Noqvtates PUBLISHED by the AMERICAN MUSEUM of NATURAL HISTORY CENTRAL PARK WEST at 79TH STREET, NEW YORK, N.Y

AMERICAN MUSEUM Noqvtates PUBLISHED BY THE AMERICAN MUSEUM OF NATURAL HISTORY CENTRAL PARK WEST AT 79TH STREET, NEW YORK, N.Y. 10024 Number 2828, pp. 1-24, figs. 1-11 October 21, 1985

Stethacanthid Elasmobranch Remains from the Bear Gulch Limestone (Namurian E2b) of Montana

RICHARD LUND'

ABSTRACT Four chondrichthyan species assigned to the atoquadrate and mandible, numbers oftooth fam- cladodontid genus Stethacanthus are described ilies and pectoral prearticular basals, morphology from the Bear Gulch Limestone ofMontana. Spec- ofthe pelvic girdle and areas ofsquamation. Com- imens are referred to Stethacanthus cf. S. alto- parison with other Bear Gulch stethacanthids nensis and S. cf. S. productus. Two other species strongly suggests that the presence of specialized are too immature to assign with certainty to known cranial and first dorsal fin squamation, with the stethacanthid spine species. The histology and presence of the first dorsal fin and spine, are sec- morphology ofthree isolated cladodont tooth types ondary sexual characters ofmature males. Clado- is described, one of which is referred to Cladodus selache is indicated as the sister group of the Ste- robustus. The species of Stethacanthidae are dis- thacanthidae, with the "Symmoriidae" being the tinguishable on the shapes and proportions ofpal- sister group of Stethacanthus altonensis.

INTRODUCTION The Upper Mississippian marine Bear taxa have also provided excellent data on the Gulch limestone member of the Heath For- relative values of the spines and other mor- mation has yielded approximately seven phologicalparameters forgrossdiagnostic dif- species ofcladodont elasmobranchs referable ferentiation, and have particularly demon- to the family Stethacanthidae (Lund, 1974). strated that striking sexual dimorphism within Several ofthese have been found in sufficient the family may pose insurmountable diffi- abundance or completeness to allow ade- culties in the recognition and diagnosis of quate holomorphic diagnosis and description incomplete or inadequately represented (Lund, 1974, 1984, 1985, in press). These forms. Causing particular difficulty are the

' Professor, Biology Department, Adelphi University, Garden City, N.Y. 11530; Research Associate, Department of Ichthyology, American Museum of Natural History.

Copyright K American Museum of Natural History 1985 ISSN 0003-0082 / Price $2.25 2 AMERICAN MUSEUM NOVITATES NO. 2828 observations that only males may possess the The specimen CM 23654 was originally characteristic first dorsal spines and fins, and referred to as a female S. altonensis (Lund, only at sexual maturity (Lund, 1984; Fal- 1974). The spine ofthe specimen is immature catus falcatus, Lund, 1985; probably and therefore proximally incomplete. The Stethacanthus altonensis, Lund, 1974 and apical angle ofthe spine is approximately 240 below; S. cf. S. productus, below). In one oth- and the structure of the spine suggests that er, "Physonemus" attenuatus (Lund, in press) the apex was a posterodorsal continuation of there is considerable change of form of the the base (fig. 2D). The apical angle of the first dorsal spine and braincase with growth, spine of S. altonensis is 340 and the apex but females are not known for this species. projects dorsally in relation to the base (New- Additionally, marked differences occur in the berry, 1889, 1897; Lund, 1974). No traces of abundances of males versus females in all dorsal fin elements are preserved. groups with striking sexual dimorphism In dorsal view, the neurocranium is pro- (Ghiselin, 1974; also see Lund, 1982). Fur- portioned essentially as in S. altonensis, but ther, sexually dimorphic living elasmo- with broader supraorbital crests. The pala- branchs ofdifferent sexes may occupy differ- toquadrate has grooves for 6-8 tooth fami- ent ranges at different times (Branstetter, lies, as in S. altonensis, but the palatine por- 1981), suggesting that there may be difficul- tion is dorsally concave and the postorbital ties in locating or recognizing fossil chon- expansion is rounded posterodorsally. There drichthyans ofthe sex opposite to that of the is no visible extension of the palatoquadrate type specimen. Following from this infor- anterior to the mesially inturned ethmoid mation, the purported absence of claspers in process. A quadrate shelf is absent, and the Cladoselache, for instance, may more rea- quadrate condyle extends below the level of sonably be attributed to either an absence of the tooth row. Meckel's cartilage is unknown, mature males at the localities searched or an but clearly had relatively few tooth families inability to recognize males than to the re- and an anteriorly placed coronoid process. mote possibility that this rather specialized The hyomandibula is stout and held in close chondrichthyan lacked claspers (Harris, 1938, association with the palatoquadrate, partic- 1951; Zangerl, 1981; Rosen et al., 1981). ularly near the quadrate condyle (fig. 1). This report will describe Bear Gulch shark "Stemmatodus"-type branchial denticles are specimens that are too rare or too incomplete evident. to assign with certainty to a previously named The pectoral fin is supported by 12 or 13 species, but that are critical to understanding prearticular basals, and eight radials are sup- the morphological differences between species ported on the metapterygial plate. The me- within the Bear Gulch Stethacanthidae. tapterygial axis is poorly preserved. The pel- ABBREVIATIONS: Specimens have been de- vic girdles are triangular, with six conspicuous posited either with the Section of Vertebrate diazonal foramina, and a very small metap- Fossils, Carnegie Museum of Natural His- terygial plate. Each girdle supports at least 12 tory, Pittsburgh, Pa. (CM), or with the Geo- single-jointed radials. Clasper elements are logic Museum ofthe University ofMontana, unknown. Missoula, Montana (MV). The second dorsal fin consists of a minimum of 19 separate but closely spaced, one- DESCRIPTIONS jointed radials, the distal segments of which are quite elongate. The posterior portion of Stethacanthus cf. S. productus the fin is followed by a thin triangular plate, Newberry, 1897 but it cannot be determined whether this plate Figures 1, 2, 7F supported fin radials as in Falcatus falcatus PREvIous REFERENCES: Stethacanthus al- (Lund, 1985). The caudal fin, as illustrated tonensis Lund, 1974, fig. 10, CM 23654. in Lund (1974), is heterocercal, and of rela- Stethacanthus cf. S. productus Lund, 1984. tively high angle, with separate neural arches REFERRED SPECIMENS: CM 23654, whole, and spines and discrete dorsal and ventral immature, disturbed specimen; CM 35695, lobes. Precaudal neural arches are paired, long palatoquadrate; CM 37669, palatoquadrate; and thin, and lack median neural spines. MV 6160, pectoral and pelvic girdles. Squamation is unknown, although loose 1985 LUND: STETHACANTHID REMAINS 3 denticles are present in MV 6160. Although no specimen is complete, and therefore no accurate length measurements can be made, the length ofCM 23654 approaches 1 m, and the palatoquadrate is 11.6 cm in length. The palatoquadrate CM 37669 is 24.5 cm long, suggesting a total length of around 2 m for an intact individual. S. cf. S. productus is thus significantly larger than S. altonensis, and the largest known vertebrate in the Bear Gulch Limestone. This size differential is reflected in the sizes ofknown spines ofthe two species (Newberry, 1889, 1897). Stethacanthus cf. S. productus differs from S. altonensis in the shape of the pterygoid portion of the palatoquadrate, the higher number of prearticular basals of the pectoral fin, the morphology of the pelvic girdle and fin, and in the relative length of the radials ofthe second dorsal fin, as well as in the shape LI and size of the first dorsal fin spine. The two FIG. 1. Stethacanthus cf. S. productus New- species are significantly similar, however, in berry, 1897. CM 23654. Neurocranium, palato- braincase shape, general plan of the pectoral quadrate, and hyomandibular, mesial view. Scale fin, and particularly in lacking a large metap- in cm; hyo, hyomandibular; sn, spiracular notch. terygial plate of the pelvic fin. The imma- turity ofthe first dorsal fin spine ofCM 23654, a moderately large specimen, is striking, sug- The contents of the digestive tract consist gesting that as in Falcatus falcatus (Lund, of an entire, although fragmented, conodon- 1985) the first dorsal fin spine is a secondary tochordate (Melton and Scott, 1973; Scott, sexual character present only in mature 1973). (probably male) individuals. The neurocranium has broad supraorbital crests and a very strong postorbital arcade. Stethacanthus cf. S. The otic region is broad and flat ventrally, altonensis narrower and rounded dorsally, with prom- Figures 3 and 4 inent ridges for anterior and posterior vertical REFERRED SPECIMEN: CM 37680. semicircular canals. Serial supraorbital fo- This specimen consists of jaws, eye pig- ramina for Ramus Ophthalmicus Superfici- ments, branchial and pectoral skeleton, ver- alis fibers supplying the supraorbital lateral- tebral axis with 54 segments, dislocated line canals are present dorsally, and a median braincase, and digestive tract contents. Cal- basal foramen, probably for the hypophysis cification ofthe neurocranium ofthe 10.8 cm and internal carotid arteries (Jarvik, 1977), fossil is incomplete anteriorly. The specimen is present midventrally. is referred to S. altonensis on the basis ofthe The mandible has a high coronoid process following: palatoquadrate has an angular slope slightly posterior to its midpoint, and indis- of the dorsal margin of the palatine portion tinct grooves for seven tooth families. Teeth to the pterygoid portion; the posterior margin in both upper and lower jaws are large and of the palatoquadrate is almost squared off five-cusped, with the medial and most lateral relative to the posterodorsal outline; the pos- cusps strongly fluted. The teeth are predom- terior margin has a tall, narrow quadrate shelf inantly of orthodentine in visual inspection, (fig. 3; Lund, 1974) as well as a pectoral fin osteodentine being absent from the pulp cav- supported by three anterior unjointed radials ities ofthe cusps. A variety ofbranchial den- and seven single-jointed radials plus the me- ticles is present, including those ofthe "Stem- tapterygial plate (fig. 4). The metapterygial matodus" type. plate supports eight radials. The axial skeleton contains 54 paired neu- 4 AMERICAN MUSEUM NOVITATES NO. 2828

