PaleoBios20(3):ll-37, December 15, 2000 £ 2000 University of California Museum of Paleontology Neuralethopterids (Trigonocarpalean Pteridosperms) from the Early Pennsylvania!! of eastern North America PIERRE GOUBET1, HERMANN W. PFEFFERKORN1*, and WILLIAM H. GILLESPIE2 1 Department of Earth and Environmental Science, University of Pennsylvania, Philadelphia, PA 19104-6316 hpfcffcrffl sas.upenn.edu. 2West Virginia University, 916 Churchill Circle, Charleston, 25314-1747 The genus Neuralethopteris Cremcr ex Laveinc is common in terrestrial strata of Late Mississippian and Early Pennsylvanian age in the Euramcrian Floral Realm. This is the first study of the genus as a group in North America where it has biostratigraphic, phylogenetic, paleoclimatologic, and paleoecologic significance. Most of the material examined was collected from the area of the "proposed Pennsylvanian System stratotype" project (l'SS) in West Virginia ,uk{ Virginia, but some additional material and records from other areas of North America have been included. Six species can be recognized in the North American material: Neuralethopteris schlehanii (Stur, 1877) Laveine (1967); N. pocahontas(White, 1900) n. comb.; N. smithsii(Lesquereux, 1880) n. comb.; N. sergiorum n. sp.; N. pentias(White, 1900) n. comb.; and N. biformis(Lesquereux, 1880) n. comb. Paleobiogcographically, the distribution of the six species shows differences. Neuralethopteris pocahotttas has been reported in West Virginia and in the Southern Anthracite field, N. pentias'm the Anthracite fields, N. schlehanii throughout North America. N. smithsii throughout the Appalachian basin, and N. sergiorum and N. biformisin West Virginia and Alabama. The consistent occurrence of foliage of N. pocahontas with Aulacotbeca-Yikc prcpollen organs and Tr(go»ocarpus-\ike ovules allows for a partial plant reconstruction. These results are a step towards better species delimitation within the genus Neuralethopteris .\m\ a better understanding of its natural affinities, including its relationship to the genera Alethopteris, Cardioneuropteris, and Crossopteris. INTRODUCTION The genus NeuralethopterisCremer ex Laveine (1967) is a common component found in Namurian and I.angsettian (al. Wcsiphalian A) floras throughout Europe. It occurs regularly in I.ate Mississippian and Early Pennsylvanian floras in North America, but it has not been formally recognized on this continent. Species of Neuralethopteris often dominate plant fossil assemblages. They are biostratigraphically important and the extinction of the genus coincides with that of at least one other significant ptcridosperm group, the Lyginopterids, at the end of the I.angsettian. Neuralctbopterish&s been reported from western Europe (Laveine 1967, Josten 1991) to the Caucasus (Anisimova 1979). It is characterized by its frond architecture (Laveine el al. 1993a) and the alcthoptcrid-like venation of its neuropterid pinnules. According to the European Carbon- iferous literature, five species are currently included in the genus. These are Neuralethopteris schlehanii (Stur 1877), N. rectinervis (Kidston 1888), N. larischii(Susta 1930), N. joiupuansiiLaveine (1967), and N. densifolia Josten (1983). Two species, N. larischii and N. schlehanii are used in the mcgafloral zonation of Wagner (1984). In North America, "Neuroptcris schlehanii" has been reported (Jongmans 1937, Bell 1944, Arnold 1949) to- gether with two other species, "Neuroptcris" smithsii Lesquereux (1880) and "Neuroptcris" pocahontas White ' author for correspondence (1900). Both species closely resemble N. schlehanii and have been considered by some authors as possibly belong- ing to the genus (Gastaldo 1985, Gillespie and Rhcams 1985, Lyons et al. 1985). These forms occur in the Lower Pennsylvanian of Utah (Tidwell 1967), Alabama and Geor- gia (Lesquereux 1880, 1884, 1888, Gillespie and Rhcams 1985, Gillespie and Crawford 1985, Gastaldo 1985), West Virginia (Gillespie and Pfefferkorn 1979, Blake et al. 1995), the Anthracite fields of western Pennsylvania (White 1900), and the Canadian Maritime Provinces (Bell 1944) (Eig. 1). Read and Mamay's (1964) mcgafloral zone 4 in North America is based on the occurrence of "Neuroptcris'" pocahontas and "Mariopteris" eremopteroides. Williams (1937) proposed to synonymize "Neuroptcris" pocahontas and "Neuroptcris" smithsii with "Neuroptcris" schlehanii, but subsequent authors have not followed this suggestion. Crookall (1959) also lists N. pocahontas as a synonym for N. schlehanii. No monographic work has been done on the genus in North America although the stratigraphic ranges of the different forms in the "Pennsylvanian System stratotype" (PSS) area (Gillespie and Pfefferkorn 1979) have been improved. Some palcoccological data have been repotted by Gastaldo (1987). The purpose of this study is to reassess current species concepts within the genus Neuralethopteris and similar forms, and to determine the specific occurrences of this genus in North America. This taxonomic study is based only on a morphological approach. However, as much as possible, it was carried out while considering information 12 PALEOBIOS, VOL. 20, NUMBER 3, DECEMBER 2000 COLORADO I ~ —------1 MICHIGAN COAL FIELD • SOUTHERN ANTHRACITE FIELD 1 Pran coal group Mary Lee coal group Black Creek coal group J coal group Unnamed coalbed j NOVA SCOTIA Brookwood coal group Ullcy Coal group Gwia coal group Cobb coal group Iaeger Castle Sewell Welch Beckley Fire Creek Little Fire Creek Pocahoni as No. 8 Pocahom as No. 6 Pocahoni as No. 5 Pocahoni as No. 3 Pocahoni as No. 2 Pocahom as No 1 Fig. 1. Map of the major Karly Pcnnsylvanian fossil plain localities in North America and lithostratigraphy of the Black Warrior (1, from Gillespie and Rheams, 1985) and West Virginian coal fields (2, from Gillespie et al., 1989). derived from biostratigraphy, palcobiogeography, and pa- leoecology. Type and figured specimens have been checked, and European collections have been studied for compari- son. Comments on the biostratigraphy and paleo biogcography of Neuralethopteris are based on published sources. The North American material of Neuralethopteris an be split into six species, Neuralethopteris schlchanii, N. pocahontas n. comb., N. smithsii n. comb., N. pentias n. comb., N. sergiorum n. sp., and N. biformisa. comb. We have defined the stratigraphic ranges of these taxa in the PSS area, but it should be noted that stratigraphy in the PSS area is currently being reevaluated (B.M. Blake Jr., West Virginia Geological Survey, personal communication). Therefore, some ranges presented here may be extended to higher strata in the near future. The probable male and female reproductive structures of N. pocahontas, Aulacotheca and Trir/onocarpus, are described in detail. The recognition of this natural association is an important step towards the reconstruction of the whole plant. Originally, the authors had planned to include the genus Crossopteris Tidwell (1967) in this study. However, it was soon recognized that the representatives of the Neuralcthoptcris-group from western North America re- quire a separate study based on more material than we had available at this time. MATERIALS AND METHODS Materials The specimens examined in this study come mainly from the proposed Pcnnsylvanian System stratotype collection of the U.S. Geological Survey, the main part of which is temporarily housed at the University of Pennsylvania (Phila- delphia). Specimens come from localities throughout the Pocahontas and New River Formations (Early Pcnnsylva- nian). Most of the material is represented by isolated pinna fragments. In order to study larger, more complete speci- mens, additional research was conducted in Alabama. This involved collecting specimens from known localities as well as studying collections made by Robert A. Gastaldo in the Plateau and Warrior coal fields. Type and figured specimens of the North American material were examined in the collections of the National Museum of Natural History of GOUBETet al.