For over 150 years, lower actinopterygian, or palaeoniscoid, fishes have been described from the Albert shales of southeastern New Brunswick, Canada (Fig. 1). The term Albert shales has been used informally to refer to the Albert shale formation, the middle formation of the Horton Group between the basal Memramcook Formation and the overlying Weldon Formation (Gussow, 1953; Greiner, 1962) (Fig. 2). Greiner (1962) and Utting (1987) present the Albert Formation itself as being composed of three members – the Dawson Settlement, Frederick Brook, and Hiram Brook members (Fig. 2). St. Peter (1993) presents the Albert Formation as being composed of six stacked lithofacies – conglomerate, sandstone, mudstone, mudstone/sandstone, kerogenous mudstone, and evaporate facies.

Since the 1800s, there has been controversy over the age of the Albert Formation. While an Early Carboniferous age was supported by the fossil fish and plants (Bailey and Ells, 1878; Lambe, 1909, 1910), some cautioned that the Albert Formation could be Devonian in age (Bailey et al., 1880; Ells, 1903). This trend continued into the late 1900s. Greiner (1962, 1974) had originally described the Albert Formation as Lower Carboniferous in age, but after the description of an osteolepid, †Latvius porosus from basal beds, he considered the lower part of the Albert Formation to be Devonian (Greiner, 1977). Recent reassessments of the Albert Formation sarcopterygian material do not support Greiner's assignment of the material to the Devonian genus †Latvius, but rather suggest megalichthyid and Carboniferous affinities (Miller and Brazeau, 2007). This, combined with spore analyses, has put an end to the controversy regarding the age of the Albert Formation. The Albert Formation is Tournaisian (Lower Carboniferous) in age, near the Devonian and Lower Carboniferous boundary (Utting, 1987; St. Peter, 1993; Miller and Brazeau, 2007).

The Albert Formation is paleontologically famous for its countless articulated lower actinopterygian specimens (Greiner, 1977). These actinopterygians are important because with the supported Early Carboniferous age of the Albert Formation, they potentially bridge morphological gaps between Devonian and Carboniferous forms. Unfortunately, these fishes have not been dealt with in great detail for over a hundred years. The taxonomic history of the Albert Formation fishes is discussed here.

The palaeoniscoid fishes from the Albert Formation of New Brunswick have been a taxonomic nightmare for over 150 years. Lambe (1909, 1910) provided a taxonomic history of these fishes in his redescription of some of the New Brunswick fishes. This information is reviewed and updated below.

In 1851, Jackson described the first palaeoniscoid fishes from the Albert shales in papers entitled “Report on the Albert Coal Mine” (Jackson, 1851a) and “Descriptions of five new species of fossil fishes” (Jackson, 1851b). Though the second title claims to include the descriptions of five new taxa, only three new species were described and named – †Palaeoniscus[m] alberti, †P. brownii, and †P. cairnsii. Four additional specimens were described but never named. No type material was designated, and though plates and figures are referenced in this publication, they were never included with the text (Lambe, 1909, 1910).

Eastman (1908) and Lambe (1910) concluded that though Jackson's plates and figures were never published with the original descriptions, a few must have existed and been distributed to paleontologists because the plates were referenced by other scientists (see Traquair, 1877, p. 49; Dawson, 1877, p. 338). Dawson (1877) described two new palaeoniscoid species from the Albert Formation – †Palaeoniscus[m] modulus and †Palaeoniscus[m] jacksonii. Again, no type material was designated. Dawson (1877) also provided additional comments on Jackson's original species and referenced particular specimens figured by Jackson (1851a, b).

In the late 1800s and early 1900s, many scientists commented on how the Albert Formation palaeoniscoids were more similar to species within the genera †Rhadinichthys and †Elonichthys. As detailed by Lambe (1910), Traquair (1877) commented that †Palaeoniscus[m] alberti and †P. cairnsii are closely allied to †Rhadinichthys carinatus (Agassiz). This reassignment was later upheld by Traquair (1911). Traquair (1877) referred †Palaeoniscus[m] brownii to †Elonichthys brownii. Newberry (1899) upheld Traquair's (1877) reassignments. In Woodward's catalogue (1891), †Palaeoniscus[m] alberti, †P. cairnsii, and †P. modulus were referred to †Rhadinichthys alberti, †R. cairnsii, and †R. jacksoni (1891). Woodward (1891) also assigned †Palaeoniscus[m] brownii to †Elonichthys brownii, but also noted that he felt this was a “doubtful” and ill-defined species (Lambe, 1910). Eastman (1908) also referred the New Brunswick species to the genera †Rhadinichthys and †Elonichthys. In the same publication, Eastman described a new species from the Albert shales – †Elonichthys elegantulus – but no type specimen was designated (1908). All of the reassignments by Traquair (1877), Newberry (1889), Woodward (1891), and Eastman (1908) were done without justifications for these reidentifications.