A B

LI I I FIG. 2. Stethacanthus cf. S. productus Newberry, 1897. CM 23654. (A) pectoral fin; (B) pelvic girdle; (C) second dorsal fin; (D) first dorsal spine, anterior to the right. A-C, scale in cm; D, maximum width 12 mm. 1985 LUND: STETHACANTHID REMAINS 5

L l l l I

- I\ "I;f . V.

N N N

FIG. 3. Stethacanthus cf. S. altonensis (St. John and Worthen, 1875). CM 37680, palatoquadrate and mandible. Scale in mm. ral arches fused in the dorsal midline and also characterize other Bear Gulch stetha- bearing slight neural processes dorsally. Dor- canthids (Lund, 1984, 1985, in press). Other sal root foramina are included in all visible squamation is absent. Further, there is no neural arches, while ventral root foramina indication of specialized cranial squamation appear to be intersegmental. There are no of the sort that characterizes known adult traces of intercalaries, arcualia, or basapo- males of the stethacanthids within (Lund, physes. The axial skeleton is downfaulted into 1984, 1985, in press) or outside of the Bear the matrix at a region of marked change in Gulch Limestone (Zangerl, 1981). the form ofthe neurals; they become considerably narrower, more closely spaced, more Stethacanthus sp. 1 posteriorly inclined, and more elongate than Figure 5 the anterior elements. This transition marks REFERRED SPECIMENS: CM 35456, CM the proximity of the caudal portion of the 35487, two partial specimens of approxi- axis (fig. 4B). mately 360 mm total length (260 mm length A portion of the anterior trunk lateral-line to the posterior pelvic margin) and 162 mm canal is present as well as melanin-derived length to the posterior pelvic margin, respec- traces of former skin pigmentation. The lat- tively. eral-line canal is supported by highly modi- The palatoquadrate has a high, stout eth- fied scales in the form of half rings, which moid process with a slight forwardly pro- 6 AMERICAN MUSEUM NOVITATES NO. 2828

,,.,2A; (A o_

FIG. 4. Stethacanthus cf. S. altonensis (St. John and Worthen, 1875). CM 37680. (A) neurocranium; (B) neural elements of segments 1-6, 20-30, 49-54; (C) pectoral fin. m, metapterygium; sc, scapulocor- acoid. Scale in mm.

jecting absymphysial process. The palatine pterygoid region. The posterodorsal margin region is almost straight, its dorsal margin ofthe pterygoid region descends in a straight curving only slightly to meet the prominent line to meet the prominent short vertical 1985 LUND: STETHACANTHID REMAINS 7 quadrate shelf. The quadrate condyle is at the level of the tooth row. There are grooves for 10-11 tooth families. The tooth-bearing portion ofthe mandible is 73 percent of its total length, contains grooves for 11 tooth families, and the coronoid process is not prominent. The mandible is moderately deep throughout. Teeth are uniformly small and strongly fluted. The pectoral fin contains seven or eight prearticular basals, a basal plate supporting eight radials, and three visible axial elements with two radials per axial. The coracoid is directed ventrally, while the scapula has a short anterodorsal extension and lacks a posterodorsal extension. The pelvic girdles, from either juveniles or females, each consists of an elongate plate with calcified parallel ridges and a series of four diazonal foramina mesially. Posteriorly, the small metapterygium supports the last three of fifteen unjointed, well-spaced radials. The two girdles are in extensive contact in the midline. The dorsal fins are poorly preserved. The 39 precaudal neural arches are paired, slender, and only the anterior arches have included dorsal root foramina. Other details are FIG. 5. Stethacanthus sp. 1. CM 35487. (A) obscure. The caudal fin is low-angle hetero- palatoquadrate and mandible; (B) pelvic fin and cercal, but has a strong series of radials sup- girdle. m, metapterygium; sn, spiracular notch. porting the ventral lobe. Scale in mm. Squamation consists of strong belts of enlarged ventrolateral flank denticles between acters of palatoquadrate, mandible, and pel- the pectoral and pelvic girdles. There is some vic girdles is unique in the fauna. evidence for finer, pointed denticles higher The palatine portion ofthe palatoquadrate on the flanks. is slender and dorsally concave, and there are The two specimens are united on the com- grooves for 11 tooth families. The tooth- mon possession of a unique pelvic girdle, as bearing portion ofthe mandible is 72 percent well as the common morphologies ofthe pal- ofits length and contains grooves for 11 tooth atoquadrate and mandible. In neither speci- families; the coronoid process is relatively men is there any indication of calcification prominent, with the articular facet located of the neurocranium. below the level ofthe highest part ofthe tooth row. The basibranchial skeleton is indicated Stethacanthus sp. 2 by encrustations of tooth aggregations, and 6 there are sparse, long "Styptobasis"-type teeth Figure on the ceratohyal as well as on the cerato- REFERRED SPECIMEN: CM 37671, small branchials. Two long, slender basibranchial dorsoventrally preserved individual (86 mm plates are indicated. Hypobranchials appar- to the posterior pelvic margin) lacking caudal ently are short, the anterior three directed region. forward while the fourth is transversely ori- This specimen is quite incomplete, lacking ented. Correspondence between tooth aggre- a braincase, dorsal fins, calcified axial skel- gations and underlying branchial elements is eton and caudal fin, as well as a complete not assured, however (Nelson, 1970). palatoquadrate. The combination of char- Prearticular basals are missing from the 8 AMERICAN MUSEUM NOVITATES NO. 2828