-EARLY l'KNNSYLVANIAN PTF.RIDOSPF.RMS 13 the Smithsonian Institution (USNM, Washington, DC), the Alabama Museum of Natural History (Tuscaloosa, Alabama), and the Geological Survey of Alabama (AGS SM, Tuscaloosa). The collections of the Laboratoire de Palcobotanique de Lille (Prance) were consulted for com- parison. Figured specimens of this study have been deposited in the USNM type collection. For specimens belonging to the PSS collection, both the PSS number and USNM number arc provided because the PSS number contains locality information. A list of the pertinent PSS localities accompa- nies the figured specimens in the USNM. This list contains both PSS numbers and U'l'M coordinates and/or locality descriptions. Several hundred specimens from more than 70 localities in West Virginia, Virginia, Alabama, and Pennsylvania were studied (Fig. 1). All are adpressions (compression-impres- sions), preserved in different types of matrix, and show different contexts of deposition. Methods Morphological terms used in the description of Neuralethopteris species are shown in Figure 2. Whole frond reconstructions (I.aveine et al. 1993a) have highlighted the in-frond variability of some pteridospermous plants and the advisability of having large specimens for final clarification. When these are not avail able, it is necessary to use the least variable pinnules, in this case mid pinnae pinnules of ultimate pinnae to define a species. In this study, it became apparent that morphological trends from the terminal pinnule to the most proximal ones in the ultimate and penultimate pinnae are valuable for species characterization in 'Neuralethopteris. This morpho- genetic differentiation is herein referred to as the distal to proximal morphological change (DPMC). In Neuralethopteris, the pinnules are very similar to each other in the proximal part of the pinna, in contrast to differences in shape and venation that occur in the distal part of the pinna. The zone in which the distal to proximal morpho- logical change occurs can be "narrow," the change can be observed over a distance of one or two pinnules (Fig. 3.1) or "wide," the change is spread over six or more pinnules (Fig. 3.2). A DPMC zone is also developed near the tip of pinnae of lower orders (Fig. 3.3). Fhe size of the DPMC zone implies two related conse- quences: narrow zones are associated with lower pinnule polymorphism and a smaller number of lobate pinnules compared with the number of entire pinnules. This could explain field observations of the constant size of dispersed pinnules of N. biformis. In the case of a wide DPMC, a distinction may be useful between the pinnule shape and the venation. For example, in N. smithsii the venation DPMC zone is wider than the one based on pinnule shape. In N. pocahontas the DPMC zone is wide for both pinnule shape and venation. However, the DPMC" zone may be different in terminal parts of a large frond (J.-P. Laveine, written communica- tion, March, 2000). We could not check this observation in the material at our disposal. If this observation of differ- niermcdioK pan of ihc front) vl "1 the Iioiitl Fig. 2. Morphological terms used in the description <>t Neuralethopteris species. Mm LIILLHI Fig. 3. Distal to proximal morphological change (DPMC) in ultimate and penultimate pinnae. 1. Short DPMC zone in an ultimate pinna. 2. Long DPMC' zone in an ultimate pinna. 3. DPMC" in a penultimate pinna. 14 PALEOBIOS, VOL. 20, NUMBER 3, DECEMBER 2000 ences of the DPMC within one species holds, the concept of the DPMC" may be less useful than we assume. Diagnoses are provided even if they are not formally required as many of the species studied were originally published with description and figure. In the synonymy lists wc have added several symbols as suggested by Matthews (1973): v—"vide," at least one author has seen the spccimen(s); *—recognized by us as basionym; p—only part of the specimens figured under this name in the publication belong to this species; ?—assignment subject to doubt because of the method of presentation. However, the synonymies are restricted to figured specimens from eastern North America. The photographs of specimens are in natu- ral size, except where a different scale is indicated. The species are presented and discussed as two morpho- logical groups. Group one is comprised of N. schlehanii, N. pocahontas, and N. smitbsii. These species are characterized by the occurrence of both larger and smaller pinnules. Group two is comprised of N. sertjiorum n. sp., N. pentias, and N. biformis. These species are characterized by larger pinnules that show less variation in size and shape. systematic: paleobotany Class: Pteridospermopsida Oliver and Scott 1904 Order: Trigonocarpales Meyen 1987 Family: Neuralethospermaceae Laveinc 1967 Genus: Neuralethopteris Crcmer ex Laveine 1967 Type— N. schlehanii Stur Diagnosis (translated from Laveine (1967) with modifica- tions)—Pinnules usually tongue-shaped, with a cordate base at least in some parts of the ultimate pinnae, occasionally stalked in the proximal part of the ultimate pinnae, more or less attached by the entire base in the distal part of the pinnae, apex more or less rounded. Pinnules generally closely spaced. Terminal pinnule strongly varying in shape and size between the different species: ovate, lanceolate or linear. Venation alethopteroid. Midvein strong, reaching nearly to the apex of the pinnule. Literal veins departing at an acute angle, bending more or less rapidly to reach the margin, i.e. appearing distinctly perpendicular or sub-per- pendicular to the margin; forking two or three times. First bifurcation close to the midvein; the second about halfway between the midvein and the margin, often directly follow- ing the major curvature of the veins; the third, if present, near the margin. Veins typically remain parallel after the second fork. Ultimate pinnae elongated triangular or elongated ob- long. Penultimate pinnae elongated