Sometime before 1908, Jackson's original unpublished plates were discovered by Eastman in the Yale Peabody Museum (Eastman, 1908; Lambe, 1910). Using these plates, it was possible for Lambe (1909, 1910) to identify some of Jackson's original type and figured specimens in the collections of the Museum of Comparative Zoology, Harvard, and the Boston Society of Natural History. Using this new information, Lambe (1909, 1910) was able to redescribe the Albert Formation fishes and describe a new taxon, †Elonichthys ellsi. Agreeing with Traquair (1877, 1911), Woodward (1891), and Newberry (1908), Lambe (1909, 1910) referred †Palaeoniscum alberti to †Rhadinichthys alberti and †P. brownii to †Elonichthys brownii. Lambe (1909, 1910) also determined that †Palaeoniscum cairnsii (Jackson, 1851a, b) was not a valid species, nor was †Elonichthys (†Palaeoniscum) jacksoni (Dawson, 1877). Lambe (1909, 1910) also referred †Palaeoniscum modulus (Dawson, 1877) to †Canobius modulus, and the validity of †Elonichthys elegantulus (Eastman, 1908) was called into question.

Reassignments of Albert shale fishes continued in the later 1900s. Moy-Thomas (1938) commented that Westoll considered †Canobius modulus to be †R. alberti and agreed with Westoll that †C. modulus is synonymous with †R. alberti. Sternberg (1939) studied newly collected specimens from the Albert Formation and assigned these specimens to †R. alberti. Sternberg (1939) cautioned though that there were differences in measurements and ratios between the specimens designated as †R. alberti, suggesting that there is more than one species within this taxon or that the species is characterized by a great degree of variation. Gardiner commented on the Albert Formation fishes in his catalog of Canadian fossil fishes (Gardiner, 1966). Miller and McGovern (1996) published a preliminary report describing palaeoniscoids tentatively identified as †Elonichthys from the Albert shales in Norton, New Brunswick, Canada. This has been the last work that has investigated the actinopterygian fauna of the Albert shale formation until this current work.

Jackson's original taxa – †Rhadinichthys alberti, †R. cairnsii, and †Elonichthys brownii – are discussed here. Some of these specimens have a long history of moving from one museum collection to another and have changed catalog numbers over the years. This makes identifying specimens mentioned in older literature difficult. Tables 1 and 2 detail the numerous identities of the original Jackson and Lambe specimens, respectively. Two questionable taxa from the Albert Formation – †Rhadinichthys elegantulus (Gardiner, 1966, states that the type is in the Museum of Comparative Zoology, Harvard (MCZ); no number given) and †Canobius modulus (Gardiner, 1966, states that the type is in Redpath Museum, McGill University; no number given) have never had type specimens designated. The statuses of these taxa are examined here as well.

The majority of the actinopterygians described from the Albert Formation have been placed within the poorly defined and paraphyletic genera †Palaeoniscum, †Elonichthys, or †Rhadinichthys. This problem is intensified by the fact that the type specimens of some of these genera are poorly preserved and uninformative. The specific problems with these genera are described below, as well as suggestions on how to deal with these problems.

Overall, the original taxa described from the Albert Formation are represented by poorly preserved type specimens, undiagnostic descriptions, and reassignments from one paraphyletic genus to another paraphyletic genus. These difficulties are reviewed below for †Rhadinichthys alberti, †Rhadinichthys cairnsii, †Elonichthys ellsi, †Elonichthys (†Rhadinichthys) elegantulus, and †Canobius modulus.

Six species have been described from the Albert Formation, but the validity of the majority of these species has been questioned. The type specimens are either unknown, or poorly preserved. On top of this, the Albert Formation palaeoniscoids have been assigned to genera that are poorly understood, represented by poor type material, and are known to be paraphyletic. Even more problematic is how every described species has been moved from one poorly described genus to another without much justification. What we are left with is a situation where the palaeoniscoids from the Albert shale formation cannot be redescribed at this time.

There are hundreds of specimens of Albert Formation palaeoniscoids in museums such as the Museum of Comparative Zoology, Harvard; the Yale Peabody Museum; the Canadian Museum of Nature; and the New Brunswick Natural History Museum. Some of these specimens are well preserved and have been previously placed within †Rhadinichthys alberti, †Elonichthys brownii, or even †Rhadinichthys cairnsii. The problem with this is that these better preserved specimens show features that are not visible in the types of the species they have been assigned to. Compounding the problem is that the features they do preserve have been used in redescriptions of the taxa. It is not possible to confidently determine if these forms belong to any species within †Rhadinichthys, †Palaeoniscum, or †Elonichthys.