each pelvic plate. There is a continuous fine shagreen of conical denticles, enlarged in a belt along the ventrolateral flank between pectoral and pelvic girdles and onto the anterior edge of the pelvic fin. Lateral-line ring scales are present. Stethacanthus sp. 2 resembles Stethacan- thus cf. S. productus in the shape and proportions ofthe pelvic girdles, but differs from it in the proportions ofthejaws, and the numbers of tooth grooves. Stethacanthus sp. 2 resembles Stethacanthus sp. 1 in the strong ventrolateral denticle belts, but the two differ in other available characters including the prominence of other squamation. TEETH Stethacanthus sp. Figures 8A-D, 9, 1OA-E B REFERRED SPECIMENS: CM 35458, six teeth ofone disrupted tooth family, and CM 35459, seven teeth in one tooth family in articulation, both specimens with either one or two intermediate cusps round in frontal section and both specimens found in immediate proximity to each other; CM 37524, 41063, 41064, 81-63026, isolated teeth. There are no specimens of either S. productus or S. altonensis with sufficient numbers of fully exposed teeth to permit char- acterization ofspecies differences. These teeth conform to the visible teeth of S. altonensis (Lund, 1974). The single-layered enameloid ofall large cusps is arranged superficially into fine vertical fluting, with the fluting of distal and proximal cusps somewhat sparser than FIG. 6. Stethacanthus sp. 2. CM 37671. (A) that of the central cusp. All teeth have rela- visceral skeleton, dorsal view; (B) pectoral fin; (C) tively prominent paired ventrolabial and pelvic fins, girdles and lateral denticles. Scale in buccolingual tubercles and rounded lingual mm. plate margins (fig. 8A-D). Nutrient foramina are abundant along the lingual margin of the pectoral fin, but the narrow metapterygial base, with the largest vascular canals tribu- plate supported eight unjointed radials. Five tary to the principal cusps and large vascular axial elements are visible, the first two sup- canals also located lingual to the bases ofthe porting two radials each, the following one tubercles. supporting one radial, and the remaining bas- Transverse thin sections were taken of a als lacking radials. The pelvic girdles are tri- series of seven articulated teeth (CM 35459; angular plates, lacking evidence for a prom- figs. 9, 1OA-C), and longitudinal sections were inent metapterygium and supporting a prepared of another tooth (CM 41064; fig. minimum of 12 slender, well-spaced one- 1 OD, E). In section, the enameloid is single jointed radials. Several large diazonal foram- layered and isotropic, lacking obvious bire- ina or uncalcified areas occupy the center of fringence, indicating a very fine foliate tex- 1985 LUND: STETHACANTHID REMAINS 9

I I l A

E l l I j m

FIG. 7. Pelvic girdles of Stethacanthidae. (A) Falcatusfalcatus, MV 5386, female; (B) Stethacanthus altonensis, MV 2830, male; (C) Stethacanthus sp. 1, CM 35487, juvenile; (D) "Cobelodus" aculeatus, from Zangerl and Case, 1976, fig. 246; (E) Orestiacanthusfergusi (Lund, 1984), CM 37525, female; (F) Stethacanthus cf. S. productus, MV 6160, probable male. A, C, E, F, scales in mm; B, D, scales in cm. a, axial element; m, metapterygium. ture (durodentine, Schmidt and Keil, 1971). junction. The enameloid layer is underlain The superficial fluting pattern is not reflected with a thick orthodentine mantle, character- in the topography of the dentinoenameloid ized by parallel Tomes' processes perpendic- 10 AMERICAN MUSEUM NOVITATES NO. 2828

''1k

.1 / /f.

E I I I I I

GI FIG. 8. Stethacanthus sp., CM 41063, tooth in oral (A) and labial (B) views. Stethacanthus sp., CM 37524, tooth in labial (C) and aboral (D) views. (E) Cladodus robustus, 82-72304B, basal plate in oral view. (F) Cladodus sp., CM 37506, partial tooth in labial view. (G) Cladodus sp., CM 37507, tooth in oral view. Abbreviations: S, lingual sulcus; T, tubercle. Scales in mm. ular to the enameloid boundary. The orthodentine can be seen to line some vascular dentine shows no circumferential banding canals. There are no significant differences in under polarized light. There is an abrupt coronal tissues among the articulated teeth change from orthodentine to the osteoden- of the radially sectioned tooth family. tine core, which continues vertically into the Tissues of the lingual plate -and ventrola- bases of the teeth (fig. IOA, D). Peritubular bial tubercle differ from osteodentine in both 1985 LUND: STETHACANTHID REMAINS I1I

w I I wI I

FIG. 9. Stethacanthus sp., CM 35459, successional tooth family in radial section. Letters correspond to illustrations in figure 10. Scale in mm.

Cladodus robustus plain and polarized light; the tissues are more darkly stained, have differing optical quali- Newberry and Worthen, 1866 Figure 8E ties, and lack Tomes' processes. The tissue of the lingual plate is acellular bone of the REFERRED SPECIMENS: CM 41090, 82- tooth pedestal, as is found in the bases of 72304, isolated teeth. some Recent Selachii (Reif, 1979b; Moss, Two teeth are referred to Cladodus robus- 1970). Two features characterize the bases of tus (Newberry and Worthen, 1866). They are the less mature teeth ofCM 35459: dorsolin- characterized by having many small inter- gually-ventrolabially arching mineralization mediate cusps and relatively small proximal of Sharpey's fibers (fibers of the periodontal and distal cusps by comparison to the robust ligament), and a similarly oriented regular set central cusp, by slightly concave proximo- of vascular canals connected in three dimen- and distolingual margins of the lingual plate, sions by rectangular cross canals (fig. l OB, C). and by the presence of many fine denticula- In the more mature teeth of the specimen, tions along the base of the labial margin of the vascular canals have undergone signifi- the crown. Intermediate cusps are variably cant diminution, Sharpey's fiber pathways no angular in frontal section and their long axes longer contact the walls of the vascular ca- and positions are irregularly arranged upon nals, and peritubular dentine is visible as a the tooth base. In the largest tooth, CM 41090, lining of some of the canals (fig. lOC). Sec- five intermediate cusps are present to the right, tions of cladodont tooth bases from the Bur- and four to the left of the central cusp in a lington Limestone of Iowa (CM 26302) re- cusp row length of 27.2 mm. The cusp row veal the same pattern ofSharpey's fibers with length is 1.9-1.97 times the maximum width a somewhat different vascular pattern super- of the lingual plate. imposed, and significant peritubular dentine Cladodus robustus teeth are the only teeth deposition. Radial sections of the lingual in the fauna large enough to have fit into the plates of teeth of Orthacanthus texensis (CM tooth grooves of the palatoquadrate of B6/1937) show the same pattern. Neither the Stethacanthusproductus (CM 37669, above). Burlington Limestone cladodont teeth nor There are no associations between the teeth those of Orthacanthus have enameloid. and the palatoquadrate, however. 12 AMERICAN MUSEUM NOVITATES NO. 2828

FIG. 10. Stethacanthus sp., CM 35459: (A) central cusp, tangential section. (B) lingual plate, early tooth. (C) lingual plate, more mature tooth. Stethacanthus sp., CM 41064: (D) osteodentine base of central cusp, longitudinal section. (E) bone, lingual plate, longitudinal section. (F) Cladodus sp., CM 1985 LUND: STETHACANTHID REMAINS 13