triangular. Rachises longitudinally striated, without intercalary ele- ments and without cyclopterid elements at the base. Main dichotomy in lower third of frond. Description -The shape and venation of the pinnules are important as diagnostic features. However, each form included in the genus shows, at least in some parts of the frond, typical Neuralethopteris pinnules, even if they are not the most common form. These pinnules are tongue-shaped and cordate, with a rounded apex. The clearly visible vena- tion is characterized by a well-defined midvein extending at least two thirds of the length of the pinnule. Literal veins depart at an acute angle, are more or less curved, and are usually grouped in fours, but this latter characteristic is not obvious in N. senfiorum or N. biformis where the vein density is high. Frond architecture is another useful character of Neuralethopteris. Fronds have a main dichotomy in the proximal part (Laveine et al. 1993a), and neither the peti- oles nor the primary rachises bear rachial elements, such as pinnae, cyclopterid elements, or pinnules. In that aspect, the frond of Neuralethopteris is similar to the frond of Alethopteris and Cardioneuropteris (Laveine et al. 1993a). In the North American material, no specimens were found that would improve knowledge of Neuralethopteris frond architecture or change the reconstruction proposed by Laveine et al. (1993a). Comparison with related taxa—Based on its associated reproductive structures, Neuralethopteris belongs to the trigonocarpalean pteridosperms and it seems to be closely related to Cardioneuropteris Goganova, Laveine, Lemoigne and Durante, Alethopteris Sternberg, Lonchopteris Brongniart and Crossoptcris Tidwell. Cardioneuropteris is best known from a Visean locality in Kazakhstan that yielded an association of complete fronds, stems, prepollen organs and ovules (Goganova et al. 1993). The frond architecture of Cardioneuropteris shows close affinities with Neuralethopteris and the two genera share Aulacothcca-Wkc and Trigonocarpus-Wkc reproductive structures. Some Alethopteris and some Lonchopteris species have the same frond architecture as Neuralethopteris. The reproductive structures reported for Alethopteris belong in some cases to the Whittleseya-Aulacotheca complex (prepollen organ) and the Trigonocarpus-Pachytesta type ovules (Stidd 1981). The genus Crossoptcris Tidwell shows close affinities with Neuralethopteris. On the basis of figured specimens, the only differences between the two genera are the pro- portion of dccurrent pinnules in the ultimate pinnae and the dccurrent catadrome lobe, even in the proximal part of the ultimate pinnae of Crossoptcris. 'The stem anatomy of Neuralethopteris is unknown. However, Alethopteris foliage grows on Medullosa-type stems (Pryor 1989) and as a working hypothesis we postu- late that the same will be true for Neuralethopteris. Discussion—The historical development of the generic concept of Neuralethopteris has been summarized by Laveine (1967). Subsequently, the genus has been used in European monographic works (Jostcn 1983, 1991). deal and Shute (1995) reviewed the genus but included only European species in their consideration. Our knowledge of frond architecture has been improved recently by the dis- cover)' of a specimen showing a main dichotomy in the proximal part of the frond (Laveine ct al. 1993a). As GOUBET et al.-EAKLY PENNSYLVANIAN PTERIDOSPERMS 15 suggested by Laveine et al.( 1993a), it becomes clearer with new data that Neuralethopteris, Cardioneuropteris and some species of Alethopteris and Lonchopteris form a distinct group within the Carboniferous trigonocarpalean pteri- dosperms to which wc can add Crossopteris. Material belonging to the Neuralethopteris schlehanii group has been reported from a number of North Ameri- can localities. Lesquereux (1880) discussed the similarities between "Neuropteris'" biformis and "Neuropteris" dluhoschiiStur; Laveine (1967) synonymi/.ed the latter with Neuralethopteris schlehanii. Williams (1937) suggested com- bining N. pucabontas and N. smithsii with N. schlehanii. Bell (1944) noted that "Neuropteris" smithsii shows close analogies and may have close affinities with "Neuropteris" schlehanii. (ongmans (1937) mentioned the similarity be- tween "Neuropteris'1'' pocahontas and "Neuropteris schlehanii". He noted that "Neuropteris"'' pocahontas was "in many respects comparable with other specimens of the same locality, which however would be named N. schlehanii" and that: "Although some of the forms of N. pocahontas can easily be separated from N. schlehanii, it cannot be denied that they are closely related and in fact passing into each other." White (1900) also noted the similar aspect among some other "Neuropteris" species from the lower Pennsylvanian. He wrote in the description of "Neuropteris" pocahontas: "like the other forms from the basal portion of the Pottsville series, it (N. pocahontas) is essentially a Ncurocallipteris"... "certain other neuropterids of the type of N. biformis, which is a typical Neuralethopteris, similarly serve as connecting links be- tween the genera Neuropteris and Alethopteris." In his considerations. White based his conclusions only on the characteristics of the pinnules. Recent workers have pointed out that the neuropterids from the lower Pennsylvanian probably belong to the genus Neuralethopteris. Gillespie et al. (1978, p. 103, Blake et al. in press) recognized the genus, but deferred formal assign- ment until a revision could be prepared. Gillespie and Rheams (1985), in discussing the need for taxonomic revisions of Carboniferous foliage in North America, wrote: "...study of the types (of the early neuropterids) in the U.S. National Museum, along with hundreds of specimens col- lected in the central and southern Appalachians, indicates that it will not be a simple problem to solve. It is, however, definite that all of the early neuropterid complex belong to the genus Neuralethopteris And that at least one and possi- bly two additional genera may be combined with this genus.''' Gastaldo (1987), in dealing with paleoecological reconstructions, posited that Neuropteris smithsii and N. pocahontas may belong to Neuralethopteris. Lyons et al. (1985), in their study of the lower Pennsylvanian bio- stratigraphy in Alabama, also pointed out the taxonomic problems within the neuropterid forms from this interval. Assignment of species to the genus—The assignment of species to Neuralethopteris is based on two primary characteristics. One is the shape and venation of the pin- nules, the other is frond architecture. The two are comple- mentary and both should be considered whenever possible, since the distinction between Neuralethopteris and Neuropteris is difficult. The study of frond architecture requires large fragments of relatively complete specimens, a preservation that is rare due to apparent pre-prcscrvation breakup of the frond. We have seen large specimens of Neuralethopteris