Problems with the paraphyly of †Rhadinichthys, †Elonichthys, and †Palaeoniscum have far-reaching effects beyond understanding the palaeoniscoids from the Albert Formation. It also impairs our understanding of the systematics of lower actinopterygians overall. Multiple species have been placed within these genera, meaning we do not understand a large portion of the lower actinopterygian diversity. Also, though many species have been described as belonging to †Rhadinichthys, †Elonichthys, and †Palaeoniscum, investigations into the relationships of lower actinopterygian fishes may include one representative species from each genus, if any. Including only one taxon from a genus known to be paraphyletic excludes a large amount of diversity and impedes recognizing problems with these genera. Future work on lower actinopterygians will necessitate redescriptions of †Rhadinichthys, †Elonichthys, and †Palaeoniscum. Until such work is done, we do not understand a large portion of the Carboniferous and Permian actinopterygian diversity.

Problems with these genera have been known since the late 1800s. A concerted effort must be made to address the problems with †Rhadinichthys, †Elonichthys, and †Palaeoniscum. The condition of the type specimens may necessitate that the types are the type and only specimen of each genus. The other species placed within †Rhadinichthys, †Elonichthys, and †Palaeoniscum may need to be reevaluated and placed in new genera that can be described on the basis of informative type species. This may be the only prudent way to deal with large cosmopolitan genera described on the basis of uninformative type species. Such an undertaking is beyond the scope of this study but must be done in the near future.

Rhadinichthys, Elonichthys, and Palaeoniscum were the three main Paleozoic palaeoniscoid genera in the mid-1800s. Palaeoniscum is a Permian genus and should not be considered for Early Carboniferous specimens independent of the problems of defining this genus. As for Rhadinichthys, some of the specimens from the Albert shales may belong to this genus; nevertheless, the preservation of the type specimen is so poor that assignment of this specimen to Rhadinichthys is not prudent. Future work should identify a new well-preserved specimen from museum collections and start from the beginning with the description of a new taxon.

Though the fishes from the Albert Formation subscribed to the genera †Rhadinichthys, †Elonichthys, or †Palaeoniscum cannot be commented on further here, a new specimen and taxon can be described. While examining specimens of palaeoniscoids from the Albert Formation, a specimen that represents a form quite different from the type specimens of the previously described fishes was uncovered. This specimen is well preserved and can be differentiated from the previously described taxa, regardless of the condition of those type specimens. One specimen, which constitutes a new taxon from the Albert Formation, is described below.

The new taxon is represented by a single specimen collected from the Albert Formation and housed in the Yale Peabody Museum (YPM). The fossil specimen and its latex peel were examined by stereomicroscopy. The latex peel allowed for three-dimensional views of the specimen preserved as a negative impression. The fossil and peel were examined side by side. Photographs were taken with a Canon XSi digital camera equipped with a macro lens. Illustrations were prepared using a camera lucida, and digital illustrations were prepared using Adobe Photoshop and Illustrator programs.

Two naming conventions have been used in reference to the two paired bones in the skull roof of actinopterygians – frontal and parietal or parietal and post-parietal. The first set of names is based on tradition, whereas the second is based on homology (Jollie, 1962; Schultze, 2008; Wiley, 2008). For further discussion on the problems the traditional naming convention poses, especially to phylogenetic analyses, see Schultze (2008) and Wiley (2008). The naming convention based on homology, parietal and post-parietal, is used here in the taxonomic description. Bones are identified as the dermosphenotic(s) and dermopterotic following the criteria of Poplin (2004). The bones of the snout are identified following the nomenclatural scheme presented by Mickle (2015).

Ao, antorbital; aop, accessory opercular bone; bsi, inserted body scales; d, dentary; df, dorsal fin; dh, dermohyal; dpt, dermopterotic; drs, dorsal ridge scales; dsp, dermosphenotic; ex, extrascapular; io, infraorbital; lg, lateral gular; mdr, median dorsal rostral; mg, median gular; n, nasal; op, operculum; pop, preoperculum; p, parietal; pp, post-parietal; ps, presupracleithrum; pt, posttemporal bone; sc, sclerotic; scl, supracleithrum; so, suborbital; sop, suboperculum; sup, supraorbital; vr-pmx, ventral rostro-premaxilla.

BSNH, Boston Society of Natural History; CMN, Canadian Museum of Nature; MCZ, Museum of Comparative Zoology; YPM, Yale Peabody Museum, Yale University, New Haven, Connecticut.

The snout is prominent and heavily tuberculated. A median dorsal rostral bone forms the anterior-most portion of the snout (Figs. 5–6) The posterior margin of the median rostral bone contacts the parietal, whereas the ventral margin contacts the ventral rostro-premaxilla (Figs. 5–6). The lateral margin of the median dorsal rostral is notched. This notch forms the median margin of the anterior narial opening. The median dorsal rostral bears tubercles ventrally and short ridges of ganoine dorsally. These ridges are longitudinal to diagonal in orientation (Figs. 5–6).