Cladodont indeterminate ulocoracoid and its processes might prove to Figure 8F, G be of systematic significance, preservational REFERRED SPECIMENS: CM 37506, 37507, difficulties limit the usefulness of these characters. Axial elements of the pectoral fin are 37675. also rarely well preserved. A form of cladodont tooth that does not The form ofthe pelvic girdle and fin, while conform to those of the known Stethacan- relatively constant within species, is the most thidae is known only from three isolated variable interspecies character in the Bear specimens. The proximal and distal margins Gulch stethacanthids. Pelvic girdles fall into ofeach cusp are drawn into enameloid blades, two broad types. The first, and almost un- the central cusp is compressed linguolabially, doubtedly primitive (Rosen et al., 1981), bears there is only one, stout intermediate cusp be- virtually the entire pelvic fin, with only the tween central and extreme cusps, and the la- terminal radials being supported by a very bial surface of each cusp bears one or two short, prominent, sometimes basally bifur- small, outtumed metapterygial element. In- cluded in this category are all the cating enameloid ridges that often extend to specimens described above as well as Falcatus falcatus the crown-base border on the intermediate (Lund, 1985) and S. altonensis (fig. 7). The cusps. The tooth bases extend lingually be- yond the labial margins of the teeth, and the second type (Lund, 1984, in press) consists of species with a prominent metapterygial lingual margin of the base roughly parallels plate (fig. 7E). the tooth row. There are numerous nutrient Squamation varies from sparse but total foramina in several rows dorsally, the largest with specialized areas in males while absent of which are concentrated near the lingual in females (Lund, 1984), to present only in margins. Two tubercles are located ventrally along the labial margin, directly under the specialized areas of males such as the dorsal surface of the head and first dorsal spine intermediate cusps. The height of the central (Lund, 1974; Zangerl, 1981). Only in Stetha- cusp differs by two times in proportion to the canthus sp. 1 and sp. 2 (above) is there evi- remaining cusps among the three teeth. This dence that some specialized squamation may difference cannot be attributed strictly to wear. be present injuveniles or females. A complete, DISCUSSION unspecialized squamation, the primitive condition for Chondrichthyes, is unknown among SYSTEMATIC CHARACTERS: The shapes and the Stethacanthidae. Squamation, however, proportions of the palatoquadrate and man- is uncertain or unknown in many species and dible remain constant within each ofthe Bear differs between the sexes where known. It has Gulch stethacanthid species described above limited utility as a supraspecific character at or cited, across a significant span of growth. this time. Additionally, numbers of tooth families and The axial skeleton, where known, provides fin radials are relatively invariant within each useful information on the presence, numbers, species. This relative constancy of form aids and forms of the elements; the locations of in ready identification of the less well rep- foramina; fusions across the midline; and the resented specimens described above. "Stem- form ofthe caudal skeleton. Numbers of pre- matodus" and "Styptobasis" type denticles, caudal segments seem to differ between all contrary to the assertions ofZangerl and Case species and only one specimen, S. cf. S. al- (1976), are not diagnostic of a single taxon tonensis (above) shows fusion ofneural arch- but are found on many, ifnot all, Bear Gulch es across the midline. In no case are separate Stethacanthidae. While the form of the scap- neural spines known of the kind attributed

26302, Burlington Ls., lingual plate, radial section. All sections with crossed polars. See figure 8 for scale. Abbreviations: E, enameloid; 0, orthodentine; OS, osteodentine; P, peritubular dentine; S, Shar- pey's fibers. Shaded arrows indicate oral direction. 14 AMERICAN MUSEUM NOVITATES NO. 2828 to S. altonensis by Zangerl (1981). Only the RELATIONSHIPS form of the caudal fin, whether heterocercal (primitive) or homocercal (derived; Lund, Analysis of the interrelationships of early 1985, in press), is ofuse in supraspecific anal- elasmobranchs is hindered by a deficiency of ysis because of disparities in preservation whole, thoroughly described specimens. Any among the available specimens. analysis, therefore, must be regarded in great The morphology ofthe first dorsal fin-spine part as an attempt to refine the questions that complex is unique to each described taxon of must be asked. The Bear Gulch material, the Stethacanthidae, where known (Lund, however incomplete, suggests a diverse series 1984). Spines are known from a male S. al- of characters useful in determining these in- tonensis (Lund, 1974), only from mature terrelationships. males of Falcatusfalcatus and a new stetha- The primitive condition ofthe squamation canthid (Lund, 1984, 1985), and an imma- of the Chondrichthyes is taken to be that of ture spine from S. cf. S. productus. Only "P." a complete, densely packed dermal covering, attenuatus spines show growth over a signif- as is found in Cladoselache clarki, C. kepleri icant size range of males at differing stages of (Claypole, 1893), and Diademodus (Harris, maturity (Lund, in press) and while females 1951) as well as other gnathostome classes. are unknown for this species, specialized cra- The absence ofa complete squamation, as in nial and first dorsal squamation correspond stethacanthids and symmoriids (Zangerl, to those ofother members ofthe family. There 1981) is a derived state. The absence (or pres- is absolutely no evidence to support Zangerl's ence ofonly specialized areas) of squamation (1981) contention that the Field Museum's in females and juveniles, as in Steth- S. altonensis (FMNH PF 2207) is a female. acanthus sp. 1 and 2 (above) is a derived state It cannot be established at this time how that cannot be adequately distinguished from prevalent among the Stethacanthidae this a condition of complete absence without form of sexual dimorphism is, but it is now knowing the condition of well-preserved known for every stethacanthid for which ad- males. The concomitant derived state, there- equate specimen numbers and both sexes are fore, in which specialized areas of squama- known. There is thus evidence to suggest that tion are secondary sexual characters, may be the first dorsal fin and spine, with specialized a more critical derived character than the cranial and first dorsal squamation, occurs mere absence of complete squamation in fe- only in sexually mature males of the Stetha- males and juveniles. There are insufficient canthidae. data upon which to judge whether the two While these are the only instances of the character states, in the sharks under consid- first dorsal spine and fin being secondary sex- eration, are autapomorphous or serially de- ual characters among the known Elasmo- rived. branchii, Recent or fossil, similar phenom- Primitively, lateral-line scales are little- ena involving head claspers have been re- modified from the normal whole-body squa- ported for Recent Holocephali (Dean, 1909), mation and support the edges of lateral-line Squaloraja (Woodward, 1886), and a chon- canals (Dean, 1894; Smith, 1937). The pres- drenchelyid (Lund, 1982); further, consid- ence offine ring scales supporting the lateral- erable sexual dimorphism in the form of the line system is a derived condition known only first dorsal fin, and spine and cranial squa- in Stethacanthidae, Symmoriidae, and cer- mation has been documented for Echinochi- tain Holocephali (Lund, 1982; Zangerl and maera (Lund, 1977a). There are many ste- Case, 1976, fig. 34) among Paleozoic Chon- thacanthid spines remaining from the Late drichthyes. Devonian and Mississippian for which no The teeth of the mandibular arch, that are holomorphs are known. The forms of these replaced in linguolabial series in all known spines, however, convey no morphological Elasmobranchii, can be considered to be information about their bearers. primitively directly underlain by a basal ped- 1985 LUND: STETHACANTHID REMAINS 15 estal of bone, as is the case with certain Re- tiana Cope (1891); teeth virtually identical cent Selachii (Moss, 1970; Reif, 1979b), Ac- to those of S. knightiana have been found tinopterygii (Moy-Thomas, 1934; Kerebel et lateral to typical dentitional teeth of Venus- al., 1978), and Dipnoi (Kemp, 1979). Ex- todus argutus (St. John and Worthen, 1875; tended basal lingual platforms and imbricat- CM 41097). If the coronodont condition is ing tooth bases are found in teeth of Clado- plesiomorphous for the Elasmobranchii, the dus, Carcharopsis, Orestiacanthus (Lund, development ofrelatively low, subequal cusps 1984), the xenacanths, and Chlamydosela- of the tooth crown, as in Hybocladodus and chus(Traquair, 1888). Orthacanthusteethdif- Protacrodus, is a derived state. The clado- fer from those of Cladodus in having strongly dont (based on Cladodus mirabilis Agassiz, banded orthodentine and in lacking osteo- 1837) and xenacanth (based on "Diplodus" dentine in the cores of the cusps. The struc- gibbosus Agassiz, 1837) conditions of the ture ofthe lingual platforms is essentially the cusps, where principal and subsidiary cusps same between the tooth types, however, and of different sizes are developed upon a syn- is considered a synapomorphy. Teeth of apomorphous base, are thus seen as alternate Chlamydoselachus sectioned in my labora- derived conditions ofthe crown. A final com- tory differ strongly from cladodont teeth in monality ofthe crowns ofcladodont and xen- having at least two-layered enameloid, highly acanth teeth is the apparent lack of terminal branching Tomes' processes in the ortho- membrane enameloid (Glikman, 1967; Duf- dentine, fiber trajectories in the bases gen- fin and Ward, 1983). Terminal membrane erally parallel to the vascular pattern, in the enameloid seems to be present in the dermal acute angle between the cusps and the long, skeletons of the agnathans Tremataspis narrow bases, and in the nature of the over- schmidti and Psammolepis paradoxa (Gross, lap between successional teeth. Strong inter- 1935, cited in Schmidt and Keil, 1971), As- dental ligaments connect the ventrolabial and traspis desiderata (Reif, 1979a), in ctena- dorsolingual tubercles; the presence of these canths, hybodonts, and Recent Selachii (Duf- interdental ligaments may be inferred for fin and Ward, 1983; Schmidt and Keil, 1971), cladodont and xenacanth teeth as well. These and Petalodontiformes (Lund, 1983). The data indicate that the teeth of Chlamydosela- enameloid of the teeth of Stethacanthus sp., chus are only grossly convergent upon the however, is anomalous in lacking birefrin- cladodont-xenacanth condition. The primi- gence. This may be an artifact of the tangen- tive tooth crown condition, by outgroup tial section, of the presumably finely foliate comparison, is a single, simple cone as is nature of the crystal matrix, of the nature of found in Osteichthyes. The simplest condi- the mineralization, or, least likely, of diage- tion known among mandibular teeth ofchon- netic changes (Schmidt and Keil, 1971). Most drichthyans is the coronodont state, in which ofthe possible explanations indicate an outer a distal-proximal series ofsubequal cusps are tissue similar to terminal membrane enam- fused into a multicuspid unit. This first-order eloid in structure but not precisely homolo- derived condition is known for Diademodus, gous. While ultrastructural details ofterminal Coronodus (Harris, 1951), and Squatinactis membrane enameloid are for the most part (Lund and Zangerl, 1974). The interpretation unknown, the absence of this layer is herein of "Styptobasis" teeth as teeth of the man- considered to be a derived condition shared dibular arch (Zangerl and Case, 1976), while with Holocephali and Bradyodonti. possible as a further derived state, is rendered The evidence for the primitive state ofdor- considerably less likely by the presence of sal fin numbers is presently equivocal both "Styptobasis" teeth lining the buccal cavity among the Chondrichthyes and in other mesial to the mandibular arch in the Stetha- gnathostome classes. Arguments of equal canthidae (Lund, 1985, in press). These teeth, weight can be made for two fins, as in Clad- however, are not morphologically similar to oselache (Harris, 1938, 1951), or a single fin those of the type species, Styptobasis knigh- as in the Xenacanthiformes (Dick, 1981), 16 AMERICAN MUSEUM NOVITATES NO. 2828