schlehanii and N. smithsii only. Another aspect of frond architecture is the absence of cyclopterid elements in Neuralethopteris. The lack of cyclopterid elements in the Lower Pennsylvanian is one argument that increases the probability of assigning ncuroptcroid foliage of that time interval to Neuralethopteris. Neuralethopteris schlehanii (Stur 1877) Laveine 1967 Eigs. 4, 5.1-5.2, 5.7-5.8 Neuropteris schlehanii *STUR, 1877, PI. 27, Fig. 7-8; JONGMANS, 1937, PI. 21, Fig. 44; PI. 22, Fig. 48- 49; BELL, 1944, PI. 30, Fig. 3; PI. 32; PI. 33, Fig. 5- 6; 'LAVEINE, 1967, PI. 5-8; JENNINGS, 1984, PI. 4, Fig. 1; LYONS, ZEISSNER, BARWOOD AND ADINOLFI, 1985, PI. 10, Fig. C; PI. 11, Fig. G; Neuropteris elrodii VLESQUEREUX, 1879, PI. 13, Fig. 4; Neuropteris dluhoschii "READ, 1934, PI. 17, Fig. 5; Neuropteris pocahontas cf. var. inaequalis 'LYONS, ZEISSNER, BARWOOD AND ADINOLFI, 1985, PI. 6, Fig. G; Neuropteris pocahontas 'GILLESPIE AND RHEAMS, 1985, PI. 2, Fig. 7. A synonymy of additional taxa found in Europe is pre- sented by Cleal and Shute (1995, p. 24). Diagnosis (translated from Laveine 1967)—Pinnules usually tongue-shaped, sides of the lamina parallel along the lower two thirds of the pinnule, and rapidly convergent to a rounded apex, or sometimes triangular with an acute apex. Pinnule base cordate, broadly attached only in the distal part of the pinna. Pinnules alternate or sometimes subopposite, with space between margins or close together. Terminal pinnule usually linear with a rounded apex. Venation sometimes coarse, midvein distinct through three quarters of the length of the pinnule. Secondary veins forking usually twice, relatively far apart, distinctly grouped into bundles of four, reaching the margin almost at a right angle. Ultimate pinnae very elongated, alternate, touching each other. Description— Neuralethopteris schlehanii is often repre- sented in the fossil record by fragments of ultimate and penultimate pinnae and pinnules. Several large specimens have been collected in Europe (Leggewie and Schonefeld 1957, Laveine 1967, Laveine et al. 1993a). Pinnules of this species show a wide variation in size being 5-20 mm long and 3-10 mm wide. The average 16 PAI.EOBIOS, VOL. 20, NUMBER 3, DECEMBER 2000 »- - Fig. 4. Neuralethopterisschlehanii(Stur) Laveine. 1, 2. Bipinnate fragment, American coal (Pratt Group), Dora, Alabama (Dora 7.5' Quadrangle T5S R4W), USNM 489961. GOUBETct al.-EARLY PI-NNSYLVANIAN PTKRI OOSPERMS \7 Fig. 5. Natvaktboptcrissclilchanii(Stur) Laveine and N. pocahontas(White) n. comb. 1,2. M schkhanii, bipinnate fragment, American coal (Pratt Group), Dora, Alabama (Dora 7.5' Quadrangle T5S R4VV), USNM 489962. 3,4. N. pocahontas, bipinnate fragment showing the typical aspect of the pinnules in the intermediate part of the pinnule, the enlargement shows the venation and the attachment at the base, Pocahontas No. 1 coal, PSS area. West Virginia, USNM 489963. 5, 6. N. pocahontas, distal secondary, or proximal tertiary pinna, Pocahontas No. 1 coal. PSS-2. Garwood, West Virginia, USNM 489964. 7,8. N. schkhanii, distal secondary, or proximal tertiary pinna, Pocahontas No. 4 coal . PSS-342. Ammonatc, West Virginia, USNM 489965. 18 PALEOBIOS, VOL. 20, NUMBER 3, DECEMBER 2000 length is close to 10 mm while the average width is 5 mm. The zone of distal to proximal morphological change of the pinnules in a pinna is narrow and involves only few pin- nules. The pinnules arc tongue shaped, elongated, and at least twice as long as wide. The pinnule base is cordate, except in the distal part of the pinnae. The midvein is rather wide compared to the width of the base of the pinnule. In some cases, due to preservation bias, it appears as if the pinnules were attached at the base, but slightly constricted. This is an appearance that is seen in Neurakthopteris schlehanii, but not in N. jongmansii where the midvein is thin com- pared to the width of the pinnule. The pinnules are, compared to some other forms of Neurakthopteris, sym- metrical with respect to the midvein. Venation is always visible and sometimes even coarse. The secondary veins emerge at an acute angle and arc regularly curved (compared with N. rectinervis) up to the second fork. The first fork occurs in the first quarter of the course, the second in the middle. After the second fork, the veins remain straight and parallel, reaching the margin at nearly a right angle. Veins appear to be grouped in fours. Sometimes there is no second fork in one of the two dichotomies, or there is a third fork near the margin. Three characteristics separate this species from others: the cordate base, the curvature of the venation, and the narrow zone of distal to proximal morphological change. Discussion—In his study of the species of Neuraktbopteris from northern France, Laveine (1967) recognized two species in material originally assigned to N. schlehanii: N. schlehanii (Stur) Laveine and N. jongmansii Laveine. These two forms are distinguished based on the shape of the pinnules, which is more opulent in N. jongmansii, and on the venation, which is very regular in N. jongmansii, both in terms of curvature and density. Apart from these morphological characteristics, the second spe- cies was created because of its particular stratigraphic range in Western Europe. In North America there are no speci- mens showing the characteristic aspect of N. jongmansii. Occurrence—Reports of Neurakthopteris schkhanii in North America arc rare. Jongmans (1937) figured some specimens from West Virginia. Bell (1944) described and figured specimens of N. schlehanii from Nova Scotia and New Brunswick in the Riversdale and Cumberland Groups. Arnold (1949) also recorded specimens near the base of the Pennsylvanian section from the Michigan Coal Basin which clearly belong to this species. Jennings (1984) figured a specimen from southern Illinois. A small, badly preserved specimen was illustrated by Read (1934) from the base of the Weber Formation (Pennsylvanian) in Colorado. Lyons et al. (1985) figured and described specimens from Ala- bama under the name of N. schkhanii; some belong to this species, others arc too fragmentary for assignment, and others are clearly other species (for instance the specimen PI. 14, figure C is Neurakthopteris biformis). Gillespie and Rheams (1985) figured two specimens of N. schlehanii, but these may be Neurakthopteris biformis. In the PSS area Neurakthopteris schkhanii occurs in the sediments immedi- ately above and below the Pocahontas No. 3 coal bed (PSS 342), and in the roof shale of the Sewell coal bed (PSS 20), but is not common between these two coals. In the Black Warrior Basin of Alabama, specimens of N. schlehanii were found by the senior author in clastic rocks associated with the Pratt and Mary Lee coal groups. According to figured specimens the species occurs in Alabama also in the Cobb and the Brookwood coal zones (localities 24 and 33 in Lyons et al. 1985). Thus the reported stratigraphic range in Alabama is from the