12 and a small Neoceratodus ofabout 5 centimeters.13 Needless to say,we record them because of their similarityto behaviors of aquatic urodeles. (1) Swim-ming is accomplished by undulations of thecaudal peduncle while the paired fins areheld loosely near the body rather than beingused as maneuvering planes, although Neo-ceratodus will sometimes extend the pecto-rals somewhat. (2) Stopping is accomplishedby sinking to the substrate on the extendedpaired fins; the pectorals are moved forwardto an angle of 90 degrees, or slightly more,with the body axis, and the pelvics are oftenmoved forward sharply (as much as 1200).Thus at rest, with the pelvics firmly down onthe substrate, the fish may raise the entireprecaudal region, sitting in this raised posi-tion for some minutes, with the pectoralshanging loosely downward. (3) When pro-gressing forward slowly, the weight of thebody is transferred to the paired fins whichlift the body and pull it forward by movingin a side-to-side alternating diagonal pattern.During each cycle of fin movement, the bodychanges in curvature (when LF-RH fins areprotracted the right side of the body is con-cave, whereas when RF-LH fins are pro-tracted the right side of the body is convex).(4) When performing special movementsacross the substrate, such as turning or back-ing, right and left fins can be moved forwardand backward synchronously or one fin canbe moved independently of the other three.(5) When feeding from the substrate, the fore-

12 In the New York Aquarium of the New York Zoo-logical Society.

13 In the Department of Ichthyology of the AmericanMuseum of Natural History.

1981 263


part of the body is elevated on the pectoralfins and the head is bent down toward thefood; if the food consists of live worms inthe substrate, small Neoceratodus will graspthe worms in its mouth and push backwardwith its pectorals until the worms are dis-lodged. None of the above observations isoriginal with us; the same or similar behav-iors were noted and recorded many times bylate nineteenth-century and early twentieth-century biologists who were no less im-pressed than we by the remarkable behav-ioral likenesses of lungfishes and urodeles.We record our own fascination with theselikenesses for a particular reason-the con-viction that, had the study of lungfish rela-tionships developed directly from compari-sons among living gnathostomes withoutinterruption by futile paleontological search-es for ancestors, there would have been no"rhipidistian barrier" to break through, thisreview would have been written decadesago, and we would all have been spared ageneration of anatomical misconstruction,dispensable scenarios about the fish-amphib-ian transition, and hapless appeals to plesio-morphy. Finally, it should be obvious fromthe nature of this review that our argumentis not with the use of fossils, but with the useof the traditional paleontological method.

ACKNOWLEDGMENTSWe especially thank Dr. Gareth Nelson for

numerous discussions of the general prob-lem, Drs. Mahala Andrews, James Atz, Eu-

gene Gaffney, Philippe Janvier, ArnoldKluge, John Maisey, Roger Miles, AngelaMilner, Charles Myers, Alec Panchen, BobPresley, Wolf-Ernst Reif, Bobb Schaeffer,Hans-Peter Schultze, Keith Thomson, E. 0.Wiley, and Richard Zweifel for providing vari-ous significant kinds of information and ad-vice. Mr. Guido Dingerkus, Ms. Norma Fein-berg, Ms. Lynne Parenti, Ms. Amy Rudnickand, in particular, Mr. Peter Whybrow provid-ed countless hours of technical assistance. Weare grateful to Dr. Anne Kemp for supplyingliving and preserved juvenile specimens ofNeoceratodus, and to Mr. Guido Dingerkusand the New York Aquarium of the NewYork Zoological Society for making possibleobservations of large, living African andAustralian lungfishes. We also thank Dr. P.Humphry Greenwood and Mr. GordonHowes, Division of Fishes, British Museum(Natural History) (BMNH), Dr. WilliamFink, Museum of Comparative Zoology,Harvard University, Dr. Stanley Weitzman,United States National Museum, Drs. BobbSchaeffer, John Maisey, and Richard Zwei-fel, all of the American Museum of NaturalHistory (AMNH), Dr. William N. Eschmeyer,California Academy of Sciences, and Dr. E.C. Boterenbrood, Hubrecht Laboratory,Utrecht, for lending specimens in their care.The photographs were taken by Mr. TimParmenter, British Museum (Natural His-tory) and the staff of the Department ofPhotography, American Museum of NaturalHistory.

LITERATURE CITED

Agassiz, J. L. R.1838. Recherches sur les poissons fossiles.

Neuchatel, vol. 3, p. 129.1843. Recherches sur les poissons fossiles.

Neuchatel, vol. 2, pp. 2-46.Aldinger, H.

1937. Permische Ganoidfische aus Ostgr0n-land. Meddel. Gr0nland, vol. 102, pp.1-392.

Allis, E. P., Jr.1919. The lips and the nasal apertures in the

gnathostome fishes. Jour. Morph., vol.32, no. 1, pp. 145-205.

1922. The cranial anatomy ofPolypterus, with

special reference to Polypterus bichir.Jour. Anat., vol. 56, pts. 3 and 4, pp.190-294.

1923. The cranial anatomy of Chlamydosela-chus anguineus. Acta Zool. Stockholm,vol. 4, pp. 123-221.

1932a. Concerning the nasal apertures, thelachrymal canal and the buccopharyn-geal upper lip. Jour. Anat., vol. 66, pt.4, pp. 650-658.

1932b. The pre-oral gut, the buccal cavity andthe buccopharyngeal opening in Cerat-odus. Ibid., vol. 66, pt. 4, pp. 659-668.

264 VOL. 167


Andrews, S. M.1973. Interrelationships of crossopterygians.

In Greenwood, P. H., R. S. Miles andC. Patterson (eds.), Interrelationshipsof fishes. London, Academic Press, pp.137-177.

1977. The axial skeleton of the coelacanthLatimeria. In Andrews, S. M., R. S.Miles and A. D. Walker (eds.), Prob-lems in vertebrate evolution. London,Academic Press, pp. 271-288.

Andrews, S. M., and T. S. Westoll1970a. The postcranial skeleton of Eusthenop-

teron foordi Whiteaves. Trans. Roy.Soc. Edinburgh, vol. 68, no. 9, pp. 207-329.

1970b. The postcranial skeleton of rhipidistianfishes excluding Eusthenopteron. Ibid.,vol. 68, no. 12, pp. 391-489.

Archer, R., J. Chauvet, and M. T. Chauvet1970. A tetrapod neurohypophyseal hormone

in African lungfishes. Nature, vol. 227,no. 5254, pp. 186-187.

Asmuss, H.1856. Das vollkommenste Hautskelet der

bisher bekannten Thierreihe. Abhandl.Erlangung der Magisterwurde, Dorpat.,40 pp.

Baur, G.1896. The Stegocephali. A phylogenetic study.

Anat. Anz., vol. 11, pp. 657-675.Beaumont, E.

1977. Cranial morphology of the Loxomma-tidae (Amphibia: Labyrinthodontia).Phil. Trans. Roy. Soc. London (B), vol.280, pp. 29-101.

Beer, G. R., de1937. The development of the vertebrate

skull. Oxford Univ. Press, pp. xxiv +552.

Bertmar, G.1965. The olfactory organ and upper lips in

Dipnoi, an embryological study. ActaZool. Stockholm, vol. 46, pp. 1-10.