Heteropetalus (Lund, 1977b), and the Chon- character state regardless of the morphology drenchelyiformes (Lund, 1982). Evidence for of the primitive condition. Further derived the primitive morphology of dorsal fin en- conditions of the first dorsal fin of Stetha- doskeletons is also equivocal. There are three canthidae involve the sagittal narrowing of possibilities: (1) that the primitive chon- the base of the fin (stenobasal dorsal fin; drichthyan dorsal fin was a membranous flap Stethacanthus), the narrowing ofthe body of between spine and dorsum, as in Acanthodii; the fin (Orestiacanthus fergusi Lund, 1984), (2) that the fin was supported by a basal plate and the loss ofcontact between the fin, mod- and radials, as in ctenacanths, hybodonts, and ified into a rod, and the dorsum of the body some euselachians; or (3) that the fin was (F. falcatus). The total absence of the first supported by serial, segmentally arranged dorsal fin of"P. " attenuatus is a final derived basidorsals and radials, as in Cladoselache, state (as in symmoriids), but it is not pres- Xenacanthiformes, Heteropetalus, Chon- ently possible to determine whether the fin drenchelyiformes, the second dorsal fins of actually fused with the first dorsal spine dur- stethacanthids and symmoriids, and certain ing an early growth stage, or was indepen- euselachians (Dingerkus and DeFino, 1983). dently lost. Conversely, the absence of a first The latter condition is also shared primi- dorsal fin in symmoriids could be considered tively by Actinopterygii among the Osteich- primitive and the Cladoselache-stethacan- thyes, which would render it a synapomorphy thid condition derived except for the ex- of Chondrichthyes and Osteichthyes at the tremely close morphological resemblances in gnathostome level. There are few other syn- all other derived characters of median and apomorphies uniting the euselachians and ac- paired fins, suspensorium, jaws, and teeth. tinopterygians with the remaining chon- The primitive condition ofthe dorsal spine drichthyans having basidorsal radial dorsal is single, superficial, small, and thomlike, with fins, evidence that while the condition may superficial dentinal ornamentation. This con- be synapomorphic among several of the Pa- dition is known from Antarctilamna (Young, leozoic shark groups, it is also subject to con- 1982) and Heteropetalus (Lund, 1977). The vergence and not necessarily an indicator of extension of the spine into a superficially at- relationships between, for instance, the tached rod, as in the later xenacanths (Dick, Hemiscyllidae (Dingerkus and DeFino, 1983) 1981) is an apomorphic state. The develop- and the Stethacanthidae (Lund, 1984, 1985, ment of a deep sagittal insertion, as in the in press). Possibility 2 occurs in a very limited ctenacanths and many Neoselachii (Maisey, suite of chondrichthyans that are also united 1981) is another apomorphic state, as is the by other synapomorphies (Duffin and Ward, presence oftwo dorsal fin spines. The sagittal 1983; Young, 1982) and cannot be consid- enlargement of a small superficially inserted ered primitive for the Chondrichthyes. There spine as in Stethacanthidae and Cladoselache is no further objective evidence upon which (Harris, 1951), and the loss of dermal orna- to base a decision on whether possibility 1 or mentation, are separate derived conditions. 3 constitutes the primitive condition, or The primitive condition for the Stethacan- whether the two possibilities are mutually ex- thidae is represented by the spine of Stetha- clusive. In either case, however, a dorsal fin canthus productus, that is posterodorsally ex- supported by serial, two-jointed radials, as in tended, only moderately sagittally enlarged, Osteichthyes, is a synapomorphy at least for and not present in sexually immature speci- Stethacanthidae, Cladoselache, and the sym- mens. Intermediate derived states ofthe spine moriids. It is also critical to note that the are seen in the spines of S. altonensis and primitive state involves no sexual dimor- Orestiacanthusfergusi, where the base is sag- phism. Thus, the transformation of the first ittally enlarged and the apex of the spine is dorsal fin to a secondary sexual character in oriented vertical to the long axis ofthe body. males with its loss (absence) in juveniles and Differential growth of the base of the spine, females, assuming that two dorsal fins, as in producing a forwardly inclined apex, as in F. Cladoselache, represent the primitive con- falcatus, is an apomorphous condition. Dis- dition for the Elasmobranchii, is a derived tinction between species with forwardly in- 1985 LUND: STETHACANTHID REMAINS 17 clined apex-es is- on the basis of degree of in- a significant increase in radial numbers, con- clination (Lund, 1984, 1985, in press). Major stitutes two derived character states uniquely differences between the F. falcatus complex held in common between S. altonensis and and "P." attenuatus include an extended pe- symmoriids. An S-curvature of the metap- riod of growth and the great anterior extent terygial axis laterad from the long axis of the ofthe spine in the latter genus, and significant pelvic plate also characterizes the Stethacan- form differences between the bases of the thus-symmoriid group. Male Stethacanthus spines. "P." attenuatus also lacks any trace have multisegmented metapterygial axes of a dorsal fin component. (Lund, 1974; fig. 7B). Male symmoriids, As with the first dorsal fins, the presence however, have a single, enlarged, S-shaped ofan anterior dorsal spine in all growth stages metapterygium rather than a series of ele- of both sexes is considered to be a primitive ments, a further derived state; no metapter- character. The loss of the spine in females ygial elements are described in females (Zan- and juveniles with secondary sexually cor- gerl, 1981). While small diazonal foramina related development in males is therefore a in the pelvic plate are characteristic and ap- derived character, and shape differentiation parently primitive, the presence of large dia- of the spine into a highly modified structure zonal foramina in S. productus seems to be provides a small suite offurther derived states. an autapomorphous character. Among the The apparent absence of a dorsal fin spine in stethacanthids with a single metapterygial male symmoriids is a derived condition that element, the lengthening or condensation of may be either convergent upon the condi- the metapterygium to support 50 to 60 per- tions seen in all known female and juvenile cent of the fin radials is a derived condition. Stethacanthidae, or a shared derived state. Two species share this character in the Steth- The primitive number of pectoral prear- acanthidae, "P." attenuatus, where the fe- ticular basals is greater than the derived tri- male condition is unknown, and Orestiacan- basal condition that characterizes the ctena- thus fergusi, in which the metapterygium of canths, hybodonts, and Neoselachii (Duffin the female supports 60 percent of the fin ra- and Ward, 1983). The number of pectoral dials, a derived condition paralleling that of metapterygial axials, while also not precisely ctenacanths, hybodonts, and Neoselachii. The known, is primitively two to three (Rosen et metapterygium of male "P." attenuatus is a al., 1981), as in Cladoselache and Denaea triangular plate supporting 50 percent of the (Bendix-Almgreen, 1975); the increased unjointed, dorsoventrally flattened, closely numbers and extension of metapterygial ax- packed radials; it appears to be uniquely de- ials into an axial "whip," as in Stethacan- rived from the condition found in F. falcatus thidae and Symmoriidae, is a synapomorphy. (fig. 7). All Euselachii and Holocephali have The primitive state of the pelvic plate, re- a prominent metapterygium (Dean, 1909; gardless of sex, is a subtriangular, paired Dick, 1978), although the form of this ele- structure with a metapterygial axis of one to ment and its relationship to the clasper axis two segments; more than 75 percent of the in males of the two subclasses suggest that fin radials are borne directly upon the pelvic the two metapterygia are not strictly homol- plate. The primitive condition is found in ogous. several Osteichthyes (Rosen et al., 1981) as Jointed pelvic radials are considered prim- well as several of the Stethacanthidae and itive. The condition of Stethacanthus sp. 1 is symmoriids. For the purposes of