A large nasal bone lies lateral to the median dorsal rostral and anterior to the orbit (Figs. 5–6). The anterior margin of the nasal is notched. This notch forms the medial margin of the anterior narial opening. The posterior border of the nasal is also notched in two different locations. The ventral-most notch forms the anterior margin of the lateral/posterior narial opening. Dorsal to this notch, there is a protuberance that extends off the posterior margin. Dorsal to this protuberance is the second notch on the posterior margin of the nasal for the supraorbital bone (Figs. 5–6). The nasal bears a complex pattern of ganoine. Unlike many other Carboniferous palaeoniscoids that bear long vertical ridges of ganoine, †Lambeia pectinatus has a nasal bone with tubercles and short ridges ventrally and longer diagonally oriented ridges mid-bone (Figs. 5–6).

There is a heavily tuberculated bone ventral to the median rostral and nasal bones (Figs. 5–6). This bone is identified here as the ventral rostro-premaxilla following the terminology of Mickle (2015). The anterior tip of the maxilla is ventral to the posteroventral margin of the ventral rostro-premaxilla. The ventral rostro-premaxilla does not bear teeth, though it is possible that small teeth are obscured by the heavy amount of tuberculations. No canal is visible in this bone, but this could also be because of the heavy ganoine tuberculations. It is termed a ventral rostro-premaxilla because the only criterion that can be used to identify this bone is its placement, information on the placement of canals in any of the snout bones is lacking because of the heavy ganoine ornamentation. The bone lies anterior to the maxilla, ventral to a median rostral bone, and separate from an antorbital bone. The size of this bone and placement suggest that it is not simply a premaxillary bone. This bone is physically located in the area where premaxillary and rostral bones are found.

Posterior to the ventral rostro-premaxilla, posteroventral to the nasal and dorsal to the anterior tip of the maxilla is an antorbital bone. This bone is referred to as an antorbital because of its position and the putative canals in this bone (Figs. 5–6). The antorbital is roughly triangular in shape and there is a row of sensory pores, illustrated with filled gray circles in Fig. 6. The antorbital forms the anteroventral margin of the orbit.

Ventral to the orbit is a thin rectangular infraorbital bone (Figs. 5–6). This rectangular bone is large enough to bear small tubercles of ganoine. This infraorbital bone contacts the posterior margin of the antorbital bone and the anterior margin of a large crescent-shaped infraorbital bone in the posteroventral corner of the orbit (Figs. 5–6). The crescent-shaped infraorbital bone bears tubercles of ganoine and traces of the main infraorbital canal near the anterior border of the bone. There are pore canals that branch off the main infraorbital canal preserved near the posteroventral margin of the bone. This bone is disturbed and broken in half by the inward collapse of the dorsal half, but it can be reconstructed to its original crescent shape (Figs. 5–6).

A single dermosphenotic is located in the posterodorsal corner of the orbit (Figs. 5–6). The anterodorsal and posteroventral margins of this bone cannot be made out with any confidence, but it appears that this bone is narrower anterodorsally than posteroventrally. There are thin short ridges of ganoine at about mid-bone and elongated tubercles anterodorsally.

The dermosphenotic abuts against a large dermopterotic (Figs. 5–6). The dermopterotic is ventral to the parietal and post-parietal bones and approximately the length of these bones combined. Dorsal of the orbit and posteroventral to the nasal bone is a rectangular bone (Figs. 5–6). This bone is not a sclerotic bone – a separate sclerotic is preserved ventral to this bone in question (Figs. 5–6). This rectangular bone bears short ridges of ganoine that are different in orientation and size from those on the nasal. This bone fills in the space created by the dorsal-most notch on the posterior margin of the nasal bone. This bone is identified as a supraorbital. The posterior margin of the supraorbital comes in contact with the anteroventral margin of the parietal.

The margins of the parietal bones are difficult to determine. The parietal contacts the nasal, dorsal rostral, and supraorbital bones anteriorly, the dermopterotic laterally, and the post-parietal posteriorly (Figs. 5–6). The parietal bears short ridges of ganoine along the length of the bone and a few elongated tubercles. The post-parietal is trapezoidal in shape, with the medial margin of the bone being longer than the lateral margin (Figs. 5–6). Pit lines are not apparent because of the heavy ganoine ridges present on this bone. The post-parietal is about a third of the length of the parietal.

A thin band of bone posterior to the post-parietal and the dermopterotic is an extrascapular bone (Figs. 5–6). There is no ganoine ornamentation on this bone, nor can it be determined if this is a series of bones or a single bone.