Mar)' Lee to the Brookwood coal zone. In Europe, N. schkhanii ranges from the middle of the lower Namurian (Serpukhovian) to the top of the Langsettian (formerly Westphalian A). Neurakthopteris pocahontas (White, 1900) n. comb. Figs. 5.3-5.6, 6, 7.3-7.7 Neuropteris pocahontas¦ "'WHITE, 1900, p. 888, PI. 189, Fig. 4; PI. 191, Fig. 5; JONGMANS, 1937, PI. 13, Fig. 15; PI. 14, Figs. 16-20; PI. 16, Fig. 27; PI. 17, Fig. 30; VREAD AND MAM AY, 1964, PI. 4, Fig. 2; 'GILLESPIE, CLENDENING AND PFEFFERKORN, 1978, PI. 2, Fig. 1; PI. 47, Figs. I- 3, 5, 7-8; 'GILLESPIE AND PFEFFERKORN, 1979, PI. 1, Fig. 1; Neuropteris pocahontas var. inaequalis 'WHITE, 1900, PI. 188, Fig. 5; PI. 190, Fig. 7-8; PI. 191, Fig. 1-4; Neuropteris pocahontas var. pentias 'GILLESPIE AND PFEFFERKORN, 1979, PL 1, Figs. 8-9. Holotype— Neuropteris pocahontas, White, 1900, Pottsvillc, PI. 191, Fig. 5, rcfigurcd here, Fig. 7.3. Depos- ited in the USNM, collection number USNM 40751. Diagnosis (revised)—Typically represented by two types of pinnules, with intermediate representatives. (1) Pinnules either oblong, tongue-shaped with a cordate base or (2) small compared to the former, rounded, and broadly at- tached; both forms have a rounded apex, a well marked venation and are asymmetrical. Pinnule shape varying ac- cording to placement on the pinna: smaller pinnules occur- ring in the intermediate part of the ultimate pinnae. Midvein not clearly differentiated in the smaller pinnules, with sev- eral veins merging directly from the rachis, forking gener- ally twice. In elongated pinnules, the midvein is well marked, almost reaching the apex, secondary veins gener- ally fork twice, once near the midvein, and again between the midvein and the margin, rarely a third time near the margin. Ultimate pinnae rather short. Description— Neurakthopteris pocahontas (White) comb. nov. is known from fragments of ultimate pinnae and occasionally from long penultimate pinnae. Pinnules are 5-20 mm long and 3-7 mm wide (Fig. 6). Within N. pocahontas, there are two morphogroups occurring in the same localities and in similar numbers and GOUBETet a/.-EARLY PENNSYLVANIAN PTERIDOSPERMS 19 there arc several intermediate stages. One morphogroup is represented by large, elongate pinnules (10-15 mm long) with a cordate base and with a well marked mid vein. Pinnules in this morphogroup are suggestive of those of N. scblehanii and resemble N. jonijmansii (Fig. 7.5, 7.7). These pinnules are found in the distal part of the penultimate pinnae and probably in the proximal part of the ultimate pinnae situated in the proximal and intermedi- ate parts of the frond. The second morphogroup is repre- sented by rounded pinnules that are smaller than those of the other morphogroup (Fig. 5). Pinnules of this type are 5-9 mm long, show a broad basal attachment, the midvein is often obscure, making its characterization difficult. Sec- ondary veins emerge from the midrib at an open angle, fork two or three times, remaining straight as they reach the lateral margins at an open angle. Some veins emerge di- rectly from the pinna axis (Fig. 7.5, 7.7) as rachial veins. For both morphogroups the apex of the pinnule is rounded, venation is well marked, and the pinnules arc obliquely inserted on the rachis at an angle between 50" and 70" ( Figs. 5.3-5.4, 7.3-7.7). The lamina of the pinnules of N. pocabontas seems to be thin compared to N. schkbanii, but this may be a prcscrvational artifact. This species is com- monly associated with Aulacotbcca-Wkc pollen organs and 1 virion oca rpus-likc ovu les. Discussion—There is the possibility that this form is a variety of N. schkbanii. As discussed by Jongmans (1937), pinnules of the large-pinnuled morphogroup are difficult to distinguish from N. scblehanii. However, pinnules of the small-pinnuled morphogroup which are characteristic of N. pocabontas are not represented in other species of Nettralethopteris, except for N. smitbsii. Sometimes, and particularly in the Pocahontas No. 3 coal zone, the distinc- tion between N. pocabontas and N. smitbsii, if based only on the smallest pinnules of the two forms, is not clear. Rounded pinnules with a cordate base, and a well marked midvein, similar to those of N. smitbsii, but in the size range of N. pocabontas have been found in beds of the Pocahontas No. 3 coal interval. The pinnule variability of N. pocabontas is due to its very wide zone of distal to proximal morphological change (Fig. 7.3-7.7). In some specimens, lateral pinnules in the distal part of the pinna exhibit venation almost in continuity with the venation of the terminal pinnule (Fig. 7.5, 7.7). There- is no midvein, but several parallel primary veins, equivalent to the secondary veins of the terminal. This contrasts with the other forms of Ncuraktboptcris, and particularly with N. smitbsii (Fig. 9.2) and N. biformis (Fig. 12.7). Figured specimens of this species represent only small fragments of the distal part of pinnae. Unfortunately, no large specimens of this species have been found in the collections, preventing a better understanding of the frond architecture. Thus, nothing is known about the number of divisions of the frond, a character state that would be very important in understanding the relationship of this species to other species of Ncuraktboptcris. Fig. 6. Ultimate pinna of Ncuraktboptcris pocabontas. White (1900) distinguished N. pocabontas var. inaequalis from Ncuraktboptcris pocabontas based mainly on pinnule size. The variety is, according to White himself (1900, p. 893) very difficult to distinguished from N. pocabontas. All of White's (1900) figured specimens of this variety are from the Southern Anthracite Field in eastern Pennsylvania, whereas the figured specimens of Ncuraktboptcris pocabontas all come from West Virginia. The large size variation in the specimens of the PSS collec- tion indicates that specimens of N. pocabontas var. inaequalis are just larger specimens of the normal form of N. pocabontas. Stratigraphic range—In the PSS area, we found N. pocabontas from the base of the Pennsylvanian (PSS 43) to the level of the Pocahontas No. 6 coal interval (PSS 16, PSS 123, PSS 143, PSS 236). This species is very common and occurs in nearly all plant fossil localities from this strati- graphic interval in this area. Outside of the PSS area, the only unquestionable documentation of Ncuraktboptcris pocabontas comes from the lower part of the Pottsville Formation in the Southern Anthracite Field (White 1900). Ncuraktboptcris smitbsii (Lesquereux, 1880) n. comb. Figs. 7.1-7.2, 8, 9.1-4 Neuropterissmitbsiivr*\,¥SQUV.KY.