this paper, atypical in that it appears to result from the a single metapterygial element supporting fusion of the proximal radial segments with only one or two radials is considered prim- the girdle, as well as the apparent broad con- itive. As around 12 radials supported on the tact of the two girdles in the midline. pelvic plate seem to be the modal quantity, Zangerl (1981) maintains that the pelvic numbers significantly higher (18-21), as in plates of the symmoriids were positioned Stethacanthus and the symmoriids, are con- vertically in the body wall, an assertion from sidered derived. Broadening ofthe pelvic plate Dean (1909) that the diazonal foramina and into a more rounded form, accompanied by plate-radial relationships ofthe specimens do 18 AMERICAN MUSEUM NOVITATES NO. 2828 not support (Bendix-Almgreen, 1975). The contrary, that segmentally arranged calcified pelvic plates, as in other chondrichthyans ex- or ossified vertebral arcualia are primitively cept for Holocephali and Iniopterygii (Lund, absent, and their presence, as in Chondren- 1977a; Zangerl and Case, 1973) were posi- chelyiformes (Lund, 1982), F. falcatus, and tioned horizontally in the ventral body wall. "P." attenuatus, is a derived character. Cladoselache has not been found with Two characters displayed by the Stetha- claspers, negative evidence that has com- canthidae are worthy of note, although they monly been transmuted into a definite state- play no part in this analysis. The ventral ment ofabsence (Rosen et al., 1981). It should branchial structure of two stethacanthids, F. also be noted that the pelvic fins and girdles falcatus and Stethacanthus sp. 2, indicates of Cladoselache and Diademodus are almost that the primitive condition for the Chon- as poorly known (Harris, 1951; Bendix-Alm- drichthyes is to have two narrow basibran- green, 1975), leading Harris to suggest that chials and posterolaterally directed ventral claspers preceded well-defined pelvic fins and branchial elements with short, laterally or girdles. Further, Harris noted that specimens posterolaterally directed hypobranchials. As of Cladoselache in which the dorsal fins are this condition also typifies Osteichthyes and well displayed do not all have a dorsal spine. can be supported for the Acanthodii, it can Aside from the dangers ofusing negative evi- be hypothesized to be the primitive gnatho- dence when claspers are known in Diade- stome condition (Nelson, 1968). The pha- modus, a "uniquely primitive" chondrich- ryngobranchial condition as well, consisting thyan (Harris, 1951), there are alternative of simple, posteriorly directed pharyngo- explanations for this absence that must be branchials lacking calcified interconnections, considered. One is that, as in some Recent also supports Nelson's hypothesis of primi- Elasmobranchii, mature males and females tive design. There is also evidence for an- may have occupied separate ranges for part terodorsally directed suprapharyngobranchi- or most of the year (Branstetter, 1981). al ligaments inserting upon the basicranium Another, and not exclusive, possibility is that of F. falcatus (Lund, 1985). claspers ofCladoselache are inadequately cal- The hyoid arch, unlike the branchial arch- cified or ossified and not subject to ready es, is not primitive. The hyomandibula ar- preservation, either generally or depending ticulates on the postorbital process mesial to upon the state of maturity of the specimens the palatoquadrate articulation, with little or at death. As claspers do not commonly calcify no room for a spiracle (figs. 1, 5; Jarvik, 1977; until sexual maturity, and virtually nothing Lund, 1985, in press). The suspensorium of is known about the growth of Cladoselache, the symmoriids appears to be the same as to assume that claspers are primitively absent that of the stethacanthids, rather than aphe- only in Cladoselache among known chon- tohyoidean, as stated by Zangerl and Wil- drichthyans of the same clade is insupport- liams (1975). Hypohyals have not been found. able. It would be more parsimonious to sug- The question ofbone, and bones, in sharks, gest that specimens of Cladoselache may be once clearly stated as absent, is no longer as displaying size and sexual dimorphism both straightforward. Bone tissue lacking cells has in the absence of claspers and the irregular been identified in the bases of scales, teeth, occurrence of dorsal spines. and vertebrae of both modem and fossil An anal fin is present in Osteichthyes, chondrichthyans (Reif, 1979b; Orvig, 1966) Acanthodii, Euselachii, Xenacanthodii, and and is found in the teeth of Stethacanthidae Chimaeriformes. The absence of an anal fin as well. However, the first dorsal spine and in Cladoselache, Stethacanthidae, and sym- fin, the sclerotics, and distal clasper elements moriids is a synapomorphy. are apparently composed of bone in the It is generally accepted that a heterocercal Stethacanthidae. The capacity to produce caudal skeleton is primitive for the Elasmo- acellular bone has not been lost in the Chon- branchii, and that a hemiheterocercal or drichthyes; the production of cellular bone, homocercal condition is derived. It is also however, has only been cited for one chon- accepted, in the absence of evidence to the drichthyan by Zangerl (1968) and has been 1985 LUND: STETHACANTHID REMAINS 19 called mesodentine by Orvig (1966). Two 12. Broadly rounded to subtriangular pelvic uniquely derived characters define the Chon- plate. drichthyes as presently understood: prismat- Greater than 14 pelvic radials. ic calcified cartilage and internal fertilization 13. Large pelvic diazonal foramina. with copulation by copulatory organs derived Dorsal spine apex posterodorsally ori- from posterior extensions of the pelvic en- ented. doskeleton of males. 14. First dorsal fin approximates an equilat- eral triangle. Dorsal spine apex vertically oriented. 15. Unresolved trichotomy. SYNAPOMORPHY SCHEME Pelvic metapterygial axis fused or con- the densed. A hypothesis of synapomorphies for Loss of secondary sexual dorsal and Stethacanthidae is summarized below (see fig. squamation characters. 11). 15'. Parallel loss of first dorsal spine, fin. 1. Coronodont tooth. Parallel loss of squamation. 2. Tooth base lacking lingual plate. Parallel development of rounded pelvic 3. Expanded bony lingual plate with inter- plate. dental ligaments. Parallel development of S-shaped me- Loss of terminal membrane enameloid. tapterygium. 4. Diplodus tooth crown, basal plate with Parallel development of high pelvic ra- single tubercle. dial numbers. One segmentally supported dorsal fin. 16. Dorsal spine apex forwardly oriented. 5. Cladodus tooth crown, basal plate with First dorsal fin either a rod, absent, or two tubercles. fused to spine. Two segmentally supported dorsal fins. Homocercal tail. Loss of anal fin. Calcified vertebral arcualia. 6. Insufficient data. Clasper axis of five to six segments. Complete squamation (primitive). 17. Dorsal spine apex oriented from 450 to 7. Dorsal spine unornamented, outer den- 150 from horizontal. tinal layers lost. First dorsal fin a rod. Dorsal spine not present in all individ- 18. Dorsal spine apex oriented horizontally. uals. Early spine development and growth. 8. Dorsal spine sagittally expanded into tri- First dorsal fin absent or fused to spine. angle. Pelvic metapterygium supports 60 per- Abbreviated heterocercal tail. cent of flat, unjointed radials. 9. Lateral-line ring scales. Pelvic metapterygium triangular. Pectoral metapterygial axis extended into 19. Dorsal fin a narrow acute triangle. "whip." Dorsal spine apex vertical. Dorsal spine and fin lost or a secondary Inline pelvic plate metapterygial axis. sexual character. Pelvic metapterygium supports 60 per- Dorsal spine, where present, with an an- cent of radials. teroventral shoulder. Clasper axis of 10 segments. Dorsal fin stenobasal where present. Squamation complete in males. Squamation reduced or absent, under secondary sexual control. 10. Greater than one pelvic metapterygial CONCLUSIONS segment. The Stethacanthidae are cladodont elas- S curve in pelvic plate-metapterygial axis. mobranchs characterized by several sexually 11. Single pelvic metapterygial segment. dimorphic features. The unique first dorsal First dorsal fin of males sagittally com- spine, fin, and surmounted squamation, ac- pressed. companied by dorsal cranial squamation, oc- 20 AMERICAN MUSEUM NOVITATES NO. 2828