1966. On the ontogeny and homology of thechoanal tubes and choanae in Urodela.Ibid., vol. 47, pp. 43-59.

1968. Lungfish phylogeny. In 0rvig, T. (ed.),Current problems of lower vertebratephylogeny. Stockholm, Almqvist andWiksell, pp. 259-283.

Bischoff, T. L. W., von1840a. Lepidosiren paradoxa. Anatomisch un-

tersucht und Beschrieben. Leipzig, vi+ 34 pp.

1840b. Description anatomique du Lepidosirenparadoxa. Ann. Sci. Nat., ser. 2, vol.14, no. 2, pp. 116-159.

Bjerring, H. C.1967. Does a homology exist between the

basicranial muscle and the polar carti-lage? Colloques Internatl. Centre Natl.Rech. Sci., vol. 163, pp. 223-267.

1973. Relationships of coelacanthiforms. InGreenwood, P. H., R. S. Miles and C.Patterson (eds.), Interrelationships offishes. London, Academic Press, pp.179-205.

1977. A contribution to structural analysis ofthe head of craniate animals. Zool.Scripta, vol. 6, pp. 127-183.

Boas, J. E. V.1882. Uber den Conus arteriosus und die Ar-

terienbogen der Amphibien. Morph.Jahrb., vol. 7, pp. 488-572.

Boulenger, G. A.1901. Les poissons du Bassin du Congo.

Brussels, lxii + 532 pp.Braus, H.

1901. Die Muskeln und Nerven der Cerato-dusflosse. In Semon, R., ZoologischeForschungsreisen in Australien unddem Malayischen Archipelago. I. Ce-ratodus. Pt. 3. Jena, G. Fischer Verlag,pp. 139-300.

Bridge, T. W.1897. The morphology of the skull in the Par-

aguayan Lepidosiren and other Dipnoi.Trans. Zool. Soc. London, vol. 14, pp.325-376.

1904. Fishes (exclusive of the systematic ac-count of Teleostei). In Harmer, S. F.,and A. E. Shipley (eds.), Cambridgenatural history. London, Macmillan andCo., vol. 7, pp. 139-537.

Broman, I.1939. Uber die Entwicklung der Geruchsor-

gane bei den Lungenfischen. Morph.Jahrb., vol. 83, pp. 85-106.

Broom, R.1895. On the homology of the palatine process

of the mammalian premaxillary. Proc.Linnean Soc. New South Wales, vol.10, pp. 477-485.

Bryant, W. L.1919. On the structure of Eusthenopteron.

Bull. Buffalo Soc. Nat. Sci., vol. 13, no.1, pp. 4-22.

Bystrov, A. P.1935. Morphologische Untersuchungen der

1981 265


Dekknochen des Schadels der Wirbel-tiere. Acta Zool. Stockholm, vol. 16, pp.65-141.

Bystrov, A. P., and J. A. Efremov1940. Benthosuchus sushkini-a labyrintho-

dont from the Eotriassic of SharzhengaRiver. Trudy Paleozool. Inst., vol. 10,pp. 1-152. [In Russian, Englishsummary].

Carrol, R. L., and R. Holmes1980. The skull and jaw musculature as guides

to the ancestry of salamanders. Zool.Jour. Linnean Soc., vol. 68, pp. 1-40.

Cope, E. D.1872. Observations on the systematic rela-

tions of the fishes. Proc. Amer. Assoc.Adv. Sci., vol. 20, pp. 317-343.

1884. Note on the phylogeny of the Vertebra-ta. Amer. Nat., vol. 18, pp. 1255-1257.

1885. On the evolution of the Vertebrata, pro-gressive and retrogressive. Ibid., vol.19, pp. 140-148; 234-247.

1887. Geology and palaeontology. Ibid., vol.21, pp. 1014-1019.

1892. On the phylogeny of the Vertebrata.Proc. Amer. Phil. Soc., vol. 30, pp. 278-281.

Daget, J.1950. Revision des affinites phylogenetiques

des Polypterides. Mem. Inst. Fr. Afr.Noire, vol. 11, pp. 1-178.

Daniel, J. F.1928. The elasmobranch fishes. Second edi-

tion. Berkeley, Univ. California Press,xi + 332 pp.

Dean, B.1903. Obituary notice of a lung-fish. Pop. Sci.

Monthly, vol. 43, pp. 33-39.Denison, R. H.

1968a. The evolutionary significance of the ear-liest known lungfish, Uranolophus. In0rvig, T. (ed.), Current problems oflower vertebrate phylogeny. Stock-holm, Almqvist and Wiksell, pp. 247-257.

1968b. Early Devonian lungfishes from Wyo-ming, Utah and Idaho. Fieldiana: Geol.,vol. 17, pp. 351-413.

1974. The structure and evolution of teeth inlungfishes. Ibid., vol. 33, pp. 31-58.

Devillers, C.1954. Les cotes. In Grasse, P. P. (ed.), Traite

de zoologie. Paris, Masson, vol. 12, pp.673-697.

Dollo, L.1896. Sur la phylogenie des dipneustes. Bull.

Soc. Belge Geol., Paleont., Hydrologie,vol. 9, no. 2, pp. 79-128.

Dumeril, A. H. A.1870. Le Lepidosiren et le Protoptere apartien-

nent a la classe des poissons ou ils sontles types de la sous-classe des dipnes.Ann. Soc. Linneenue Maine-et-Loire,vol. 12, pp. 140-149.

Edgeworth, F. H.1935. The cranial muscles of vertebrates.

Cambridge Univ. Press, pp. ix + 493.Edwards, H. Milne

1840. Remarque sur les affinites naturelles duLepidosiren. Ann. Sci. Nat. Ser. 2, vol.14, no. 2, pp. 159-162.

Emelianov, S. W.1935. Die Morphologie der Fischrippen. Zool.

Jahrb. (Anat.), vol. 60, pp. 132-262.1936. Die Morphologie der Tetrapodenrippen.

Ibid., vol. 62, pp. 173-274.Fitzinger, L. J. F. J.

1837. Vorlaufiger Bericht uber eine h6chst in-teressante Entdeckung Dr. Natterers inBrasil. Isis, Jena, pp. 379-380.

1843. Systema reptilium. Vienna, Braumulleret Seidel, pp. vi + 106.

Forey, P. L.1980. Latimeria: a paradoxical fish. Proc. Roy.

Soc. London, B. Biol. Sci., vol. 208, pp.369-384.

[In press] The coelacanth Rhabododerma fromthe Carboniferous of the British Isles.Palaeontology, vol. 24, pt. 1, pp. 203-229.

Forster-Cooper, C.1937. The Middle Devonian fish fauna of

Achanarras. Trans. Roy. Soc. Edin-burgh, vol. 59, pp. 223-239.

Fox, H.1965. Early development of the head and

pharynx of Neoceratodus with a consid-eration of its phylogeny. Jour. Zool.London, vol. 146, pp. 470-554.

Francois, Y.1966. Structure et developpement de la ver-

tebre de Salmo et des teleosteens. Paris,Archs. Zool. Exp. Gen., vol. 107, pp.287-328.

Fritsch, A.1883-1895. Fauna der Gaskohle und der Kalk-

steine der Permformation Bohmens.Prag, 1885, vol. II, pt. 1, pp. 1-32.