Posterior to the circumorbital bones lies a series of suborbital bones (Figs. 5–6). Though this region is disturbed by the inward collapse of the infraorbitals and suborbitals, it is clear that there are numerous suborbital bones. The suborbitals are roughly arranged in vertical rows – the first row is posterior to the circumorbital bones, the second is between the first and third rows of suborbital bones, and the third row contacts the anterior margin of the preoperculum (Figs. 5–6). The third row of suborbitals is composed of two large bones (Figs. 5–6). The dorsal-most of the two large suborbitals has a rounded convex anterior margin. Ventral to this is a triangular suborbital bone with rounded corners. The anterior border of this suborbital bone is slightly concave. Both of these suborbital bones bear fine diagonal ridges of ganoine.

The second row of suborbitals is also composed of large bones. Two large bones with fine ganoine ridges are present and overlain in sections by the suborbital bones in rows 3 and 1 (Figs. 5–6). The first row of suborbital bones is the area with the most disturbance. It seems that multiple smaller suborbital bones are located posterior to the dermosphenotic and infraorbital three (Figs. 5–6).

The maxilla is a large bone with a deep and broad postorbital plate (Figs. 5–6). A posteroventral process off the postorbital plate overlaps the posterodorsal margin of the lower jaw. The maxilla tapers to a narrow arm that extends anterior to the orbit. This narrow arm terminates ventral to the heavily tuberculated ventro–rostro-premaxilla. There are fine, minute, needle-like teeth on the oral margin of this portion of the maxilla. Large conical teeth are seen on the oral rim ventral to the postorbital plate, with small conical teeth inserted between the large ones.

The ganoine ornamentation of the maxilla is complex (Figs. 5–6). The anterior-most portion of the maxilla, up to the anterior-most corner of the infraorbital in the posteroventral corner of the orbit, is heavily tuberculated with closely set tubercles. Below the crescent-shaped infraorbital, the maxilla bears short diagonal ridges of ganoine. The pattern of ganoine on the maxilla then changes to fine, faint, and more horizontally oriented ridges along the postorbital plate. Fine vertical ridges are present along the posteroventral process of the maxilla. Though the anteroventral margin of this process is disturbed, it appears that these vertical ridges of ganoine break down to fine tubercles at this margin. The ventral margin of the postorbital plate has its own ganoine pattern – here ornamentation consists of short, closely set, vermiform ridges that are horizontal to vertical in orientation. Directly ventral and posteroventral to the orbit, there is a narrow band along the dorsal-most border of the maxilla that is smooth and does not bear any ganoine ornamentation. This is an area of articulation between the maxilla and the overlying infraorbitals.

The preoperculum is hatchet shaped, anteriorly inclined, and contacts the dorsal and posterior margins of the postorbital plate of the maxilla (Figs. 5–6). Dorsal to the maxilla, the preoperculum is expanded, whereas posterior to the maxilla, the preoperculum is a tall, narrow arm arched around the posterior margin of the maxilla. The anterior margin of the preoperculum sutures with two suborbitals. These suborbitals are situated within a concavity made by the arms of the expanded region of the preoperculum. There are short ganoine ridges along the posterior margin of the preoperculum and fine horizontal ridges on the expanded region of the preoperculum, dorsal to the maxilla.

Posterior to the preoperculum and anterior to the operculum is a tall and narrow wedge-shaped dermohyal (Figs. 5–6). The dermohyal extends from the anterodorsal corner of the operculum to about half the depth of this bone. The dermohyal bears short ridges of ganoine parallel with the anteroventral and posterodorsal margins of the bone. These ridges are similar to those found on the expanded region of the preoperculum. Posterior to the preoperculum, ventral to the dermohyal, and along the anterior border of the operculum is a series of accessory opercular bones. The largest bone of the series is found near the anteroventral corner of the operculum. This bone is wider ventrally than dorsally and bears diagonal ridges of ganoine. There are three accessory opercular bones dorsal to the expanded ventral bone. The dorsal accessory opercular bones are small and rhombic and bear faint ridges of ganoine.

Individual bones that make up the lower jaw cannot be distinguished. Overall, the lower jaw is a large bone that is longer than the maxilla (Figs. 5–6). The lower jaw extends slightly anterior to the ventral rostro-premaxilla. Vertically oriented conical teeth are along the oral rim of about the anterior half of the lower jaw. In between these teeth are smaller conical teeth. There are also teeth medial to this series of conical teeth. The medially placed teeth are conical, short, and closely set. Anteriorly, the lower jaw bears ganoine tuberculations. These tubercles are present until about the posterior margin of the ventral rostro-premaxilla. Posterior to the ventral rostro-premaxilla to the posterior margin of the median gular, the lower jaw is ornamented with short diagonal ridges. At the posterior-lateral margin of the median gular, there is a change in ganoine ornamentation on the lower jaw to fine and long horizontal ridges. Near the posterior margin of the lower jaw, these horizontal ridges curve dorsally towards the posteroventral process of the maxilla.