FIG. 11. Cladogram of the Stethacanthidae. See text for explanation.

cur only in mature males; other squamation morphous at the gnathostome level. Acellular also seems to be sexually determined. Bran- bone comprises several structures. chial arches, where known, are symplesio- There are ample numbers of characters 1985 LUND: STETHACANTHID REMAINS 21 permitting interspecific discrimination among from low angle heterocercal to homocercal the Stethacanthidae. At the present level of (Lund, 1984, 1985, in press) reinforce the knowledge, however, few characters are con- potential taxonomic diversity of the Stetha- sistently available that may be used in an canthidae. analysis of suprafamilial interrelationships. Among these, braincase, squamation, first ACKNOWLEDGMENTS dorsal fin and spine data, and caudal struc- I am grateful to Wendy Lund for the illus- ture fall into the category ofpotentially useful trations and curatorial work, and to William but sporadically available information. Con- G. Melton, Jr., for his coleadership of the tinuously variable data, such as counts and field crews. The ranchers of the Grassrange- measures, provide little insight. Pelvic girdles Forest Grove area have graciously allowed us and fins provide the most readily available access to their land. I also thank the many information on suprafamilial interrelation- volunteers who served on the field crews dur- ships. The pelvic girdles of the "Symmori- ing the last 14 years. This work has been idae" are represented as being essentially funded by National Science Foundation identical to each other (Zangerl, 1981, figs. grants BMS 75-02720, DEB 77-02548, and 76, 77, 80) as well as to those of S. altonensis DEB 79-19492, and by grants from Carnegie (Zangerl, 1981, fig. 81, but compare this pa- Museum of Natural History, Pittsburgh, Pa. per, fig. 7). The near identity of these girdles calls into question whether discriminative LITERATURE CITED powers of the X-rays were insufficient to re- Agassiz, L. veal fine morphology or whether the girdles 1837-1843. Recherches sur les poissons fos- are, in fact, all but identical (Zangerl, 1981; siles. Neuchatel, 5 vols., 1420 pp. Zangerl and Case, 1976). Precisely the same Bendix-Almgreen, S. E. questions can be raised about the nearly iden- 1975. The paired fins and shoulder girdle in tical morphology ofthe braincases, jaws, and Cladoselache, their morphology and three phyletic significance. Coll. Intern. axial skeletons ofthe purported species. C.N.R.S., vol. 218, pp. 111-123, 4 pls. The assumption that, as Zangerl (1981) has Branstetter, S. stated, it is indeed not possible to discrimi- 1981. Biological notes on the sharks ofthe Gulf nate among the postcranial skeletons of the of Mexico. Contrib. Marine Sci., vol. three species leads to several interesting con- 24, pp. 13-24. sequences. The first is that generic-level sep- Claypole, E. W. aration ofthe symmoriids has not been prov- 1893. The cladodont sharks of the Cleveland en, the species being recognizable only by Shale. Amer. Geol., vol. 11, pp. 325- proportional and numerical differences. Be- 331, 1 pI. cause a separate derivation of the Symmo- 1895. On a new specimen of Cladodus clarki. riidae and Stethacanthidae would require nu- Amer. Geol., vol. 15, pp. 1-7, 2 pls. Cope, E. D. merous parallelisms (fig. 11, point 15'), all 1891. On the characters of some Paleozoic morphological data suggest that the "Sym- fishes. Proc. U.S. Natl. Mus., vol. 14, moriidae" are derivable most parsimonious- pp. 447-463, 5 pls. ly as a sister taxon ofS. altonensis among the Dean, B. Stethacanthidae. Consequently, a generic- 1894. Contributions to the morphology of level separation of S. altonensis and S. pro- Cladoselache(Cladodus). Jour. Morph., ductus becomes a virtual necessity. The taxo- vol. 9, pp. 85-115. nomic structure ofthe group thus either must 1909. Studies on fossil fishes (sharks, chimae- be expanded to a suprafamilial level, or the roids, and arthrodires). Memoirs of the American Museum of Natural History, family Symmoriidae must be considered a vol. 9, pp. 211-287. nomen nudum, or both. The discontinuous Dick, J. R. F. variation of pelvic structures within the 1978. On the Carboniferous shark Tristychius Stethacanthidae, ranging from plesiomor- arcuatus Agassiz from Scotland. Trans. phous for the class to selachian level (Schaef- Roy. Soc. Edinburgh, vol. 70, pp. 63- fer and Williams, 1977; Young, 1982) and 109. discontinuous variation in caudal structures 1981. Diplodoselache woodi gen. et sp. nov., 22 AMERICAN MUSEUM NOVITATES NO. 2828