Fiurbringer, K.1904. Beitriige zur Morphologie des Skelettes

der Dipnoer, nebst Bemerkungen uberPleuracanthiden, Holocephalen und

266 VOL. 167


Squaliden. Denkschr. Med.-Naturwiss.Gesell. Jena, vol. 4, pp. 425-510.

Gadow, H. F.1933. The evolution of the vertebral column.

Cambridge Univ. Press, 356 pp.Gaffney, E. S.

1979a. Tetrapod monophyly: a phylogeneticanalysis. Bull. Carnegie Mus. Nat.Hist., vol. 13, pp. 92-105.

1979b. An introduction to the logic of phylog-eny reconstruction. In Cracraft, J., andN. Eldredge (eds.), Phylogenetic anal-ysis and paleontology. Columbia Univ.Press, pp. 79-111.

Gardiner, B. G.1973. Interrelationships of teleostomes. In

Greenwood, P. H., R. S. Miles and C.Patterson (eds.), Interrelationships offishes. London, Academic Press, pp.105-135.

Gardiner, B. G., and A. W. H. Bartram1977. The homologies of ventral cranial fis-

sures in osteichthyans. In Andrews, S.M., R. S. Miles and A. D. Walker(eds.), Problems in vertebrate evolu-tion. London, Academic Press, pp. 227-245.

Gegenbaur, C.1895. Das Flossenskelet der Crossopterygier

und das Archipterygium der Fische.Morph. Jahrb., vol. 22, pp. 119-160.

Gilbert, S. G.1973. Pictorial anatomy of the Necturus.

Univ. Washington Press, 47 pp.Gill, T.

1872. Arrangement of the families of fishes.Smithsonian Misc. Coll., vol. 247, pp.xlvi + 49.

Goodey, T.1910. A contribution to the skeletal anatomy

of the frilled shark, Chlamydoselachusanguineus Gar. Proc. Zool. Soc. Lon-don, vol. 19, pp. 340-571.

Goodrich, E. S.1901. On the pelvic girdle and fin of Neoce-

ratodus. Quart. Jour. Micros. Sci., vol.45, pp. 311-324.

1909. Vertebrata craniata. First fascicle: cy-clostomes and fishes. In Lankester, E.R. (ed.), A treatise on zoology. London,A. and C. Black, vol. 9, xvi + 518 pp.

1919. Restorations of the head of Osteolepis.Jour. Linnean Soc. (Zool.), vol..34, pp.181-188.

1924. The origin of land vertebrates. Nature,London, vol. 114, pp. 935-936.

1925. Cranial roofing-bones in Dipnoi. Jour.Linnean Soc. (Zool.), vol. 36, pp. 79-86.

1930. Studies on the structure and develop-ment of vertebrates. London, Macmil-lan and Co., pp. xxx + 837.

Gray, J. E.1856. Observations on a living African Lepido-

siren in the Crystal Palace. Proc. Zool.Soc. London, Pt. 24, pp. 342-346.

Gregory, W. K.1904. The relations of the anterior visceral

arches to the chondrocranium. Biol.Bull. Marine Biol. Lab. Woods Hole,vol. 7, pp. 55-69.

1911. The limbs of Eryops and the origin ofpaired limbs from fins. Science, newser., vol. 33, pp. 508-509.

1915. Present status of the problem of the or-igin of the Tetrapoda, with special ref-erence to the skull and paired limbs.Ann. New York Acad. Sci., vol. 26, pp.317-383.

Gregory, W. K., H. Rockwell, and F. G. Evans1939. Structure of vertebral column in Eus-

thenopteron foordi Whiteaves. Jour.Paleont., vol. 13, pp. 126-129.

Gross, W.1956. Uber Crossopterygier und Dipnoer aus

dem baltischen Oberdevon im Zusam-menhang einer vergleichenden Unter-suchung des Porenkanalsystems paliio-zoischer Agnathen und Fische. K.Svenska Vetensk. Akad. Handl., ser. 4,vol. 5, no. 6, pp. 1-140.

1962. Neuuntersuchung der Stensioellida (Ar-throdira, Unterdevon). Notizbl. Hess.Landesanst. Bedenforsch. Wiesbaden,vol. 90, pp. 48-86.

Gunther, A. C. L. G.1871. Description of Ceratodus, a genus of

ganoid fishes, recently discovered inrivers of Queensland, Australia. Phil.Trans. Roy. Soc. London, vol. 161, pp.511-571.

Gunther, A. E.1975. A Century of zoology at the British Mu-

seum. London, Dawsons, 533 pp.Haeckel, E.

1866. Generelle Morphologie der Organis-men. 2 vols. Berlin, G. Reimer, 574 +447 pp.

1889. Naturliche SchopfungsGeschichte.Eighth edition. Berlin, G. Reimer, xxx+ 832 pp.

1981 267


Heaton, M. J.1979. Cranial anatomy of primitive captorhi-

nid reptiles from the late Pennsylvanianand early Permian ofOklahoma and Tex-as. Bull. Oklahoma Geol. Surv., no.127, iv + 84 pp.

Heckel, J.1845. Bemerkung uber Lepidosiren paradoxa.

Arch. Anat. Physiol., pp. 534-535.1851. Uber einer neue Fisch-Species aus dem

Weissen Nil, Protopterus aethiopicus.Sitzber. Akad. Wiss. Wien, vol. 6, pp.685-689.

Helbing, H.1904. Beitrage zur Anatomie und Systematik

der Laemargiden. Nova Acta Leopol-dina, vol. 82, pp. 335-524.

Higgins, G. M.1920. The nasal organ in Amphibia. III. Biol.

Monogr., vol. 6, no. 1, 91 pp.Hogg, J.

1841. On the existence of branchiae in theyoung Caeciliae; and on a modificationand extension of the branchial classifi-cation of the Amphibia. Ann. Mag. Nat.Hist., vol. 7, pp. 353-363.

Holmgren, N.1933. On the origin of the tetrapod limb. Acta

Zool. Stockholm, vol. 14, pp. 185-295.1939. Contribution to the question of the ori-

gin of the tetrapod limb. Ibid., vol. 20,pp. 89-124.

1949a. Contributions to the question of the or-

igin of tetrapods. Ibid., vol. 30, pp.459-484.

1949b. On the tetrapod limb problem-again.Ibid., vol. 30, pp. 485-508.

Holmgren, N., and E. Stensio1936. Kranium und Visceral skelett der Ak-

ranier, Cyclostomen und Fische. In L.Bolk (ed.), Handb. Vergl. Anat., vol. 4,Berlin, Urban u. Schwarzenberg, pp.233-500.

Huxley, T. H.1861. Preliminary essay upon the systematic

arrangement of the fishes of the Devo-nian epoch. Mem. Geol. Surv. UnitedKingdom, decade 10, pp. 1-40.

1876. Contributions to morphology. Ichthy-opsida No. 1. On Ceratodus forsteri,with observations on the classificationof fishes. Proc. Zool. Soc. London, pp.24-59.

1880. On the application of the laws of evo-lution to the arrangement of the Verte-brata and more particularly of the Mam-malia. Ibid., pp. 649-662.