\JX 1880, p.106, PI. 13, Fig. 1-3; 'ADAMS, BUTTS, COOKE AND WYTHE, 1926, PI. 70C, Fig. 1-3; JONGMANS, 1937, PI. 20, Fig. 42-43; BELL, 1944, p. 79, PI. 29, Fig. 2; PI. 30, Fig. 2; PI. 31, Fig. 1, 4; PI. 33, Fig. 3- 4; PI. 67, Fig. 4; VGILLESPIE AND PFEFFERKORN, 1979, PI. 1, Fig. 5-6; ''LYONS, /CEISSNER, BARWOOD AND ADINOLFI, 1985, PI. 2, Fig. A; PI. 6, Fig. A-B; Ncuroptcris (Ncuraktboptcris?) smitbsii "GASTALDO, 1985, PI. 3, Fig. B; Ncuroptcris smithii GILLESPIE AND CRAWFORD, 1985, PL 2, Fig. 6. Lectotype—AGS SM 25c, Ncuropteris smitbsii Lesqucreux, 1880, Coal Flora, PI. 13, Fig. 2. Despite a careful search, the specimen shown on the same plate as Fig. 1 could not be found in the AGS collection, and the specimen cited above is here designated as the type. 20 PALEOBIOS, VOL 20, NUMBER 3, DECEMBER 2000 GOUBETet a/.-KARLY PENNSYLVANIAN PTKRIDOSPERMS 21 Diagnosis (revised)—Pinnules small, often ovoid or round, sometimes triangular, cordate even in the distal part of the pinna; alternate. Venation coarse, widely spaced, midvein well marked, often appearing as a furrow, second- ary veins twice forked, reaching the margin at a right angle. Terminal pinnules often elongated, with a very well-marked midvein, either dense, parallel and well marked secondary veins, or spaced and curved. Distal pinnules in the penultimate pinnae larger than the others, with venation similar to the one of the terminal pinnules. Ultimate pinnae relatively short, often consisting of less than six pairs of lateral pinnules, triangular elongated. Description—Although large frond fragments of Neuraletboptetis smithsii are known from Alabama (for instance Adams et al. 1926, PI. 70C, Fig. 1-3; mentioned by Gastaldo 1987) the form is commonly represented by small isolated ultimate pinnae (Pig. 9.4). Analysis of a relatively complete frond specimen of this species figured by Adams et al. (1926) confirms the reconstruction pro- posed by Laveine et al. (1993a, p. 158). The pinnules of N. smithsii are small, 2-5 mm long, cordate at the base, ovate, often rounded, occasionally triangular (Lesquereux 1880, PI. 13, figure 1), inserted perpendicular on the rachis. The venation is widely spaced (Fig. 7.2), except in the terminal pinnule where the veins occasionally appear more crowded (Fig. 9.4). The midvein is well-marked and often appears like a furrow on the basal half of the pinnule (Fig. 9.1-4). The secondary veins emerge from the midvein at a relatively open angle com- pared to other species, fork twice, and reach the margin at a right angle. The veins are either widely spaced or dense; when widely spaced they lose the appearance of straight parallel lines and display a more flexuose aspect (compare for instance the Figs. 7.2 and 9.4). The ultimate pinnae are perpendicularly inserted on the rachis and often consist of a long terminal pinnule with one to four pairs of small pinnules (Figs. 7.2, 9.4). Bell (1944) noted that there were usually five or six pairs in the specimens from Canada; however, the specimen figured here in 9.1 has more pairs of pinnules. Pinnules and pinnae are commonly perpendicu- larly inserted on the axis. The two main aspects of the pinnules of N. smithsii are illustrated in Fig. 8. Ncuralcthoptcris smithsii is often associated with Anlncotheca. Discussion—Lesquereux first mentioned Neuropteris smithsii in 1876 but did not provide figures. He then later formally described and figured Neuropteris smithsii in two volumes of the "Coal Flora" (volume 1, 1879 and 1880; E 5 o Fig. 8. 1,2. Penultimate pinnae of Ncuralcthoptcris smithsii. volume 3, 1884). Four specimens are figured, three in the first volume (PI. XIII, Fig. 1-3), which were deposited in the Alabama Museum of Natural History, one in the third volume, that was deposited in the USNM in Washington, -^ Fig. 7. Ncuralcthoptcris smithsii'(Lesquereux) n. comb, and Neuraletboptetis pocahontas (White) n. comb. 1,2. N. smithsii, tripinnate fragment showing the short penultimate pinnae, perpendicularly attached to the rachis, enlargement showing the shape of the ultimate pinnae and the venation, Blount County, Alabama, USNM 489966. 3. N. pocahontas, bipinnate fragment, figured by White (1900), pi. 191, fig. 5., Pocahontas coal undetermined, Gilliam, West Virginia, USNM 40751. 4, 5. N. pocahontas, bipinnate fragment, figured by White (1900) as Neuropterispocahotitaswxr. inacqualis, pi. 188, fig. 5. Pottsville gap, Pennsylvania. USNM 40744. 6,7. N. pocahontas, bipinnate fragment showing the gradual mode of differentiation, enlargement showing the venation in the distal and proximal pinnules, Pocahontas No. 4 coal. PSS 142, Pluto Church, West Virginia, USNM 489967. 22 PALEOBIOS, VOL. 20, NUMBER 3, DECEMBER 2000 Fig. 9. Neuralethopteris smithsii (Lesquereux) n. comb. 1, 2. Bipinnate fragment, enlargement showing the shape and venation of the pinnules, Sewell coal, PSS-221, Black Mine I, West Virginia, USNM 489968. 3, 4. Slab showing dissociated pinnae and pinnules, enlargement showing the venation of the terminal pinnules, Beckley coal, PSS-9 214, Caretta, West Virginia, USNM 489969. GOUBETet al.-EARLY PENNSYLVANIAN PTERIDOSPERMS 23 DC. In the third volume of the "Coal Flora," Ixrsquereux figured (PI. XCVI, Fig. 3) a different specimen from the original specimens he figured in 1879, and chose to emend the diagnosis in order to include the new form in the species. Thus, in the literature, within the concept of Ncuralcthopteris smithsii two morphogroups have been de- fined. The first is the one we have defined in the description section, and is also the one considered to be the "true" N. smithsii by White in 1900. It is characterized by pinnules with a well marked midvein, cordate base even in the distal part of the pinnae and an elongated terminal pinnule. All these features are in accordance with the form figured and described by Lcsqucreux (1879, 1880) and Bell (1944). The second form is characterized by the lack of a midvein, the broad attachment of the pinnule base and the obtuse terminal pinnule. Specimens of this form have been de- scribed and figured by Lcsqucreux (1884) and Gillespie and Rhcams (1985, PI. 1, Fig. 1) and Lyons et al. (1985, PI. 10, Fig. A). Wc have not examined enough specimens of the second form to give an opinion of the relationship between these two distinct forms. The second form has not been found in West Virginia and is currently known only from Alabama. We propose to keep the few specimens of the second form separate from Ncuralcthopteris smithsii. Concerning the relationship of N. smithsii, among the other forms of Ncuralcthopteris, it has been well established that the three forms, Ncuralcthopteris schlehanii, N. pocahontas, and N. smithsii arc morphologically similar. Williams (1937) proposed to synonymize all three forms as one species. However, in our opinion, the forms arc mor- phologically distinct and show both geographic and strati- graphic differences, demonstrating that the separation is not only justified, but useful. Occurrence—Ncuralcthopteris smithsii is particularly abundant in the proposed Pennsylvania!! System stratotypc area in the upper part of the Pocahontas Formation and the lower and middle part of the New River Formation. Speci- mens have been found in the Pocahontas No. 2 and Pocahontas No. 3 coal intervals, where it is difficult to c Fig. 10. Pinnule of Ncuralcthopteris serjjiorum n. sp distinguish from