The operculum is a rectangular bone that is anteriorly inclined and about twice the depth of the suboperculum (Figs. 5–6). The operculum bears fine diagonal ridges of ganoine that are not as closely packed as the ganoine ornamentation on other bones. The suboperculum is vertically oriented and rhombic in shape (Figs. 5–6). There are short diagonal and vertical ganoine ridges on this bone. The suboperculum is taller posterodorsally than anterodorsally.

Gulars and branchiostegal rays are visible in lateral view (Figs. 5–6). The proximal portion of the median gular closest to the lower jaw is missing so that only the distal tip of this bone can be commented upon. The distal portion of the median gular bears short ridges of ganoine and a few tubercles. Posterior to the median gular lies the first of two lateral gulars. Both lateral gulars are teardrop shaped and bear short ridges of ganoine. When seen in lateral view, these ridges are diagonal on the first lateral gular but more horizontally oriented on the second lateral gular. The second lateral gular bears a prominent pit line. Posterior to the lateral gulars follows a series of branchiostegal rays (Figs. 5–6). The branchiostegals are obscured just anterior to the posteroventral corner of the lower jaw, making determinations of their number and shape difficult. Anterior to this point, there are two branchiostegal rays. Posterior to this point, there are seven rays. If the size of the branchiostegal rays anterior to the posteroventral corner of the lower jaw are consistent with those directly anterior and posterior to this area of ambiguity, the total number of branchiostegal rays is estimated to be around 12.

The posttemporal is a large bone with a rounded posterior margin (Figs. 5–6). The posttemporal bears prominent ridges of ganoine that extend to the posterior border of the bone to form a serrated posterior margin. The posttemporal overlaps the dorsal border of the ventrally located supracleithrum.

A rounded presupracleithrum is situated near the posterodorsal corner of the operculum, ventral to the posttemporal and overlapping the anterior margin of the supracleithrum (Figs. 5–6). The presupracleithrum bears diagonal ridges of ganoine that extend to the posterior end of the bone, giving the presupracleithrum a serrated posterior margin. The supracleithrum lies posterior to the operculum, overlapped by the presupracleithrum and posttemporal. The supracleithrum is about the same depth as the operculum (Figs. 5–6). At about two-thirds down the depth of the bone, the posterior margin of the supracleithrum is concave. The posterodorsal and posteroventral margins of the supracleithrum are convex. The supracleithrum bears strong ridges of ganoine. The ridges in the posterodorsal portion of the bone are curved, whereas the ornamentation on the anterodorsal portion of the bone consists of straight diagonal ridges. These ridges are more vertically oriented near the anterior border of the bone. The ridges in the ventral portion of the supracleithrum are slightly diagonal to vertical in orientation.

Ventral to the supracleithrum is a tall but narrow crescent-shaped postcleithrum (Figs. 5–6). The postcleithrum bears short diagonal ridges on the dorsal half of the bone and vertical ridges down the ventral half. In the dorsal half of the bone, these ridges extend to the posterior margin, giving the bone a pectinated posterior margin. The smooth area anterior to the postcleithrum and posterior to the suboperculum is the cleithrum, but the shape of this bone cannot be determined.

There is a series of large dorsal ridge scales anterior to the origin of the dorsal fin (Figs. 4, 7). The dorsal ridge scales are not continuous to the occiput; rather, they begin at scale row 20. The dorsal ridge scale series consists of six large scales. The posterior border of the preceding scale overlaps the anterior margin of the subsequent scale. The first dorsal ridge scale has more of an acuminate posterior margin compared to the subsequent scales with blunt and rounded posterior borders (Fig. 7). The first three dorsal ridge scales have serrated posterior margins (Fig. 7). These pectinations are formed by ridges of ganoine that run down the center of the ridge scales. The ganoine ridges on the lateral margin of these ridge scales are curved to follow the convex lateral margin of the scale.

The first and second dorsal ridge scales correspond to two ventrally placed rows of body scales (Fig. 7). Because of the overlap of the dorsal ridge scales, the second body scale row that is ventral to the posteroventral margin of the first dorsal ridge scale also overlaps the posterior margin of the second dorsal ridge scale. This gives the appearance of the body scale rows being inserted between the dorsal ridge scales (Fig. 7). This pattern is seen between dorsal ridge scale one and two, as well as two and three. This may indicate that the two vertical scale rows correspond to one myomere rather than a 1 : 1 ratio of scale rows to myomeres. The third dorsal ridge scale is unique in that it is the only one of the six to correspond to just one body scale row. Dorsal ridge scales four through six all correspond to two body scale rows (Fig. 7). There are no pectinations on the posterior edges of dorsal ridge scales four through six. These ridge scales also differ in shape from the anterior three ridge scales. Ridge scales four through six do not have the same rounded appearance as the first three, and they have more pronounced convex posterior margins (Fig. 7).