Hyrtl, J.1845. Lepidosiren paradoxa: Monographie.

Abhandl. Bohmens Ges. Wiss., vol. 3,no. 5, pp. 1-64.

Jardine, N.1970. The observational and theoretical com-

ponents of homology: a study based onthe morphology of the dermal skull-roofs of rhipidistian fishes. Biol. Jour.Linnean Soc., vol. 1, pp. 327-361.

Jarvik, E.1937. On the species of Eusthenopteron found

in Russia and the Baltic States. Bull.Geol. Inst. Uppsala, vol. 27, pp. 63-127.

1942. On the structure of the snout of cros-sopterygians and lower gnathostomes ingeneral. Zool. Bidrag frtn Uppsala, vol.21, pp. 235-675.

1944a. On the dermal bones, sensory canalsand pit-lines of the skull in Eusthenop-teron foordi Whiteaves, with some re-marks on E. save-soderberghi Jarvik.K. Svenska Vetensk. Akad. Handl.,ser. 3, vol. 21, no. 3, pp. 1-48.

1944b. On the exoskeletal shoulder-girdle of te-leostomian fishes, with special refer-ence to Eusthenopteron foordi White-aves. Ibid., ser. 3, vol. 21, no. 7, pp. 1-32.

1948. On the morphology and taxonomy ofthe Middle Devonian osteolepid fishesof Scotland. Ibid., ser. 3, vol. 25, no. 3,pp. 1-301.

1950. Middle Devonian vertebrates from Can-ning Land and Wegeners Halvo (EastGreenland). Pt. II. Crossopterygii.Meddel. Gr0nland, vol. 96, no. 4, pp.1-132.

1952. On the fish-like tail in the ichthyostegidstegocephalians with descriptions of anew stegocephalian and a new crossop-terygian from the Upper Devonian ofEast Greenland. Ibid., vol. 114, no. 12,pp. 1-90.

1954. On the visceral skeleton in Eusthenop-teron with a discussion of the parasphe-noid and palatoquadrate in fishes. K.Svenska Vetensk. Akad. Handl., ser. 4,vol. 5, no. 1, 104 pp.

1960. Theories de l'evolution des vertebresreconsiderees 'a la lumiere des recentesdecouvertes sur les vertebres inferieurs.Paris, Masson, 104 pp.

1963. The composition of the intermandibulardivision of the head in fish and tetrapodsand the diphyletic origin of the tetrapod

268 VOL. 167


tongue. K. Svenska Vetensk. Akad.Handl., ser. 4, vol. 9, no. 4, pp. 1-74.

1964. Specializations in early vertebrates.Ann. Soc. Roy. Zool. Belgique, vol. 94,no. 1, pp. 11-95.

1%5. On the origin of girdles and paired fins.Israel Jour. Zool., vol. 14, pp. 141-172.

1966. Remarks on the structure of the snoutin Megalichthys and certain other rhip-idistid crossopterygians. Arkiv Zool.,vol. 19, no. 2, pp. 41-98.

1967. The homologies of frontal and parietalbones in fishes and tetrapods. ColloquesInternatl. Centre Natl. Rech. Sci., vol.163, pp. 181-213.

1968a. The systematic position of the Dipnoi.In 0rvig, T. (ed.), Current problems oflower vertebrate phylogeny. Stock-holm, Almqvist and Wiksell, pp. 223-245.

1968b. Aspects of vertebrate phylogeny. In0rvig, T. (ed.), Ibid., pp. 497-527.

1972. Middle and upper Devonian Porolepi-formes from East Greenland with specialreference to Glyptolepis groenlandican.sp. and a discussion on the structureof the head in the Porolepiformes. Med-del. Gr0nland, vol. 187, no. 2,307 pp.

1975. On the saccus endolymphaticus and ad-jacent structures in osteolepiforms, an-urans and urodeles. Colloques Internatl.Centre Natl. Rech. Sci., vol. 218, pp.191-211.

Jessen, H.1966. Die Crossopterygier des Oberen Plat-

tenkalkes (Devon) der Bergisch-Glad-bach-Paffrather Mulde (RheinischesSchiefergebirge) unter Beriicksichti-gung von amerikanischem und euro-piiischem Onychodus-Material. ArkivZool., vol. 18, no. 2, pp. 305-389.

1967. The position of the Struniiformes Strun-ius and Onychodus among the crossop-terygians. Colloques Internatl. CentreNatl. Rech. Sci., vol. 163, pp. 173-180.

1972. Schultergiirtel und Pectoralflosse beiActinopterygien. Fossils and Strata,vol. 1, pp. 1-101.

1975. A new choanate fish, Powichthys thor-steinssoni n.g., n.sp., from the earlyLower Devonian of the Canadian ArcticArchipelago. Colloques Internatl. CentreNatl. Rech. Sci., vol. 218, pp. 213-225.

Johansen, K.1970. Air breathing in fishes. In Hoar, W. S.,

and D. J. Randall (eds.), Fish physiol-

ogy. vol. 4, London and New York,Academic Press, pp. 361-413.

Jollie, M.1962. Chordate morphology. New York, Rein-

hold, 478 pp.Jordan, D. S.

1905. A guide to the study of fishes. 2 vols.,New York, H. Holt and Co. 624 + 599PP.

Jurgens, J. D.1971. The morphology of the nasal region of

Kerr, J.

Amphibia and its bearing on the phylog-eny of the group. Ann. Univ. Stellen-bosch, ser. A, vol. 46, no. 2, 146 pp.G.

1932. Archaic fishes-Lepidosiren, Protopte-rus, Polypterus-and their bearing uponproblems of vertebrate morphology.Jena Zeitschr. Naturwiss., vol. 67, pp.419-433.

Kesteven, H. L.1950. The origin of the tetrapods. Proc. Roy.

Soc. Victoria, new ser., vol. 59, no. 2,pp. 93-138.

Kingsley, J. S.1892. The head of an embryo Amphiuma.

Amer. Nat., vol. 26, pp. 671-680.Klaatsch, H.

1896. Die Brustflosse der Crossopterygier.Ein Beitrag zur Anwendung der Archip-terygium-Theorie auf die Gliedmassender Landwirbelthiere. In Festschriftzum siebenzigsten Geburtstage von CarlGegenbaur. vol. 1, Leipzig, pp. 261-391.

Krefft, J.1870. Description of a gigantic amphibian al-

lied to the genus Lepidosiren, from theWide Bay District, Queensland. Proc.Zool. Soc. London, pp. 221-224.

Lauder, G. V.1979. Review of J. Millot, J. Anthony and

D. Rabineau, Anatomie de Latimeriachalumnae, vol. 3 (1978). Copeia, no.3, pp. 560-562.

Lehman, J.-P.1956. L'evolution des dipneustes et l'origine

des urodeles. Colloques Internatl. CentreNatl. Rech. Sci., vol. 60, pp. 69-76.

1959. Les Dipneustes du Devonien superieurdu Groenland. Meddel. Gr0nland, vol.160, pp. 1-58.

Lehmann, W. M., and T. S. Westoll1952. A primitive dipnoan fish from the Lower

Devonian of Germany. Proc. Roy. Soc.,B, vol. 140, pp. 403-421.