Ncuralcthopteris pocahontas suggesting re- cent divergence of the two forms. Ncuralcthopteris smithsii is abundant from the Pocahontas No. 6 to the Sewcll coal interval. An upward extension of this range to the Castle interval results from a rcinterpretation of stratigraphic rela- tionships (B.M. Blake, oral communication). Neuralethopteris smithsii has been reported throughout the Appalachian basin, Nova Scotia and New Brunswick (Bell 1944), West Virginia (Gillespie and Pfcffcrkorn 1979), in Georgia (Gillespie and Crawford 1985), and Alabama (Gastaldo 1985, Lyons et al. 1985). White (1900) reported an occurrence in the Southern Anthracite Field, but unfor- tunately did not figure any specimens. Nevertheless, the description of N. smithsii is very precise, and White's species concept was the same as Lesqucrcux's in 1880. It may be worth noting that this typical American species has been reported in the Caucasus (Anisimova 1979, PI. 38, Fig. 3); however, the figure is too poor to test the validity of the identification. Neuralethopteris serjjiorum n. spec. Figs. 10, 11, 12.1-12.4 Neuropteris pocahontas var. inaequalis White 'GILLESPIE AND PFEFFERKORN, 1979, PI. 2, Fig. 2-3. Holotype-USNM 489970, Fig. 11.1. Below Pocahontas No. 8 coal, PSS 10, Garwood Section. Etymology—The species is named in honor of three- great paleobotanists, Sergio Archangclsky, Sergius H. Mamay, and Sergei V. Meyen, the "three Scrgeis" who met in 1979 at the International Carboniferous Congress in Illinois where preliminary papers on the "proposed Pcnn- sylvanian System stratotypc" were presented. Diagnosis—Pinnules large, elongated, even in the distal part of the pinnae, often broadly attached by the proximal side conferring a decurrcnt aspect, distal base contracted as if cut off obliquely, rarely cordate. Apex somewhat acute or rounded. Basal part of pinnules often asymmetric, usually inserted perpendicularly on the rachis, slightly inclined in the distal part of the pinnae. Venation thin and dense, sometimes coarser and well spaced, midvein well-marked. Secondary veins leave midvein at an acute angle, curving near the midvein and then remaining straight, reaching the margin at a angle of close to 60°, forking two or three times. In the broadly attached pinnules, some veins emerge directly from the rachis. Ultimate pinnae oblong, tapering in the distal part of the pinna. Description—Specimens of Neuralethopteris serjfiorum n. sp. often consist of pinnules and long uni-pinnate pinnae (Figs. 10, 11.1, 11.3), rarely of bipinnatc pinnae (Figs. 11.6, 11.7). Generally pinnules are 20-25 mm long, com- monly reaching 50 mm in length. Pinnules shorter than 15 mm are rare. The pinnules arc much longer than broad, even in the distal part of the pinna. The apex is cither somewhat acute (Fig. 11.1) or rounded (Fig. 11.4). The pinnule attach- 24 PALEOBIOS, VOL. 20, NUMBER 3, DECEMBER 2000 GOUKETct al.-EARLY PF.NNSYLVANIAN PTF.RI OOSPERMS 25 merit is a distinguishing feature for N. scrjjiorum. The distal lobe is slightly cordate or appears to be cut off whereas the proximal side is broadlv attached, with rachial veins emerg- ing directly from the pinna rachis (Fig. 12.2). However, in some cases, this characteristic is not that obvious (Fig. 11.5). The degree of attachment at the base seems to depend on placement within the frond rather than the location of the pinnule within the pinna. The lack of larger frond fragments prevents a detailed analysis of this varia- tion. Available material often displays the same degree of pinnule attachment over the entire length. Venation is also diagnostic. The midvein is well-marked, often appearing like a furrow over two thirds of the pinnule length (Fig. 11.3). Secondary veins are thin and are slightly curved when close to the midvein. They fork two or three times, the initial fork occurring close to the midvein as in the other species of the genus. The second fork occurs midway from the departure point. When present, the third fork occurs close to the margin. The secondary veins appear to be almost perfectly parallel (Fig. 12.4). The /.one of distal to proximal morphological change is very narrow in the ultimate pinnae of N. scrjjiorum (Fig. 11.6). For instance. Fig. 12.2 shows a terminal pinnule with a large basal lobe. The same specimen shows the asymmetrical shape of the pinnules just below the terminal pinnule. The pinnules are inserted perpendicularly on the rachis (Figs. 11.1, 11.2), except in the distal part of the pinna where they are often obliquely inserted (Figs. 11.6, 12.1). A diagrammatic drawing of the pinnule shape and the venation of Neurakthoptcris scrjjiorum is illustrated Fig. 10. Discussion— Neurakthoptcris scrjjiorum is very distinct in several aspects. However, some of the pinnules resemble those ol" N. schlehanii (e.g., compare Gillespie and Pfefferkorn 1979, PI. 2, Fig. 2, and the specimens figured here. Fig. 6.2 or 6.8). These pinnules seemingly link the two species and is the main reason we assign N. scrjjiorum to the genus Neuraktbopteris, even if some characters such as the venation and the attachment at the base are some- times atypical of Naimletbopteris. Gillespie and Pfefferkorn (1979) described and figured specimens of this form under the name of Neuropteris pocahontas var. inacqualis. Although there are some simi- larities, there are also strong differences. In the description of this variety, White (1900) did not mention the decurrent proximal part of the pinnules. White's figures (PI. 188, figure 5 and PI. 191 figures 1, 3, 4) show pinnae similar to those of N. pocnhontas, with the particular mode ot differ- entiation of this species. These features are not represented in Neurakthoptcris scrjjiorum. Occurrence— Neuraktbopteris scrjjiorum occurs in the Pocahontas No. 6 coal interval (PSS 16, PSS 70, PSS 236, PSS 237) and just above and below the Pineville Sandstone in the Garwood Section (PSS 10, PSS 12). A specimen collected by Robert A. Gastaldo in the Rosa seam interval in Alabama, is shown in Fig. 11.2 and 11.3. As far we know, it is the only record of the species outside of the proposed Pennsylvania!! System stratotype area. Neuraktbopterispentias (White, 1900) n. comb. Fig. 12.5-12.6 Neuropteris pocahontas var. pentias "WHITE, 1900, p. 892, PI. 188, Fig. 2-4; PI. 189, Fig. 5. Holotype— USNM 40741. Neuropteris pocahontas var. pentias White (1900), PI. 188, Fig. 2. Refigurcd here. Fig. 12.5-12.6. Description—This species is represented only by the four pinna fragments figured by White (1900) and a few other fragmentary specimens. White (1900, p. 892) studied a larger collection and from his description and our obser- vation of the type specimens, three points are obvious: (1) The pinnules of this form are of two main types; one is comprised of large, elongated, cordate pinnules with a well marked midvein, that conform to the general definition of pinnules of this genus; the other has small, rounded, broadly attached pinnules with rachial veins emerging from