It is questionable whether ridge scales are present between the pectoral and pelvic fins. Between the pectoral and pelvic fins there is an area of ambiguity caused by the body of a smaller palaeoniscoid overlapping the ventral margin of the larger specimen in YPM 8664. Partially visible are two rounded structures that do not bear pectinated posterior margins or ganoine ornamentation. These could potentially be ventral ridge scales between the pectoral and pelvic fins. There are enlarged scales anterior to the anal fin and potentially the vent of the fish.

In order to describe the body scales, the body has been broken down into different regions. These regions are detailed and illustrated in Fig. 8. In YPM 8664, scales in region A1 (dorsally placed scales posterior to the skull roof) are heavily ornamented with ridges of ganoine and have strongly pectinated posterior margins. Posterior to the posttemporal, the posteroventral margins of the scales have four to five serrations, but at the level of the supracleithrum, there is a change so that the entire posterior margin is pectinated. In region A2 (mid-body scales posterior to the pectoral girdle), scales are generally pectinated and bear diagonal ridges of closely set ganoine. These scales are rhombic at the level of the supracleithrum. At the level of the supracleithrum, at scale rows 1 through 3, there are diagonal ridges of ganoine dorsally and curved ridges of ganoine near the ventral border of the scale. Scales ventral to the supracleithrum at about the level of the postcleithrum are taller, rectangular, and narrower. These scales only bear diagonal ridges of ganoine. The scales bearing the lateral line are notched posteriorly. The notch is more superiorly placed in this region then in regions B2 or B3. The lateral-line-bearing scales are pectinated.

In region A3 (ventrally placed scales posterior to the pectoral girdle), the scales are rhombic, but there is a trend of the scales being narrower and more rectangular in shape the more ventral they are on the body. Regardless of their placement, all of the scales in the A3 region bear close-set diagonal ridges of ganoine.

Scales in region B commence at scale row 6. Scales in region B2 (mid-body scales from scale rows 6–12) are tall, narrow, and rectangular in shape. The posterior margins are pectinated; the pectinations are formed by the closely set diagonal ridges of ganoine that ornament the scales. The lateral-line-bearing scales are notched posteriorly.

Scales in region B3 (ventrally placed scales from scale rows 6–12) are very unique. Here, the ventral margins of the scales, as well as the posterior margins, are pectinated (Fig. 9). These scales also bear ridges of ganoine.

Scale region C starts at scale row 13. Scales in region C1 (dorsally placed scales from scale row 13 to origin of dorsal fin) are more teardrop shaped in appearance as compared to the anteriorly placed rhombic and rectangular scales. These scales bear diagonal ridges of ganoine. The posterior edges are sometimes pectinated with five to six serrations.

The posterior borders of scales in region C2 (mid-body scales from scale row 13 to origin of dorsal fin) are pectinated with closely packed fine serrations. The exact number is hard to determine because of how close the serrations are, but there are at least a dozen serrations per scale. The scales that are located at the level of the ventral portion of the supracleithrum are more rectangular in shape than the rhombic scales above and below this point. The lateral-line-bearing scales are noticeably notched posteriorly. This notch is located mid-scale.

Like the scales in region B3, the scales in region C3 (ventrally placed scales from row 13 to origin of dorsal fin) are serrated on both the posterior and ventral borders. These scales are ornamented with fine diagonal ridges of ganoine. The more ventrally placed scales are shorter and more rectangular in shape than the more dorsally placed rhombic scales in this region.

Scale region D begins at the origin of the dorsal fin. The scales in region D1 (dorsally placed scales from origin of dorsal fin to preserved end of specimen), are short but rectangular in shape. They are closely packed and overlapping. There are faint signs of horizontal ridges of ganoine. The first four scales ventral to the dorsal fin are slightly different – these scales are about 2 times the depth of the others and rhombic in shape.

Scales in region D2 (mid-body scales from origin of dorsal fin to preserved end of specimen) are rhombic and not as tall as the more anteriorly placed scales. The posterior edges of scales in region D2 are pectinated with six to seven serrations. The ganoine ornamentation is not as dramatic as that of the more anteriorly placed scales but there are faint horizontal to diagonal ridges. Scales in the region D3 (ventrally placed scales from origin of dorsal fin to preserved end of specimen) are similar to those described for region C3, but the scales in D3 lack the serrations on the ventral margin. There are a few scales in D3 that have serrations on the posterior margin.

The pectoral fin is large, fan-shaped, and formed by highly bifurcated and closely packed fin rays (Fig. 10). Fringing fulcra line the leading edge of the pectoral fin. The fulcra are thicker and longer proximally and decrease in size down the height of the fin so that the distal-most fringing fulcra are fine and short. Proximally, the fringing fulcra from the opposite side of the fin are seen. The two sides of the fulcra meet in midline to form a V-shaped unit.