1981 269


Lindsey, C. C.1978. Form, function, and locomotory habits

in fish. In Hoar, W. S., and D. J. Ran-dall (eds.), Fish physiology. vol. 7. Lon-don and New York, Academic Press,pp. 1-100.

L0vtrup, S.1977. The phylogeny of Vertebrata. London,

J. Wiley, xii + 330 pp.M'Donnell, R.

1860. Observations on the habits and anatomyof the Lepidosiren annectens. Jour.Roy. Dublin Soc., vol. 2, pp. 3-20.

Medvedeva, I. M.1959. The naso-lachrymal duct and its rela-

tionships to the lachrymal and septo-maxillary dermal bones in Ranodon si-biricus. Dokl. Akad. Nauk SSSR, vol.128, pp. 425-428.

1960. On the relationships of the developingnaso-lachrymal duct to the lachrymaland septomaxillary dermal bones in Hy-nobius keyserlingii. Ibid., vol. 131, pp.1209-1212.

1968. Die Homologie des Jacobsonschen Or-gans bei Anura und Urodela. In 0rvig,T. (ed.), Current problems of lower ver-tebrate phylogeny. Stockholm, Alm-qvist and Wiksell, pp. 331-340.

Melville, G.1848. On the Lepidosiren. Rept. Brit. Assoc.

Adv. Sci., vol. 17, p. 78.1860. Appendix to M'Donnell's paper. Jour.

Roy. Dublin Soc., vol. 2, pp. 405-406.Miles, R. S.

1964. A reinterpretation of the visceral skel-eton of Acanthodes. Nature, London,vol. 204, pp. 457-459.

1965. Some features in the cranial morpholo-gy of acanthodians and the relationshipsof the Acanthodii. Acta Zool. Stock-holm, vol. 46, pp. 233-255.

1973. Relationships of acanthodians. In Green-wood, P. H., R. S. Miles and C. Patter-son (eds.), Interrelationships of fishes.London, Academic Press, pp. 63-104.

1975. The relationships of the Dipnoi. Col-loques Internatl. Centre NatI. Rech.Sci., vol. 218, pp. 133-148.

1977. Dipnoan (lungfish) skulls and the rela-tionships of the group: a study based onnew species from the Devonian of Aus-tralia. Zool. Jour. Linnean Soc., vol. 61,pp. 1-328.

Miles, R. S., and G. C. Young1977. Placoderm interrelationships reconsid-

ered in the light of new ptyctodontidsfrom Gogo, Western Australia. In An-drews, S. M., R. S. Miles and A. DWalker (eds.), Problems in vertebrateevolution. London, Academic Press,pp. 123-198.

Millot, J., and J. Anthony1958. Anatomie de Latimeria chalumnae.

Tome I. Squelette, muscles et forma-tions de soutien. Paris, Centre Natl.Rech. Sci., 122 pp.

1965. Anatomie de Latimeria chalumnae.Tome II. Systeme nerveux et organesdes sens. Ibid., 120 pp.

Miner, R. W.1925. Pectoral limb ofEryops, and other prim-

itive tetrapods. Bull. Amer. Mus. Nat.Hist., vol. 5 1, pp. 145-3 12.

Mookerjee, H. K., D. N. Ganguly, and S. K.Brahma

1954. On the development of the centrum andits arches in the Dipnoi Protopterus an-nectens. Anat. Anz., vol. 100, pp. 217-230.

Moy-Thomas, J. A., and R. S. Miles1971. Palaeozoic fishes. Second edition. Phila-

delphia and Toronto, W. B. SaundersCo., 259 pp.

Muller, J.1840. Bericht uber die Fortschritte der ver-

gleichenden Anatomie im Jahre 1839.Arch. Anat. Physiol., pp. 159-198.

1844. Ueber den Bau und die Grenzen derGanoiden. Ber. Akad. Wiss. Berlin,1846, pp. 67-85.

Natterer, J.1837. Lepidosiren paradoxa, eine neue Gat-

tung aus der Familie der fischahnlichenReptilien. Ann. Naturhist. Mus. Wien,vol. 2, pp. 165-170.

Nelson, G. J.1968. Gill-arch structure in Acanthodes. In

0rvig, T. (ed.), Current problems oflower vertebrate phylogeny. Stock-holm, Almqvist and Wiksell, pp. 129-143.

1969. Gill arches and the phylogeny of fishes,with notes on the classification of ver-tebrates. Bull. Amer. Mus. Nat. Hist.,vol. 141, art. 4, pp. 475-552.

1970. Pharyngeal denticles (placoid scales) ofsharks, with notes on the dermal skel-eton of vertebrates. Amer. Mus. Novi-tates, no. 2415, pp. 1-26.

270 VOL. 167


Newman, F.1857. Note on Lepidosiren annectens. Proc.

Linnean Soc. London, 1856, pp. 73-74.Nielsen, E.

1936. Some few preliminary remarks on Tri-assic fishes from East Greenland. Med-del. Gr0nland, vol. 112, no. 3, pp. 1-55.

1942. Studies on Triassic fishes from EastGreenland. I. Glaucolepis and Boreo-somus. Palaeozool. Groenlandica 1, 403PP.

1949. Studies on Triassic fishes from EastGreenland. II. Australosomus and Bir-geria. Ibid., 3, 309 pp.

Noble, G. K.1931. The biology of the Amphibia. New

York, London, McGraw-Hill Book Co.,577 pp.

Norris, H. W.1913. Cranial nerves of Siren lacertina. Jour.

Morph., vol. 24, no. 2, pp. 245-338.Oken, L.

1841. Lepidosiren paradoxa. Isis, vol. 34, pp.462-469.

0rvig, T.1969a. Cosmine and cosmine growth. Lethaia,

vol. 2, pp. 241-260.1969b. Vertebrates from the Wood Bay Group

and the position of the Emsian-Eifelianboundary in the Devonian of Vestspits-bergen. Ibid., vol. 2, pp. 273-328.

1976. The interpretation of osteodentine, withremarks on the dentition in the Devo-nian dipnoan Griphognathus. Zool.Scripta, vol. 5, pp. 79-96.

Owen, R.1839. On a new species of the genus Lepido-

siren of Fitzinger and Natterer. Proc.Linnean Soc. London, vol. 1, pp. 27-32.

1841. Description of the Lepidosiren annec-tens. Trans. Linnean Soc. London, vol.18, pp. 327-361.

1846. Lectures on the comparative anatomyand physiology of the vertebrate ani-mals. Pt. I. Fishes. London, Longman,Brown, Green and Longmans, xi + 308PP.

1859. Palaeontology. In Encyclopaedia Bri-tannica. Edinburgh, vol. 17, pp. 91-176.

1866. The anatomy of vertebrates. Vol. I.Fishes and reptiles. London, Long-mans, Green and Co., xlii + 600 pp.

Panchen, A. L.1964. The cranial anatomy of two Coal Mea-

sure anthracosaurs. Phil. Trans. Roy.Soc. London (B), vol. 247, pp. 593-637.

1967. The nostrils of choanate fishes and earlytetrapods. Biol. Rev., vol. 42, pp. 374-420.

1970. Anthracosauria. In Kuhn, 0. (ed.),Handbuch der Palaoherpetologie, Stutt-gart, Fischer, pt. 5a, pp. vii + 84.

1972a. The skull and skeleton of Eogyrinus at-theyi Watson (Amphibia: Labyrintho-dontia). Phil. Trans. Roy. Soc. London(B), vol. 263, pp. 279-326.