an early Carboniferous shark from the Mississippian of Montana. Ann. Car- Midland Valley of Scotland. Trans. negie Mus., vol. 46, pp. 195-221. Royal Soc. Edinburgh: Earth Sci., vol. 1977b. A new petalodont (Chondrichthyes, 72, pp. 99-113. Bradyodonti) from the Upper Missis- Dingerkus, G., and T. DeFino sippian of Montana. Ann. Carnegie 1983. A revision of the orectolobiform shark Mus., vol. 46, pp. 129-155. family Hemiscyllidae (Chondrichthyes, 1982. Harpagofututor volsellorhinus new ge- Selachii). Bull. Amer. Mus. Nat. Hist., nus and species (Chondrichthyes, vol. 176, pp. 1-94. Chondrenchelyiformes) from the Na- Duffin, C. J., and D. J. Ward murian Bear Gulch Limestone, Chon- 1983. Neoselachian sharks teeth from the low- drenchelys problematica Traquair (Vi- er Carboniferous ofBritain and the low- sean) and their sexual dimorphism. Jour. er Permian ofthe U.S.A. Palaeontology, Paleontol., vol. 56, pp. 938-958. vol. 26, pp. 93-1 10. 1983. On a dentition of Polyrhizodus (Chon- Ghiselin, M. T. drichthyes, Petalodontiformes) from the 1974. The economy of nature and the evolu- Namurian Bear Gulch Limestone of tion of sex. Univ. of California Press, Montana. Jour. Vert. Paleontol., vol. 3, Berkeley, 346 pp. pp. 1-6. Glikman, L. S. 1984. Spines of the Stethacanthidae (Chon- 1967. Subclass Elasmobranchii. In Obruchev, drichthyes), with a description of a new D. V., ed., Fundamentals of Paleontol- genus from the Mississippian Bear Gulch ogy, vol. 11, Agnatha, Pisces, pp. 292- Limestone. Geobios, 17, fasc., vol. 3, 352. Israel Program for Scientific Trans- pp. 281-295. lations, Jerusalem. 1985. The morphology ofFalcatusfalcatus (St. Harris, J. E. John and Worthen), a Mississippian 1938. The dorsal spine of Cladoselache. The stethacanthid chondrichthyan from the neurocranium and jaws of Cladose- Bear Gulch Limestone ofMontana. Jour. lache. Scient. Publ. of the Cleveland Vert. Paleontol., vol. 5, pp. 1-19. Museum ofNat. Hist., vol. 8, pp. 1-12. [In press] On the identity of Physonemus at- 1951. Diademodus hydei, a new fossil shark tenuatus Davis (Elasmobranchii, Clad- from the Cleveland Shale. Proc. Zool. odontida). Jour. Vert. Paleontol.,10Ims. Soc. London, vol. 120, pp. 683-697. pp., 8 figs. Jarvik, E. Lund, R., and R. Zangerl 1977. The systematic position ofacanthodian 1974. Squatinactis montanus, a new elasmo- fishes. In Andrews, S. M., R. S. Miles, branch from the Upper Mississippian of and A. D. Walker, eds., Problems in Montana. Ann. Carnegie Mus., vol. 45, Vertebrate Evolution, Linnean Society pp. 43-54. Symposium Series, vol.4, pp. 199-225. Maisey, J. G. Kemp, A. 1981. Studies on the Paleozoic selachian ge- 1979. The histology of tooth formation in the nus Ctenacanthus Agassiz: No. 2. By- Australian lungfish, Neoceratodus for- thiacanthus St. John and Worthen, steri Krefft. Zool. Jour. Linnean Soc., Amelacanthus, new genus, Eunemacan- vol. 66, pp. 251-287. thus St. John and Worthen, Sphenacan- Kerebel, B., M.-T. Le Cabellac, G. Daculsi, and thus Agassiz, and Wodnika Munster. L.-M. Kerebel Amer. Mus. Novitates, no. 2722, pp. 1- 1978. Osteodentine and vascular osteodentine 24. of Anarhichas lupus (L.). Cell. Tissue Melton, W. G., Jr., and H. W. Scott Res., vol. 187, pp. 135-146. 1973. Conodont animals from the Bear Gulch Lund, R. Limestone, Montana. In Conodont Pa- 1974. Stethacanthus altonensis (Elasmobran- leoecology, Geol. Soc. America spec. chii) from the Bear Gulch Limestone of paper 141. Montana. Ann. Carnegie Mus., vol. 45, Moss, L. M. pp. 161-178. 1970. Enamel and bone in shark teeth: with a 1977a. Echinochimaera meltoni new genus and note on fibrous enamel in fishes. Acta species (Chimaeriformes), from the Anat., vol. 77, pp. 162-187. 1985 LUND: STETHACANTHID REMAINS 23

Moy-Thomas, J. A. plesiomorphy. Bull. Amer. Mus. Nat. 1934. On the teeth of the larval Belone vul- Hist., vol. 167, pp. 150-276. garis and the attachment ofteeth in fish- St. John, 0. H., and A. H. Worthen es. Quart. Jour. Micros. Sci., vol. 76, 1875. Descriptions offossil fishes. Geol. Surv. pp. 481-497. Ill., vol. 6, pp. 245-488. 1939. The early evolution and relationships of Schaeffer, B. the elasmobranchs. Biol. Rev., vol. 14, 1967. Comments on elasmobranch evolution. pp. 1-26. In Gilbert, P. W. et al., Sharks, Skates Moy-Thomas, J. A., and R. S. Miles and Rays. Baltimore, Johns Hopkins 1971. Paleozoic Fishes, 2nd ed. Philadelphia, Press, pp. 3-35. Saunders. 259 pp. Schaeffer, B., and M. Williams Nelson, G. J. 1977. Relationships of fossil and living elas- 1968. Gill arch structure in Acanthodes. Nobel mobranchs. Amer. Zool., vol. 17, pp. Symp. 4, pp. 128-143. 293-302. 1970. Pharyngeal denticles of sharks, with Schmidt, W. J., and A. Keil notes on the dermal skeleton of verte- 1971. Polarizing microscopy ofdental tissues. brates. Amer. Mus. Nat. Hist. Novi- New York, Pergamon Press, 577 pp. tates, no. 2415, pp. 1-26. Scott, H. Newberry, J. S. 1973. New Conodontochordata from the Bear 1889. Paleozoic fishes ofNorth America. U.S. Gulch Limestone (Namurian, Mon- Geol. Survey Monogr., vol. 16, 340 pp., tana). Mus. Paleontol. Michigan State 53 pls. Univ., Paleontol., vol. 1, pp. 81-100. 1897. New species and a new genus of Amer- Smith, B. G. ican Paleozoic fishes, together with notes 1937. The anatomy of the frilled shark, on the genera Oracanthus, Dactylodus, Chlamydoselachus anguineus Garman. Polyrhizodus, Sandalodus, Deltodus. New York, Bashford Dean Memorial, Trans. N.Y. Acad. Sci., vol. 16, pp. 282- vol. 6, pp. 333-505. 304. Traquair, R. H. Newberry, J. S., and A. H. Worthen 1888. Notes on Carboniferous Selachii. Geol. 1866. Descriptions of new species of verte- Mag., ser. 3, vol. 5, pp. 81-86. brates, mainly from the Subcarbonifer- Woodward, A. S. ous limestone and Coal Measures of Il- 1886. On the anatomy and systematic posi- linois. Geol. Surv. Ill., vol.2, pp. 9-134, tion ofthe Liassic selachian Squaloraja 13 pls. polyspondyla Agassiz. Proc. Zool. Soc. Orvig, T. London, no. 35, pp. 527-538. 1966. Histological studies of ostracoderms, Young, G. C. placoderms and fossil elasmobranchs. 1982. Devonian sharks from southeastern 2. On the dermal skeleton of two late Australia and Antarctica. Palaeontolo- Paleozoic elasmobranchs. Arkiv for gy, vol. 25, pp. 817-843. Zoologi, vol. 19, pp. 1-30, 5 pls. Zangerl, R. Radinsky, L. 1968. The morphology and developmental 1961. Tooth histology as a taxonomic crite- history of the scales of the Paleozoic rion for cartilaginous fishes. J. Morph., sharks Holmesella sp. and Orodus. No- vol. 109, pp. 73-92. bel Symp. 4, pp. 399-412, Stockholm, Reif, W. E. Almqvist and Wiksell. 1979a. Structural convergences between enam- 1973. Interrelationships of early chondrich- eloid of actinopterygian teeth and of thyans. In Greenwood, P. H., R. S. shark teeth. Scanning Electron Micros. Miles, and C. Patterson (eds.), Interre- 1979, pp. 546-554. lationships of Fishes, Suppl. 1, Zool. 1979b. Morphogenesis and histology of large Jour. Linnean Soc. London, vol. 53, pp. scales of batoids. Palaeontol. Zeitschr., 1-14. vol. 53, pp. 26-37. 1981. Chondrichthyes I. Paleozoic Elasmo- Rosen, D. E., P. L. Forey, B. G. Gardiner, and C. branchii. Handbook of Paleoichthyol- Patterson ogy, vol. 3A, Stuttgart, Gustav Fischer 1981. Lungfishes, tetrapods, paleontology, and Verlag, 115 pp. 24 AMERICAN MUSEUM NOVITATES NO. 2828

Zangerl, R., and G. R. Case shales of North America. Palaeonto- 1973. Iniopterygia, a new order ofchondrich- graphica, Abt. A, vol. 154, pp. 107-157. thyan fishes from the Pennsylvanian of Zangerl, R., and M. E. Williams North America. Fieldiana Geol., mem. 1975. New evidence on the nature of the jaw 6, pp. 1-67. suspension in Paleozoic anacanthous 1976. Cobelodus aculeatus (Cope), an anacan- sharks. Palaeontology, vol. 18, pp. 333- thous shark from Pennsylvanian black 341.

Recent issues of the Novitates may be purchased from the Museum. Lists of back issues of the Bulletin, Novitates, and Anthropological Papers published during the last five years are available free of charge. Address orders to: American Museum of Natural History Library, Department D, Central Park West at 79th St., New York, N.Y. 10024.