the axis. (2) The secondary veins in the larger pinnules are thin and oblique with respect to the midvein and the margin. (3) The larger pinnules and the terminal pinnules usually have an acute apex. Discussion— Neurakthoptcris pentias \\. comb, is similar to N. pocahontas in all characteristics but the shape of the distal part of the larger pinnules and the venation. We consider these features to be sufficient for the separation of the two forms. The venation and apical region of the pinnules of N. pentias are similar to N. scrjjiorum. However, the two can be distinguished based on pinnule size and attachment. White (1900) wrote that N. pentias is smaller than N. pocahontas, whereas N. scrjjiorum has by far the largest pinnules within the North American species of Neurakthoptcris. The attachment of the pinnules on the rachis is not decurrent as in N. scrjjiorum. The distal to proximal morphological change zone of N. pentias is simi- lar to the one of N. pocahontas. ¦^ Fig. 11. Neurakthoptcris serjjiorum n. sp. 1. I lolotype, secondary or tertiary pinnae showing the two different aspect of the pinnules, below Pocahontas coal 8, PSS-10, Garwood, West Virginia, USNM 489970. 2, 3. Ultimate pinna with an enlargement showing the perfect similarity to the specimens from West Virginia, Rosa seam, Blount County, Alabama (Clarence 7.5' Quadrangle, section 21, Tl IS R2H), USNM 489971. 4, 5. Two pinnae showing the homogeneity of the pinnule shape within the pinnae, enlargement showing the thin and oblique venation, above Pocahontas No. 8 coal, PSS-12, Garwood, West Virginia, USNM 489972. 6, 7. Bipinnate fragment, showing the distal to proximal morphological change (DPMC) and the venation, below Pocahontas No. 8 coal, PSS-10, Garwood, West Virginia, USNM 489973. 26 PALEOBIOS, VOL. 20, NUMBER 3, DECEMBER 2000 GOUBET et «/.-EARLY PENNSYLVANIAN PTERIDOSPERMS 27 Purkynova (1970) described Neuropteris multivenosa, a species that shows some apparent similarity to Neuralethopteris pentias, from the Poruba Formation (Namurian "A") in the Upper Silesian Basin, Poland and Czech Republic. However, Laveine et al. (1977, p.779) consider this species to be near the root of the Margaritopteris - Callipteridium series and Cleal and Shute (1995) transferred it to the genus Margaritopteris. Occurrence— Neuralethopteris pentias has not been found in the collection from the proposed Pennsylvanian System stratotype. The only figured specimens of this spe- cies come from the Southern Anthracite Field, Pennsylva- nia. Some specimens from West Virginia (Gillespie and Pfefferkorn 1979) and Alabama (Lyons et al. 1985) have been figured as White's (1900) variety, but they differ from the type specimens and resemble Neuralethopterisschlehanii. Neuralethopteris biformis (\xsqucrcu\, 1880) n. comb. Figs. 12.7, 13, 14 Neuropteris biformis T.ESQUEREUX, 1880, p. 121; Atlas 1879, PI. 13, Fig. 7; Neuropteris elrodii V*I.ESQUEREUX, 1884, p. 107, PI. 94, Fig. 1-2, T.ESQUEREUX, 1888, PI. 29, Fig. 1-3; Neuropteris rectinenns JONGMANS, 1937, PI. 23, Fig. 50-51; Neuralethopteris schlehanii-GILLESPIE AND RHEAMS, 1985, PI. 2, Fig. 9-10; LYONS, ZEISSNER, BARWOOD AND ADINOLFI, 1985, PI. 14, Fig. C; Neuralethopteris cf. schlehanii 'GILLESPIE AND RHEAMS, 1985, PI. 3, Fig. 7; Neuralethopteris larischii G11.1 .FSPIE, CRAWFORD AND WATERS, 1989, PI. 1, Fig. 15. Holotypc—AGS SM 34. Neuropteris biformis Ixsquercux (1880), p. 121; Lesquereux (1879) PI. 13, Fig. 7. Diagnosis—Pinnules large, always elongated and cor- date, even in the distal part of the pinnae, sometimes stalked, alternate. Literal margins tapering gradually in the distal two thirds of the pinnule. Apex usually roundly acute. Terminal pinnule fusiform. Venation coarse, always with a well marked midvein often appearing like a furrow in the lower two thirds of the pinnule. Secondary veins twice forked, occasionally forking a third time near the margin. Secondary veins curving upon leaving the midvein, then running straight from the second fork to the margin; angle with margin about 60 degrees. Ultimate pinnae oblong, slightly constricted in the proximal part, tapering in the 10 mm Fig. 13. Pinnule of Neuralethopteris biformis distal part. Penultimate pinnae somewhat triangular in the distal part. Description— Neuralethopteris biformis n. comb, is rep- resented primarily by pinnules, sometimes ultimate pinnae, and only rarely, bipinnate fragments. The size of the pin- nules is similar to N. sergiorum, i.e., 10-40 mm long, usually 20-25 mm. Pinnule shape and venation arc con- stant in Neuralethopteris biformis, only the length/width ratio changes; thus there is not a large range of variability. The pinnules are elongated, an appearance enhanced by the curved lamina margins that disappear into the rock matrix. For instance, the specimen in Fig. 14.3, shows the relief induced by the in-rolling of the margin. The longest pinnules are sometimes slightly flcxuosc (Figs. 12.7, 14.1). The sides of the lamina usually taper slightly from the base to the apex (Figs. 12.7, 14.1, 14.5). The base is always clearly cordate, even in the distal part of the pinnae. The pinnules often appear to be asymmetrical because of the differing angles under which the secondary veins approach the margin on opposite sides of a pinnule (Fig. 14.6). A typical pinnule of N. biformis is illustrated in Fig. 13. The distal to proximal morphological change zone is quite narrow and contrasts with the one of N. pocahontas, N. smithsii and N. pentias, but is similar to that of Neuralethopteris sergiorum. Specimen Fig. 12.7 shows the abrupt change between a pinnule and its neighbor on a well differentiated pinna. This lack of relatively small pinnules on the pinnae explains why most of the common specimens encountered in the sediments are large isolated pinnules. Discussion—I.esquereux figured specimens conspccific to the one wc arc dealing with under both Neuropteris elrodii And Neuropteris biformis \n several of his publications (Ixsquereux 1879, 1880, 1884, 1888). Both names were published in the first part of the Coal Flora (Atlas in 1879; Text in 1880). The original figures of the two new species -+ Fig. 12. Neuralethopteris sergiorum n. sp., N. pentias (White) n. comb, and N. biformis (Lesquereux) n. comb. 1, 2. N. sergiorum, ultimate pinna showing the shape of the pinnules in the distal part of the pinna, below Pocahontas No. 8 coal, PSS-10, Garwood, West Virginia, USNM 489974. 3, 4. N. sergiorum, ultimate pinna showing the attachment at the base and the veins merging from the rachis, Pocahontas No. 6 coal, PSS 70, Odd, West Virginia, USNM 489975. 5, 6. N. pentias, pinna showing the shape of the pinnule in the distal part of the pinna, specimen figured by White (1900), pi. 188, fig. 2,1.ykens coal No. 6, Brookside, Pennsylvania. USNM 40741. 7. N. biformis, bipinnate fragment showing the abrupt differentiation of pinnules into pinnae, specimen figured by Lesquereux (1884), pi. 96, fig. 1, Dade County, Tennessee, USNM 11840. 2