1972b. The interrelationships of the earliest tet-rapods. In Joysey, K. A., and T. S.Kemp (eds.), Studies in vertebrate evo-lution. Edinburgh, Oliver and Boyd, pp.65-87.

1977. The origin and early evolution of tetra-pod vertebrae. In Andrews, S. M., R.S. Miles, and A. D. Walker (eds.), Prob-lems in vertebrate evolution. London,Academic Press, pp. 289-318.

1979. The cladistic debate continued. Nature,vol. 280, p. 541.

Parrington, F. R.1967. The identification of the dermal bones

of the head. Zool. Jour. Linnean Soc.,vol. 47, pp. 231-239.

Parrington, F. R., and T. S. Westoll1940. On the evolution of the mammalian pal-

ate. Phil. Trans. Roy. Soc. London (B),vol. 230, pp. 305-355.

Patterson, C.1968. The caudal skeleton in Lower Liassic

pholidophorid fishes. Bull. Brit. Mus.Nat. Hist. (Geol.), vol. 16, pp. 201-239.

1973. Interrelationships of holosteans. InGreenwood, P. H., R. S. Miles, and C.Patterson (eds.), Interrelationships offishes. London, Academic Press, pp.233-305.

1977. Cartilage bones, dermal bones andmembrane bones, or the exoskeletonversus the endoskeleton. In Andrews,S. M., R. S. Miles, and A. D. Walker(eds.), Problems in vertebrate evolu-tion. London, Academic Press, pp. 77-121.

Patterson, C., and D. E. Rosen1977. Review of ichthyodectiform and other

Mesozoic teleost fishes and the theoryand practice of classifying fossils. Bull.Amer. Mus. Nat. Hist., vol. 158, pp.81-172.

Pearson, D. M.[In press] Functional aspects of the integument

1981 271


in polypterid fishes. Zool. Jour. LinneanSoc.

Pearson, D. M., and T. S. Westoll1979. The Devonian actinopterygian Cheiro-'

lepis Agassiz. Trans. Roy. Soc. Edin-burgh, vol. 70, pp. 337-399.

Pehrson, T.1922. Some points in the cranial development

of teleostomian fishes. Acta Zool.Stockholm, vol. 3, pp. 1-63.

1947. Some new interpretations of the skull inPolypterus. Ibid., vol. 28, pp. 399-455.

Perks, A. M.1969. The neurohypophysis. In Hoar, W. S.,

and D. J. Randall (eds.), Fish physiol-ogy. Vol. 2. London and New York,Academic Press, pp. 112-206.

Pollard, H. B.1891. On the anatomy and phylogenetic po-

sition of Polypterus. Anat. Anz., vol. 6,pp. 338-344.

1892. On the anatomy and phylogenetic po-sition of Polypterus. Zool. Jahrb., vol.5, pp. 387-428.

Presley, R., and F. L. D. Steel1978. The pterygoid and ectopterygoid in

mammals. Anat. Embryol., vol. 154,pp. 95-110.

Rackoff, J. S.1976. The osteology of Sterropterygion (Cros-

sopterygii: Osteolepidae) and the originof tetrapod locomotion. Ph.D. thesisYale Univ., 292 pp.

1980. The origin of the tetrapod limb and theancestry of tetrapods. In Panchen, A. L.(ed.), The terrestrial environment andthe origin of land vertebrates. London,Academic Press, pp. 255-292.

Ramanujam, S. G. M.1929. The study of the development of the

vertebral column in teleosts, as shownin the life-history of the herring. Proc.Zool. Soc. London, pp. 365-414.

Rayner, D. H.1951. On the cranial structure of an early pa-

leoniscid, Kentuckia, gen. nov. Trans.Roy. Soc. Edinburgh, vol. 62, pp. 53-83.

Reynolds, S. H.1913. The vertebrate skeleton. Second ed.

Cambridge Univ. Press, xvi + 535 pp.Romer, A. S.

1924. Pectoral limb and shoulder girdle in fishand tetrapods. Anat. Rec., vol. 27, no.2, pp. 119-143.

1933. Vertebrate paleontology. First edition.Univ. Chicago Press, 491 pp.

1936. The dipnoan cranial roof. Amer. Jour.Sci., vol. 32, pp. 241-256.

1937. The braincase of the Carboniferouscrossopterygian Megalichthys nitidus.Bull. Mus. Comp. Zool., vol. 82, no. 1,73 pp.

1941. Notes on the crossopterygian hyoman-dibular and braincase. Jour. Morph.,vol. 69, no. 1, pp. 141-160.

1945. Vertebrate paleontology. Second edi-tion. Univ. Chicago Press, 687 pp.

1956. The early evolution of land vertebrates.Proc. Amer. Phil. Soc., vol. 100, pp.157-167.

1957. The appendicular skeleton of the Perm-ian embolomerous amphibian Archeria.Contrib. Mus. Paleo. Univ. Michigan,vol. 13, no. 5, pp. 103-159.

1966. Vertebrate paleontology. Third edition.Univ. Chicago Press, ix + 468 pp.

1968. Notes and comments on vertebrate pa-leontology. Univ. Chicago Press, 304PP.

Romer, A. S., and F. Byrne1931. The pes of Diadectes: notes on the primi-

tive tetrapod limb. Palaeobiology, vol. 4,pp. 25-48.

Rudebeck, B.1945. Contributions to forebrain morphology

in Dipnoi. Acta Zool., vol. 26, pp. 9-156.

Save-Soderbergh, G.1932. Preliminary note on Devonian stegoce-

phalians from East Greenland. Meddel.Gr0nland, vol. 94, no. 7, pp. 1-107.

1933. The dermal bones of the head and thelateral line system in Osteolepis macro-lepidotus Ag. with remarks on the ter-minology of the lateral line system andon the dermal bones of certain othercrossopterygians. Nova Acta Roy. Soc.Sci. Uppsala, ser. 4, vol. 9, no. 2, pp.1-130.

1934. Some points of view concerning theevolution of the vertebrates and theclassification of this group. ArkivZool., vol. 26, no. 17, pp. 1-20.

1935. On the dermal bones of the head in lab-yrinthodont stegocephalians and primi-tive Reptilia. Meddel. Gr0nland, vol.98, no. 3, pp. 1-211.

1936. On the morphology of Triassic stego-cephalians from Spitzbergen, and theinterpretation of the endocranium in theLabyrinthodonta. K. Svenska Vetensk.Akad. Handl., ser. 3, vol. 16, no. 1,pp. 1-181.

272 VOL. 167


1937. On the dermal skulls of Lyrocephalus,Aphaneramma, and Benthosaurus, lab-yrinthodonts from the Triassic of Spitz-bergen and N. Russia. Bull. Geol. Inst.Uppsala, vol. 27, pp. 189-208.

1946. On the fossa hypophyseos and the at-tachment of the retractor bulbi group inSphenodon, Varanus, and Lacerta. Ar-kiv Zool., vol. 38 (A), no. 11, pp. 1-24.