All of the fin rays that make up the pectoral fin are segmented. The fin rays are bifurcated numerous times. The first bifurcation occurs in the proximal quarter of the fin. The fin rays bifurcate at least one more time distally so that the distal-most portion of the fin is made up of fine delicate fin rays. This makes determining how many times the fin rays bifurcate and detailed illustrations of the distal portion of the fin difficult. The first two fin rays do not extend down the entire depth of the fin to contribute to the distal margin, instead, the highly bifurcated fin rays posterior to the first two fin rays fill in and form the distal margin of the fin.

There is a large triangular pelvic fin that spans four scale rows and contains 25 fin rays (Fig. 10). The pelvic fin originates at scale row 12. Like the pectoral fin, the fin rays that form the pelvic fin are highly bifurcated. The first bifurcation occurs close to the proximal margin of the fin at about the second or third segment. At about mid-depth of the pelvic fin, the fin rays bifurcate again. Because of all the bifurcations, the distal region of the pelvic fin is made of very fine closely packed fin rays. There are fringing fulcra along the leading edge of the pelvic fin, but the fulcra are not as large or dramatic as that of the pectoral fin. The most proximal structure on the leading edge of the pelvic fin is different from the distal fringing fulcra – it is a single median structure that looks more like a ridge scale than the start of the fringing fulcra series.

The triangular anal fin originates posterior to the origin of the dorsal. The anal fin is large – spanning about 12 scale rows and containing 42 fin rays (Fig. 10). Like the other fins, the anal fin has highly bifurcated fin rays and fringing fulcra. The fringing fulcra are clearest on the leading edge of the distal portion of the fin, though it is presumed to be continuous along the entire anterior margin. The caudal fin is not preserved.

The new Albert Formation fish differs from the other actinopterygians described from this site in regard to scale, cranial, and fin characteristics. The scales with pectinated posterior and ventral margins are unique and set †Lambeia pectinatus apart from all other Albert Formation fishes, as well as Carboniferous fishes from other localities. The ganoine ornamentation on the nasal bones is also different from the other previously described Albert Formation fishes, which all show vertical continuous ridges of ganoine. The presence of multiple suborbital bones, a large tuberculated rostro-premaxilla, a single supraorbital bone, the size and shape of the dermopterotic, the presence of accessory opercular bones, and the body shape also distinguish †Lambeia pectinatus from the other Albert Formation fishes. Other Albert Formation fishes have a series of dorsal ridge scales beginning at or just behind the occiput, whereas †Lambeia pectinatus has six dorsal ridge scales that start at about mid-body at scale row 20.

†Elonichthys brownii is of a similar size as the new fish. Though specimens of †E. brownii do not preserve much cranial information, there are enough characteristics regarding scales and fins to support separating †Lambeia pectinatus from †E. brownii. These include the placement of the dorsal fin relative to the anal fin, with the dorsal fin originating slightly anterior to the anal fin in †Lambeia pectinatus and far in advance of the anal fin in †E. brownii, and the presence of scales with serrated posterior and ventral margins in †Lambeia pectinatus and the absence of this scale type in †E. brownii.

While describing †Lambeia pectinatus, certain characters were uncovered that deserve more discussion. These characters include dorsal ridge scales, suborbital bones, and supraorbital bones.

The Albert Formation of New Brunswick, Canada, preserves an array of palaeoniscoid fishes from the Lower Carboniferous. Much of this diversity is not understood because of the need for redescriptions. Albert Formation taxa have been assigned and reassigned to numerous paraphyletic taxa. Attempts at redescribing the previously described taxa from the Albert Formation will not be successful until the genera these taxa have been described as belonging to are redescribed. The taxonomic review of the Albert shale formation fishes has brought attention to the fundamental need for descriptive and revisionary work. In order to form a stronger understanding of lower actinopterygian fishes, we must have a firm foundation when it comes to the taxonomy and systematics of this group of fishes.

On top of revisionary work, there is a need for descriptions of new taxa. Undescribed diversity remains to be uncovered. Here, a new genus and species is described from the Albert shale formation. This new taxon shows affinities to Carboniferous fishes. Many of the morphological features of this new taxon are typically seen in Late Carboniferous forms, not Devonian forms. The presence of a separate antorbital bone and multiple suborbital bones arranged in many rows are two such features. This said, the presence of a single supraorbital bone is only seen in the Devonian †Cheirolepis trailli (Pearson and Westoll, 1979; Pearson, 1982). It was thought that fishes from the Albert Formation, which is right above the boundary between the Devonian and the Carboniferous, may preserve forms that help bridge morphological gaps between Devonian and Carboniferous lower actinopterygians. While there are features seen in Devonian fishes, most notably the single supraorbital bone, †Lambeia pectinatus does preserve characters that are more commonly seen in later Carboniferous forms. †Lambeia pectinatus presents a mixture of Devonian and Carboniferous characters, suggesting these morphological features were present early in the Carboniferous.