Schaeffer, B.1965. The evolution of concepts related to the

origin of the Amphibia. Syst. Zool., vol.14, pp. 115-118.

1967. Osteichthyan vertebrae. Zool. Jour.Linnean Soc., vol. 47, pp. 185-195.

1968. The origin and basic radiation of theOsteichthyes. In 0rvig, T. (ed.), Cur-rent problems in lower vertebrate phy-logeny. Stockholm, Almqvist and Wik-sell, pp. 207-222.

1975. Comments on the origin and basic ra-diation of the gnathostome fishes withparticular reference to the feedingmechanism. Colloques Internatl. CentreNatl. Rech. Sci., vol. 218, pp. 101-109.

Schaeffer, B., and M. Williams1977. Relationships of fossil and living elas-

mobranchs. Amer. Zool., vol. 17, pp.293-302.

Schauinsland, H. H.1906. Die Entwickelung der Wirbelsaule nebst

Rippen und Brustbein. In Hertwig, 0.(ed.), Handbuch der Vergleichendenund experimentellen Entwickelungs-lehre der Wirbeltiere. Jena, GustavFischer, vol. 3, pt. 2, pp. 339-572.

Schmalhausen, I. I.1917. On the dermal bones of the shoulder-

girdle of the Amphibia. Revue Zool.Russe, no. 2, pp. 84-110.

1950. On the homology of the skull roof bonesof fishes and terrestrial vertebrates.Zool. Jour. Akad. Nauk SSSR, vol. 29,no. 2, pp. 176-186.

1968. The origin of terrestrial vertebrates.New York, Academic Press, xxi + 314PP.

Schultze, H.-P.1969a. Griphognathus Gross, ein langschnau-

ziger Dipnoer aus dem Oberdevon vonBergisch-Gladbach (Rheinisches Schie-fergebirge) und von Lettland. Geol. etPaleont., vol. 3, pp. 21-79.

1969b. Die Faltenzahne der rhipidistiiden Cros-sopterygier, der Tetrapoden und derActinopterygien-Gattung Lepisosteus.

Palaeontographia Ital., vol. 65, pp. 60-137.

1970a. Folded teeth and the monophyletic ori-gin of tetrapods. Amer. Mus. Novitates,no. 2408, 10 pp.

1970b. Die Histologie der Wirbelkorper derDipnoer. Neues Jahrb. Geol. u. Palaeont.Abhandl., vol. 135, no. 3, pp. 311-336.

1977a. The origin of the tetrapod limb withinthe rhipidistian fishes. In Hecht, M. K.,P. C. Goody, and B. M. Hecht (eds.),Major patterns in vertebrate evolution,New York, Plenum Press, pp. 541-544.

1977b. Ausgangsform und Entwicklung derrhombischen Schuppen der Osteichthy-es (Pisces). Palaeont. Zeitschr., vol. 51,pp. 152-168.

Semon, R.1898. Die Entwickelung der paarigen Flossen

des Ceratodus forsteri. In Semon, R.,Zoologische Forschungsreisen in Aus-tralien und dem Malayischen Archipel-ago. I. Ceratodus, pt. 2. Jena, G. Fisch-er Verlag, pp. 59-111.

1901. Uber das verwandtschaftsverhaltnissder Dipnoer und Amphibien. Zool.Anz., vol. 24, pp. 180-188.

Sewertzoff, A. N.1927. Der Ursprung der Quadrupeda. Palae-

ont. Zeitschr., vol. 8, pp. 75-95.Smith, M. M.

1977. The microstructure of the dentition anddermal ornament of three dipnoans fromthe Devonian of western Australia: acontribution towards dipnoan interrela-tions and morphogenesis, growth andadaptation of the skeletal tissues. Phil.Trans. Roy. Soc. London (B), vol. 281,pp. 29-72.

1979. Structure and histogenesis of toothplates in Sagenodus inaequalis Owenconsidered in relation to the phylogenyof post-Devonian dipnoans. Proc. Roy.Soc. London (B), vol. 204, pp. 15-39.

Stensi6, E. A.1918. Notes on a crossopterygian fish from

the Upper Devonian of Spitzbergen.Bull. Geol. Inst. Uppsala, vol. 16, pp.115-124.

1921. Triassic fishes from Spitzbergen. Pt. I.Vienna, A. Holzhausen, 307 pp.

1922. Uber zwei Coelacanthiden aus demOberdevon von Wildungen. Palaeont.Zeitschr., vol. 4, nos. 2/3, pp. 167-210.

1925. Triassic fishes from Spitzbergen. Pt. II.K. Svenska Vetensk. Akad. Handl.,ser. 3, vol. 2, no. 1, 261 pp.

1981 273


1932. Triassic fishes from East Greenland col-lected by the Danish Expeditions in1929-1931. Meddel. Gr0nland, vol. 83,no. 3, pp. 1-305.

1947. The sensory lines and dermal bones ofthe cheek in fishes and amphibians. K.Svenska Vetensk. Akad. Handl., ser. 3,vol. 22, no. 3, pp. 1-70.

1959. On the pectoral fin and shoulder girdleof the arthrodires. Ibid., ser. 4, vol. 8,no. 4, pp. 1-229.

Strong, 0. S.1895. Cranial nerves of Amphibia. Jour.

Morph., vol. 10, no. 1, pp. 101-230.Szarski, H.

1977. Sarcopterygii and the origin of tetra-pods. In Hecht, M. K., P. C. Goody,and B. M. Hecht (eds.), Major patternsin vertebrate evolution. New York, Ple-num Press, pp. 517-540.

Taverne, L.1979. Osteologie, phylogenese et systema-

tique des teleosteens fossiles et actuelsdu super-ordre des Osteoglosso-morphes. 3. Evolution des structuresosteologiques et conclusions generalesrelatives a la phylogenese et a la syste-matique du super-ordre. Mem. Acad.Roy. Belg. Cl. Sci., 8°, ser. 2, vol. 43,no. 3, pp. 1-168.

Thomson, K. S.1964. The comparative anatomy of the snout

in rhipidistian fishes. Bull. Mus. Comp.Zool., vol. 131, pp. 313-357.

1965. The nasal apparatus in Dipnoi, with spe-cial reference to Protopterus. Proc.Zool. Soc. London, vol. 145, pp. 207-238.

1966. Intracranial mobility in the coelacanth.Science, vol. 153, pp. 999-1000.

1968. A critical review of the diphyletic the-ory of rhipidistian-amphibian relation-ships. In 0rvig, T. (ed.), Current prob-lems in lower vertebrate phylogeny.Stockholm, Almqvist and Wiksell, pp.285-305.

1972. New evidence on the evolution of thepaired fins of Rhipidistia and the originof the tetrapod limb, with description ofa new genus of Osteolepidae. Postilla(Yale Peabody Mus.), no. 157, 7 pp.

1975. On the biology of cosmine. Bull. Pea-body Mus. Nat. Hist., vol. 40, pp. 1-59.

1977. On the individual history of cosmineand a possible electroreceptive functionof the pore-canal system in fossil fishes.

In Andrews, S. M., R. S. Miles, and A.D. Walker (eds.), Problems in verte-brate evolution. London, AcademicPress, pp. 247-270.

Thomson, K. S., and K. S. W. Campbell1971. The structure and relationships of the

primitive dipnoan lungfish-Dipnorhyn-chus sussmilchi (Etheridge). Bull. Pea-body Mus. Nat. Hist., no. 38, pp. vi +109.

Thomson, K. S., and P. P. Vaughn1968. Vertebral structure in Rhipidistia (Os-

teichthyes, Crdssopterygii) with de-scription of a new Permian genus. Pos-tilla (Yale Peabody Mus.), no. 127, pp.1-19.

Trewavas, E., E. I. White, N. B. Marshall, andD. W. Tucker1955. Reply to A. S. Romer. Nature, London,

vol. 176, pp. 126-127.Vogt, C.

1845. Translator's comments on Muller's pa-per. Ann. Sci. Nat. (Zool.), ser. 3, vol.4, no. 3, p. 31.

1851. Zoologische Briefe. Frankfurt-a.-M.,vol. 2, 640 pp.

Vorobyeva, E. I.1962. Rhizodontid crossopterygians of the

main Devonian field in the USSR. Tru-dy Paleont. Inst. SSSR, vol. 94, pp. 1-139. [In Russian.]

1973. Einige Besonderheiten im Schadelbauvon Panderichthys rhombolepis (Gross)(Pisces, Crossopterygii). Palaeonto-graphica, vol. 142A, pp. 221-229.

1974. On the formation of tetrapod features incrossopterygians. Paleont. Zh., no. 2,pp. 110-116.

1975. Bemerkungen zu Panderichthys rhom-bolepis (Gross) aus Lode im Lettland(Gauja-Schichten, Oberdevon). NeuesJahrb. Geol. Palaeont. Mh., pp. 315-320.

1977a. Morfologiya i osobennosti evolyutsiikisteperykh ryb. Trudy Paleont. Inst.SSSR, vol. 163, pp. 1-240.

1977b. Filogeneticheskie svazi osteolepiform-nykh Crossopterygii i ikh polozhenie vsisteme. In Ocherki po filogenii i siste-matike iskopaemykh ryb i beschelyustt-nykh. Paleont. Inst. Akad. Nauk SSSR,pp. 71-88.

1977c. Evolutionary modifications of the teethstructure in the Palaeozoic Crossopte-rygii. Jour. Paleont. Soc. India, vol. 20,pp. 16-20.

274 VOL. 167


1980. Observations on two rhipidistian fishesfrom the Upper Devonian of Lode, Lat-via (Lode). Zool. Jour. Linnean Soc.,vol. 70, no. 2, pp. 191-201.

Wahlert, G., von1968. Latimeria und die Geschichte der Wir-

belthiere. Eine evolutionsbiologischeUntersuchung. Stuttgart, G. FischerVerlag, vii + 125 pp.

Watson, D. M. S.1912. The larger Coal Measure Amphibia.

Mem. Proc. Manchester Lit. Phil. Soc.,vol. 57, pp. 1-14.

1913. On the primitive tetrapod limb. Anat.Anz., vol. 44, nos. 1/2, pp. 24-27.

1919. The structure, evolution and origin ofthe Amphibia. The "Orders" Rhachi-tomi and Stereospondyli. Phil. Trans.Roy. Soc. London (B), vol. 209, pp. 1-73.

1925. The structure of certain palaeoniscidsand the relationships of that group withother bony fish. Proc. Zool. Soc. Lon-don, 1925, pp. 815-870.

1926. The evolution and origin of the Am-phibia. Phil. Trans. Roy. Soc. London(B), vol. 214, pp. 189-257.

1940. The origin of frogs. Trans. Roy. Soc.Edinburgh, vol. 60, pp. 195-231.

Watson, D. M. S., and H. Day1916. Notes on some Palaeozoic fishes. Mem.

Proc. Manchester Lit. Phil. Soc., vol.60, pp. 1-52.

Watson, D. M. S., and E. L. Gill1923. The structure of certain Palaeozoic Dip-

noi. Jour. Linnean Soc. (Zool.), vol. 35,pp. 163-216.

Westoll, T. S.1936. On the structure of the dermal ethmoid

shield of Osteolepis. Geol. Mag., vol.73, pp. 157-171.

1937. On a specimen of Eusthenopteron fromthe Old Red Sandstone of Scotland.Ibid., vol. 74, pp. 507-524.

1938. Ancestry of the tetrapods. Nature, Lon-don, vol. 141, pp. 127-128.

1943. The origin of the tetrapods. Biol. Rev.,vol. 18, pp. 78-98.

1949. On the evolution of the Dipnoi. In Jep-sen, G. L., G. G. Simpson, and E. Mayr(eds.), Genetics, paleontology and evo-lution. Princeton Univ. Press, pp. 121-184.

1958a. The lateral fin-fold theory and the pec-

toral fins of ostracoderms and early fish-es. In Westoll, T. S. (ed.), Studies onfossil vertebrates. Univ. London Press,pp. 180-211.

1958b. The origin of continental vertebrate fau-nas. Trans. Geol. Soc. Glasgow, vol.23, pp. 79-105.

1961. A crucial stage in vertebrate evolution:fish to land animal. Proc. Roy. Inst.Great Britain, vol. 38, pp. 600-618.

1979. Devonian fish biostratigraphy. In TheDevonian system. Spec. Papers in Pa-laeontology, no. 23, pp. 341-353.

White, E. I.1962. A dipnoan from the Assise de Mazy of

Hingeon. Bull. Inst. Roy. Sci. Nat. Bel-gique, vol. 38, no. 50, pp. 1-8.

1965. The head of Dipterus valenciennesiSedgwick and Murchison. Bull. Brit.Mus. Nat. Hist. (Geol.), vol. 11, pp. 1-45.

Whiting, H. P., and Q. Bone1980. Ciliary cells in the epidermis of the lar-

val Australian dipnoan Neoceratodus.Zool. Jour. Linnean Soc., vol. 68, pp.125-137.

Wiedersheim, R. E. E.1892. Das Gliedmassenskelet der Wirbel-

thiere, mit besonderer Berucksichtigungdes Schulter- und Beckengurtels beiFischen, Amphibien und Reptilien.Jena, 266 pp.

Wiley, E. 0.1979. Ventral gill arch muscles and the inter-

relationships of gnathostomes, with anew classification of the Vertebrata.Zool. Jour. Linnean Soc., vol. 67, pp.149-179.

Woodward, A. S.1891. Catalogue of the fossil fishes in the Brit-

ish Museum (Natural History), vol. 2.London, British Museum (Natural His-tory), xliv + 567 pp.

1898. Outlines of vertebrate palaeontology forstudents of zoology. Cambridge Univ.Press, xxiv + 470 pp.

1922. Observations on crossopterygians andarthrodiran fishes. Proc. Linnean Soc.London, vol. 134, pp. 27-36.

Zangerl, R.[In press] Elasmobranchii. In Schultze, H.-P.

(ed.), Handbook of paleoichthyology.Vol. 3. Stuttgart, Gustav Fischer Verlag.

1981 275

Documents

LUNGFISHES, TETRAPODS, PALEONTOLOGY, AND · ABSTRACT Weconclude that the internal (excurrent) nos- tril of Recent lungfishes is a true choana, as judgedbyits comparisonwith(1) theinternal